Commodore Amiga Bench

Welcome!

As the title suggests, these posts will be like a journal of my interest in the Commodore Amiga.

Produced by Commodore International, the Amiga line of home computers started with the A1000 in 1985 during their domination of the low cost home computer market the C64 filled.  Many models were produced with version changes until ending with their last model, the A1200 in 1996.  The models in order of release are:

A1000 - 1985 - Small case size offered little in expansion options but offered the Keyboard Garage!

A2000 - 1987 - A full-size offering greater expansion capabilities.  An A2000HD hard drive equipped model was offered later.  First Amiga to offer IBM-PC capability with the A2088XT or A2286AT bridgeboard.

A500   - 1987 - Released a couple of months after the A2000 its form factor was somewhere between a C64 aand a C128.  An A500+ variant was released later.

A1500 - 1987

A2500 - 1988

A3000 - 1990

A4000 - 1992

A600   - 1992

A1200 - 1992 - Original Commodore model.  After Commodore's bankruptcy, Escom rereleased a version in 1995 but was discontinued in 1996 thus ending the Commodore brand market presence.

I have a Commodore 64 blog for my rediscovery journey, here if you're interested.

Continue reading to join me on my journey to rediscovering my interest in Amiga.

_________________________

February 19, 2023

I got started with the Amiga 500 in 1987 as the A2000 was waaaaay too expensive for me.  My wife and I used it for playing games (what else?).  We did some word processing but that was the extent.  I tried a hard drive I borrowed from a friend but since copying games and utilities requiring the diskette, I didn't find much use for it.  As for pre-internet activities, I was still dialing BBS's with the C64 and a PC at the time.  I never bought a modem to put on the Amiga.  Even after getting a used A2000 a couple of years later, I was so heavily vested in PC due to the job I had at the time with Electronics Boutique and later, CompuAdd.  It would be ten years later when my wife told me she was pregnant.  From that point, we needed a bigger house and thus nearly everything went into storage and mostly forgotten over the years as we prepared for raising our daughter.

Fast-forward to 2016 and we had moved three times since, built a shed in the back yard and transferred everything in storage for the last 18 years to save on costs.  Should have done it sooner but life had us disorganized after 9-11-01.  Nothing was 'normal' after that.  I had lost the inventory list of what was in storage and not interested in doing another one, I did what I had to do to keep the mobile home now.  I've been doing Jeep things and vacuum tube radio things up to May-2022 when my father passed away unexpectedly.  Having lost much of the motivation to work on the vacuum tube(valve) things I've been dabbling in other interests without much intention.

That is until I happened across a video on YT repairing a Commodore 64.  Adrian's Digital Basement snagged my interest in old computer technology and I started reminiscing about the great 8-Bit Age of the 1970's and 80's.  A few more videos on his channel and I was pretty hooked.  There was a repair series on an Apple ][ Plus a viewer had donated to him.  He was excited about it as he had one as a kid.  He was really into it and his enthusiasm really impressed upon me.  After that series I saw a video about an Amiga 2000 he was given and could he bring it back to life?  Now I was very interested as I used to have one!

The video opened up the question of can one of those be repaired after the battery had leaked.  I remember the clock system needing a battery if you didn't want to keep entering the date and time in after turning it off.  PC's had the venerable DALLAS real-time clock chip block people could order as an option but later main boards came with one built in and used a coin cell like the 2025 or 2032.  That got me to thinking.  I remembered coming across my A2000 and a CRT monitor in the very back of the shed when I was looking for other stuff and wondered.  So the next weekend I went digging and found the computer and monitor.  In the process, I also found nine other boxes of Commodore stuff I didn't think I still had!  Nine freaking boxes!  So I pulled them out too!  I hadn't seen what was in these boxes since June of 1998 mostly due to life just happening.  I thought I maybe had a box or two, but nine?  Life changing events do reprioritize your thinking.

My A2000 with 1084S monitor.
Will be here in a special place on
my desk after evaluation and
hopefully, repair.

The Nine Freaking Boxes!
The two on the right contain the A500, a C64,
a 1541 disk drive, and a 1541 compatible drive
by BlueChip.  The other boxes contain various
hardware and software as well as media.

When I got the A2000 on the bench I had to open it.  With the preknowledge of what to expect, to my mock horror, I too suffered a leaky battery and ensuing corrosion.  Looks worse than Adrian's Digital Basement video.

Just a bit of dirt.  It'll buff out.

Still just some dust.  It'll buff out.

Oh no!  Looks bad but it should buff out.

Um. Well...

I think I'm gonna need a bigger buffer!

All kidding aside, the corrosion is pretty bad.  After cutting the battery off the board, I laid white vinegar-soaked paper towels on the board and CPU as Adrian had done.  Since the NiCad battery is an alkaline, I needed a mild acid to counter it.  I left the towels on for two hours straight to see if any corrosion would come off.  I was pleasantly surprised it did a pretty good job.  The corrosion left is down to scraping off the solder mask, removing the CPU, KickStart ROM and their associated sockets.  All of the electrolytic capacitors in the corrosion zone will be replaced.  I probably should replace all of them regardless.  I do this on all of my vacuum tube radio projects.

This motherboard is REV 4.4.

So below are photos of some of the work done thus far.

After the first vinegar soaking I got the CPU out.

After the second round of vinegar.  Note two
contacts are pretty much gone.

Much light scrubbing with a small brass brush.

You can see how the white milky residue
extends back to the Gary chip.  I pulled the
Fat Lady chip out and did a local vinegar
treatment.  Seems odd that it doesn't
look corroded the same as the KickStart
socket was.

Trying to scrape off the solder mask.  There's
just too much corrosion underneath to trust
it won't continue to react with the copper
traces.  I'll need to break out the Dremel
and the brass wire wheel next.  Desoldering
of the electrolytic capacitors and resistor packs
are next on the list.

So there you have it.  This is as far as I've gotten with the repair after a week.  Next is to run by the hardware store and grab a brass wire wheel for the Dremel and start working on the mask.

I've started the continuity checks and I have not found a break between Via's or socket traces through the corrosion zone.  Getting the sockets off was difficult in that trying to keep from breaking or lifting solder pads.  I lifted one pad but it wasn't connected on the underside.  The Via is still intact though.  Adrian has a hot air rework station so I may get one.  I did cut the CPU socket into two pieces to make it easier to remove.  Refreshing the old solder with new helped a lot as usual but much work remains to be done.

Here's to hoping it can be repaired.

_______________________

February 21, 2023

Flash Update!

The order of 4 machined-pin 64 pin sockets for the Amiga 2000 came in today!  True I only need one at the very least but spares were cheap.  Now one more step closer to repairing this machine.

___________________

February 24, 2023

Two more orders of parts have arrived today.  One, DIP sockets and a couple of 5 pin DIN sockets.  The second order contained a Motorola 68000P8 processor.  I got one just in case the original Signetics version installed on the motherboard failed to work.

I will run by a hardware store to pick up a couple of brass wire wheels for my Dremel.  That will help to remove the solder mask over a larger area of the ground plane.  Scraping the tiny traces of the mask requires a bit more finesse.  This weekend looks to be a productive one.

___________

February 25, 2023

Started today with a trip to a hardware store for a couple of Dremel wire brushes.  I used the cup type brush and it has done okay.  Stainless steel isn't as aggressive as carbon steel so I figured I would have a chance to clean the mask rather than the copper off the fiberglass substrate.  As usual the wire filaments of the wheel break off and after wiping the board down I turned it over and tapped on it to knock any loose strands off.  Several came off onto the bench top.  The board has been inspected and I haven't found any left over.

Next were the rest of the continuity tests and found nothing has been severed from where it should go.  Pulling the three electrolytic capacitors in the corrosion zone was next and my trusty Aoyue 701A++ rework station came through nicely.  No lifted traces and three undamaged capacitors removed. 

Two of these, C404 and C407 are 47uF/16V tested way high above their 20% tolerance.  C404 tested at over 65uF while C407 tested at 63uF.  C811 near the battery tested above its 4.7uF rating at 6.2uF (6278nF).  I have a 50V replacement but since I still need the 47uF caps I'll make up an order of extras I should have on-hand.  Since I still need to solder the CPU and Kickstart ROM sockets in place I've covered the exposed copper with clear nail polish to prevent further oxidizing.  Solder points are unpainted and will be protected with the Kester leaded solder.

Painted with Insta-Dri nail polish.

The 4.7uF showing a 3.4ohm ESR.  Much too high even
if it weren't testing at 6.2uF capacitance.

73uF is what C407 tested at.  ESR was 1.4ohms.

C404 testing at 63uF.  Unacceptable at best.

In case I hadn't mentioned it yet, I did pull the Fat Lady chip and checked the contacts.  No corrosion was found so I may have dodged a bullet there.  Resistor packs seem okay but I'm hesitant to remove them now so I'll test with them in place.  If anything seems off I'll test them next if diagnosis points to something on the address or data bus.

Forward to 2032 hours and I've just ordered the capacitors (multiples of each value) and 5 bussed resistor arrays.  Of the caps I ordered low ESR types.  Only the 4.7uF was rated 35V.  All the others are 16V.  I got 25V-50V depending on price and availability of the same manufacturer, Panasonic.

The reason for the resistor arrays was I was unable to get continuity on one trace that had the array in between.  After checking the trace up to the array from the CPU I was still unable to get continuity even though it looked like it was connected.  Some scraping with the Dremel revealed more corrosion at the solder pad of the last pin.  The trace was broken.  It was so thin I couldn't see it without 10x magnification.  So I decided to forego any fun foodies this week to get these parts.  The Amiga 500 is similar in architecture so any cap rework needed there should already be taken care of.  Estimated time of arrival is Wednesday.  More like Friday or Monday after that.  That's fine.  I have other things to check as I dig deeper into the board.

I pulled the A500 out and removed the memory expansion board from it as it also has a battery backed up clock.  Fortunately this board doesn't have nearly the copper traces and ground plane material as the main board does so the corrosion was limited to the positive post and an adjacent electrolytic capacitor.  There are two so I should have enough when the order arrives to effect repairs and see if it works!

____________________

May 18, 2023

Wow!  Time flies as one ages.  Since the previous post I've stopped working on the Amiga 2000 main board to focus on my wife's A500 R5.  It's in far better shape than the A2000.  First power up and I got ten short and one long flash of the power LED which is stating "Bad Agnus".  I noticed one corner of her socket was cracked so I pulled the chip, cleaned the contacts of the chip and the socket, applied DeoxIt D100 then reinstalled Fat Agnus.  It still came up with the same error.  I figured the removal of the bad trap-door RAM expansion board might help and low and behold, it booted up!  Since I don't have a working 15kHz horizontal frequency monitor, I could only use the monochrome composite port.  Clear and sharp, it tested good on all fronts but color in SYSINFO including, every-freaking-keyboard-key!!!  I am shocked beyond words!  I plugged the A1011 external diskette drive and after a cleaning disk, attempted to load the Amiga Extra's drawer.  Everything worked.  Though the boot disk is a copy of the master WB 1.2 that used to work, I formatted that disk and copied the contents to it.  It booted up fine and thus I was able to continue on.  Or so I thought.  After several minutes I caught the acrid odor of burning electronics and saw a whisp of smoke from a resistor adjacent to a filter capacitor in the video circuitry.

Shutting off the machine immediately I hoped I had stopped the destruction in time.  The color bands were pretty much gone so I had to desolder it and test it.  I checked the value and compared the measurement to what the schematic showed.  Finding it was still well within tolerances I reinstalled it.  I felt pretty lucky at this point.  Next I pulled its companion filter capacitor and tested it.  It tested as a dead short.  This probably happened as it tried to reform during the several minutes power was applied.  It was on the verge of failure right out of storage and I nearly screwed the pooch on this one.  I know better and got impatient with the trouble I was having with the A2000 main board.  I desoldered the other filter capacitor and it tested good but for how long?  I wasn't going to trust it so I replaced both. There are several other smaller value electrolytic capacitors spread throughout the board.  I'll replace those later.  They aren't as critical as the two for the Graphics Hybrid.  If the hybrid goes bad it's done.  There won't be any usable 23 pin graphics port after that and the only usable interface will be the RGB2HDMI adapter.  I have a 1084S-D (Daewoo manufactured) monitor but it requires work.

From the last post above I've gotten the system up and running in color on a Vizio LCD TV/Monitor with an RGB2HDMI adapter I purchased from jaakko-o on ebay in Finland.  At the time he was the only one with stock on hand so I paid the extra shipping from overseas and got it within 3 weeks.  Not bad!  I had to buy a Raspberry Pi Zero W in a starter kit and received it only a few days later.  With the OS loaded to drive the video adapter after assembling I installed the setup and tested it.  The color was vivid, clear and the text was sharp.  I'm very happy with the result.

So next I started assembling the A501 R5 memory expansion re-make jond5, also of eBay, had for sale.  It's basically a near-exact replica of the original.  His Bill Of Material (BOM) is included in the product description.  He did do a video of assembling it too on YT.  He also recreated the A501 R6 model as well.  I probably should have gotten that one but it uses four DRAM chips instead of 16 like my old one and I wanted to reuse those.

Speaking of which, after assembling the board and moving the DRAM chips over, I got the same LED flashing sequence but with a Gray to Yellow error screen.  This basically says bad RAM.  So suspect a couple of these old chips are bad.  The Varta battery on the old board leaked and corroded many pins on the chips closest to it.  I've since ordered 18 of the 41256A-12 DRAM chips from JameCo.  I hope to have them by the 25th.  Without a DRAM tester I don't have a clue as to which ones are truly bad as I don't have a working machine to test them in.  Unless I get the A2000 working that is. 

That about catches up the events to this point.  I hope to be more timely in my updates on these machines in the future.

TTFN!

________________________

July 25, 2024

A lot has happened since February of 2023 so here's a summary update on the Amiga's.

A2000: Finished finding all the broken traces.  One lost its via and required a through-wire as a jumper.  Six others had corroded and couldn't be repaired with drag soldering techniques.  After initial power up I finally got a red screen.  Pulling the Agnus chip and inspected it under a 10x magnifier I did see some corrosion.  Inspecting the socket more closely than before also yielded some corrosion.  A pin was bent and in trying to bend it back, broke off.  That was a fun moment and at that moment I knew, I'd F'd up.  So I replaced the socket, cleaned Aggie real good and popped her back in.  Tadaaah!  It worked.

Next I installed the DiagRom and ran through everything there.  Nothing came up with an error so I did the only other thing to do.  I reinstalled the KickStart 1.3 ROM and booted up on an AmigaKit 1.16 disk.  Sure enough, it came right up.  The only test that failed was the Tick signal.  So after powering off and switching jumper J300 to pins 2-3, it tested good then.  It was late and I decided to head to bed.  About a week later I got back to it only to find it wouldn't power up.  After some poking around I saw no 5V or 12V rail.

I took the PS out and tore it down to do some checks.  Nothing seemed out of the ordinary but I suspected it would be the transformer.  So I dug around and found a LiteOn power supply to replace the other one with.  Found it here in the US and grabbed it.  It needed some cleanup work and a new fan but all voltages tested good under load.  This power supply does have the two bottom-rear screws in a different location as they are at least 1mm (likely 2mm) off.  Only the two top-rear screws and internal front lower screws line up.  So what!  Send it!  And it has performed well since.  Oh, and the Noctua fan is so quiet I can't hear it unless the room is quiet. Nice!

With power restored I finished testing round three.  No hiccups.  Even the RAM tested flawless after a full night of looping tests.  Couldn't be more happy.  After getting the GoTek set up internally and working as DF0: it was time to install the SCSI controller.  Turns out I don't have a working SCSI drive left in my stash of 8.  None would initialize with two failing to start even with the START jumper added to force it to.  So bad to the buying board.  I'd seen an IDE controller with 4 drive capability but it's a bit pricey so I went for an IDE68k with GottaGo FastRAM built on.  It's a neat little package.  I bought the CPU relocator board and installed the IDE-RAM board onto that.  Connecting the jumper wires to grab INT/2 and fulfill its Auto Config function, it came right up and everything tested good!

After spending several days testing IDE drives directly, including the CF Card adapter, DVD Drive as a CDROM drive, and utilizing a SATA2IDE adapter, I have plenty of drives to pick from now.  Ultimately I went with a 160GB drive and the DVD drive.  I also learned a lot about connectivity and the order in which stuff has to work.  I also found out that even with the KS 3.2.2 ROM the OS 3.2 CD won't boot.  Can't tell you why.  It just won't.  So I had to copy all the disk images from the CD to the GoTek and load the OS.  It took a long time.  But, once it was done it was a matter of getting the system to recognize the optical drive.  And that made the system fairly complete.  I would liked to have had the extra IDE port for the CF card for data transfers between the PC and the A2000.  The serial link is only 19200, 8, N, 1.  Too slow to be effective.  A faster solution is being looked at but again, pricey stuff.

The Amiga 500 is pretty much as it was back then.  All my time has been spent on the A2000 and getting it up and running reliably.  The A500 just doesn't have any issues, save for the 15kHz horizontal refresh no modern monitor supports.  So I either get the 1084S fixed or get another RGB2HDMI adapter.  The adapters are easier to get as is the Pi Zero 2W.

I purchased a 1541 drive from Raven-Wolf Retro Tech as mine has the infamous Mitsumi mechanism.  So it may be the parts donor to the one purchased.

July 27th I'll be participating at a local community Maker/Thing Faire and friend of mine invited me to.  It's the first one and it will be held at his church so he wants as many as can participate there.  I'll talk about my passion for vintage equipment and computers by bringing the A2000, C64, 1541, A1011, Pi1541 (single drive), and two monitors with me.  I'll also have some 3D print examples to show how they help enhance/repair vintage equipment along with some of the tools of the hobby.  The A2000 will run some demos and the C64 will have a couple of games that can be played.  I hope it will be a success and generate more interest in home-brewing electronics projects through preservation.

That's about it for this post.

__________________________

June 14, 2025

I'm baaaaaack!

Wow!  Hard to believe another year has gone by!  Time seems to go faster the older you get.  When I hit 60 in December, will I get to warp 1?

Let's get to business.  I'm having trouble with the 2000 again.  Back last year I added some upgrades like the 1MB 8372A Fat Agnus chip.  With a jumper setting and cutting a jumper pad all of the onboard RAM is seen as a single block of ChipRAM.  This is shared between the CPU and custom chips but managed by Agnus.  Before there was 512K of CPU RAM called FastRAM.  Only the CPU had access to it.  Graphics, sound and other things used the other half of the 1MB total RAM of 512K ChipRAM.  I could play anything I could before.  I could run any apps that worked before as well and pursued the 2MB upgrade option.  I purchased one of those from Independant Computers over in Germany.  i tried to get one from a US supplier.  After asking about one he said he could do it and that was the last I heard from him for three weeks.  About ten days into that time I ordered it from Icomp.de.  It came in about three weeks later due to customs but it was really packed well so no damage.

I did the modifications for use with the 2MB enhancement and was impressed with the performance!  I had a hard drive, CDROM, 8MB of RAM plus the 2MB of ChipRAM.  I booted up F/A-18 Interceptor and wow!  had a few changes in the graphics since it had more memory to work with and it ran really smooth....for a 7.16 MHz machine.  Then something happened.

I let the intro-demo loop while I did a few other things.  When I checked on the system I noticed the screen was blanking out a bit.  The demo was running so I thought maybe my RGB2HDMI upgrade was getting hot and failing.  Nope.  It was warm but not hot.  The screen continued to blank out but increased in frequency.  Then it started blanking out longer and longer until there was no more screen.  The demo was running but the video was gone!

After some diagnostics and reverting back to the 1MB ChipRAM mod, the demo ran fine for about an hour before starting the blanking routine again.  It was consistent every time I tested it from a cold start.  After taking it apart again I used my magnifying headset to try and see if there was any damage or something.  I happen to catch something familiar about the coloring of the board and realized the corrosion had migrated deeper toward the rear of the board.  I suspect one of the electrolytic capacitors started leaking and added to what the battery already offered.  So I was battling corrosion on multiple fronts.  I did the vinegar wash and IPA afterwards as a wash.  

Things were really busy at work and I was becoming more run down so I decided to put the system away and come back to it later.  I was planning on heading out to Florida in July anyway and the Jeep needed my attention.  A lot of it!  An ark-floating deluge of rain happened in the early afternoon and didn't quite for three hours.  Long story short, my wiring harness was loaded up with water from the vents I cut into the hood for improved cooling.  The ignition coil suffered an internal short, the MAP sensor was bad as was the crankshaft sensor.  Thing started fine when I was ready to go home but the engine just quite.  Took me the whole week I was supposed to be in Florida to figure out all the problems.  In the end I flooded the harnesses with WD40 to force the water out.  Once the bad parts were replaced, she ran great and it was time to go back to work.  Some vacation!

I came across a brand new Amiga 2000 board on one of the Amiga retailers in Europe.  After digging a little I found out it was a GitHub project by floppie209.  Finding a video on YT by Retro Anachronist, I watched him assemble the board with interest.  John "Chucky" Hertel, designer of Amiga computers did a short blog on his assembly of one.  He and another person built a component locator for this project as well.  Between Sprint Layout schematic software and the component locator, I would be able to get down to the nitty gritty of it all!

So I went looking for this board in August only to find the thing wasn't in stock anywhere!  I thought maybe it was a limited run and I missed out like I usually do.  Nobody could tell me when they'd get them back in and mentioned if I needed one badly I could just submit the Gerber files to PCBWay or one of the other circuit board makers and get one!  There's a 5 board minimum order.  A bit pricey to be sure and I'd have to sell the other four.  Maybe I could but I didn't have the funds to get that started.  So I put myself on a few waiting lists to be notified when it would come back in.  I looked at some accelerator boards for both the Amiga 500 and the 2000 and found the SpitFire a really inexpensive option that did a really decent job with compatibility.  GadgetUK built one as did others and they seemed to work good.  I looked into the Terrible Fire line of accelerators and really liked the TF534 and TF536.  The TF534 offered only 2MB or 4MB of RAM but did have an FPU.  Not much software uses an FPU so I looked at the TF536.  It has 64MB of RAM available plus the 8MB of Zorro II RAM the other board has!  Now I think I can get some added performance from the computer I could only dream of back in the late '80s!

Fast forward to today...

I have acquired a Plip Box to connect to my network, a TF536 accelerator card, a CF card adapter for the accelerator so I can have a virtual hard drive and the CDROM installed on a single controller.  The IDE68K with 8MB Zorro II RAM should stay in place and run the internal hard drive.  

For video I bought the adapter board that plugs into the video slot proper so I can move the RGB2HDMI mod over to it.  I'm hoping this will help with my video problem.

But before that happened, a month ago something else did.  I got an email and I couldn't order it fast enough.

floppie209's Amiga 2000 ReMake was again available.  Only problem was, it wasn't here in the US.  About two months ago I got the email I had forgotten about signing up for.  And when I read it I nearly dropped my phone in surprise!  Amigstore.eu was the only one with stock to order from and I wasn't going to wait for a UK or US seller to show up weeks later.  If I'd been on my toes I would have seen there was one guy in Europe that sells parts for Amigas as well as building the board using your board's parts.  Generally just the custom chips are reused but the fact he sells parts kits got me to thinking I could have gotten a board from him with parts having been assembled and shipped to me for about $700.  All I would need to do is transfer my own chips over and it's done.  I'm glad I didn't see him as he was going on vacation for two weeks a couple of days after we exchanged emails.  So I needed another source anyway.

I found one here in the US and got the starter parts kit quickly and just needed to wait for the motherboard to arrive from Spain.

Also in my haste after receiving the board, I neglected to take a detailed inventory of the parts kit I got.  Was wrapped up in prepping for assembly that I missed the fact there were parts in there I thought I needed to transfer over from my original motherboard.  I bought some crystals for the Real Time Clock chip if I could find the chip.  Turns out there's one Epson had made with an internal crystal oscillator and no need for the crystals I bought.  Good thing as I was still trying to find the variable capacitor with the right range value (4.5pf-47pf) to get it clocking correctly.  So I have spare parts.  Just need to find the Epson clock chip and a crystal for the CPU clock.

It was then I decided to go into the parts kit and see what chips there were so I could put together an order for the missing parts.  The seller did state it was a Starter Components Kit so it would have some parts I'm expected to transfer over.  Boy was I surprised when I found the Epson chip AND the CPU crystal already in there!  Wow!  Honestly this isn't a starter kit.  It's everything but the custom chips, video hybrid, and CPU!  This seller is God Level!

All of the port connectors are there too.  RCA A/V jacks, 23 pin D-shell female jack for the floppy drive, 23 pin D-shell male for the video port!  Everything!

Speaking of video hybrid, this is also available.  I ordered one from Bob's Bits, who incidentally also sells the motherboard which he had in stock for $20 less than I paid from from Spain but shipping was more so it balanced out fairly even.  The Video Hybrid, also known as the Vidiot, converts the digital video signal into an analog signal for the typical CRT technology of the day.  I don't know if Vidiot is an acronym or just something made up by enthusiasts but it's there and is required when connecting to a CRT monitor capable of syncing with a 15kHz horizontal frequency.  This happens to coincide with television set tech as well.  Most Multi-Sync monitors will sync 15Khz, even LCD and Plasma types but for the most part, 15kHz is a dead tech and no modern regular monitor will work.  So some of us keep our CRTs even if they don't work in the hopes of repairing them to service again.  I'm one of those but I have two CRTs.  And neither work thus the need for the RGS2HDMI mod.  Which is ok but is too sharp an image in some cases where the dot pitch of a CRT was taken advantage of by programmers.  Colors blend better and make a more fluid image due to electron beams hitting a phosphoric display through a color mask.  The finer the pitch the higher the resolution (not pixels!) and the crisper the image.  I had a Mitsubishi MultiSync and not knowing its circuitry thought the tube was going bad as it aged because it had become so dim.  Little did I know back then that I could have been remedied that with a plastic adjusting tool as the flyback aged.  I saw it on Adrian's Digital Basement on YT.  Man was I dumb!  But I digress...again.

So now I have lots of parts, a main board, more parts on the way plus some custom chips so I don't have to use the ones on my original main board.  I am trying to keep my original board as complete as I can.  There is one custom chip I have to wait to afford.  Buster, which is a bus master controller.  It's pricier than Paula, Gary, and Denise combined!  I'll move it from my old board if need be but at least I have a source for it for now.  There are two CIA chips I also ordered but relatively cheap by comparison.  They handle most of the I/O interfacing like, serial, and parallel ports. I also ordered a Paula chip too.  That chip makes the sounds but also handles the floppy drives and interrupts for other functions like for serial port.  Denise is the graphics chip but also handles digital joysticks and the keyboard.  I have my original and the Super Denise to work with the RGS2HDMI mod so no need to get one.  Gary is the gate array chip and I needed one for this new board too.  All the custom chips have a hand in everything.  This cut down on the number of chips required to control the ports and other I/O functions.  What seems to be overlap is simply coverage that another chip isn't supporting.  The only really hang up is the serial port.  The Kickstart ROM and associated WorkBench don't allow faster than 19.2K bit rates.  There is no FIFO buffer so it's just a raw serial receiver/transmitter chip pair (1488 and 1489).  I guess the limits are due to the CPU speed being 7MHz.  If the TF-536 had a conventional RS-232 serial port it would run up to 115kbps.  Not networking speeds but simple transfers between systems would be tolerable. 

Aside from Buster I'll have everything I need.  I have three 68000 CPUs as they are very inexpensive.  At least they were last year.  Since they've been discontinued I'm sure the prices will rise.

As of today I have installed all of the resistors needed.  Other passive components installed were the four axial fuses, eight axial Ferrite beads, two 1N4001 diodes, four 1N4148 diodes.  An odd kind of signal line filter had to be assembled with a resistor in series with a capacitor the Rev 6.2 had.  This remake is, for all intents and purposes a Rev 6.2.  My component tester really came in handy as the writing or color coding on some of these parts were just too hard to see.

So that brings you up to date on what's been going on with the Amiga.  All it took was an upgrade to expose a fault and down the rabbit hole I go!  LOL!  But it's been a fun journey so far.

Here are a few pics and a short video.

___________________________

June 15, 2025

No video update as Blogger has a 100MB limit and I'm not willing to lower my video standards below 1080P for something as ludicrous as a mere 100MB.

Today's work session began with a lot, and I mean a lot, of capacitors!  Those little ones with the axial leads.  Yeah.  Those buggers.  

At any rate, I had fun setting them then soldering them.  They were placed all over so I kept the bus capacitors on the side until next work session.  Meticulous attention to detail was needed with every component tested for rated value.  There were a few capacitors that were .090uF instead of .1uF as called for.  I put those in less critical locations like the audio circuit and buffer/drivers for the ISA bus.  The ones rated closer to their stated value got the nod for anything tied to the CPU or support chips.  Fortunately all of the capacitors for the memory banks tested within .001uF of their values.  The others for the Zorro II bus buffer/drivers also were very close in spec to each other.  I'm hoping this will be the trend for the remaining 139 axial capacitors.

As for soldering, I'm getting the hang of it again.  I haven't done any since last year while working on the original main board repairs.  I learned from those experiences that the ground planes and power planes suck up a lot of heat quickly so running the soldering iron at 420c was the right call.  I probably could have gone to 460c to reduce the dwell time maybe a little but the iron doesn't have a large heat capacity and cools it pretty quick.  The dwell time isn't reduced that much and I didn't want to damage the components or risk delaminating the rings and traces.  I had a few issues where the component didn't stay put and rose from the surface during soldering.  I had clipped the leads only to discover upon flipping the board over that had happened.  Apply heat to the top makes for an ugly component so I inspected components I soldered only one lead per part.  Once it was set properly I soldered the other lead then clipped the excess.  So far so good.  

There are a few other parts left like Ferrite beads, resistor packs (RP), and EMI filters before the sockets can be soldered in place.  I may do the power up tests without the sockets first just to eliminate them as a problem.  Then if it doesn't work I can look at those knowing everything was correct prior to their installation.  Sometimes solder doesn't flow through the hole to the other side due to heat migration away from it or not enough solder applied.  I won't have the option to apply the iron to the socket side lest the plastic melts and makes a mess.

I took a look back at my previous work on the original board way back and realized how bad the corrosion from the battery really was.  I'm still surprised I was able to get it running again and set up with a hard drive, a CDROM and a CF card acting as a hard drive with a cable swap until this latest issue arose.  

What's next?  More component placement and soldering.  I'll have a total count of actual soldering points from my estimate I made a couple of weeks ago.  Estimated total solder points if all of the component locations in the silkscreen are used: 2916.  Estimated actual solder points for only those parts used: 2896.  Let's see how far off I was on the count when complete!

That's it for now so until next time, Commodore made it possible.  Amiga made it happen!

__________________________

June 16, 2025

I got home today to find my daughter's text was correct.  The A/C in the house was nonfunctional.  The blower would run but the compressor and fan on the condenser wouldn't run.  So as usual I replaced the contactor thinking ants had crawled in there again and fouled the contacts.  Nope.  But I replaced it anyway.  Next was some wire tracing to make sure it was wired correctly as it was working this morning and noting obvious stood out.  So a call to the HVAC guy was made and I hope to hear from him in the morning.  Could be anything in there but without proper tools I'm not going to just go in there and start messing with the logic board. 

So what this means is no update on the Amiga remake project tonight.  It's too hot in here to run the soldering iron and too sweaty to not be distracted with it dripping on the board.  So no-go until the A/C is repaired and running again.  Yep!  It was Monday alright!

Until next time!

___________________________

June 17, 2025

Despite the lack of air conditioning due to a blown controller board that may not be available, I have set the remaining 139 axial ceramic capacitors.  It's late so I'm going to wait on soldering them until later.  I have about a dozen remaining components before the electrolytic capacitors are applied.  These, incidentally, are solid polymer types so there's now electrolyte to dry up or leak out.  These would have been nice to have on the original boards but glad this remake has the advantage of newer tech.

__________________________

June 20, 2025

Time flies when you're really busy!  Waiting on the HVAC service for a couple of system options as I'm in need of complete system replacement.  I wish I could get a 4 Zone mini-split system but two rooms don't have the wall space for the air handlers.  Would be really nice to make my office lab like the McMurdo Station if I want without freezing everyone else in the home.  Alas, I need something right away and I'm pretty sure I'm going to be looking at an $8K-ish bill.  Ugh.  Just can't catch a break.

Amiga ReMake:

The last couple of evenings I've soldered in the aforementioned 139 ceramic capacitors.  Unfortunately two didn't set correctly and in the attempt to correct them, broke a lead off each.  One was a .1uF which I already had spares for from other projects in the past.  The other, wasn't exactly the same type or style capacitor.  I don't remember what type it is but it is a 1000pF value rated at 160V.  Though the voltage isn't matched it doesn't really matter.  Kinda like throwing a huge green tree onto a small campfire.  It's overkill for what's needed even though it's wood.

So with the repairs made I went on to soldering the resistor packs in.  You can view my video on my Facebook page here--> https://www.facebook.com/share/v/1BwBaXFE2f/

It is a tad over 7 minutes of me soldering resistor packs.  But to mitigate the boring magic behind making a working replica of an Amiga 2000, I did include a short video of peeling blue painter's tape off of a couple of packs for some ASMR goodness.  LOL!

So with the packs soldered in I consulted "Chucky" Hertel's blog about his assembly experience of this board.  In it I discovered I don't need to worry about when the electrolytic capacitors are installed.  According to his advice, don't install them unless the next step requires them.  This makes sense as there's no need for all of them to be installed for the power and RESET tests.  

What's left to get to the RESET test?  EMI filters and a power LED header.  The first power test should light up the LED kind of dim like.  Voltage checks around the board are done for the four main voltage: +5VDC, -5VD, +12VDC, and -12VDC.  The "tick" signal is about 2.5V and drives the vertical video circuit sync line since the power supply has that signal built in.  An ATX power supply substitute does not carry this signal and a jumper change on the board is required.

The second power test really tests the RESET circuit function.  The appropriate IC and supporting components are installed now.  A successful RESET test will have the LED start out dim and then go bright about 1 second later.  If that happens, go to next step and so on.  Mr. Hertel's blog entry can be found here--> Building the Amiga 2000 Replica

So that's where I'm at right now.  It has been just too hot to get in much progress but I can't just stop and wait for an A/C system to show up and be installed.  The project must go on so break a lead!  Or don't.  It's less expensive.

A moment of silence please as it gave it's lead in sacrifice
to it's motherboard's continuation.  Though it never saw the
first electron of power by this board, it will not be forgotten.

Hi!  I'm a 1000pF capacitor and I do the same thing as the others do
even though I'm blue!  I'll do the best filtering I can!

____________________

July 1, 2025

The HVAC system is supposed to be installed today.  The office lab is cool enough this morning so I pulled two capacitors and prepped the board to install replacements when they arrive from Mouser.

I discovered I had missed installing a capacitor and misplaced another into the wrong pair of through holes while doing a final inspection before applying power.  I also broke a third trying to get it settled in a better position.  So waiting on parts.  

A PC card slot bracket mounted CF card adapter arrived and hopefully it's the right model for use with real IDE type CF cards like the type known as Industrial.  I can't really use the SD card type CF card as it doesn't fully support IDE protocols.  I found this out with another adapter that won't work with my industrial card but does work with my Windows PC.  That's when I discovered the difference.  I put in an SD type CF card but it too failed to be recognized by the Amiga.  So I bought one Sordan over in Ireland spec'd out from Amazon and viola!  It worked great!  But I have to open the case every time I want to change the card out thus the bracket mounted version.  It'll be a while before I know for sure with the Amiga down and the new board unfinished.

All told when the new machine is done I hope to have all of the upgrades I wanted.  Buddah Plus One IDE Zorro II card with CF card adapter built in.  A Terrible Fire TF536 68030/50 accelerator with lot of RAM, a secondary IDE adapter with integrated 8MB of Zorro II RAM, an Amiga video slot mounted RGB2HDMI adapter for graphics, and the typical IDE hard drive.  Honestly I have it in there only for the period correct sounds of a real hard drive but I probably shouldn't use it as it's the only working drive I have.  I do have a proper SATA to IDE adapter and can use any of my many SATA hard drives.  I'm thinking this might push the limits of the 200 watt power supply but we'll see.  It all hinges on the main board being built correctly and working reliably.

_______________________

July 2, 2025

I received the parts today and got to work on installing them after work.  C800 (.1uF) was removed and the solder vias were cleared.  The capacitor was reinstalled into the correct position and the active flux cleaned up.  Next C920 (100pF) was installed after verifying I had ordered the correct part, Mouser sent the correct part, and the tester showed it was the right part.  I ordered ten of them figuring I might need to do repairs in the future.  C800 and 920 are Class 1 or NP0 (C0G) capacitors having a capacitance thermal stability of ±0.54% within the total temperature range of -55 to +125 °C  (-67 to +257 °F).  For data and address bus designs these types were most commonly used due to their cost to performance ratio.  

I am happy with the results and spent some time to remove the power supply from the chassis.  With it out on the bench I will be able to move forward in the next work session.  For now, I'm tired.  Tonight will be the second night of decent sleep in two weeks since the air conditioning failed.  The new system is in and running really well!  It's so quiet I have to listen carefully for the air flow through the registers.  Awesome!

What's next?

 As mentioned in the June 20th posting, power ON test and verification.  I did opt to install the +5V and +12V rail onboard indicators using a red and yellow LED respectively.  It's not a pretty installation as I'm gun shy about using the hot air station still.  I haven't used it very much and I'm out of even that little amount of practice.  If they work while powered ON, then I won't mess with it.  I'm not confident with surface mount stuff and probably should get a couple of kits using it to get in practice time.  This was the only SMD components on the board and they still lit up with a 3V battery so I'm satisfied.

________________________

July 6, 2025

Well.  The power ON test passed.  All voltage points had the correct voltage and the power LED lit up kinda dim.  So glad this first test passed!

I installed the _RESET circuit next and powered the board ON.  Same result.  So next is to test for H-Sync on the video port.  The 84 pin PLCC socket for Fat AGnus and the socket for GAry were set and taped in place.  Solder was then applied to secure them before moving on to the DB23M connector.  This went quickly including soldering the anchors down to keep the connector in place.

The recently acquired GAry chip was plugged into its socket and I went to pull Fat AGnus from the original board.  That was right up until I couldn't find the PLCC extractor.  I searched everywhere I could think of where I had put it.  Turned my workbench upside down too.  No joy there.  So pissed at myself for not putting where it should be, I resigned to buying another one.  Only this time I'm splurging on a good one.  None of the ones sold on Amazon or ebay are worth a crap.  The fingers either bend or break the corners off the chip making it that much harder to extract.  Nope.  Done with crap tools from China!

Enter Hakko.  They make incredible soldering and rework stations!  For a price.  I can't afford their stations but I will afford their 84 pin PLCC extractor.  That was one extractor what should never have been a reduced cost design into the cheap garbage Aliexpress, Amazon, ebay, and others sell.  Sure some diehards will sing their praises because they're so cheap.  What they don't tell you is how many chips have been damaged because of the cheap design.  Nope.  Still done with the crap tools.  Klein Tools, Tektronics, and several others I do have and love their quality and performance.  With my retro computing interests I can't afford cheap and risk damaging chips.  Next purchase I make after this Hakko piece will be a Jonard IC extractor.

__________________________

July 13, 2025

Today is the first day of the rest of this project.  After installing the Paula chip, a 1488 RS232 serial transmit chip, and the RAM bus multiplexer ICs, I have a rudimentary working Amiga 2000.  I shot a video and posted on my Facebook page because Blogger is cheap and won't let me post a video larger than 10MB.  Maybe if I use 260p video I might be able to post it but then you couldn't see anything with pixels as big as your finger nail.

So in lieu of Blogger's, ahem, short comings, I present stills of progress with captions.

Part of the Amiga custom chipset.  
Top Left: Gary (Gate Array) - Right: CIA (Complex Interface Adapter)
Bottom Left: Paula (mix of audio, serial, floppy) - Right: CIA.
I bought these from a seller here in the States.  He has one other chip
I would like to have a spare of but I need to save up a bit first.

Each Amiga 500 and 2000 uses two CIA chips (8520) for interfacing with the outside world.  Across the architecture are other chips required to complete sub-systems to make it all work so there is interdependency between the custom chipset.  Denise (not shown) came in two varieties: regular and super.  The Super Denise offered a few more graphics and text modes for later models that my 2000 Rev 4.4 could be upgraded to use, partly.  There has been an upgrade path available to get 2MB of ChipRAM to take advantage of higher level graphics modes.  Though likely not required with an accelerator card with a video port of its own, a stock Amiga 2000 could make some limited use with only 1MB using an ECS Fat Agnus (Enhanced ChipSet).  

These are Compact Flash card adapters for
IDE interfaces.  With it attached with a bracket
I can have access to it without opening up
the case every time.

Looking for a H-Sync signal and found it!

The old TV standard 15kHz Horizontal Sync
signal.  Some call it a Line signal but either way
this is a very important test.  True that modern
monitors won't sync to this any longer, though
there are rare cases where one does, but I have
a workaround for this.  It's called RGB2HDMI.

Now that its second test is complete it's time to install the CPU,
Boot ROM (DiagROM for now), and serial portcircuitry.

Serial port circuit almost in.  Just the 1488, DB25M, and a few more support ICs are left.

This is a top of the line PLCC extractor for 84 pin/contact
chips in the square format.
PLCC: Plastic Leaded Chip Carrier
These are a real PITA when you don't have the right or
a good tool to remove them with. Hakko is one of the best!

Ready for testing if it runs code from the ROM.  I have an old
LapLink serial cable connected to the board.  The other
end connects to a Serial to USB cable using a Prolific
chip and puTTy as the Telnet terminal.  Back in the day
we used programs with protocols called Qmodem,
Zmodem, Kermit, and several other "modem" programs
to connect to BBSs (Bulletin Board System)
over the phone lines.
Many had a second line in the house just for modems
(Modulator/Demodulator).  Most began with 300 baud
but by the time modems fell out of favor they ended
with 56k baud speeds.

Data on the data bus of the board!  Actively running code on the Motorola 68000 CPU.

DiagROM 1.3 for Amiga 500/600/2000

Yes, there is no RAM.  This is on purpose.  Adding RAM before this
functional test and have a problem would mean longer diagnosis and
likely backtracking to figure out why code would not run.

With the functional test complete I will install the RAM circuitry and test again.  

This ends another successful work session.  I really wish I had vacation time to build this in a week but I do what I can with what I have.  Until next time!

___________________

July 16, 2025

I got to installing the rest of the video circuit last night.  All of the IC sockets are installed that I'm going to install.  I started with about three ICs that I did not install sockets for since I was going to pick which ones don't have a history of failing.  Well, that idea pretty much went out the window but since those first three are already soldered in, I'll leave them alone until they fail.

Joystick/Mouse ports are installed as are the sockets for Denise, RTC, Buster, the Even CIA, the CPU expansion slot, video expansion slot, and all other support ICs.  Basically the only things left are the external floppy drive connector, printer port connector, the ISA card slots, and the Zorro-II card slots.  I am thinking about installing the coin cell battery for the Real Time Clock for now.  I don't know how often I'll power it up requiring the date and time but it is nice to have that already programmed when I boot up.  Since the Amiga didn't suffer from the Y2K bug even though it only encoded a 2 digit year, the RTC and Workbench OS are good for up to the year 2078.  At that point the date will rollover back to January 1, 1978 and I'll not be around to witness it.

Video:  The Video Hybrid, Vidiot or Videot is really a Digital To Analog Converter for DAC.  It converts the 4 bit digital color data for each of the three RGB color drivers for screen output to analog RGB for the then monitor technology.  Unfortunately it was designed for 15kHz H-Sync allowing for use on older monitors and televisions the monitors are based on.  Really it was because of the PAL and NTSC standards in use at the time.  31kHz monitors were already available but due to the Amiga 2000 falling out of favor due to the advancements made with the Amiga 3000 and later, 4000 models, Commodore did not correct this.  There were a few graphics card manufacturers making cards that would allow use on the newer monitors.  Video Toaster is but one example.   There is a composite conversion with C-Sync for gray scale or monochrome video as well.

The Video Hybrid, Denise, and supporting ICs are all that's needed for the next step.  Since I do not have another Super Denise (ECS) I will have to use the one my my old A2000 system.  Buster and the RGB2HDMI parts will also need to be transferred over when the time comes.

The RGS2HDMI will get switched over to the video slot via the board.  Though this doesn't truly test the video port it will at least validate the video circuit is working and can move from the serial port driven menu's to the graphics driven ones.  Though I'll probably keep the serial port connections just in case I run into trouble with the graphics.

__________________________

July 17, 2025

Took a look at the Video Hybrid tonight and yes, it does sit too high with the header pins I soldered to it.  So the next approach is to lay it down on top of a set of straight header pins and install a socket header on the main board.  So in addition to last night's posting of "all that's needed..." statement now includes a female square pin header.  And that means I get to wait until next week until it shows up! Yay.  Woop.  More delays.

The above image shows how it will look when installed "flat".  I didn't have any round machined pin or "turned pin" type male headers and used the square pin type only to find I don't have square pin female headers either.  Ironic that I do have machined round pin female headers though.

So for now this system's progress will be on hold until next week when the parts arrive. 

____________________

July 25, 2025

I received the square pin inline socket order.  I trimmed one to the correct number of pins without issue and soldered onto the main board.  Next was to remove the 90 degree header pins I soldered onto the Hybrid and only had an issue with two of the solder points.  Once cleaned up the new straight header pins were installed and the assembly was plugged into the socket.

A monitor with composite input was connected and the OCS (Original Chip Set) Denise IC was inserted into the proper socket.  The PSU was reconnected and was turned on.  The screen is monochrome but the video showed the DiagROM's initialization screens and diagnostic results.  The final screen shown was the menu to select which subsystem you want to test.  Success!  Very happy with the results!

After a couple of minutes the screen showed random selection of menu options due to, I think, the lack of a complete keyboard circuit.  I'm hoping that's what it is but it could possibly the Fat Agnus IC.  I'll not try to diagnose this since Buster and the Mouse/Joystick port controllers are not installed along with bus multiplexers so as a whole the board is not complete.  

A video of the demonstration is available on my Facebook page here --> Smitty's FB Page

Other things have occupied my time lately regarding my Windows 10 machine and Linux test systems.  I think I had a near-strike of lightning that took them out.  The Windows 10 machine was catastrophic losing the power supply, CPU, RAM, and the main board.  The ASUS GTX1060 graphics card appears to be OK.  My daughter gave me her retired main board and PSU so I installed the board to hopefully get Windows back up.  After setting the BIOS settings to turn off Secure Boot and boot Other OS instead of UEFI, Windows booted up with two restarts.  No reinstallation required!   Without hesitation I broke the soft RAID 1 (mirror) between the two 4TB HDDs so I have two independent copies now.  The boot SSD and another 500GB SSD also checked out OK.

The Linux machine now contains an old Asus AMD Phenom-II board but won't run the GTX1060 graphics so it looks like I'm forced to get a new main board no matter what.  I'm considering an MSI or ASRock board for AMD Ryzen 7000 series CPUs.  Even still I'll be spending about $600 for board, CPU, and RAM.  As much as I like my M700 Tiny system from Lenovo I cannot play games much on it.  The onboard graphics are inadequate so I'll need a big box machine again for the horsepower and expandability.

That's it for this session.

______________________

August 3, 2025

Not Amiga but has affected the project:  With the failure of my 11 year-old Z97A main board in my Windows machine, I have been scrambling to get it back up while simultaneously get my Linux rig running again for a similar failure.  Amiga work had to HALT in order to get my other stuff sorted out and back up.  My daughter had retired her main board set up to move to Linux and improved graphics capabilities so I used her old board in my Windows box.  Once powered up Windows did adjust the drivers accordingly as I used a Core i7-7440 as my i7- 4970K seems to have been damaged from the Z97A's failure.  The RAM appears to be OK so it was added the RAM she left onboard.  With that system running again I turned my attention to the Linux system again.

I was going to use her old board for my new-er build and just add a radeon RX6600 to the graphics but with being left with only a older 2010 Phenom-II powered system board I felt it wasn't justified for the hug dispirit between the two hardware pieces.  So I searched, researched, questioned, and evaluated some budget options.  It's really amazing how expensive building a new system from the main board on up is nowadays!  I spent a total of $800 for upper crust stuff in 2014.  That won't even get me into the lower midrange today.  The Z97 was new for 2014 as was the nVidia GTX1060 graphics card.  Today I can afford a four year old main board design and equivalent RX6600 graphics card.  Yes, I'm switched back to AMD for linux.  Intel was solid and reliable in Windows.  nVidia was a major innovator in graphics chips.  Today, one is in disarray in its attempt to define its place in the market while the other is redefining GPU performance and pushing the limits of AI.  Both are failing to keep the user in mind as they interpret what we are supposed to need or tell us what to buy.  Nope.  Done with that crap!  I build my machine for me and no one else.  So I have a high side of the low end main board with something equivalent to nVidia's RTX2060.   It's what I could afford but then, I'm running Linux now so Windows bloat isn't a factor any longer.  Linux is more efficient at resource management and so I don't worry about it.  Here's what I got for my "L" machine.

OS: CachyOS - An ARCH based Linux build.

Hardware:

ASRock B850 Pro RS Wifi main board

Crucial 64GB DDR 5 RAM

AMD Ryzen 7 7700X CPU

ASRock RX6600 Challenger 8GB

Crucial T710 Gen 5x4 1TB NVMe SSD stick - Boot

Crucial P3 Plus Gen 4x4 4TB NVMe SSD stick - Data

Hitachi 4TB SATA-III HDD - Backup storage

Crucial P3 Gen 3x4 1TB NVMe SSD stick - Not installed yet.  Disables PCIe x4 slot.

Fourth m.2 slot will run at only x2 speeds when SATA drive is connected.  ASRock has their own SATA ports SATA_A1 and A2 but not sure if they work.  Many have had issues with them glitching and disconnecting in Windows.  Will attempt to test this in Linux at some later date.

So as these two issues settle down with some effort I am seeing the light at the end of the tunnel once again to get back to my Amiga.

That's it for this update.  Good night!

_________________________

August 9, 2025

The Linux machine is complete and running.  I'm using it now to write this blog entry, not that you'd notice or anything, but it is nice to have a fully capable machine running Linux instead of Windows.  Speaking of which, my daughter gave me her retired main board and power supply so I still have  a Windows box just in case I run into a software issue and need it.  

Now on to the Amiga project!

I finally got started on it tonight with the intention of getting the sound working.  I was so close.  

After confirming the video was working as before, I installed the keyboard/Mouse/Joystick components as well as the Serial-IN chip.  Got out the keyboard and mouse and booted up on the DiagROM.  The arrow keys worked to select menu items but the mouse wouldn't.  It only moves the highlighted menu item down so I don't know if there's a real problem or if there's something else it's dependent one that isn't installed yet.  The system responds to the input but not in the way I expected.  I'll keep an eye on it as I progress to final assembly and in-chassis testing with a WorkBench boot disk.

So as I said, the got those items tested and it passed the serial loop-back test too!  Very happy!  Next was to install the rest of the audio circuit.  Just as I was pulling the two electrolytic capacitors needed out I heard a "Snap!" sound.  I looked up and saw the solder station showed Err on the display.  I powered it off then back up and noticed the iron's temp was dropping.  I disconnected the iron and checked the heating element and it tested good.  So I pulled the main unit from the shelf, opened it up and found burnt electronics with the associated odor.  Odd I didn't smell it until the case opened up.  Guess it was minor and went quickly.  The fuse was still good but the part is a total loss.  I went to the mfr's website and couldn't find any repair parts.  I couldn't find anything!  No replacement heaters, soldering tips, accessories, nothing!  Everything was out of stock.  So now I'm stuck with useless stuff.  I probably could reverse engineer a temperature controller for this thing but you know?  It's already 9 years old.  I don't have the time for that so out it comes.  This is a dual station containing a soldering iron and a de-soldering gun to suck up solder.  I've had terrible luck with it as it clogs easily and the filters have to be changed quite often so this was the last straw keeping it on the bench.

I'm getting a Yihua 939D+ digital unit for about $60.  It has temp stability control and a smoothing power transformer.  It has a huge range of temps to work with and less than half the cost of a Hakko or Weller, which are excellent products but too expensive for me.  It should be here tomorrow.  It'll go well with my Yihua 959D hot air rework station.

That's it for this session!

______________________________

August 17, 2025

Check out my FB page for a short I took of the system's current status dated today.  It shows the system coming ready with the Kickstart 1.3 disk screen.  Then I insert a disk to boot FA/18 Interceptor to its own Workbench 1.2 desktop to launch the game from.  Knowing it is doing this without any electrolytic capacitors installed (save for the audio circuit) to filter the power rails, I'm quite happy with the results.

The Yihua 939D+ soldering station came in and I'm loving it!  I used the pencil tip and got better results than the chisel tip on the Aoyue so that's a big gain there.  Temp stable and better joints.  

The main board had a few things done to it to get to this point since last update.  I also needed to check the mouse on the old board to make sure it still worked correctly as it would not in the DiagROM.  After confirming the music on the ROM also made popping noises I suspect that's how it was recorded when it was programmed in just to give you confirmation the audio worked.  The mouse did work correctly on the old board telling me I had an issue that needed to be resolved.  I hoped the IC that handles mouse and joystick input for the system wasn't bad.  After some serious scrutiny under a 10X monocle on my head-mounted magnifier, I found what looked to be a metal hair bridging a pin to ground.  I also found some solder balls strewn about that needed cleaning off.  Not sure what that was about but after another wash and brush with 99% IPA I put the DiagROM back in, started it up with the mouse and keyboard attached and it worked as it should.

Next was to move Buster from the old board to the new one for Zorro bus control and testing later.  The Super Denise I had on the old board as an upgrade was removed with the RGB2HDMI adapter and put aside for later.  I want to make sure this is a stable system in stock form, albeit monochromatic, before any upgrades or add-ons are applied.  Once proven then those will be added.

So the next step was to make sure the IRQ system was complete to test the floppy drive function.  Again, pulling the drive from the old machine and attaching to the new one proved I missed something.  I forgot to jumper the drive for DF0: and sat there waiting for it to even start up and ask for the boot disc.  Hmmm.  I had to lookup the drive's jumper settings as I couldn't remember what orientation, vertical or horizontal, the jumper block was to be for DF0:.  The Chinon FB-354 uses horizontal jumper orientation.  Left and middle or middle and right.  Set the jumper to middle and right then tried it.  Success!  The system came up to the Kickstart 1.3 screen.  So thus began the video opportunity described above.  So this brings me another step closer to completion and installation into the chassis.

What's next?

Install the Zorro bus card slots and the ISA bus slots.  Then install the rest of the electrolytic capacitors and retest everything before removing all the ICs and washing the board.  As for the shield on the bottom of the original board?  It will stay on the original board.  This was something the FCC required on machines sold in the US for whatever financial gains that lined pockets rather than any real RF interference issues it actually generated, which were none by the way.  So the new board will not get a shield.  It will be inside a grounded metal chassis with an insulating plastic sheet just like the original board.

What's in the future of expansion?  Mentioned earlier will be a Terrible Fire TF-536 accelerator, Super Denise with RGS2HDMI adapter for the video slot, GoTek USB virtual floppy drive, optical drive, internal and external Compact Flash cards on IDE interfaces as well as an IDE Seagate Barracuda 100GB HDD (spinning rust).  The external CF will allow me to change out OS's or bring files over from my Linux or Windows machines without having to transfer the stuff over serial.  I have a PlipBox that will allow me to connect the Amiga to my network so that will help but I've never done it.  Lastly, a 2MB ChipRAM upgrade may be installed if the TF-536 doesn't automatically configure it within its own 64MB of onboard memory.  We'll see.  That was the upgrade that exposed an issue with the old board and got me started on this path.

That's it for this session.  See you next round!

_____________________________________

August 20, 2025

Update Blast:

Tonight's work session got the ISA bus card slots installed.  I did these instead of the main slots on the Zorro bus as I needed to get a feel for the new soldering iron and its characteristics.  I've got a better feel for it now and feel more confident installing the Zorro card slots as this bus is the main one.  I doubt I'll ever get the chance to get hold of a working 386 or more rare 486 bridgeboard to use IBM PC compatible cards in the ISA slots.  

What's next:

The Zorro bus card slot installation as mentioned above and the rest of the electrolytic capacitors for filtering the power rails.  Once done and confirmed the board still boots up it'll get a wash of electronic cleaner to get all that rosin off.  The final wash will consist of 99% IPA and a thorough drying.

__________________________

August 24, 2025

Zorro slots installed, electrolytic capacitors installed.  The board has been cleaned, dried, and tested again.

I installed C0pperdragon's RGB2HDMI build from the old main board so I could test the graphic and color quality.  I used the OCS Denise chip to get a solid base line even though I think that by board version 6 the 2000 main boards were equipped with ECS chips.  Either way it worked and I got a clear, crisp color image of the KS 1.3 boot screen.  After powering down I switched the jumper on the dual ROM adapter to select KS 3.1.  This is a nice way to test original and "last version" of original Amiga ROM code Cloanto got hold of after the shutdown.  The system came up with the new style boot screen and so I booted F/A-18 Interceptor and let it run for about a half hour.  I did swap out the Denise with the Super Denise (8373R3) and installed the RGB2HDMI adapter from LinuxJedi.  I got it assembled, Pi'd, and coded with an Micro SD card so it was ready to plug in.  My only qualm was that it came with a mini-HDMI instead of going from that to a standard HDMI out to the slot bracket.  I'd rather not stress the onboard connector on the Pi like that but that's the way he built it.  After booting and running the demo mode of the combat simulator I shut it down and am ready for the next upgrade.

____________________

August 28, 2025

A bit has happened in the last four days.  The Terrible Fire TF-536 accelerator was installed with its bootable CF card and it booted right up!  It just worked!  As Adrian Black would say:"It freakin' works!"  This was a really nice ending to the build itself.  True it booted just fine in factory stock form but to add something that wasn't built back in the day, the new tech is amazing.  Very happy so far.

Last night I installed the Buddah Plus One IDE card and the 100GB Seagate Barracuda hard disk drive to see if it would boot the AmigaOS 3.2 on it.  Sure enough it did with one caveat.  The OS was installed on a stock Amiga 2000 Rev 4.4 with Motorola MC68000 CPU which means it doesn't have any libraries (drivers) for the 68030 CPU and thus produced an error stating as such.  But the few things I did do after boot up and OS loading it performed very fast.  I did run the F/A-18 Interceptor simulator and it ran through the demo about three times faster.  Obviously partly due to the increased CPU speed but I wonder if this game is tied to the CPU clock speed instead of being programmed with asynchronous timing to avoid the playability issue on faster computers.  If it is then I'll not be able to play it without reverting back to the stock speed and slower frame rates.  I did, however start up Frontier.  It's space combat game using rendered polygons like F/A-18 does but I think it did it differently.  In stock form it's playable but seems to improve that with faster CPUs.  Frame rates are higher making it run more smoothly but it also plays a little faster possibly making it more challenging to complete missions.

Tonight's work session tested CF card adapters on the Buddah card and discovered something not in the documentation provided with it.  There are three sockets.  One is for a CF card.  The second is a 40 pin IDE socket for one drive.  The last is also a 40 pin socket with a DOM plugged into it.  The DOM is short for Disk On Memory and is basically a solid state drive for IDE interfaces.  Without any bootable drives in the system this DOM will boot up and offer a choice to install the OS onto a connected drive or exit to a CLI prompt or reboot.  So I wanted to make use of a expansion card slot bracket for a CF card reader I can access without opening up the case.  Just reach around back and swap the CF card.  Well it didn't go as planned and in about an hour I discovered why.  

I figured the reader wasn't compatible with the Buddah and switched to a different reader I did have working on the old set up.  It didn't work either.  The system just sat there waiting at the "Insert Disk" screen.  I did finally get it worked out by removing it and plugging in the CF card from the accelerator directly into the Buddah card.  It booted up fine without anything else connected to that first IDE socket.  The DOM appeared on the Workbench desktop and it showed I could run the installer for it from there.  Well, after swapping out another CF card reader into the mix I tried the one that had worked before again but in "slave" mode.  Now I could see it in the list of available drives but was blank.  No sweat.  Making progress.

In the end the onboard CF card slot on the Buddah was the Master or Device 0 drive and the cabled CF reader had to be in the slave or Device 1 mode.  So the discovery I made that isn't in the document packaged with the Buddah is the IDE socket 1 is Device 1 only.  I cannot put two drives on that cable.  I was a bit bummed out but at least I understood the issue.  I think there's a more detailed manual but it was in German last I looked where icomp.de is located.  The have some good stuff even if it is overpriced.  There's a network card I would love to get my hands on but for $175US I'll wait and use the PlipBox network adapter I bought for the parallel port.

Side note  : As I was writing this I did a quick check on icomp's web site and found the description in English.  It does state a total of 5 drives can be connected to the card.  One onboard CF card and four cabled drives.  I could be a default configuration that's causing the issue so when I'm done experimenting with addons and mods I'll do a formal installation and run the configuration program.

So that's where I'm at right now.  Some more experimentation and waiting for a few things to expand the power connectors shortage so I can power all the drives I want.  I have two dead PSUs that I've robbed one power connector strand from.  It has three Molex 4 pin "PC" drive power connectors and a single "3.5 inch" 4 pin power connector I may splice into the Amiga's PSU harness.  If and when the original PSU fails I will install a semi modern PSU board that still contains a +5V, -5V, +12V, and -12V rail.  I'll lose the "tick" signal the Amiga uses for timings (50/60Hz) but that's was remedied with a jumper Amiga graciously included in the schematic in case the signal is lost.  Jumpering it will use the timing signal from the Agnus chip instead.  It's got something to do with PAL/NTSC timings or something like that.

That's it for tonight's post.  Good night!

________________________________

September 15, 2025

Umm.  Yeah.  Been kinda busy and I apologize for not offering any updates so here's a somewhat final update to the Amiga ReMake project.

Amiga ReMake Project: Final Update:

Yes.  It's finally together!  The last part needed to get an optical drive working arrived.  There's obviously some tweaking of the configuration but it's fully assembled and running.  I had a blast with this project even with the few hiccups experienced and overall I'm quite satisfied with the outcome.

A 100GB hard drive has AmigOS 3.2 installed.  The 8GB Compact Flash card has Amiga OS 3.1 as the default boot device.  I can switch between them as needed.  The cable came from Cables Online here ----> MK-C18D .  It has a female 44 pin with Molex 4 pin power connector to supply power to a 2.5" laptop drive (not for 50 pin drives).  Obviously this is backwards if you are connecting to a 44 pin controller port to convert to standard 40 pin IDE drives.  So DO NOT plug in the power connector!!!

The cable offers on a TF-536, the ability to connect a hard drive and optical drive, a CF card (true IDE or Industrial type card with PIO support) and an optical drive, OR two CF cards of the same true IDE type.  You cannot mix a mechanical hard drive with a CF card as far as I can tell.  I have not gotten any combination of this to work.  Could be my hardware or it could be that you just can't mix the two technologies.  Be that as it may, I have the CF card and the optical drive connected and working just fine.  I want the optical drive to run as fast as it can instead of being hamstrung on a 7MHz bus.  The mechanical hard drive will be instead.

The light colored optical drive did fail and pressed a black faceplated LG drive, circa 2009, back into service.  The black faceplate clashes with the vintage beige of the machine and I don't like it much.  It is nice to see the old hardware running again though.  The hard drive hardly makes any noise and hope it will last a long time to come.

My current configuration:

Amiga 2000 rev 6.2 ReMake main board by floppie209

2MB ChipRAM conversion via iComp.de ACE2b board and 8375 (Super Fat Agnus)

Terrible Fire TF-536 Accelerator (with CPU card slot adapter)

                     68030 CPU at 50MHz + 64MB FastRAM with MMU

                     Onboard 44 pin IDE interface

8GB CF card loaded with Amiga OS 3.1 optimized for the TF-536

Buddha Plue One IDE interface for Zorro II bus - includes CF card slot and two IDE ports

                     plus the Buddha Plus One installer on a DOM

                     Added 4GB CF card to the slot loaded with AmigaOS 3.2

Aditional CF card slot located in rear card slot space for external access to CF card storage swapping

Added Seagate Barracuda 100GB IDE hard disk drive for additional storage

GoTek floppy drive emulator as DF1: and original floppy drive as DF0:

LG CD / DVD / CDRW drive as slave configured to 8GB CF card on TF-536

RGS2HDMI converter adapter to use modern monitors

Original Amiga 1351 mouse

Converted Amiga 500 Mitsumi keyboard sans case

My original rev 4.4 main board might get a VIP treatment with a frame and hung on the wall in a place of honor.  I might put in a mfr date and the date of it's failure.  Would read something like 1990-2024 with a hiatus of 1998-2022 with Varta Barf repairs.

Anyway, with this system working I am able to play games, write up documents and such.  The next step is to get the network adapter working and try to set up a file transfer option between it and my other Linux and Windows machines.

So what's next on the project list?

While the Amiga is technically complete, I'll post updates for significant events in the future here.  In the meantime, I would like to get my Kenwood TS-450S/AT HF transceiver working again with that replacement processor board I acquired earlier this year.  Oh there's going to be a lot of screws to keep track of with this one.  The pain, the pain...

That's it for tonight!

_________________________