VIC 20 - Black Screen that's really fighting me

[USA_Joe]

I have a VIC-20 (not CR) that has a black screen upon boot. Here's what I've done so far:

Per Ray Carlsen's VIC-20 diagnostics document for black/blank screen
UAB1, UAB3 6522 VIA – PENDING CHECKING FOR SHORT OR FAILURE
UB4 7406 LOGIC - REPLACED
UB7 6560-101 VIC – Produces signal of “NTSC” on screen. Will swap.
UC4, UC5 74LS138 LOGIC MEMORY CONTROL - REPLACED
U14, U15 MB8416A RAM MEMORY – REPLACED with HM6116P-3
UD8 & UE8 74LS245 TRI-STATE LOGIC - REPLACED
UE10 6502 MICROPROCESSOR – Good on scope & probe. Will swap.
UE12 901486-06 ROM KERNAL – Was bad – REPLACED with modern replacement.
UF8 74LS245 TRI-STATE LOGIC - REPLACED

Moreover, I've checked:
- Voltage - OK
- Blind Disk Access - NO RESPONSE
- Swapped BASIC ROM, tested in working VIC-20 as well - OK
- When replacing KERNAL, I tested the bad one in a working VIC-20, it failed. New one tested fine.
- Tested Logic chips (replaced and new) 7406, 74LS138, and 74LS245 in eprom burner that can test login chips - All OK
- Used YouTube video of a gentlemen testing the CPU and VIC with a oscilloscope, did the same - Both look good. CPU held high. Clock good. IRQ good.

To make it even more difficult, each chip was soldered before I started pulling chips and installing sockets. Granted I plan on making this my "chip swap" VIC-20 unit, but I'd like to see it working soon. The vendors I've been buying supplies from love this VIC-20 for the sales it has produced for them to date.

Anyways, I have a question. Can any of the three SRAM chips (UE1, UE2, UD2) cause a "black screen"? I see nothing on Ray's documentation as to this being a possibility, and I have found nothing on the internet that references these chips and a "black screen".

Not counting the 3 SRAM chips, I have 4 suspect chips left on the list. The VIC, CPU, and two VIA chips. If the SRAM chips are a potential suspect as well, that makes 7. I do have a new 6502, but no VIC or VIA chips. So the CPU will be next on the list for obvious reasons. But I will eventually pull the VIC and VIA's, swap them into a working VIC-20 if needed, and install sockets from where the chips were pulled.

Any suggestions and/or comments are appreciated.

Have a great weekend!

Joe

[kickstand]

Hi USA Joe!!

I am in the same predicament, x6. The 8-But Guy kindly gave me 6 VIC-20 boards as part of the VCF Southwest lab he ran. I'm currently focusing on one board where I have socketed every chip but UB9 (the clock generator). I have lots of spares, as I pulled a lot of components off other boards.

The problem: Black Screen (but does have sync).

Important note: I do not have a 9VAC supply, I am using a USB Adapter with a voltage regulator to produce a solid +5v. According to Ray Carlsen's VIC-20 docs, this should be fine save for the cassette relay.

UB7 6560-101 VIC – No idea if any of the 3 I have work short of putting them in a good board (which I don't have).
UE10 6502 MICROPROCESSOR – Looks good, and breadboarded out a "NOP" generator where all the address lines checked out. I did find one bad CPU this way.
U14, U15 TM2116 2Kx8 State RAM – Breadboarded with a Mega (or in my case an AVR128DA64) a way to test these. All seem to work fine save for a pair of Toshiba TC5517APs
UD2, UE2, UE1 2114 1Kx4 Static RAM - Breadboarded with an UNO a way to test these. Most worked fine, and I still have lots of spares.

UAB1, UAB3 6522 VIA – Pulled, and no idea how to check these short of putting them in a good board. Not critical though, as according to Ray Carlsen's VIC-20 docs video should still come up.

UD1 4066 - Checked in TL-866 and replaced
UB4 7406 - Checked in TL-866 and swapped
UC4, UC5 74LS138 Decoder - Checked in TL-866 and swapped
UF8, UD8 & UE8 74LS245 Bus Controller - Checked in TL-866 and swapped
UC2 7404 Inverter - Checked in TL-866 and swapped
UC3 7402 2 Input NAND - Checked in TL-866 and swapped.
UD9 74133 13 Input NAND - Swapped with one from another board that I checked in the TL-866. When I pulled this off the board, the chip cracked two legs off. That chip was either bad or headed that way.
U13 7408 2 Input NAND - Checked in TL-866 and swapped.
555 - I checked Reset on the CPU and it works fine.

ROMs: I am working on code that dumps via a Mega (or in my case an AVR128DA64) the code on the ROM. None of my checksums match, so that's a work in progress.

UE12 901486-06 ROM KERNAL – Replaced with a 27C128 modded to fit in the 24 pin slot.
UE11 901486-01 BASIC ROM - Checked OK?
UD7 Character ROM - Checked OK, I did get the chance to check this chip out on a known working machine.

Since I was getting boards anyway, I ordered 5 ROM/RAM/Diag boards. I burnt the 27256 with 4kx2 DeadTest, Battlezone, Centipede, DigDug. I'll happily trade a diag board for parts or help.

My goal is to get 5 of the 6 VIC-20 boards working, one fully functional and give the other 4 away (the 6th is the newer board that is easy to pull traces on).

I can post my code for testing the RAM (modified from other's work) and scope captures of my signals: my clocks work through the CPU, and I have sync at the emitter of Q2.

[USA_Joe]

Wow! You got lots of work to do too!

I just got home from a weekend at a Hawaiian Music Festival here in southern California. I live where it's been 103 degrees lately here in southern California. The festival was close to Los Angeles airport area which is on the coast, so it was much cooler. It was near Hawthorne, where Elon Musk's SPACEX headquarters is located.

Anyways, I'll be looking more at my problem VIC-20 tomorrow (Monday) and using my oscilloscope to see what I can find while I'm still learning how to use it and interpret the readings. I'll be comparing the readings I get to those on an identical VIC-20 that is fully functional. I also printed out the block diagram for the VIC-20 bus that is available on zimmers (http://www.zimmers.net/anonftp/pub/cbm/ ... index.html. The file is VIC20BUS.GIF). I'll be using this to trace the bus from chip to chip. I had already pulled up four traces, and had to install four bodge wires to fix my mistakes. I know those wires are working correctly. The board is really easy to pull traces on, so I really take my time now. Oxidation is a real problem. Doxit helps some, but a rework heat gun is still needed to get chips out.

Thanks for offering the code, but I'm going to order a Chip Tester Pro V2, with accessories, from BackBit. It will make my chip testing easier. If I find anything of technical information value to you on my journey through this VIC-20, I'll let you know.

Good luck!

Joe

[kickstand]

Hey USA-Joe!!

You must have the darker green (later) RC board. I have found that on the one I have it's really easy to lift traces. Would you like one of the other boards I have? If so, I'd like to trade a test of a couple of VIC-1's I have. I'm pretty sure that's where my problem is.

I'll look at posting (or referring to them) waveform captures at various places on my board. For sure you want to follow the clock signals and verify the reset line is working. FB-4 and FB-5 for clock into the VIC-1, FB-10 and FB-11 for clock into UB-4. FB-11 and UE10 pin 3. Reset, IRQ and NMI on the CPU should be all high most of the time, CR/W looks funny on board 1, so please post that. Also Pin 4 on the VIC-1 (VR/W) and Pin 9 of UD9 would be helpful to see what it looks like working. Your average data line and several address lines on the CPU would be interesting as well. I do have a CPU that several of the address lines failed when I did a NOP test on it.

That chip tester-pro looks like it will check everything on that board save the RF circuit out of the VIC-1. Unless someone has been to your board previously, you're bound to find out what's gone wrong. I'm pretty much down to not having a working VIC-1. That being the case, I'm pretty much stuck as getting a new one seems like a crap shoot.

[kickstand]

Update:

I got garbage on the screen once (random characters), so I think my VIC-1 might be good, or at least intermittant. Progress!!

I'll post my pics to my website: www.itdiscovery.info/vic20...which I will create right now.

[USA_Joe]

I hope your VIC is good!

I'm taking a break from working on my VIC-20 since I do my soldering/desoldering work in my garage with the door open in an effort to keep the air flowing with the help of a fan. It's been hot here, and my garage is about 95 degrees right now, and over 100 outside. So I might try again this evening if things get cooler.

I'm struggling on learning how to use my oscilloscope. Some videos are helping, others are not. I don't feel that I can trust much of what I'm seeing one the scope without knowing how to use it properly. So I'll get back to you when I can actually provide some factual help or insight.

I did make a contact with a guy that lives near me that is selling a Commodore PET. He's out of town right now, but he's also retired. He did a full restore of the PET. So I'm assuming he has extensive skills that I can learn from. Plus, I might buy the PET from him.

Good luck!

Joe

[kickstand]

Being in Texas, I completely empathize with doing your soldering in the heat. My workshop is also sudo outside, and so I am limited to quick sessions before 9AM at the latest.

I'm working on posting the first waveforms downloaded from my FNIRSI scope. Since you're new to scopes, duplicating the first few waveforms should be pretty easy: the clocks going to the VIC-1 and the CPU. You don't have to do two channels as I did, just find FB4 and FB4 just to the right of the VIC-1 chip inside the RF Shield (UB7). Set Channel 1 to 2 Volts per division (if your probes are x1 or x10 this may make a difference, and new scopes can be set for the probe value) and then use the +/- knob if the waveform is too big or too small. Now set the timebase to 50nSec/division. Again +/- to see the waveform. Digital scopes will show the frequency right on the display. If you're old school, you'll have to do the math. You should see a 4v (ish) peak to peak sine wave at 14.3Mhz. If you put both FB4 and FB5 on the screen at the same time, you'll see they are 180 degrees out of phase.

What comes out of the VIC-1 chip is a 1.02MHz (NTSC) square wave, and is easy to pick up on FB10 and FB11. Same Volts/Div, but change your timebase to 500nSec/div.

Then make sure that clock and getting to the 6502 and good clocks are coming out. FB11 goes in, and look on UE10 Pin3 and UE10 Pin 39 for outputs.

I've been checking how the VR/W signal is derived via U2 and U3. I followed it all the way from Pin 34 of the CPU to Pin 4 of the VIC-1. I had to hammer on reset to get activity, and pressing on my board made it croak...tapping lightly around UD7 made it go away, so I am replacing the factory socket and the caps around it in the AM.

Hope that helps, and I look forward to seeing your results.

[kickstand]

Update:

Replacing the socket and caps didn't help, but the signal still croaked when tapping lightly. Re-flowing R15 and the signal stayed up. I could see that AD0-4 ran after reset and then stopped. That smelled like it was running code, but bogus code. I then swapped back in the original kernel ROM, and...

Success!!

I'll post a few more waveforms as needed, but before I move onto the other boards, I need to make sure that my USB Keyboard adapter for these machines works correctly (the original intent of the project).

[USA_Joe]

Great job! Thank you for letting me know about the scope screenshots. Let me know where I can find them when they're ready.

I'm still fighting my VIC. I desoldered the CPU, tested it in the working VIC-20, and it's fine. So I installed a socket, and put it back in. So I'm tediously checking each trace for the pins on numerous chips. This is a "dark green" board where lifting a trace without even making an error is easy.

I did email Ray Carlsen about the SRAM ships being a potential cause. As I suspected, he stated that any chip that is shorted can cause a black screen. Before I remove any more chips, I need to double check my work on installing sockets by checking for continuity in the trace paths. It's lots of time consuming work, but it's really helping me learn more about the architecture of the VIC-20. And besides, it's only getting hotter here.

As an aside, I did visit the guy with a Commodore PET 4016 for sale. He doesn't tackle the Commodore systems like we do, but he did repair the keyboard so that it works. The system is amazing, and is very clean. I'm going to buy it today for my home computer museum that I have in my den.

Joe

[joshuadenmark]

I'm taking a break from working on my VIC-20 since I do my soldering/desoldering work in my garage with the door open in an effort to keep the air flowing with the help of a fan. It's been hot here, and my garage is about 95 degrees right now, and over 100 outside. So I might try again this evening if things get cooler.

Feel free to come to Denmark to do your soldering 59 degrees and it's my summer hollyday...

Hope you get your Vic 100% working.

[kickstand]

Hey Joe!!

Sorry to hear that replacing the CPU didn't fix the problem.

The scope shots are now on my website (my apologies if I am not supposed to post links):

https://www.itdiscovery.info/vic20

I would be happy to take any shots on the board you want and post them, now that my VIC-20 works (at least board 1). You also want to make sure that you aren't inadvertently connecting two address lines or data lines together. Remember, my board had a mechanical issue, a cold solder joint on a through hole or via, so do some tapping and see if stuff stops working.

My offer still stands for a non-dark green board if you get fed up with yours.

...and yes Joshua, I'd love to visit Denmark! My weatherstation is showing 110F (43C).

I'll be working on my USB2AVR keyboard converter for a day or so, and then on to board 2.

[USA_Joe]

Thank you for the link! I did put the missing "o" on the "inf" portion of the link to get there.

I appreciate the offer on the board, and we'll see what the future holds. I don't surrender well at all to problems.

As for Denmark, I actually do have Danish blood from my mom's side of the family. According to what I've been told, it goes back a few hundred years to Danish royalty. Now we're just "serfs" like everyone else. But I'd still like to visit when the world gets to be a safer place, if ever in my lifetime.

Back to the VIC-20 while listening to Hawaiian slack key music on a streaming service.

Joe

[USA_Joe]

I found it!

When I replaced the chip and installed a socket for the 74LS245 at UF8, pin 1 apparently lost a trace to UC3 pin 2 during the process. I had a feeling it had to do with something from my previous pulling up traces problems. I connected a bodge wire from one pin to the other on the back of the board, and it booted right up! It even passed the diagnostics without an issue.

Despite the heat in the garage, I had to do the repair because I was excited that I found a problem, and was hoping it was the cause. I got lucky, it was. Time to clean up the soldering / desoldering station in the garage, and call this repair case closed.

Thanks for everyone's input and assistance!

Time to go jump into the pool!

Joe