NTSC Video Issue After C64R MK2 is Powered On For Extended Time

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser Please create an account and log in to see all content by default. This is a limitation of the forum software.
  • I'm experiencing a problem with my C64R MK2 board, which is only a couple months old at this point. After the unit is powered on for an extended amount of time and allowed to get warm, the Sharp LCD TV I have attached to switches PAL mode, and the picture goes off center and mostly monochrome. I have a VIC-II 6567R9 installed. I have heat sinks from Ray Carlsen on the VIC-II (6567R9), SID (6582) and CPU (6510). I've tried another VIC-II (6567R8) that experiences the same thing, and I've got another version (8562) on the way from eBay, but I anticipate the same issue with it as well. If I turn off the unit and allow it to cool for only a short time, the picture returns to normal.


    Everything is housed in a new C64C case that I received this week. It was happening in a bread bin case, and it's still happening in the C64C case.


    Any assistance with this would be greatly appreciated! I'm very excited about owning this board since I only just recently purchased it. However, it is a bit disappointing to see the screen change into something that isn't right.


    Regards,

    Neil Forker

  • I have basically the same issue. Using a PAL VIC chip seems to have no problems. The issue there is I have to use a PAL compatible monitor, and those aren't particularly common in the US.


    I chatted back and forth with Jens via email, and he concluded that something in the NTSC timing circuitry is having trouble.


    From what I can tell, it goes something like this: there are two clocks on the Reloaded motherboard: the NTSC and the PAL clock. They apparently both always operate, and there's a component on the board that switches back and forth. So, in my case, one of the components involved in the NTSC clock generation is failing after it gets warm, but the equivalent PAL component is not.


    Since the clock also generates timing signals for the CIA chips, and the CIAs are used for the serial port to talk to the floppy drive, I found that fast loaders were also affected by this; the one built into the Final Cartridge caused data corruption when I saved files to disk. I was using TFC to write some BASIC programs at the time, so I lost several hours worth of work. (I did save often, but didn't verify the saved files, and all the saves were corrupt.)


    At Jens's advice, I took the top off of my 64C case and touched the different chips to see if any were getting hot or maybe had cold solder joint issues. When I touched U23, the image changed. Moving my finger around caused the image to change on the screen, and touching it with a non-conductive object had no effect - so it was my body's capacitance, not a loose electrical contact, altering the image.


    So at this point, i believe my Mk2 has a manufacturer's defect. I need to get back together with Jens by email and see what I can do about a replacement.

  • I've been in e-mail contact with at least one person in august, and this has not been forgotten (tomxp411, that might have been you). I yet have to reproduce the problem here, as I currently believe it to be an edge case that can be fixed with a singla resistor value change.


    I confirm that cases like this will be handled under warranty.


    Jens

  • I've been in e-mail contact with at least one person in august, and this has not been forgotten (tomxp411, that might have been you). I yet have to reproduce the problem here, as I currently believe it to be an edge case that can be fixed with a singla resistor value change.


    I confirm that cases like this will be handled under warranty.


    Jens

    Jens - Should I reach out to you via email for this?

  • Hi Crew,


    I'm having a similar issue to tomxp411 and in fact have chimed in on his lemon64 thread.


    I'm connected from std video out via S-Video+RGB conversion cable, connected to Framemeister Mini, connected to TCL LCD TV though HDMI


    For me, after about an hour, the screen starts flashing, then a minute or two later, I will lose signal and just blue screen from there -- I suspect this is the FM giving up and just blue-screening.


    I also get disk drive read errors during this time (I was doing some copy sessions to pull disks onto SD) -- Some times the read errors would present before the video errors i.e. using Maverick disk compare to verify copies would produce random failures before the video crash.


    There was discussion of being able to spray some cold air on a certain chip(s) when the issue happens and seeing if the momentarily corrects.


    Jens Can you confirm and/or give a little more detail. I have purchased a can of compressed air for the experiment.


    Thanks for any help and also for activating the forums !

  • I've been in e-mail contact with at least one person in august, and this has not been forgotten (tomxp411, that might have been you). I yet have to reproduce the problem here, as I currently believe it to be an edge case that can be fixed with a singla resistor value change.


    I confirm that cases like this will be handled under warranty.


    Jens

    Thanks, Jens. Yes, that was me. I have also been a bit busy, with the kid going back to school and life in general.

  • I have this issue as well using the NTSC versions of the VIC-II (the issue coming up after extended uptime), and another issue to it that is related, that being only PLCC versions of the 6567R8 seem to be detected. If I use a ceramic version the board can rarely detect the chip is present, sometimes taking 6-7 power on/off cycles to get a ceramic/gold plated version to detect and it will promptly resort to not being detected on a power cycle.

    I have tried with (2) different PLCC 6567R8s and (2) different ceramic 6567R8s.

    I am updated to the most current firmware.

  • Chip regocnition is a different topic; the last changes to the MCU code (which can be updated through USB using the Windows-Tool from the Wiki) were all related to chip recognition, especially for the older revisions. Please make sure that you're running the latest MCU code, and also make sure that your power supply outputs exactly 12V (to be on the safe side, it can be up to 12.2V). If the voltage is any lower, the recognition may fail and revert to a "safe" voltage in order to protect the precious VIC-II chip.


    Jens

  • Oh. I'm sorry everyone is sick, that's rough. I hope everyone feels better soon.


    I can be patient, as I've switched over to using a Keyrah+VICE system until I can get the glitches with the Reloaded worked out.

  • Totally forgot to write here that we haven't been able to reproduce the issue. The test-board ran for almost two weeks without any issues using an NTSC 6567 VIC-II chip.


    So to really get to the bottom of this, I need to have a machine here that will let me reproduce the problem. However, it may still be a power supply issue, and to really replicate the setup that you have there, I also need to have a 110V/60Hz source and your power supply (I should not exclude anything at this point!).


    I may be able to get a converter from 230V to 110V, but the AC frequency will still be 50Hz. Let's hope this is enough to replicate the problem.


    So can *one* of you who still has the issue send the board back to Germany for closer investigation? Or did one of you get the Commodore 12V PSU by now and maybe resolved it with that?

  • I have used several power supplies with mine, so I am certain it's not a power issue; what convinced me was when I hooked it up to my Wescom 8 amp test bench power supply and had the same problems. =)


    I've actually got a lot of power-related problems with my board: when I use the Ultimate II+, I need to attach a USB power adapter, or the U2 is very unstable. Likewise, when I use Jim Drew's WiModem, the video glitches - a lot. And when I try to run the U2, an SD2IEC, and the WiModem all at the same time, one of the three things simply won't boot properly.) (These all work together just fine on my 64C, so this seems to be a Reloaded problem.)


    I should also note that this doesn't happen when the board is out in the open. It does need to heat up a little bit to have the problem happen.


    Since my Reloaded isn't installed right now, I can ship it back. I just need the shipping information. I can't really ship the power supply, since I use it for other things.


    Should I ship it with the chips installed, so you can use those to help reproduce and/or isolate the problem?

  • Yes, please include the chips. We only have that one company address - find it in the imprint or on the invoice.


    As for Jim Drew's WiModem, I already discussed this directly with him: That's a design error on his side that just happens to work with the old C64, but not with the C64R. Of course he does not second that opinion, but the internet is full of sources that describe power-related problems with ESP32 (or the smaller variant of it)-based products.


    I'd have to check power consumption of the 1541U2: If that's also pulsed, then a simple coil could do the trick. In any case, I'd like to resolve this, but before I can do that, I need to reproduce the issue.

  • Okay. I will ship it with my NTSC VIC installed. It might be a day or two before I can get everything in the mail; I haven't shipped internationally before, so I need to go down to the post office and find out how to do that - and I can only do that on Friday or Saturday.


    Also, it's worth noting that this only happens when the board gets warm. Sitting out on a test bench in an air conditioned room, there's just enough air circulation to keep the problem from happening. Also, as I noted in our prior email, my HDMI converter is actually able to recover the video sync., but my NTSC/PAL monitor does not.


    So the actual failure is easy to miss with good display hardware. I didn't even know anything was wrong until I took the computer and my 1702 monitor out for an event, and the display broke up in the middle of a demo.

  • The board was indeed not in a case when it ran for two weeks, but I did put a piece of styrofoam on the PLL part (I call this "negative heat sink", but beware: This is not a scientific name!).


    Further, we use a flat screen with S-Video input for testing, which may actually have the recovering properties you're describing. My problem is now to find a CRT that's capable of displaying NTSC - this is extremely rare in Germany. A few years ago, NTSC parts used to be available on eBay Germany, as GIs who were working in Germany wanted to reduce the amount of stuff that they return to the states. However, I have no idea how to find the people who have bought those components back then (well over 15 years ago).


    I'll ask around - I depend on a "lucky find" here.

  • So I put my Mk2 in a case today to test it, before sending it off to you, and of course it's stopped doing it now that the weather is cooled off. It will do this in 10 minutes on an 85 degree (30c) day, but now that it's below 70 (21c), the board is staying cool enough that it won't fail. Right now, I've got the system just running a BASIC program to let it repro the problem.


    I did get it to glitch once, but a power cycle fixed it, and it hasn't glitched again. It's a little frustrating that the problem is this temperature sensitive; considering the cost of shipping, I'd like to make this 100% reproducible before I send the board off.