Chameleon v2 not booting with a C64C

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.


Also users that are not logged in can not create new threads. This is a, unfortunately needed, counter measure against spam. Please create an account and log in to start new threads.

Don't Panic. Please wash hands.
  • When the Chameleon v2 is inserted into the expansion slot of my C64C, it (almost always) refuses to boot. After about 50 power-cycles, and many reinstertions of the cartridge, the C64C booted successfully with the Chameleon inserted only once. No other accessories are connected. The power LED of the C64C turns on, but the screen remains black, or sometimes displays a white screen. The LEDs on the Chameleon remain off. The power supply is a T40/E with 1.7 A on 5 V and 1 A on 9 V.


    Both the C64C and the Chameleon always boot successfully when disconnected. I have noticed that the C64C on a few occasions displays a grayscale image, suggesting that its CT1 trimmer capacitor near the VIC may need to be replaced. Is the Chameleon sensitive to CT1 deviations?


    Finally, it would be useful to forward keyboard commands via the IEC connector (from a C64 keyboard) or the ChaCo USB connector (from a PC keyboard), as an option, and not only the PS/2 connector. The amount of electronic waste can be reduced by not requiring a PS/2 keyboard, just for the Chameleon. :)

  • The (colour) clock being off could be a problem indeed.


    In any case, i would start with cleaning the contacts at the cartridge and the expansion port. also perhaps bend them a little bit to make it a tighter fit.

  • Cleaning the contacts with IPA (chemically pure isopropanol) did not help, unfortunately. The cartridge fits rather snugly in the expansion port. I suppose that one possibility is that the soldering between the connector and the mainboard is loose, but I don't have any evidence of that.


    Is the Chameleon supposed to blink anything with its LEDs when it boots? They remain off when the C64C is powered on.

  • If the C64's clock is off, Chameleon cannot sync to the board and probably fails to initialize before it can show anything on screen. This is not a Chameleon issue, but a "flaky C64" issue (if the clock is really off).


    If your C64 main board has a 8701 clock generation chip, we have a replacement available for that:


    https://icomp.de/shop-icomp/en…/csg8701-replacement.html


    If you're unsure about the clock generation of your C64, just post a picture of the main board.

  • Yes, the clock generation chip is marked K6 8701 2186 S. What is the best approach to determine whether the 8701, possibly only the CT1, needs to be replaced? A logic analyser?

  • A storage oscilloscope should be fine - even an old analogue oscilloscope with trigger-delay function should be fine (I used to have a 2-channel Hameg 60MHz scope with that function). The idea is to measure the Phi0 clock (1MHz that go to the VIC chip) and see how far it jitters after a few milliseconds. Switch on one channel only, set trigger to "falling edge" (that's sharper than the rising edge), trigger level somewhere around 1.4V and make the clock visible for starters. Then start delaying the trigger and see if the waveform gets distorted; if it does, you may improve things by turning the trimmer-cap near the xtal. If that doesn't help, replacing the Xtal, caps and 8701 (or placing our replacement in that socket) will help.

  • It turned out that the Chameleon is somehow incompatible with this C64C PSU. We tried another PSU, and then the Chameleon started without apparent issues. Now things get mysterious:

    • The oscilloscope confirms that the video clock is nice and stable under all circumstances, with either PSU, and with or without the Chameleon inserted in the expansion port. The clock runs nicely despite a crashed Chameleon. We did a minor adjustment to CT1 to have it in the middle of the acceptable clock interval, rather than at the edge near grayscale problems. This didn’t help the Chameleon to boot, though.
    • VCC is 5.27 V with either PSU, so the voltage appears to be good.

    In short, with the oscilloscope we were unable to measure any meaningful difference between the PSUs. Still, the Chameleon refuses to boot with one of them.


    One idea is that there is a difference in the PSU power-up sequence, and that the Chameleon starts executing too early, before power is stable enough, leading to an immediate crash. The oscilloscope measured the power-up from 0 to 5 V to 1.4 ms in both cases, and they appeared to behave about same.


    Any idea what the problem could be?

  • Any idea what the problem could be?

    It may be the voltage itself: 5.27V is slightly above the 5V rating, and the Chameleon has a Z-diode at it's input that starts burning power above 5.1V. This extra load may get one of the PSU's regulation circuits into oscillation.


    One idea is that there is a difference in the PSU power-up sequence, and that the Chameleon starts executing too early, before power is stable enough, leading to an immediate crash. The oscilloscope measured the power-up from 0 to 5 V to 1.4 ms in both cases, and they appeared to behave about same.

    If you believe that it's the startup behaviour of the PSU, try a reset after switching on.

  • Both PSUs have approximately 5.27 V, and unfortunately pressing the reset button doesn’t help.


    One peculiar observation is that the red power LED on the C64C is at about half the intensity when booting fails, as compared to a successful boot when the LED is at full intensity.


    I have now also verified that the Ultimate II+ has similar issues with the PSU that the Chameleon fails with.


    What are the best PSU replacement options? It would be nice to replace the PSU internals and reuse its external casing, if possible.

  • What are the best PSU replacement options?

    There is a company from poland that offers PSUs for C64 and Amiga, but I strongly recommend to avoid those, as previous support cases have shown that the parts they're using are not suitable for the advertised task.


    The best advice I can give is to use an original PSU - not the brown wedge-shaped one, but the white cuboid model with the imitations of heat sink finns on the outside.

  • Hmm... In fact I’m currently using an original cuboid PSU looking exactly like that, and buying another one for EUR 30 or so seems like a bit of a lottery.


    I recently replaced the PSU for an external SF314 floppy unit for the Atari ST with this Mean Well RPD-60A dual 5 V and 12 V PSU, and it fit nicely inside its original PSU case with minimal soldering:


    https://www.meanwell-web.com/e…ame-power-supply-rpd--60a


    Is it possible to find something similar to replace the C64C PSU? Admittedly, dual 5 V DC and 9 V AC could be a rare combination.

  • The one thing I'd avoid is MeanWell chassis - all "commercial" solutions for Amiga and C64 use them, and they cause trouble all along.

    Would you be able to elaborate on the kind of problems? Modern PSUs, such as Mean Well, seem to comply with significantly higher standards than the original PSUs from the 1980’s, when measuring ripple, and so on. For 9 V AC it seems transformers from Breve Tufvassons are somewhat popular in home-built PSUs, and for 5 V DC just about any modern components.


    I understand that the cheapest stuff likely is crap, but surely there must exist plenty of modern and reasonably priced PSU components that by a wide margin exceeds the power requirements of a C64?

  • Would you be able to elaborate on the kind of problems?

    Just search for PSU problems in this support forum, mainly Amiga-related: Whenever there were problems "hard to explain", they were eliminated after using an original PSU.


    Modern PSUs, such as Mean Well, seem to comply with significantly higher standards than the original PSUs from the 1980’s, when measuring ripple, and so on.

    That's not all you need to look at. Commodore PSUs are far from the power sink, so you have to look at a chain of losses between source and sink: The cable, the connector, the switch (in case of the C64) and the fuse. So the search term for a proper chassis would be "sense wire", as that would allow for automatically compensating these losses. You can also pre-compensate by raising the output voltage with the current that flows - that's what Commodore PSUs do. However, the commercially available solutions do simply nothing in terms of compensating for these losses, hence they are simply not suitable for the advertised application.

    but surely there must exist plenty of modern and reasonably priced PSU components that by a wide margin exceeds the power requirements of a C64?

    You are right that the power requirements of the C64 are low, but since "modern" and the large distance from power source to sink don't go together, the only thing to really go for would be a point-of-load regulator placed very close to the C64, or even inside the C64. I do have a concept for that, but won't have the time to take it to the prototype stage before the end of the year.

  • Thanks, Jens, that was helpful to learn. I suppose one could try to cut down the power cable from 1.6 m to 0.2 m or so, but if the original Commodore PSU is preconditioned for a certain length (and thereby a certain voltage drop), then I suppose the voltage may become too high. Nevermind having to put a large power brick with such a short cable on the desk. The switch and fuse could be replaced, of course, and the connector could be cleaned and resoldered.


    Your regulator concept seems like a promising idea, especially if it could be built in. Would you know if anyone else has successfully attempted to do this?

  • Your regulator concept seems like a promising idea, especially if it could be built in. Would you know if anyone else has successfully attempted to do this?

    I don't know of anyone having implemented this idea yet. I did so on the C64 Reloaded and the C64 Reloaded MK2 - the latter still selling to lots of satisfied customers.

  • With the V1 and the C64C NTSC, I would get the occasional no boot on power up from the pair. But this happened on occasion with a Retro Replay. I found that the C64C required a warm-up period. I would boot it without anything attached then leave it on for 30 seconds. Then off, insert cart and power up. It seemed to fix the issue.

  • But this happened on occasion with a Retro Replay.

    This is a property of the Retro Replay, in combination with the C64 - it's because the Retro Replay uses the 8MHz pixel clock and to get into more detail, the phase relation of this clock to the Phi2 clock. On a (very) cold C64, the phase relation of the two clocks confuses the logic. Leaving it on for maybe 10 seconds is mostly enough. Granted, if the room is very cold, it may take the 30 seconds you're describing here.


    This property is actually still in the "Nordic Replay" product. Since the phase relation of the two clocks stabilizes after this short warm-up period, and a proper C64 session is way longer than that, I figured I'll leave it like this, as it's still (after 18 years) the only product that also works fine on the C128 - and I really don't want to go through the hell of making the timing right for C64 and C128 at the same time!


    BTT: The Chameleon does not use the 8MHz pixel clock, but only the Phi2 clock to generate it's bus timing, so it does not depend on any warmup of the computer.

  • The last reply was more than 365 days ago, this thread is most likely obsolete. It is recommended to create a new thread instead.