Beamracer power requirements

[bodgit]

I use a 1541-U2 with my Beamracer-equipped C64 and I’ve found when entering and leaving the U2 menu that the screen can get corrupted and the machine becomes slightly unstable in general, random crashes/lockups, etc.

On a hunch I took the Beamracer out and now my C64 seems stable again. The Beamracer was inactive at the time so it should be fairly passive. My only thought was my C64 is a long board that has a separate 7805 regulator that provides a dedicated +5V for the VIC and possibly other parts of the machine.

Is the added load of the Beamracer likely to hit the current limit for the 7805? I read somewhere that the normal load is about 250-300mA, how much extra does the (inactive) Beamracer draw? In order to measure it I would have to desolder the regulator in order to put my multimeter in series.

Other than that, does anyone have any other ideas? My C64 has a component mod board and a SIDFX which makes me think I might be hitting a current limit somewhere…

[laubzega]

I use a 1541-U2 with my Beamracer-equipped C64 and I’ve found when entering and leaving the U2 menu that the screen can get corrupted and the machine becomes slightly unstable in general, random crashes/lockups, etc.

On a hunch I took the Beamracer out and now my C64 seems stable again. The Beamracer was inactive at the time so it should be fairly passive. My only thought was my C64 is a long board that has a separate 7805 regulator that provides a dedicated +5V for the VIC and possibly other parts of the machine.

Is the added load of the Beamracer likely to hit the current limit for the 7805? I read somewhere that the normal load is about 250-300mA, how much extra does the (inactive) Beamracer draw? In order to measure it I would have to desolder the regulator in order to put my multimeter in series.

Other than that, does anyone have any other ideas? My C64 has a component mod board and a SIDFX which makes me think I might be hitting a current limit somewhere…

Hi Bodgit,

Depending on what it is currently doing, BeamRacer consumes a little bit less or a little bit more than 80 mA from the +5V pin (VDD - pin 13 - is not used). When deactivated, BR's power use should be below 75 mA. NMOS VIC-II needs around 180 mA of +5V, so the total worst case should be around 260 mA. You can measure it yourself by stacking DIP-40 sockets like in the attached photo (note the removed pin 40).

There are some other components near VIC-II that rely on the +5V CAN supply, but given that the 7805 can easily provide 1A of current, power should not be a problem (and it has not been in our experience so far).

Can you please share more details about your setup:

- C64 mainboard revision,
- BeamRacer revision (see the right edge near top-right corner),
- Version of the C64 power supply,
- VIC-II version,
- Where the component mod board takes its +5V from?
- Do you use BeamRacer's VideoMod Connector or the original adapter?

Also:
- Does instability only occur after entering U2 menu or can it be triggered in other ways?
- Does instability after entering U2 menu occurs every time or from time to time?
- Can you share a photo of screen corruption after using U2 menu?
- Is your SIDFX installed permanently or can be removed for testing?
- Since you mentioned a multimeter - can you please verify that the power supply provides correct voltages?


These are many questions, but hopefully the answers will help us pinpoint the culprit.

[bodgit]

I use a 1541-U2 with my Beamracer-equipped C64 and I’ve found when entering and leaving the U2 menu that the screen can get corrupted and the machine becomes slightly unstable in general, random crashes/lockups, etc.

On a hunch I took the Beamracer out and now my C64 seems stable again. The Beamracer was inactive at the time so it should be fairly passive. My only thought was my C64 is a long board that has a separate 7805 regulator that provides a dedicated +5V for the VIC and possibly other parts of the machine.

Is the added load of the Beamracer likely to hit the current limit for the 7805? I read somewhere that the normal load is about 250-300mA, how much extra does the (inactive) Beamracer draw? In order to measure it I would have to desolder the regulator in order to put my multimeter in series.

Other than that, does anyone have any other ideas? My C64 has a component mod board and a SIDFX which makes me think I might be hitting a current limit somewhere…

Hi Bodgit,

Depending on what it is currently doing, BeamRacer consumes a little bit less or a little bit more than 80 mA from the +5V pin (VDD - pin 13 - is not used). When deactivated, BR's power use should be below 75 mA. NMOS VIC-II needs around 180 mA of +5V, so the total worst case should be around 260 mA. You can measure it yourself by stacking DIP-40 sockets like in the attached photo (note the removed pin 40).

There are some other components near VIC-II that rely on the +5V CAN supply, but given that the 7805 can easily provide 1A of current, power should not be a problem (and it has not been in our experience so far).

Can you please share more details about your setup:

- C64 mainboard revision,
- BeamRacer revision (see the right edge near top-right corner),
- Version of the C64 power supply,
- VIC-II version,
- Where the component mod board takes its +5V from?
- Do you use BeamRacer's VideoMod Connector or the original adapter?

Also:
- Does instability only occur after entering U2 menu or can it be triggered in other ways?
- Does instability after entering U2 menu occurs every time or from time to time?
- Can you share a photo of screen corruption after using U2 menu?
- Is your SIDFX installed permanently or can be removed for testing?
- Since you mentioned a multimeter - can you please verify that the power supply provides correct voltages?


These are many questions, but hopefully the answers will help us pinpoint the culprit.

C64 mainboard revision: 250425
BeamRacer revision: 1.1
C64 PSU: Aftermarket, rated 5V DC @ 2A, 9V AC @ 1A
VIC-II: 6569R3

Component mod board is installed in place of the original modulator so it's taking power from those pins. I'm using the BeamRacer connector rather than stacking VIC adapters. I'm not actually using the component output when this happens, just the normal VIC-II output over S-Video, I haven't actually compared what the component mod output looks like.

Instability can happen when playing a game, it will randomly crash, graphics will go bad, etc. I've noticed from past experience going in and out of the U2 menu will eventually trigger some sort of glitch. Sometimes it takes a while, sometimes it happens first go. Sometimes going in and out of the menu will clear the glitch and it will look normal again and sometimes it will just wedge harder the more times you do it.

Sometimes all of the graphics are scrambled like here:
   

Sometimes one or two characters will be wrong (look at the second 'o' in the oneload64 directory name):
   

This was Last Ninja Remix that crashed:
   

SIDFX can be removed if needed. I can measure the voltages although I'm not sure where the best spots are on the C64 to take the measurements from.

I should also add, I don't think the original IC sockets (at least on the VIC-II) on my C64 are very good quality now (I can pull the VIC-II chip out with just my fingers without needing an extractor tool or screwdriver), also the MOS 8701 chip next to the VIC socket stops the BeamRacer inserting fully flat, if I use a second socket as a spacer everything feels a bit loose and wobbly and I can't then close the C64 up as the heatsink I've added to the VIC-II butts up against the underside of the keyboard. My plan is to eventually remove the original socket and install a turned pin socket with longer legs and get the height set just right so the BeamRacer doesn't foul on the 8701 and the keyboard can sit over the VIC-II heatsink. I do wonder if the combination of the worn socket and the interference from the 8701 is causing a very intermittent connection on one pin?

Also, the C64 has had all of the electrolytic caps replaced.

Here's the full install:
   

[laubzega]

Thanks for the inputs, looks interesting.

Instability can happen when playing a game, it will randomly crash, graphics will go bad, etc. I've noticed from past experience going in and out of the U2 menu will eventually trigger some sort of glitch. Sometimes it takes a while, sometimes it happens first go. Sometimes going in and out of the menu will clear the glitch and it will look normal again and sometimes it will just wedge harder the more times you do it.

Just to clarify - does it mean that the glitches can only occur _after_ entering U2 menu at least once? Or that entering menu is just increasing the likelihood of crash/glitch, but is not strictly necessary?


I have been experimenting most of the day with Ultimate-II Plus 3.6a (115), a barebones 250425 board, BeamRacer, and a variety of power sources. I was repeatedly (>100x) entering and exiting U2 menu from the boot screen, with following results:

- two different original PSUs: no issues with and without BR,
- externally provided +5V (bench PSU): one screen memory corruption event (with no BR inserted at the time),
- externally provided +5V with undervoltage: the lower the voltage, the easier was to create a corruption, with 4.3 V "succeeding" in just a few attempts. I wasn't running careful statistics, but it did not appear that the presence of BR was making any difference,
- original PSUs using low-quality connection (see the photo): having jumper wires and hook clips leading from the PS plug to C64 power socket was producing screen corruption around once per twenty menu enter/exit events. (This setup corresponds to about 4.7V on +5V pin, but also adds poor ground).


The screen and color ram corruption were relatively mild, but I'd expect that having a demo/game running while memory corruption is taking place could be much more impressive visually. EDIT: after testing with a demo, it indeed does blow up spectacularly.


Things to do:

- please measure PS voltages while computer is powered on, ideally at power socket pins,
- please try another power supply, if possible,
- if U2 is not stricly required to cause a corruption, please remove it and see if problems persist,
- please remove SIDFX and see if there is change in stability, 
- please download and run VSP Lab with and without BR inserted, to see how susceptible to VSP bug your machine is (https://csdb.dk/release/?id=120810),


While it does not appear that the socket situation you describe is solely responsible for the instability, it can certainly be a part of the problem. Can you see if touching/nudging the BR while the computer is powered on causes any visual glitches and/or crashes?

Also, what is the make of your PS?