Excellent video. When I checked the resistors on my board I did notice that Commodore used higher rated resistors in several locations. Thank you for sharing this.
Excellent source for anyone that wants to do these mods! Thanks! My original C64 was an early model 5-pin video. We also later got a 1701 monitor. Back then they sold a cable that could be used with the 5-pin models and separated inputs on the back of the 1701/1702. It was a pretty well known hack and was published by Jim Butterfield in an issue of COMPUTE! or COMPUTE!'s Gazette in 1984. You basically used Luma for Luma and used the Composite for Chroma. It does work but the levels are a little off so the colors are a bit brighter. It did give you a much better picture then the standard Composite but not as good as the newer 8-pin models. Anyway, your method here is much better. Keep up the good work.
I just tried using the composite for chroma and yeah, it works but it's not great. Kind of a shame as I can think of a few cases where using composite in place of chroma would be useful.
@@TheRetroChannel For sure not as good as true separated. But back in the 80s it was still much better then composite and gave a sharper picture. That is on an unmodded C64 though.
Mark, just wanted to say THANKS for this information. I've replaced the resistor at R10 in three C64s and the output looks so much better. Keep up the great videos!
This is some great information. I have been sitting on quite a few 5-pin machines that were originally from our local school district and this seems a great way to keep them as close to original as possible but improve the video out.
I have to agree with replacing the audio in with the chroma out signal. The audio in on the C64 was almost never used for anything and I don't even know how useful it is really is. The best it could do would be to mix incoming audio with SID tunes so filters and volume get applied. It is possible a special cartridge/user port add-on may have used it, but you still wouldn't need it since you could mix that manually with the audio out yourself. In most cases, I've heard it just adds noise sitting there unused anyway.
Also, you may want to think about grounding the audio in line. It has been reported to improve audio for real SID chips, but obviously would have no effect on the ARMSID you have installed there.
@jandjrandr I've experimented with grounding the audio in (on a real sid), and while it does make a difference, I want to do further analysis before I draw a final conclusion
I'll be damned. As someone who regularly modded 326298 boards in all these same ways, the reset mod shown here is actually slightly easier and cleaner version of the one I did a few years back on my channel. Another easy mod worth looking into is replacing the SID filter capacitors as well, because the factory caps are incorrect (I believe they came with 2200pf from memory? been a few years) and should be replaced with 470pF ceramics so that the SID's filters sound the same as they do on the later boards with 6581 SID chips. Great work; I've always been a big fan of these simple mods to make use of these otherwise overlooked '82 326298 boards...
Thanks. I didn't know about the SID filter error, I don't remember it sounding any different or if it did I probably put that down to the armsid. I'll do some more tests with a real sid and the filter caps
Fun fact, the earliest silver label 326298 were actually Rev C and Rev B while Rev A took over towards the end of the silver label run and the rest of the 326298 production.
Excellent mods. I followed but with a further change. In RF can, ran the composite to the RF out jack. Then, dropped the composite from 5pin din rather than audio in.
I wonder what could happen to the SID if one were to forget which cable went with which machine... plugging a monitor's chroma in to the SID audio in. In theory, nothing...? At least your mod to go to the RF port would prevent that, but then requires a separate cable.
That looks great for a RevA. I keep my Rev As packed up and not used. Not because I want to preserve them, it's because the look terrible ;) One of my best video output machines is a 425 board that I PAL modded.
@@BellaLugoshi On my 64C I'm using separate (luma/chroma) to an S-Video-to-HDMI converter box and the display is nice - definitely a step up in quality from composite.
Why did Commodore use resistors that put the video output voltage levels so far out of spec? Has anyone asked Bil Herd? While you're at it, ask why it's "Bil" instead of "Bill."
Actually, nice. My C64 is this model, the old one without the chroma pin. I have 1541 Ultimate, and I don't know if that's supposed to reset the machine or not. It seems to be working that I know, but maybe something broken and I just assumed it was C64 random?
Interesting video. I have a C128 DCR and the video output on those are horrible!. Was wondering if getting a replacement modulator would solve the issue. Can't find much information about the video modulation on this specific model though.
Funny you should mention the DCR, I just posted a little thread on twitter about putting an RF replacement in one of those twitter.com/TheRetroChanneI/status/1651914147691630597?t=4jch5gvuZS8V7GuwUbrygg&s=19
@@TheRetroChannel Now all I need is a way to fix my R/W head of the internal drive. It did not survive a intercontinental move (forgot to put the cardboard head protector in). One side is still attached, the other is cracked and makes for a tilted head.
I have this exact board and problem. Amazing that you are working on exactly the same thing at the same time. Which Atari video cable did you use? I don't see any designed for chroma/luma.
@@TheRetroChannel cool. On the other end of the Atari video cable, are the separate chroma and luma signals coming out of one of the RCA connectors or the actual S-video plug?
@@TheRetroChannel makes sense. I ordered an Atari video cable but when you mentioned that you made yours I pulled apart my existing composite, L/R audio and rewired it for Chroma, Luma and mono audio. What a difference! Amazing. Thank you for this tutorial!
When you say the CHROMA signal is "too hot", what was its voltage level? I'm concerned that attenuating it will make it fare not-too-well with a 1701 / 1702 (where the manual specs 1Vpp for the CHROMA signal).
I don't remember but 300mV will still work just fine with Commodore monitors that expect 1V. Worst case scenario is you will need to turn the colour control up slightly
@@TheRetroChannel I finally got around to poking at this a bit more today. With a white border, on an unmodified board, the LUMA signal clocks in at ~3.2Vp-p right after its emitter follower (i.e. before passing through R9), and the CHROMA, right off the emitter follower, is at about 3.1Vp-p. Obviously that would be "too hot" without a series resistor to create the voltage divider like the LUMA has. Anyway, what's a little puzzling is that, with a 75ohm load on the display, that the output would be nearly 2Vp-p ... With he stock value in R9, it should be ~1.6Vp-p, which is still high but easier to live with. Even with the 5% resistor tolerance, it should be < 1.7Vp-p. I wouldn't expect these resistors to drift *down* over time!
Hi, excellent work. I've only worked on this revision once, and now have one in front of me that I am being paid for. Yes! It passes all tests. I am definitely going to do those mods, especially the regular first . Is that the 5 or 12 regulator? I don't really want to probe down there. 7805?
Nice video bud. I'm currently working on a shortboard. :) strange fault, but I'll get it. You are pretty smart with your mods. I always enjoy your videos
It has black screen, but with fastload, it boots to basic. Diag with harness runs perfectly for 10+ times. A cartridge works, sometimes freezing with random characters. I can load a disk no problems. I changed the kernal/basic rom chip, no change. Soldering is good. So I changed ram tonight, just because I had it. No change. I really thought it was a flaky kernal basic rom::( I'm thinking possibly the character rom, but I don't know. I'm not sure how these short boards work. 250469. I'm stumped
@@donaldblakley6796 Interesting, yeah my money would have been on the kernal/basic ROM. I'm thinking either CPU or PLA, I'd go CPU first as it's easier to get out and replace, you can even use a longboard 6510. And I'm yet to come across a bad shortboard PLA
The only thing in the 64 that relies on the mains frequency is the time of day clocks in the CIA's, and only a handful of programs use these clocks anyway. All other frequencies (CPU clock, video output) are derived from the crystal oscillator inside the 64
@@TheRetroChannel perfect, that’s amazing, I’d never realised that it didn’t rely on external timing from power, so I’ve always avoided!? Do you hang out on discord with the VCF crowd like Bil Herd and Johan Grip and all of those guys? If you don’t, you really should…! :)
There's no easy answer as it can vary even with the same board revision. The C64C 250469 shortboard is probably the safer bet for getting good video quality
After making this modification, I ran the diagnostics cart and it came back with a bad SID and a bad Control Port. But it played all of the test sounds correctly (as far as I can tell) and the control ports seems to be working fine. Can anyone else replicate this? Could the removal of C12 could be the culprit?
I think I just answered my own question. Page 1-6 of the C64 DIAGNOSTIC INSTRUCTION AND TROUBLESHOOTING MANUAL says On some of the older 5-pin video output pc's, the control port will Display 'bad" and indicate a defective 6581 SID IC, I18, each time the Control port test runs. This is due to different value capacitors, PCB Locations C48 and C93, at pins 23 and 24 of U18, which cause the digital output to fall outside of the specified range. These capacitors values should be 1800 pF. This situation will not cause any problems with paddle operation and these capacitors need not be changed if a different value cap was used at the time of manufacture.
Interesting, I didn't know about that one. Pretty sure I've run the diagnostic on it in the past and everything was ok, maybe my board already has the correct value caps
@@TheRetroChannel it does say that they say that this is true for only "some of the older 5-pin" boards. So I suppose it is possible that yours is one of these boards for which this doesn't apply. Oh, Commodore!
Commodore made a lot of shortcuts to the point they got the Monitor OEM to add a fix for the crappy C64 video out. I find my silver label rev A to be dark, limited jail bars, but the colours are trash. I'm going to look at this resistor. PAL games on a NTSC machine are HARD, they speed up.
Right now you can have Raspberry Pico for 5$ which can emulate every 8Bit home computer, also with VGA and stereo audio output. No need for original old hardware anymore.
It is a Revision 1, YES there where mistakes, but COMMODORE Fixed them in Rev 2, these mods are variations of Publicly available fixes, I remember doing these Repairs & updates as Commodore Dealer, nice to see these updates still being installed in older boards, but "Commodore F**Ks Up", is not a discovery you made ...
I just got a 326298 C64 that needs these mods. Thanks for sharing your videos.
Thanks for the super thanks Ben, I really appreciate it
Excellent video. When I checked the resistors on my board I did notice that Commodore used higher rated resistors in several locations. Thank you for sharing this.
Excellent source for anyone that wants to do these mods! Thanks! My original C64 was an early model 5-pin video. We also later got a 1701 monitor. Back then they sold a cable that could be used with the 5-pin models and separated inputs on the back of the 1701/1702. It was a pretty well known hack and was published by Jim Butterfield in an issue of COMPUTE! or COMPUTE!'s Gazette in 1984. You basically used Luma for Luma and used the Composite for Chroma. It does work but the levels are a little off so the colors are a bit brighter. It did give you a much better picture then the standard Composite but not as good as the newer 8-pin models. Anyway, your method here is much better. Keep up the good work.
I just tried using the composite for chroma and yeah, it works but it's not great. Kind of a shame as I can think of a few cases where using composite in place of chroma would be useful.
@@TheRetroChannel For sure not as good as true separated. But back in the 80s it was still much better then composite and gave a sharper picture. That is on an unmodded C64 though.
Wow... I can't believe how much better that looked after. Great job!
Mark, just wanted to say THANKS for this information. I've replaced the resistor at R10 in three C64s and the output looks so much better. Keep up the great videos!
This is some great information. I have been sitting on quite a few 5-pin machines that were originally from our local school district and this seems a great way to keep them as close to original as possible but improve the video out.
Great video, more like the absolutely necessary guide to keep your micro alive for the years to come! thanks mate!
Nice!!!, mods that are relatively easy to do and are reversible!
Great summary of upgrade mods for this board... I will do the same... as it has been neglected for the reasons you mentioned on the video!
This channel literally covers everything
I have to agree with replacing the audio in with the chroma out signal. The audio in on the C64 was almost never used for anything and I don't even know how useful it is really is. The best it could do would be to mix incoming audio with SID tunes so filters and volume get applied. It is possible a special cartridge/user port add-on may have used it, but you still wouldn't need it since you could mix that manually with the audio out yourself. In most cases, I've heard it just adds noise sitting there unused anyway.
Also, you may want to think about grounding the audio in line. It has been reported to improve audio for real SID chips, but obviously would have no effect on the ARMSID you have installed there.
@jandjrandr I've experimented with grounding the audio in (on a real sid), and while it does make a difference, I want to do further analysis before I draw a final conclusion
That’s really clever. Good job, and excellent result!
You outdid yourself today,. Thank you very much indeed.
I'll be damned. As someone who regularly modded 326298 boards in all these same ways, the reset mod shown here is actually slightly easier and cleaner version of the one I did a few years back on my channel. Another easy mod worth looking into is replacing the SID filter capacitors as well, because the factory caps are incorrect (I believe they came with 2200pf from memory? been a few years) and should be replaced with 470pF ceramics so that the SID's filters sound the same as they do on the later boards with 6581 SID chips. Great work; I've always been a big fan of these simple mods to make use of these otherwise overlooked '82 326298 boards...
Thanks. I didn't know about the SID filter error, I don't remember it sounding any different or if it did I probably put that down to the armsid. I'll do some more tests with a real sid and the filter caps
There was a similar error made with the SID filter circuit in the SX-64 machines with the wrong value caps installed.
Fun fact, the earliest silver label 326298 were actually Rev C and Rev B while Rev A took over towards the end of the silver label run and the rest of the 326298 production.
Isn't there a SID fix as well? Like a 470pf Cap to swap out the improper one? -Mark.
Yeah aphexteknol also mentioned it in the comments here. I have not tried it out yet
Excellent mods. I followed but with a further change. In RF can, ran the composite to the RF out jack. Then, dropped the composite from 5pin din rather than audio in.
I wonder what could happen to the SID if one were to forget which cable went with which machine... plugging a monitor's chroma in to the SID audio in. In theory, nothing...?
At least your mod to go to the RF port would prevent that, but then requires a separate cable.
Excellent video! Great info
@TheRetroChannel Thanks for the great insights. You were talking about the NTSC model. I have the same assy, but the PAL model. Any hints on that?
The same may apply to the PAL model. The PAL 326298 boards are very rare and I'm yet to come across one
@TheRetroChannel I see, I'll try to see what I can do with mine.
11:41 Just to make sure, "that cap" is C71, right?
C79
That looks great for a RevA. I keep my Rev As packed up and not used. Not because I want to preserve them, it's because the look terrible ;) One of my best video output machines is a 425 board that I PAL modded.
Maybe it's time to bring that rev A out of storage
Hi, thnx for all videos. One question - C64C needed this mods?
No, it only applies to the older 326298 boards
@@TheRetroChannel it's just that I also have a grid and blurring on the c64c through the composite output. so I'm thinking how to solve the problem.
@@BellaLugoshi On my 64C I'm using separate (luma/chroma) to an S-Video-to-HDMI converter box and the display is nice - definitely a step up in quality from composite.
@@TheUtuber999 thnx
@@BellaLugoshi I see you're a Bauhaus fan. 💪
Excellent work. I'm so used to playing on NTSC machines when I play on PAL the games seem so slow.
Why did Commodore use resistors that put the video output voltage levels so far out of spec? Has anyone asked Bil Herd? While you're at it, ask why it's "Bil" instead of "Bill."
Actually, nice. My C64 is this model, the old one without the chroma pin. I have 1541 Ultimate, and I don't know if that's supposed to reset the machine or not. It seems to be working that I know, but maybe something broken and I just assumed it was C64 random?
Thanks! I've got a machine that has this reset issue and I'm going to at least do the reset mod on it!
Interesting video. I have a C128 DCR and the video output on those are horrible!. Was wondering if getting a replacement modulator would solve the issue. Can't find much information about the video modulation on this specific model though.
Funny you should mention the DCR, I just posted a little thread on twitter about putting an RF replacement in one of those twitter.com/TheRetroChanneI/status/1651914147691630597?t=4jch5gvuZS8V7GuwUbrygg&s=19
@@TheRetroChannel Now all I need is a way to fix my R/W head of the internal drive. It did not survive a intercontinental move (forgot to put the cardboard head protector in). One side is still attached, the other is cracked and makes for a tilted head.
@@anotheruser9876 Damn, that sounds like it's going to be a tricky one to fix. I don't know what to suggest apart from good luck
I have this exact board and problem. Amazing that you are working on exactly the same thing at the same time. Which Atari video cable did you use? I don't see any designed for chroma/luma.
I built my own cable a long time ago when I got the Atari. You can see the pinout at 10:50
@@TheRetroChannel cool. On the other end of the Atari video cable, are the separate chroma and luma signals coming out of one of the RCA connectors or the actual S-video plug?
You would want to wire the luma and chroma to an svideo plug, unless you are using it with a Commodore monitor that has the 2 RCAs for luma and chroma
@@TheRetroChannel makes sense. I ordered an Atari video cable but when you mentioned that you made yours I pulled apart my existing composite, L/R audio and rewired it for Chroma, Luma and mono audio. What a difference! Amazing. Thank you for this tutorial!
When you say the CHROMA signal is "too hot", what was its voltage level? I'm concerned that attenuating it will make it fare not-too-well with a 1701 / 1702 (where the manual specs 1Vpp for the CHROMA signal).
I don't remember but 300mV will still work just fine with Commodore monitors that expect 1V. Worst case scenario is you will need to turn the colour control up slightly
@@TheRetroChannel I finally got around to poking at this a bit more today. With a white border, on an unmodified board, the LUMA signal clocks in at ~3.2Vp-p right after its emitter follower (i.e. before passing through R9), and the CHROMA, right off the emitter follower, is at about 3.1Vp-p. Obviously that would be "too hot" without a series resistor to create the voltage divider like the LUMA has.
Anyway, what's a little puzzling is that, with a 75ohm load on the display, that the output would be nearly 2Vp-p ... With he stock value in R9, it should be ~1.6Vp-p, which is still high but easier to live with. Even with the 5% resistor tolerance, it should be < 1.7Vp-p. I wouldn't expect these resistors to drift *down* over time!
Hi, excellent work. I've only worked on this revision once, and now have one in front of me that I am being paid for. Yes! It passes all tests. I am definitely going to do those mods, especially the regular first . Is that the 5 or 12 regulator? I don't really want to probe down there. 7805?
Yeah 7805. Good luck
Ok. Thank you. I really appreciate it (even though you are getting sick of me) what gauge wire do you recommend? Also. 2watt resistors are good?
Also... is there no filters with diag on these boards?
2W resistors would be overkill, 1/4W will be fine. And "filters"? do you mean SID filters
Yes. Sid filters, and I already have 2w resistors. Will that be ok then?
Nice video bud. I'm currently working on a shortboard. :) strange fault, but I'll get it. You are pretty smart with your mods. I always enjoy your videos
Thanks mate. What's happening with the shortboard?
It has black screen, but with fastload, it boots to basic. Diag with harness runs perfectly for 10+ times. A cartridge works, sometimes freezing with random characters. I can load a disk no problems. I changed the kernal/basic rom chip, no change. Soldering is good. So I changed ram tonight, just because I had it. No change. I really thought it was a flaky kernal basic rom::( I'm thinking possibly the character rom, but I don't know. I'm not sure how these short boards work. 250469. I'm stumped
@@donaldblakley6796 Interesting, yeah my money would have been on the kernal/basic ROM. I'm thinking either CPU or PLA, I'd go CPU first as it's easier to get out and replace, you can even use a longboard 6510. And I'm yet to come across a bad shortboard PLA
Cpu is good, swapped with my other short board. I have another kernal basic rom coming next week. I'm still convinced it's that. Super pla can go bad?
Thanks
Haaa!! In the future will unmodified Rev.A 326298's be going for a premium price????
Hey dumb question, htf does an ntsc version work with 50hz power?!
The only thing in the 64 that relies on the mains frequency is the time of day clocks in the CIA's, and only a handful of programs use these clocks anyway. All other frequencies (CPU clock, video output) are derived from the crystal oscillator inside the 64
@@TheRetroChannel perfect, that’s amazing, I’d never realised that it didn’t rely on external timing from power, so I’ve always avoided!? Do you hang out on discord with the VCF crowd like Bil Herd and Johan Grip and all of those guys? If you don’t, you really should…! :)
Does the C64 support RGB via Scart?
I wish. The best you can get out of the VIC-II is s-video
@@TheRetroChannel
So what is the best quality out of a C64 or C64C?
There's no easy answer as it can vary even with the same board revision. The C64C 250469 shortboard is probably the safer bet for getting good video quality
Nice, try out the Skate or die intro at NTSC speed :D
After making this modification, I ran the diagnostics cart and it came back with a bad SID and a bad Control Port. But it played all of the test sounds correctly (as far as I can tell) and the control ports seems to be working fine. Can anyone else replicate this? Could the removal of C12 could be the culprit?
I think I just answered my own question. Page 1-6 of the C64 DIAGNOSTIC INSTRUCTION AND TROUBLESHOOTING MANUAL says
On some of the older 5-pin video output pc's, the control port will Display 'bad" and indicate a defective 6581 SID IC, I18, each time the Control port test runs. This is due to different value capacitors, PCB Locations C48 and C93, at pins 23 and 24 of U18, which cause the digital output to fall outside of the specified range. These capacitors values should be 1800 pF. This situation will not cause any problems with paddle operation and these capacitors need not be changed if a different value cap was used at the time of manufacture.
Interesting, I didn't know about that one. Pretty sure I've run the diagnostic on it in the past and everything was ok, maybe my board already has the correct value caps
@@TheRetroChannel it does say that they say that this is true for only "some of the older 5-pin" boards. So I suppose it is possible that yours is one of these boards for which this doesn't apply. Oh, Commodore!
Nice video.
Cool video. 👍👏
Commodore made a lot of shortcuts to the point they got the Monitor OEM to add a fix for the crappy C64 video out. I find my silver label rev A to be dark, limited jail bars, but the colours are trash. I'm going to look at this resistor. PAL games on a NTSC machine are HARD, they speed up.
You big babies. When I was a kid this was new and awesome at a cost. Nobody back then expected anything different than crappy RF. That was normal.
When we were kids it looked great because there wasn't anything else to compare it to. I mean, it looked better than Apple graphics! lol
Great vid.
But. .. stop hogging all the 1084 monitors...😂
he says chroma luma or svideo bro they are the same 😆
Right now you can have Raspberry Pico for 5$ which can emulate every 8Bit home computer, also with VGA and stereo audio output. No need for original old hardware anymore.
Oh wow really? Well I guess I'll pack up all this crap and take it to the dump.
What's the video / audio latency penalty?
It is a Revision 1, YES there where mistakes, but COMMODORE Fixed them in Rev 2, these mods are variations of Publicly available fixes, I remember doing these Repairs & updates as Commodore Dealer, nice to see these updates still being installed in older boards, but "Commodore F**Ks Up", is not a discovery you made ...