Commodore 1541 II Cheap Replacement PSU Test

[@ZEROSTATIC72]

The ripple on the output of the switching type power supply is completely normal for that type of power supply and will likely not cause any trouble with the drive. The switching power supply output is far more noisy due to the high frequency switching in the supply being coupled to the outputs. The 5V for the drive will probably work at plus or minus 10% and the sasme for the 12V rail. There were 2 +5V wires in the cable and it would be best to connect both at the plug into the 1541 as they may be both used to reduce losses in the cable or possibly for the 5V feedback to the supply. The advantages with the switch mode power supply are effeciency and cost of manufacture. There is no need to make a linear supply for the C64 or drive as modern switch mode supplies are far better than what could have been made at the time back in the 80's. The only real issue is cheap capacitors that limit the lifespan of these units. Thanks for the videos. :-)

[@JanBeta]

Ah, thanks for the information. I've seen smoother output from (better and more costly) switching power supplies but this seems indeed to be sufficient for the drive. I'm still going to get (or make) something better some time soon. Did you watch Greisi's video I linked to? He's doing a kind of "best of both worlds" approach using a classic transformer and modern DC to DC converters in his PSU resulting in a very straight output and pretty good efficiency. I quite like that idea and am probably going to build one of these.

[@ZEROSTATIC72]

A switching main converter with a linear post regulator is the normal approach to get really clean power for noise sensitive circuits. A switcher with 15V and 8V outputs then inductor, capacitor filters followed by low dropout regulators on the 5V and 12V rails would give really clean power with decent effeciency. The linear followed by switching topology doesn't really make a lot of sense if you're after really low noise on the outputs. It's rather practical for generating lots of different voltages from a single output transformer but your cheap power supply is likely more than adequate for the task at hand. As for the 1541, the 6502 processor section won't care about 50mV of noise and the motor driver running from the 12V rail won't care either. The head amplifier for the drive is likely the only part that may be noise sensitive at all and should have pretty good power rail noise rejection. If you want it to last, just replace the cheap capacitors with good brand, long life, low ESR variants and make sure the power supply is used in a well ventilated area. Thanks for the reply. :-)

[@JanBeta]

+ZEROSTATIC72 Sounds convincing to me. Thanks! :)

[@danielmantione]

I did a similar thing for my 1541 as my original power supply was broken. I did use one of these "hard drive power supplies": www.ebay.nl/itm/c/231899556009 This power supply seems to be of significant less quality than the one you used. Allthough the 1541-II works perfectly on it, it makes noise appear on my monitor, which means the PSU is simply not suited for purpose. Your display did seem stable and I didn't hear you mention any issue, so I guess the power supply you did use is much better in that regard. For ripple, I don't think it is an issue for the 1541-II, there are ceramic capacitors behind the power connector that already filter the ripple a bit, but more important: The chips work on TTL voltages, the margin in TTL is waaayyy higher than the ripple. In the end I did repair/upgrade my original power supply. It's not easy, you need to get it about of its case without damage to the case, then use a saw to saw through the resin to separate the transformer and the rectifier board. After that install a new board with switching rectifier, because they stay cool no resin is needed to cool that. Worked perfectly in the end and gave me the results I needed. It would be a good topic for a KZbin video, but alas, it is already done. For the C64 I built my own power supply. For this project I made some pictures: www.freepascal.org/~daniel/c64psu/ I believe for the C64, a do it yourself power supply is the best solution, alltough it is a lot of work and not cheap to get all the components. Some people have resorted to simulated AC, or even converted their C64 to DC, but because of the clocks and user port AC, I believe real AC is by far preferable for the C64. The 5V is best implemented with a switching power supply, providing good and more than sufficient power. However, because of the work and costs involved, fixing the original C64 power supply is also a proper approach. I already fixed a few and once wrote a manual how to do it in Dutch language (contains pictures): www.freepascal.org/~daniel/c64_wigvoeding_revisie.pdf

[@danielmantione]

It might be a good idea to install a ferrite bead on the cable, such as this: nl.aliexpress.com/item/x/32626685759.html It prevents that the ripple becomes a radio source (with the cable as antenna), but also prevents that part of the ripple reaches the 1541-II. I use one on my self-built C64 PSU.

[@Dkentflyer]

Really interesting video Jan, was quite suprised how good the performance of the original psu was after all these years.

[@JanBeta]

Yes, most of them work fine and the output is handled by regulators which output very clean voltages. The problem of using the original PSUs is that when they fail they sometimes do so in violent ways (producing overvoltage) and kill the device connected to them.

[@TheDefpom]

You could also strap some more filter caps across the 12v and 5v lines to improve the ripple, a couple of different values to give the best result, maybe even a tantalum to help filter the high frequency content. I think I saw a trimmer in the psu, maybe try using it to see which voltage it adjusts.

[@JanBeta]

Maybe I'll try that. Would be a good exercise for me, too. There's indeed a little trimmer pot in there. Nothing much to lose if I try fiddling with it, I guess. ;) Thanks for the tips!

[@greisisworkbench1236]

Awesome video and thanks for the shout out :-) .

[@JanBeta]

Thanks & you're very welcome. :)

[@discoHR]

Nice find (€2 is dirty cheap for a PSU) and nice tests. I was already starting to type a "use a scope, not a multimeter" comment but then you pulled a Hameg out of nowhere. :) Both rails were too low for my taste. Personally, I wouldn't go below 4.75V because of TTLs inside but you tested it at max current which explains the voltages lower than usual. 1541-II won't draw so much so in reality the voltage is higher. Well done! Oh, regarding the alignment test... It's normal that drive which wrote the disk reads it as perfectly aligned 100%. You need an analog floppy to check the alignment, they're rare, but in a hurry a factory written floppy (original games etc) would do the trick. I wish I knew how to align the 1541 using an oscilloscope. I read the instructions somewhere but forgot to bookmark it and never found it again.

[@JanBeta]

Haha, yeah, maybe I should have used the scope in the first place. I found it easier to read the voltages from the dmm though. ;) The thing with the alignment test dawned at me just after I uploaded the video. Of course it would have 100% accuracy if it formatted the disk itself just a minute ago. D'oh. But as I mentioned, it was the first time I used the diagnostics cart so I'll know better next time. I'm actually planning to do a 1541 (the old model) alignment video. I have found instructions for some refinements and also alignment w/ a scope (in German) here: www.forum64.de/index.php?thread/46261-1541-wartungs-und-einstellarbeiten/

[@idimidodjimi6760]

Much better this time, as you said you are hobbyist but you have showed more interesting things than just a simple show how to solder in video , so this is thumbs up

[@JanBeta]

Thanks! Glad you liked it. I'm learning all the time.

[@MindFlareRetro]

That was very interesting. And 2€ for a PSU? That is cheap. I like that you did the voltage under load and ripple/noise tests. Coincidentally, I had just watched Dave's (EEVBlog) video on measuring ripple after seeing Greisis Workbench PSU build video. Even though your ripple measurement might have done "the crude way", as you say, it does demonstrate the stark difference between the cheap replacement switch mode PSU and original linear PSU. Well done! Perhaps a follow-up with adding filter caps, as ZEROSTATIC72 and GadgetUk164 suggest, would be interesting too. Looking forward to the C64 cartridge video.

[@JanBeta]

Thanks. :) I really hope I find the time to try adding some filter caps and make a short follow-up to this one. In the end, I plan to build Greisi's PSU as a permanent solution. The cartridge video is in the works. Might take some time still though.

[@MindFlareRetro]

Yeah, I haven't had time for video making this week, myself. I did find this short, yet informative video on "Power Supply Decoupling & Filtering: why we use multiple caps in different locations", goo.gl/zGopfH. Thought you might find it interesting as well, or perhaps this is old news to you.

[@JanBeta]

Added to my list. Thanks!

[@user-wj9xq7ig2v]

An Xbox 360 psu is more than sufficient if you have one lying around. Cut the end off. The red cable need to be connected to the blue to enable the psu to turn on. The orange light should go green on the psu. This red blue combo is also your 5 v. 12v are both yellow wires. Ground is the black wires. The last wire isn't used. Din4 connector and you're done.

[@dialog36]

Great video !! i like your channel. Question.. What happen if the Pin 2 is connected to GND?

[@dreamvisionary]

Better to tin cables, insert them into the solder buckets and then apply solder. Filling the solder bucket before putting in wires is more difficult as you have to keep heat on it for longer while you try to get wire in, resulting in melting wire insulation and possible the plastic that hold the pins in the connector.

[@wolcek]

Great video :). However... To your alignment test - it is of no surprise that the drive is 100% aligned to a disk formatted in this drive just a second before. You'd need to use a disk formatted in a properly aligned drive for this test to be meaningful.

[@SianaGearz]

High frequency switching noise isn't that hard to get rid of, just chain series inductors and parallel capacitors. Usually a single inductor is just fine, and with capacitors before and after it should filter out high frequencies at 18db/oct. Make sure to strap the most of capacitance on the device side rather than on supply side, because it really wants to see a low-impedance source across the whole frequency range, but you're trying to increase high-frequency impedance on the power supply side, if this makes any sense :P I'd have an urge to start with tiny little ferrite beads to be placed inside the DIN connector on those lines, that might already do quite a bit.

[@JanBeta]

Thanks for the tips! Someone pointed out in the comments before that there's some additional filtering happening inside the drive already. At least there's some caps on the power lines. I'm going to experiment with adding ferrite beads and some more caps at some point, I think.

[@rarbiart]

with all the thin wires between the connector and the load the voltage meassurements are systematically flawed. BTW: the switching PSU is probably just regulating the 5V rail, and the 12V (and the other unneeded) do just follow. This is sufficient for most applications, since 12V is just for Motors and/or feeding "internal" buck regulators.

[@mdzacharias]

Probably not enough noise in this case to cause a measurement error, but I have seen where noise can make the measured DC "appear" low using a multimeter. For example oscillation at the output of a 7812 voltage regulator, which is fairly common.

[@JanBeta]

+Mark Zacharias Good point. I'll try to keep that in mind for the next time. Probably really doesn't matter much in this case.

[@mdzacharias]

Happens all the time in the repair biz.

[@NozomuYume]

Your alignment check is kinda silly if you're testing on a disk that was formatted on the same drive. It doesn't show if your alignment is close to "standard" since the disk was formatted with the drive's alignment in the first place. That said, if it *did* show any errors that would indicate problems with the stepper motor having trouble keeping alignment even with itself.

[@JanBeta]

Yes, I realized that after uploading the video... A bit of a noob mistake. I usually use a factory produced original disk for checking alignment now. Preferably the test/demo disk which should have been recorded on a properly adjusted drive.

[@mikegravgaard340]

Really silly observation but when you took the PSU apart, couldn't you just have rewired it from inside instead rather than cutting the lead? You could have used your desoldering station to do a nice professional job. Not saying you've not done a professional job but I would'nt want the 33 volt rail present in the lead as a safety precaution or the second 5 volt rail for that matter.

[@JanBeta]

Of course you're right. I should have replaced the cable or at least cut the unused connections on the other side, too. I think I'll actually do that as a quick hack to make it safer. Thanks for the tip!

[@dominikschutz6300]

Cool, i'll try to put that circuitry into an old 1541 as they have The same voltages internally (as i remember...) :)

[@JanBeta]

Yes, it also uses 12V and 5V. I'm not quite sure about the current though so be careful about that. I would guess it draws a bit more than the newer model.

[@cbmeeks]

Love your videos!

[@JanBeta]

Thanks! :)

[@Justin-py9je]

Bought the same Power supply for my 1541 recently :)

[@JanBeta]

Seems to work so far even if it's not perfect (as shown). Did you experience any issues with it?

[@naytch2003]

Another top notch video Jan;)

[@JanBeta]

Thanks. Glad you enjoyed it. :)

[@alynicholls3230]

a good source of cheap and plentiful power supplies is the xbox 360, al the correct voltages we need are there, and the output is "clean", if your plug and lead is ok but your c64 psu is dead the cleanest way, would be to solder the c64 lead onto the 360 supply, if you dont want to open the 360 supply, you can buy an adaptor that is meant to connect an xbox 360 fat to a slim console, and wire the c64 power plug onto the adaptor, pinouts for both machines are all over the net and easy to find.

[@damienretro4416]

You could power a 1541-II with a 360 PSU but it's a very poor choice. It puts out 15-20 AMPS on the 12V line (depending on model) but only 1 AMP on the 5V line. A C64 requires 1.7AMPS on the 5V line and also 1 AMP 9VAC.

[@Markchill2]

Have you ever been making video and it goes bang

[@JanBeta]

Yep, repaired a TV a while back and the transformer produced some nice lightning and thunder. The video is Patreon only though.

[@Markchill2]

@@JanBeta omg and that is why i not toch crt displays

[@GadgetUK164]

Great video =D You could probably clean up that PSU output with some caps btw!

[@JanBeta]

Thanks! Maybe I'll do some experimentation with smoothing this (also to learn how to do it, of course). In the end, I want to build something better to power the drive I think. I like Greisi's approach (I linked to his vid in the description) using a classic transformer and modern DC to DC converters.

[@GadgetUK164]

It's not as hard as you think - you could more or less stick at 1000uF / 2200uF on there to start with and that will likely smooth the ripple a fair bit, and if you can work out what frequency the switching is at you could calculate and add a smaller cap there to help filter any noise from the switching. The reality is the 1541 II probably has capacitors in the drive itself where the power feed goes. The real test is to meaure the ripple on the DC line inside the drive. I cannot remember now, did you do it that way or just measure the output of the PSU whilst it was loaded with your dummy load?

[@JanBeta]

+GadgetUK164 I just measured the output with a dummy load so the smoothing caps inside the drive are a good point. I'll try to see how much ripple and noise actually goes into the circuitry.

[@JanBeta]

+Mr T. Guru Probably a good idea anyway. :)

[@silkwesir1444]

I have got a question, as I am looking for replacement PSUs both for my C64 as well as the 1541-II drive. I want to go for the "lazy" (and honestly, safer) route and buy complete PSUs, 12V DC, 5V DC and 9V AC, with enough Amperage to run those devices. Now the question is with the 1541, when I put on the connector, there is the ground pin (GND). However, as I have two PSUs (12V and 5V), I have two negative wires... do I solder on both to that one pin? I guess so, otherwise it wouldn't work, or would it? but wouldn't that mean that 12 Volts could be fed back into the 5V PSU, damaging it? The alternative would of course be to just use one DC (12V) PSU, and then use a voltage regulator to also make 5V from that... however that then has to be quite strong in order to run both the computer and drive, probably 3.5A or more would be a good idea speaking of that, what are your thoughts about running both the C64 as well as the floppy from one PSU like that (of course plus another one with 9V AC for the C64), is that a good idea or could it cause problems? (like, maybe i imagine when the floppy drive uses a lot of power there could be fluctuations affecting the C64, or vice versa?)

[@silkwesir1444]

I see there are PSUs for hard drives etc. which already deliver both 12V and 5V DC. However, usually those too have two GND pins, and they are not labelled in any way different from each other... in that case, would i have to solder on both, or just pick one of them and use that?

[@JanBeta]

I might be mistaken but I think the grounds are all connected to the common ground in the C64 anyway (the 1541 has a ground on the serial cable, too) so it's a good idea to connect them together. If you get power supplies with enough amperage it should be perfectly fine to run both from one psu.

[@DocMacLovin]

This guy is really good in repairing things - why is he still beta?

[@TheDefpom]

If you only need 5v and 12v that should be easy to build something, even just using 7812 and 7805.

[@JanBeta]

Of course these voltages are not too uncommon. I explicitly wanted to try this hilariously cheap PSU. Not at all ideal, as suspected. ;)

[@GadgetUK164]

The problem with regulators is getting sufficient current (and they will generate a lot of heat). Your typical 7805 is 1 amp or 1.5 amp for an S revision. It might work with a 7805S (or the LM version I forget the PN) and 7812.

[@JanBeta]

I haven't looked inside the original PSU but would imagine that that's exactly how it is constructed (at least the C64 PSU is made that way). Transformer, rectifier, smoothing caps, voltage regulators. Might be interesting to check how much current the drive really needs. I would guess that it won't need quite as much as the psu rating.

[@danielmantione]

I hate these parts a lot and would wish they get forbidden like incandescent lightbulbs. You will get really inefficient power supplies with those parts, which is nowadays considered not done, but C64 hardware shows very well what the consequences of those parts are: They generate lots of heat. The C64 PSU is the best example, which fails due to being overheated for extended periods of time and then also kills the C64. But the most likely cause why my original 1541-II power supply did fail was that the LM78xx regulators did probably kill the capacitor in the PSU. Use switching regulators whenever possible. LM2576 and XL4015 are good regulators to fix old power supplies, or build new ones. For the 1541-II, a switching power supply is in principle just fine, you just need to find one that is good enough.

[@TheDefpom]

Daniël Mantione - the efficiency depends upon how you use them, if you run a 7805 from a 30v supply then yes, very bad, but if you can run them from a voltage which is only about 3v more than their rating then they are much better.

[@SuperRetroid]

Interesting video! Sadly, seems that getting that PSU delivered to the UK adds another €10 to the price! D:

[@JanBeta]

+Paul Wood Thanks! That makes it less of a bargain, of course. Still not too bad value for money I'd say... but you'll probably also find something better for that price. There's quite a few "harddisk power supplies" with the same voltages.

[@SuperRetroid]

Aye! At least I now know they're more common than I thought - which could come in handy, I'm sure!

[@knight79rider]

@@JanBeta hi there,was looking for some of those HDD power supplies..they look fine but no mention of overvoltage protection of any sort ..would they be safe to use with 1541ii?

[@TheHighlander71]

Well the conclusion is clear enough. It's a piece of crap but at that price it's alright for a temporary piece of crap. I wonder though what the noise on the 5V and 12V rails were on the new PSU. What is causing the pattern?

[@JanBeta]

I think the noise/ripple is mostly because of the way these switching power supplies work. They are switching on and off at certain changing frequencies and have a PWM output. The output is than smoothed by various capacitors and such which in the end generates peculiar repeating patterns like the ones visible on the scope here. Some of the noise, especially some of the high frequency stuff is probably interference from my LED lamp, PC, monitor etc. For proper measurements you would have to use better probes, cancel out external noise and so on. What I did is VERY basic. ;)

[@TheHighlander71]

Well it's good enough for the purpose of course. I'll watch Dave's video as well when I'm up to it.