Hey Jan, thanks for the video! Liked it! You should have a look at the connectors of the PSUs to the Amiga. Maybe the pins are not touching the electrodes of the port correctly. 😮

@IanSlothieRolfe

Жыл бұрын

I'd vote for this.. The LED wouldn''t care if the voltage was low or ripply. You didn't test for voltages on the Rev 3 which would be my suggestion. Maybe the pins on the later power supply connectors are slightly longer and touch parts of the power connector in the Rev 3 that the "proper" one has worn away with use. Its pretty much the only component in common.

@jwvdvuurst

Жыл бұрын

Exactly what I thought.

@RetroKrazy

Жыл бұрын

Ditto. If you have even one pin in the socket slightly recessed, then there'd be no contact. If the other PSUs have slightly longer pins, they'd make contact like you see. Also, try checking the ripple current on the PSU that is powering it on as think that is a red herring.

@amigacoverdisk

Жыл бұрын

Yep, I too was thinking this, some unlucky combination of slightly out of tolerance/worn plug with a slightly out of tolerance/worn socket.

@grantfryer1

Жыл бұрын

I was thinking the same.

I'd measure the voltage with the rev3 amiga on the power supply side, if that dips under the 4.8V then there's your problem, you could probably change the 180 ohm resistors to increase the voltage a bit and see if it powers on. Also check exactly what the current draw is on the 5V line if you have a clamp amp meter. If it's way off something in the Amiga might draw a lot of current, maybe a chip that's marginal. An infra red camera should show you which chips are getting excessively hot.

@WizardNumberNext

Жыл бұрын

IR camera is expensive, while isopropyl alcohol is very cheap and would show that as well

@imqqmi

Жыл бұрын

@@WizardNumberNext yeah or a temperature probe, point is you could check what gets hot. A cheap android ir camera can be had for under 200 euros with 120x90 resolution. Well worth it if you do repairs regularly.

@WizardNumberNext

Жыл бұрын

@@imqqmi check out Louis Rossmann He definitely can afford such camera, but prefers isopropyl instead You just pour enough to make a layer. One place would get dry first Anyway I suspect Amiga itself same as you, but more mondaying, as it works on different power supply. It is probably capacitors Probably adding ceramic capacitor close to CPU would solve it.

@sjarken3979

Жыл бұрын

I suspect the same, something is drawing too much power, which is why the new amiga power supply works in rev.3 amiga and the old power supply works on the newer amiga.

I believe Rev 6 has only 4 CMOS DRAM chips while the Rev 3 has 16 NMOS ones. The difference in static power consumption between NMOS and CMOS can be very significant. Also applies to the 512k expansion. Some come with 16 NMOS, some with 4 CMOS. You can literally feel the difference in heat. I would have monitored ripple and power on behavior with the Rev 3 Amiga. You can also try to first power on the PSU without the Amiga attached and then plug it in. I bet it works that way. No pain, no gain!

Jan, you need to try putting a amp meter in line with the 5 volts and see what a working amiga pulls then try it with the non working amiga and see if it pulls more current than normal.

Hi Jan, What you call a rectifier is actually a thyristor. This is part of a protection circuit (crawbar). If the output voltage gets too high, this thyristor will conduct and short out the 5V. In the best-case scenario, the fuse will blow, but the hardware will be protected from the surge (but has nothing to do with your problem).

@borayurt66

Жыл бұрын

Yeah, came in to write the same. There are no 3 pin rectifier packets, also the schematics clearly shows what it is.

@SeanBZA

Жыл бұрын

Will not blow fuses unless the transistor Q1 is shorted, but will pull down the 5V rail and turn on the current limiting in the SG3524 to change the PWM to the minimum the chip can go to, limiting current to a few hundred milliamps, depending on the actual chip as to exact value. D7 and D8 should be replaced though, preferably to a Shottky rectifier, 5A 30V will do here, not a slow as molasses 1N5401, though you can use UF 5401 diodes there, as they are fast switching types. Incidentally D10 is typically a 6V2 or 6V8 400mW zener diode, typical for 5V crowbars.

Problem should be the power connector on the Rev3 board itself (use some contact cleaner or change connector) and/or the power plug pins from your power supply (clean with fine sandpaper). Because of bad connection ripple might be higher and causing problems maybe.

@smunaut

Жыл бұрын

Yeah, I would measure the voltage on the rev3 amiga rails directly and also look at the power up behavior. Could be a slew issue getting it into a bad state.

I'm rebuilding an Amiga 4000 power supply and came across a blog exclaiming if new capacitors with a low ESR are used, then the power supply may fail. Not sure of this comment is of value, but I'm throwing it in here anyway. Viel Glück und danke fürs Video.

Congratulations for 50k. Well deserved, and more to come in the future! Can't help here, so I'm looking forward to the next video. Hallo! M

I had the same Problem with my Rev3 Board. And the Problem was the 1 Ohm Resistor EMI406 on the A500 Rev3 Board. Maybe it's the same Problem?

Congrats on 50k, Jan! Keep up the great work!

Congrats with 50k subscribers! You deserve it, fun channel with interesting content. Hope you get the power supply working with the rev. 3 A500.

It would be interesting to measure the rise time of the +5 V from off to on, then compare it to the ‘good’ psu. Also look for any bounce or ringing during power up

Continuity in the output cable? Apologies if you checked this and i missed it or it's been suggested already. Great content, always fun to watch.

cograts on 50K - like your vids, enjoyable content in a relaxed natural manner....

Check the main filter capacitor on the input side to the buck converter. It may cause supply droop under higher load and you could be seeing ripple from it.

Check the voltages on the actual Rev3 board rather than using the broken board. Also try swapping out the DC power supply cable to eliminate that from the equation.

Love your Amiga videos!!!

I think you have made a logical step by step approach, and have made good effort. I have made my own Amiga supply using a pc power supply, which works very well.

If you have the oscilloscope connected to the 5V rail when you power on the Amiga, you could se if it drops below 5V on initial startup. It could be dropping low just for a short while during power-on. Maybe single-shot it with the scope to se how it behaves then. Another alternative is to load the 5V with some electronic load (or a resistor) to see if it can deliver the rated 2.5A of current.

That D5 is an SCR/Thyristor that may be used in some rectification circuit, HOWEVER in this circuit it is used as a so called crowbar for over-voltage protection (essentially short the +5V to GND if the +5V is too damn high) Also don't downplay the role of the choke/inductor, as that is half of the frequency-tuned circuit that stores and supply the current during the "off" phase and smoother during "on" phase.

Enjoyed this vid as much as the others, some you win and some you lose. Took me much trial and error to fix an Amiga 1000 power supply so I feel your pain

I just got an Amiga 500, built in West Germany. Going to be going through your A500 videos before I try and power it on.

Hi Jan, since the +12V seem to be a bit on the low side and the +5V are derived from the same base 25V transformer output, I suspect the transformer to be the culprit. You should measure the input to the switching part of the power supply under load. Maybe it drops too low and there isn't enough energy left for the switching circuit to feed into the 5V? Checking the control pins of the IC with the scope might give you a clue. If I'm not mistaken, the power LED on the A500 is not directly driven by the supply voltage as it may be turned on and off by some aspects of the audio system. Marcus

I noticed you could have used the second set of holes for the 2W resistors.

Yes, a big congrats on reaching 50k. Thought you had a higher subscriber base. 😁😁👍👍

Incredibile,spero che il problema si risolva

My approach would have been to measure the rails when trying to power the rev3 to determine which rail is actually failing, and then gone backwards from there, I have seen plenty of the 78xx/79xx series regulators fail, I could have been one of the 12V rails, you seemed too focused on the 5V rail, even though it was testing as being OK.

@jaycee1980

Жыл бұрын

Missing +12V would NOT stop an Amiga from booting - you would simply have no sound and possibly no disk drive. The +5V line is the important one.

I'd try bodging a 4.7pf capacitor across the offending rail. While larger capacitors will take care of lower frequencies you'll find that smaller values of ceramic disk caps will eliminate the higher frequencies quite nicely. I'd bodge it in directly across the output of the 5v rail possibly on the underside of the PCB. If that doesn't work, I'd suggest having a look around on the board that doesn't work with it. This could be an issue specific to that particular revision.

well done thanks!

I would suggest changing the inductor, it might have a crack in the core. Will cause the marginal operation you have. Incidentally the schematic you refer to has an error, the junction between C13 and D8 should also go to ground. Also check the voltages with no load, to check if the power supply is turning off with overvoltage. Output capacitor needs to be low ESR, 1R7 is way too high, and will result in marginal operation of the overvoltage trip.

Welcome to another episode of Mr. Puzzle! Oh, wait, wrong German.

Great video as always! 🙂 I'm not the first to suggest this but I agree with the folks here saying to check the power at the connector on the rev 3 board. Unless you already checked off camera, we don't actually know what voltage the chips on that board are seeing - my main thought is whether the connector is a bit marginal on both the board and the power supply, in a way that together doesn't contact properly but which when either the board of the supply are changed, it is still good enough. In which case the power supply could be fine electrically, which would agree with your measurements so far I think unless it really can't cope with the power draw of the rev 3 board! Either way it would make sense to directly measure the voltages on that board.

Have you been able to successfully power-up the targeted Amiga (Rev. 3) with any other known-good power supply? If so, perhaps a side-by-side comparison of the two power supplies. If not the ripple voltage, perhaps something is getted 'loaded' down.

Try intercepting the current from the power supply to the amiga and then compare it with a different Amiga. They maybe something on the old amiga drawing roo much current for the old supply. If you have a adjustable current limit bench supply you could then find the current level which works.

Try a larger filter cap also measure with rev3.

Check for tarnished pins on the PSU and Amiga power connectors, which could limit the current. Or check resistances between the ground and 5V traces on the PSU board and the Rev 3 board.

Dirty power connector when connected to the rev 3? Beyond that I am thinking maybe the 3300uF caps on the rev 3? I cannot remember if those were swapped?

When I bought my Amiga 500 back in the day I struggled the first year with a power supply problem that made it unusable most of the time. The power supply failed, I took it to repair and it would break again in less than a week, repeating the cycle over and over until I got a replacement one (not Commodore but very similar in appearance) and that finally fixed it. From the looks of it I think mine was 100% SMPS but I can’t be sure after so many years since I bought the replacement supply in 1990 OMG!

(not related to this video) but it is worth mentioning that the rev 3 boards take a while to charge up capacitors when you first turn them on. my Amiga didnt power on at first boot after being off for very long period of time. when i switched on (first boot) it seemed dead, after a few mins tried again and it powered on immediately. I have experienced this on two different amigas *both* rev 3.

Hi Jan, another great video! If I understand correctly, you measured across the power supply connector on the board. That basically gives you an idea what the PSU delivers to the board under load. But, as I found in C64 boards, the voltage can drop significantly along the power rails if you have more ICs or if they draw more current than expected. In one case, an input of 5V meant an IC supply of about 4.7V. In that case ICs may fail to work or get damaged. In other words, a spot-on 5V supply may not be sufficient for that board revision. It might need a bit more. Maybe you could try to measure the Vcc at several ICs on the board, opposite to the connector. It may be the case that you have a voltage drop that undercuts the permissible input voltage on some of the critical ICs. Measuring the supply current would also give more insights. Of course, it's also possible that the voltage regulators are marginal and working with a later revision, lower power board, but cut out with an elevated load. 🤔 And transient effects can also play a role, as others have pointed out in the comments.

How much current is the Rev. 3 board drawing versus the other boards? Is it possible there's a low resistance path on the Rev. 3 board and it's drawing too much power for that particular PSU?

When soldering in through hole resistors in "high" power circuits, I bend the legs to form a 270 degree loop at the resistor ends. The extra lead length improves heat dissipation and reduces the mechanical stress under thermal expansion. You also get test points for probe clips. 😊 I'm surprised that you don't have a test load box for the power supplies... With loads on each rail switchable to rated load. Just a hunch... Put a choke on the 5V output of the PSU. There might be some strange coupling on the earlier Amiga boards that was eliminated later.

If the resistors are in parallel, you could replace with a single higher wattage resistor of half value if space is too tight. In this case, you could use a single 90 ohm, especially since they are 2 watts which is up from 1/2 watt combined.

Hello! Thanks for a great video! Someone might have pointed this out already and it is of very little importance =) Gold band on the resistor in this case is 5% I believe, not 1% Correct me if I'm wrong =) BR, Nin

We want to see voltages inside the working Amiga and in the rev.3 one! Is there any additional smoothing on the Amiga board for the incoming 5V line and if so, is this in good working order on the rev.3?

Just load the PSU directly with electronic load and see how many amps it could supply on the 5V rail. Then measure how much amps the amiga needs. If all matches - inspect connectors for resistance/voltage drops.

have you tried wrapping a large ferrite core around the power supply cable to reduce high freq ripple?

Have seen transistors and diodes test OK on a simple multimeter/component tester but fail under load, I'd be checking the 1N5401 main rectifier diodes under load or replacing them.

What about the inrush current of the rev3? Maybe it can’t cope with that and the reset circuit doesn’t work properly as a result? You should be able to measure this with the scope and a small series resistor. And at the reset pin. Also compare with the rev5 current draw.

Is the plug of the power supply slightly misaligned? Could you measure the voltages inside the rev 3?

You have to check resistance on both cables of psu, they are old. Also check voltages on rev3 power conector, I think think the problem is not with psu, rev3 mbo has a problem with current draw.

@kreskogym

Жыл бұрын

As I sayed in FB comments; why bother with old psu, try power it up with new modern psu. Yes, ripple is slightly higher on meanwell's but that will not be a problem.

Checked the physical connectors vs known good? Slightly undersize pin + stretched/loose pin socket?

I've had the same problem with a C64 psu I built. The 5V rail was giving out exactly 5V's. when I connected I cartridge the voltage went down to 4.5 which was not enough to power it up. So I modified the 5V smps changing the ref voltage going to TL431 which determines the regulation voltage by changing the pull up and pulldown resistors which the TL431 is fed. I configured the ref voltage to be 5.2V then everything was ok. So, in your PSU the ref voltage is determined by two resistors, (in this case the are R5 and R1 or R4 and R2. I'm not sure) you should change them to give 5.2V or you can just add a 10k potentiometer to that circuit . I mean you should modify 5vline to supply 5.2v. that's all. I hope it works for you.

I didn't bother to try repair my 2.5A PSU since it's a bit weak if you want to add some extra hardware. I just replaced it with a Meanwell instead.

what are the replacement Resistors the big chunky ones, I'm not very familiar with that size I'm looking a one at the moment, and some what stomped what it is guessing it just big sized resistor, but there must be a reason for the size? the one I got is potted in rubber resin, so it going be a bit of a acarology dig to get at it., and it may just crumble getting it off the PCB board? what is real name? as what to call it to by one? to three etc? and are colour banding system the same the regular sized ones like ones being replaced on the board? multi metered iit not showing any resistance at all, (so guessing just dead), and it look slightly discoloured, not burned, but it may cracked or broken? and needs replacing?

Just a guess but see if there is a difference between rev3 and the later revisions on the 12v rail. they might have changed something on later revisions that does not pull as much on the 12v side. the 12+ was low if there is a difference that could be the cause of the not powering problems with that supply.

Dont plug a monitor cable in, you can check the voltages on the monitor port. No need to take that lovely machine apart again. I’ve got a bad psu too. Had some cap leakage, cleaned up, changed them all, but has some video interference. Better than before repair, but not good enough on a pvm. Luckily i have another psu thats perfect.

I don't have much idea, but for what I've seen in the past with high demanding (a lot of things connected) miggys, you could try with a much shorter cable from the PSU to the Amiga, just avoiding resistance. My guess.

Hello Jan, my guess will be that the start for psu is not good. Maybe connect it after a few seconds with the connector? Tchuss

Did you check both of the rectifier diodes from the transformer? Ripple (especially with a small load) and weak output could be a sign that one of the diodes has failed.

check the filtering input of the psu

Is there a slight variation in the power input cable to the early Amiga , have you measured what voltage is on the board in the Amiga ? Think the issue lies with the dc input plug of the old Amiga

What TTL ICs have the rev3 board? May they be mostly regular TTL, they draw more current than LS/ALS later ones. I had a spectrum clone with a lot of standard TTL chips that needed a 3A/5V while a later revision made with mostly LS-TTL ICs drawed only around 2A/5V. The PSU from the later revision was unable to power the first one. Also the memory chips were NMOS on the first version and CMOS on later ones. Theirs power consumtion are quite different. Try to power the boards with a computer ATX PSU and check how much current they draw.

I've seen too low of ESR on output caps cause supplies to not want to fire up when under a load, but I doubt that's your issue. More or less the inrush with charging cap(s) plus load ends up being too much. Again, probably not the case. I would one shot trigger capture the output on switch on and also probe the PWM chip.

Why didn't you use the holes next to the blue component to install the new resistors?

hi jan i may be wrong with my idea but cant you smooth out the ripple with a another capacitor that sits in the 5 volt output or isnt that possible?

Have you tried without the floppy connected? Isn't that Amiga the one with the converted PC floppy? Maybe that floppy draws a bit more current than s standard Amiga floppy?

Why are you not connecting it to the rev 3 and using the meter and scope to fault find. Is the IC switching what's it's supply look like on pin 15. I assume D10 is a zener if so D5 is an over voltage trip/crowbar. so the PSU would work without it, not recommended to connect to a board without it though use an incandescent car lamp as a load.

Maybe rev 3 is drawing more juice and some over-current protection kicks in and causes to PSU to shut down ?? Have You tried to measure voltages with rev3 connected ?

Have you tried a different plug? Maybe this PSU and the rev3 connectors have some slight misalignment? Long shot😬

Have you also checked the connector itself? Maybe that has a bit too much tolerance?

I guess inrush currents are limited in the later revs of the Amiga 500. Switch mode power supply often can’t handle large inrush currents (when the load is switched on) Loss of capacity in the output caps could also introduce such behaviour.

What about measuring the powersupply under load with the rev.3 Amiga.

It’s not often we see you fail. I would be interested to see you revisit this, using some of the suggestions in the comments. Definitely take apart the Rev 3 Amiga and check voltage at the board.

Why not measure the voltages on the rev3 board? And the current(s)? Or get a rev6 Amiga 500 and add some extension boards to increase the load? And see if that Amiga boots?

Nach Lage der Dinge würde ich jetzt den eckigen Powerconnector des rev3 verdächtigen, entweder das Kabel, das vom Netztteil kommt, ganz tauschen und dazu den connectorcauf dem Mainboard mit ordentlich kontaktreiniger bearbeiten...

What if the 5V on powerup rises slower than the power on reset circuit's timing? (sorry, now I should spend probably an hour on reading the schematics to see if the POR can even be the reason - most PORs just charge a capacitor off the 5V and that might let go off reset before the 5V really has stabilized whereas fancy PORs really check if the 5V has stabilized before letting go)

Try to supply the 5V rail from a solid benchpower supply.. just for testing. Or power the 5V rail from a modern DC-DC step down converter.

Try putting an inductor/capacitor low band filter on the five volt line to see if you can reduce the noise more if it works great if it doesn't that is not the path to follow

to be honest I am now also modding a Amiga power supply to house the MeanWell RT-50B , and to be honest I am not really planning to use any oldscool Amiga power supply "as is" ever again, and considering to convert them all, even if they still work. all my heavy and also the light weight bricks. unfortunately the narrow A1200 and A600 bricks does not fit the RT-50B. but in that situation i will still use the big brick converted ones.

I would start by checking that the Amiga Rev. 3 is actually getting power from this power supply at all, perhaps there is a problem with the connector.

@dick_tm

Жыл бұрын

I can only support this suggestion. If there is a pin a bit curved and the other side is also curved, but in the other way the board doesn't get any voltage. Other powersupply simply work because there is a connection.

Here's a stupid thought. Are the pins in power sockets same between new and old? Only then, if you're getting good continuity on all legs, split rails used between both supplies (shared ground), until you've identified failure.

I would check voltages on rev 3 on floppy power connector, both 12v and 5v, as it sound like to me there is no ground getting to the Amiga, if you get +12v on floppy connector and no 5v then ground is there and 12v too, but if you getting no voltage then there is a ground issue.

Given the power led is not on the power to the non working board must be virtually non existent. Is the power supply shutting down from excess current?

maybe add one more capacitor inside the powersupply to reduce ripple?

Run a load on the 5 volt rail and put a scope on it and or DMM on it.

Hi Jan check current draw of both PSUs

Check the connector😊

Please save our collective sanity and check the voltage levels on the actual rev 3 board! It could be something as simple as tolerance on the power connector not allowing the 5v pin to make a good contact. The PSU is evidently working. The Rev 3 board likewise and that pretty much limits the fault to where the two meet...

Check connector & cable. Also, make your voltage measurements inside the Amiga, not the psu. Good luck !

Jan, how many Amiga 500 do you own?

Of course, the easy solution is just to swap the power supplies and send back the one that works with the Rev 3 Amiga. For all we know, the other supply might not have been original, and almost nobody cares about the originality of a power supply, let alone if it's the exact same one that came in the box. I do understand you really want to understand the issue though!

50k 🎉

Caps on Amiga itself?

ground and sheild actually conncted? as it should but if not.. it might behave wrong

Hello. The scheme from the address indicated by you is incorrect, diode D8 does not exist, capacitor C13 is connected directly to ground. According to the catalog data, the integrated circuit has an internal output overcurrent protection circuit connected between pins 4 and 5; it is activated when a voltage of 200mV appears between these pins. In the schematic pin 4 is connected to +5V and pin 5 to resistor R9 in parallel with diode D9, the manufacturing tolerance of these components may be the cause of your problem. The circuit formed by: R10, Zener diode D10, R11 and D5 forms the overvoltage protection circuit for the output voltage of +5V. I hope this information helps you. Good luck in your work

@lincamarius7092

Жыл бұрын

I don't have access to such a source, but I found this web page where you can see the PCB. I noticed that pin 4 is connected to coil L1 and to a jumper wire, which is connected at the other end to capacitor C13. This is actually a resistor of approximately (0.2V / 2.5A = 0.08 Ohm) which determines the current consumed. So check the solder on this wire. With respect Marius.

Yep- you measured the REV5 under load, but you did not do same to REV3 unit. This would be first thing I would do before you replaced just about everything within external PS...

It certainly is quite the conundrum, works with other machines, but not that one, could be something with that particular Amiga's board that draws a higher current, which that PSU may not be able to supply for whatever reason, which makes me think that something has degraded in the supply if it is original to the computer, or it's the wrong one taken from a later and less power-hungry model... :\

My concern is that you are using a machine pin socket which only makes contact with the pins of the chip on the 4 corners. You really want to use a dual wipe socket if anything those make contact on both sides of the pins and have overall more surface area. I would only use sockets with logic chips and avoid using them on the main draw of the 5v power rail.

27:17 1.74 ohm is far too high for a 4700uF cap. It should read 0.01 ohm. Even good 470uF caps have lower ESR than that. Those capacitors are both bad.