Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: yorgle on October 25, 2013, 06:18:30 AM
-
Hey all. My Amiga 1000's power supply (dated October 1985) has just failed. It was working last night, but this evening, I went to turn it on to listen to some mods, and i was left sad instead.
It's a late 85/early 86 NTSC Amiga 1000
Here are the symptoms:
- Power cord is functional
- Dust blown out/off the board (fairly clean for a 28 year old machine)
- switch on the side only turns on the fan (110v fan)
- Continuity test on the fuse shows that the fuse is fine
- Voltage tests on the pinouts of the output cable show 0v
- No obvious burnout marks on the board, resistors, capacitors
I'd like to repair it if I can, rather than just replacing it with a switcher.
are there common fail points in these?
is there anything I should look for?
Is there a repair manual anywhere on line (with test point voltages listed, etc)
Any help at all would be awesome
-
I was hoping this would include PSU schematics but sadly it just says "if volrages are incorrect, replace PSU" - but worth checking all the same:-
http://www.devili.iki.fi/mirrors/4x4.hopto.org/A1000_ALR.zip
Several sites suggest A1000 PSU schematics have been available but all downloads fail.
-
This site offers a reconditioned A1000 PSU for $29: http://www.oldsoftware.com/Amiga.html
However, if it's an old fashioned linear PSU it should be quite easy to diagnose and fix. The form factor makes me think it's a SMPS though?
Switched PSUs typically suffer from a few common modes of failure. Without trying to teach you to suck eggs, I've found this guide to be very useful in the past:-
http://www.repairfaq.org/sam/smpsfaq.htm
I have to admit though, the last SMPS I tried to diagnose I ended up simply replacing.
If you do make any progress I'm sure your experiences (positive or negative) would be useful to others.
-
Make sure you don't have a shorted drive or motherboard loading down the supply and putting it into shutdown.
-
JimS: I've tried it in and out of the chassis. There's nothing coming out of any of the voltage outputs. I'll be doing more extensive testing and repair work soon.
TCMSLP: Thanks for the suggestions. The $29 original supply looks like a good solution if I can't repair this one. I feel like it's a linear supply, but yeah, the parts in it look more like a switcher. I can't find any details about it anywhere.
If i completely fail, and decide to go modern with it, I may put into the case, something like the PicoPSU (if it supplies enough power, otherwise, i'll make an A500-like box housing an ATX supply) along with the "Big Box ATX Adapter to provide the other voltages and tick signal...
http://www.ianstedman.co.uk/Amiga/designs/Amiga_ATX_Adaptor/amiga_atx_adaptor.html
-
Several sites suggest A1000 PSU schematics have been available but all downloads fail.
Oops, that would probably be my fault. Helps when the HTML link matches the file it's linking to. The A1000 PSU schematic download will work now:
http://amiga.serveftp.net/schematics.html
But that's for the 220V model, the American/110V one might be totally different, not sure.
Yes, of course it's a switch mode power supply, and a fairly old and simple one at that. I've written a few notes on repairing A3000 power supply common faults here (http://amiga.serveftp.net/A3000_HardwareGuide/booting-problems.html) which might help also.
Power transistor or bridge rectifier failures are common in these supplies. Though when they fail, it's usually in a short-circuit condition which blows the input fuse, but it sounds if this is OK. Would be worth checking them anyway, easy enough to do with a multimeter on diode test range. Also check low value series resistors (under about 47 Ohms) for being high resistance or open circuit. During power on, the high inrush current through these resistors can damage them. As stupid as it may sound, I've seen this many times in practice.
-
Excellent! Thanks for the pointers, Castellen.
So, I just cracked it all back open, powered it on, and probed at it with my multimeter. The bridge rectifier seemed to behave as I'd expect four diodes to behave. Everything worked in the right directions. I powered it on, checked voltages on it, and everything looked reasonable. I've never worked on such a thing, so the +80 and -80 volts coming out of it (with 110v AC going in to it) seemed somewhat reasonable. I had expected it to be +-60 volts, but whatever.
There's a line printed on the board by the smallish transformer there, so I assumed that on one side it's AC, and the other side is DC... I checked voltage from ground to the heat sinks on the two transistors on the AC side, and was getting 80v, which was interesting...
I did notice one of the ICs on the DC side had a bit of dust on it, that I neglected to remove last night, so I did that. powered it back on, and continued probing. I had the black lead of my multimeter jammed into the black (ground) pin on the motherboard connector (not connected to the motherboard), and I started probing around the DC side. I was getting 0v on most of the chips, then i tried one, and got 5v. then i checked the cable, and I was getting all expected voltages again. +5, +12, -12, ~3v (the TICK pin). I plugged it into the motherboard, turned it on, and it powered up.
I reassembled the whole thing, turned on the switch and.... nothing. No red LED anymore, nothing. I was getting frustrated but then I just decided (for whatever reason) to leave it on for a few minutes, to see what happened. After a minute or so, it came to life. That was about an hour ago, and it's still on without rebooting or powering off. I recopied the Kickstart floppy that failed a moment earlier (I think I put my magnetic screwdriver on it. oops). As an A1000 owner, you can never have too many backups of Kickstart floppies!
I'll check those resistors in the future (I meant to do it, but then it came to life and I forgot) for the next time this happens. I'm getting really quick with opening this thing.
So... do I trust it? Probably not. I'm still looking into my options, but being that I didn't really have the cash to do anything about it, it's nice that I got a bit more (admittedly, borrowed) time on it.
-
So... do I trust it? Probably not. I'm still looking into my options, but being that I didn't really have the cash to do anything about it, it's nice that I got a bit more (admittedly, borrowed) time on it.
So it's obviously an intermittent problem then. Agreed, it may stop working at any time.
Check all solder joints, particularly on larger components and components that dissipate significant heat. Resolder anything that looks cracked or suspicious.
These older SMPS designs usually work around a self-oscillating principle as opposed to newer designs that have a dedicated controller IC. There are always capacitors involved in the oscillation network and often these are aluminium electrolytic types. As these components age, the internal liquid chemical electrolyte gradually dries out and the capacitence decreases as a result. So what you may be seeing is that certain capacitors have degraded to the point where the oscillation is not self-starting reliably. The solution is of course to replace all of the aging electrolytic capacitors. Not expensive to do, but somewhat time consuming in ordering and replacing them. But in doing so, the power supply is likely to be good for another 25+ years.
-
I've never worked on such a thing, so the +80 and -80 volts coming out of it (with 110v AC going in to it) seemed somewhat reasonable. I had expected it to be +-60 volts, but whatever.
I forgot to mention....
The incoming AC mains is full wave rectified to DC by the bridge rectifier, and on the DC output of the bridge rectifier there will be one large electrolytic capacitor, or two in series.
The voltage across this reservoir capacitor will be the mains AC waveform peak voltage, less the forward voltage drop of two silicon diodes in the bridge rectifier.
i.e.
Vcapacitor = Vpeak - bridge rectifier fwd drop
Vcapacitor = Vrms / SIN(45) - (2 x 0.7V)
= 110V / SIN(45) - (2 x 0.7V)
= 155.6V - (2 x 0.7)
= 154V approx. across the reservior capacitor
If there are two reservior capacitors in series, there will be about 77V across each one. Hope that makes sense.
-
Look closely for cold cracked solder joints, replace all eletrolytic caps (normally only good 15-20 years). Bad caps will cause other parts like bridge recifiers diodes and power transistors to burn also, so check those for shorts before power up. Some you may have to remove from the board to check accurately. You might get lucky with just some cracked solder, but when I go in I get all the stuff that might go later. Good hunting.
Plaz
(Reanimator of thousands of power supplies over the years. :) )
-
You know.. I should have made the connection that it would be a capacitor thing. I've been doing cap-kits on monitors for video games, I should have made the connection. Feh to me.
Okay. I guess it's time to do an inventory of everything in there, and put together a mouser/digikey order!
-
as a quick side note; I had to desolder one of the 12(?) electrolytic caps get its value. Out of curiosity, I snipped it open, and sure enough, the paper in it was damp only on about 25% of it. :)
-
If you have a list of the cap values in the PSU, could you post it here? I have been thinking of recapping mine some day.
-
I have to get some time to triple-check this, but I believe this is correct for all of the electrolytic capacitors on a 110v NTSC Amiga 1000 Power Supply:
C11 2200uF 16V
C12 2200uF 16V
C13 2200uF 16V
C15 100uF 25v
C17 100uF 25v
C19 100uF 25v
C50 470uF 220V
C09 4.7uF 50v
C10 1uF 50v
C16 470uF 35v
C18 330uF 35v
C49 10uF 85v
C50 runs around $6.50, while the others go for $0.50-$1 US. Priced out at Mouser, it runs about $13.50. Here's the Mouser project for the above parts: http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=69b972847a
Remember to preserve polarity while installing them!
-
Related: The list of electrolytic capacitors needed for an NTSC Amiga 1000 (I don't have the board revision)
All are rated at 16v
C75 100uF (hidden near the side D9 ports)
C78 100uF
C79 100uF
C80 100uF
C49 22uF
C50 22uF
C63 22uF
C64 22uF
C65 22uF
C66 22uF
C92 47uF
C93 220uF
C95 470uF
Daughtercard:
100uF (two of them, no noticeable part number indications on the board)
-
C49 10uF 85v
Once I got this part out, I realized it was covered in dust and it really is 10uF, 35V. (It's okay to put in an 85V part in there, but it's not the value on the original.)
-
For what it's worth, I spent about a half hour last night, replacing all of the caps, and the power supply works 100% now. Installed it into my 1000, and I was playing .mods deep into the night. Huzzah.
My only note to people who decide to do this;
The one large capacitor, the 200v one, be sure to try to get one that's wide-and-short, rather than a thin-and-tall, because there is limited clearance inside the power supply case. I got a tall one, and had to have it mounted almost horizontal to allow it to fit inside there. oops!
Next up is replacing the caps on the mainboard and daughterboard... but there's no need to post updates about that. ;)
-
and the power supply works 100% now. Installed it into my 1000, and I was playing .mods deep into the night. Huzzah.
Contratulations, ready to go for another 25 years.
Plaz
-
Contratulations, ready to go for another 25 years.
Plaz
Hi, resurrecting this from the dead to see if you can help me out.
I have a similar issue with the big difference that I haveno clue what I am doing or have any experience at all reparing these...
I recently got another A1000 in better cosmetic condition than the one I already had, so I wanted to sell the first one of and took it out for testing.
I was using it yesterday all day, and today as well up to now, booting up kickstart, workbench, some games, testing the mouse, joysticks, etc.
All was running great! Inhad also opened it up because the floppy drive's button had broken, and I needed to fix it, etc.
So all was working great and I was ready to put it away again. But then I decided to test a 2MB side expansion I have that came with the second A1000.
I don't want to say/assure that this was the culprit, but I turned the computer on, all good, but thought weird free mem showed as if nothing was there. Sysinfo comfirmed this.
I thought maybe it had not been plugged correctly, and turned it off and tried again, and nothing same issue (although ofdly jow a disk I was sure workied started giving errors).
I disconnected it, and turned the Amiga back on, and now it did not come on. :(
Symptoms now are:
Turn it on, fan starts, no power led, and after some 5-15 seconds (varies) thefoppy drive led comes on (but not the drive itself seeking or anything).
Not really sure what to do. I opened up the PSU lid and don't see any "bulging" capacitors, but that's as far as my knowledge goes.
I have a multimeter if that helps.
Any help is much appreciated.
Thanks!
-
Its honesty just easier to rebuild it by replacing the inside with a pico ATX or a micro that will fit in it. Still looks original and you get a much better and efficient PSU.
-
It appears that the memory expansion was the culprit. Could you try the other A1000 power supply to rule out the PSU?
-
For what it's worth, I spent about a half hour last night, replacing all of the caps, and the power supply works 100% now. Installed it into my 1000, and I was playing .mods deep into the night. Huzzah.
My only note to people who decide to do this;
The one large capacitor, the 200v one, be sure to try to get one that's wide-and-short, rather than a thin-and-tall, because there is limited clearance inside the power supply case. I got a tall one, and had to have it mounted almost horizontal to allow it to fit inside there. oops!
Next up is replacing the caps on the mainboard and daughterboard... but there's no need to post updates about that. ;)
thanks for the info! I've replaced caps in my A4000 and A1200 but my A1000 has been running steady since 1986 but alas as you also found out even the mighty Amiga 1000 will eventually need all the caps replaced so thanks for the reminder:hammer:;)
@ Acill
unless i can find someone in Canada, I'll be sending my A1k boards and PSU to you to get capped
-
thanks for the info! I've replaced caps in my A4000 and A1200 but my A1000 has been running steady since 1986 but alas as you also found out even the mighty Amiga 1000 will eventually need all the caps replaced so thanks for the reminder:hammer:;)
@ Acill
unless i can find someone in Canada, I'll be sending my A1k boards and PSU to you to get capped
Thats fine, but unless you are okay with it taking some time to finish you may want to only send one at a time. I am in Japan the rest of this month and only home from the 4th to the 10th next month and back here again for the rest of it.
-
I have to get some time to triple-check this, but I believe this is correct for all of the electrolytic capacitors on a 110v NTSC Amiga 1000 Power Supply:
C11 2200uF 16V
C12 2200uF 16V
C13 2200uF 16V
C15 100uF 25v
C17 100uF 25v
C19 100uF 25v
C50 470uF 220V
C09 4.7uF 50v
C10 1uF 50v
C16 470uF 35v
C18 330uF 35v
C49 10uF 85v
I finally got around re-capping my A1000. Here is the caps list for the 220v PAL Amiga 1000 power supply:
3x 2200uF 10v
470uF 35v
330uF 25v
3x 100uF 25v
10uF 35v
4,7uF 50v
1uF 50v
I did not replace the big mains cap C50 as I couldn't find close enough replacement. The original in my PSU was 180uF 400v or there abouts. I believe that cap is not in a stressed position and as the smaller caps measured very good after removal, I was happy to leave the big one there for now.