Killer power supplies

[Quagliarulo]

Hello.
Did it ever happen to any of you that a Commodore power pack (those used with A500/A600/A1200) would kill your computer due to overvoltage?
It happened to me twice. The first time, I was testing a PSU I bought at a flea market with an A1200 and the computer died instantly. Yesterday, my heavy duty A500 PSU fried my Amiga 600, although I'm not 100% sure it was the PSU's fault, because I had used it with an Amiga 1200 a few weeks ago without problems. After the 600 died, I checked the +5V and +12V lines and found 13.3V where 12V was supposed to be. I also tried adding a little load (an old CD ROM drive) and I read voltages between 12.6 and 13. The 5V line read 4.88.

[pandy71]

If feedback use 5V as a reference then 12V can be over voltage - but... usual tolerance for supply is +-10% so 12 +10% = 13.2V - this should be no problem for any electronic equipment...

[source]

For frying the computer isn't the important one the 5 volt. The 12 volt just runs motors and fans?

[Hewitson]

Drives generally use 12v for their motors, it's possibly used for op amps in the audio section as well.

[Mr B]

I'm not clued in to the design of the Amiga PSU's but generally speaking, taking measurements unloaded isn't providing good data.
B!

[Hewitson]

Originally Posted by Mr B

I'm not clued in to the design of the Amiga PSU's but generally speaking, taking measurements unloaded isn't providing good data.
B!

Especially with switchers, these are generally damaged if run without a load for any more than a few seconds.

Back when I started with electronics I wasn't aware of this and I fired up a Compaq PSU to take some measurements. After a minute or so it was squealing like a pig and no longer outputted any voltage!

[Eamoe]

Incredible to find this thread now because a friend of mine found an A500 at a flea market last week and he came to me to test it. Before plugging his found PSU I proposed to check output voltages, and the + and - of 12V actually read 14V. (NOTE: th 5V line read something like 4.88V.)

His opening the PSU didn't reveal any visible damage to inner components.

I told him I would start a thread here and that it would be better not to fire his Miggy up with this PSU until we get a proper answer.

So, is it safe to consider that these readings are so because of the PSU not being loaded?

EDIT:

Originally Posted by Hewitson

Especially with switchers, these are generally damaged if run without a load for any more than a few seconds.

Really?! Cause I did run them unattached several times already...

[Mr B]

Hook your PSU to a dummy load if you want to measure voltages. Lightbulbs come to mind. Need to hook "all" positive voltages up.
I haven't heard of switched PSU's burning them self out from being unloaded, but it's a very common issue that they cant regulate voltages if you don't draw enough on the 5v line. People using PC PSU's (AT/ATX) for battery chargers (RC) regularly end up putting a dummy load, such as a fan, or a lightbulb on the 5v current. Some PSU's even need the 3.3v feed wired to something for the same reasons. Cheap PSU, 5v regulator circuits. Better stuff, more picky. Generally speaking that is.
As i said, no idea about the Amiga's PSU.
B!

[Loedown]

Originally Posted by Mr B

Hook your PSU to a dummy load if you want to measure voltages. Lightbulbs come to mind. Need to hook "all" positive voltages up.
I haven't heard of switched PSU's burning them self out from being unloaded, but it's a very common issue that they cant regulate voltages if you don't draw enough on the 5v line. People using PC PSU's (AT/ATX) for battery chargers (RC) regularly end up putting a dummy load, such as a fan, or a lightbulb on the 5v current. Some PSU's even need the 3.3v feed wired to something for the same reasons. Cheap PSU, 5v regulator circuits. Better stuff, more picky. Generally speaking that is.
As i said, no idea about the Amiga's PSU.
B!

Switch mode supplies or SMPS rely on having a load connected, the load becomes the feedback mechanism for the supply, as load is increased or decreased the internals adjust duty cycle to compensate for the required current draw, in ye olde days when circuitry wasn't well planned, they could blow themselves up because the circuitry didn't have protection against open circuit connections.

Wall warts don't generally suffer this issue because they use 78xx or 79xx regulators which will happily output voltage all day into an open load.

[Eamoe]

Originally Posted by Loedown

Wall warts don't generally suffer this issue because they use 78xx or 79xx regulators which will happily output voltage all day into an open load.

Really gone off-topic by now, but this is quite interesting. I guess that's why we are now told that removing those wall adapters saves electricity.

1. Still, I don't know if I should fear plugging my mate's Miggy to a PSU that outputs 14V.

2. On the other hand, I've tested one of mine, which feeds an A500 with power since 2 years, and I was quite surprised to find +4.88V / +11.6V / -16.2V (that was tested unloaded, though). Maybe it could explain the vibrating noise this PSU sometimes does...? Should I feel concern about this?!

[Loedown]

Originally Posted by Eamoe

Really gone off-topic by now, but this is quite interesting. I guess that's why we are now told that removing those wall adapters saves electricity.

1. Still, I don't know if I should fear plugging my mate's Miggy to a PSU that outputs 14V.

2. On the other hand, I've tested one of mine, which feeds an A500 with power since 2 years, and I was quite surprised to find +4.88V / +11.6V / -16.2V (that was tested unloaded, though). Maybe it could explain the vibrating noise this PSU sometimes does...? Should I feel concern about this?!

Vibrating noise is either the transformer has come loose from its normal position or the transformer former has come loose, in either case Araldyte or your favourite epoxy resin of choice is the solution.

In wall warts current is drawn by the circuitry inside the 7805 chip, it's minimal but measurable, some milli amp value ~2.5mA

[Eamoe]

Originally Posted by Loedown

Vibrating noise is either the transformer has come loose from its normal position or the transformer former has come loose, in either case Araldyte or your favourite epoxy resin of choice is the solution.

Ah, good, will do this!

Erm, em I taking any risk of bad shock when pulling the screws out and starting to mess with the inner parts? I always heard so...

[Loedown]

Originally Posted by Eamoe

Ah, good, will do this!

Erm, em I taking any risk of bad shock when pulling the screws out and starting to mess with the inner parts? I always heard so...

If the unit isn't plugged into the wall, but you leave it plugged into your Amiga for about 1 minute with the Amiga power switch on then it will discharge the main filter cap, after that go mad.

Mains filter caps hurt, just believe me they do, Zetr0 will concur, most likely Ian Stedman, probably Rkauer, Thomas and several others.

[Toni Wilen]

It isn't 1990 or early 2000s anymore.

Most "wall warts" have been switchers (light and >0.5A capacity) for years. Only very old, (too) cheap or crappy switchers die without load.

[Mr B]

I haven't thought much about it. Last time a "wall wart" broke here, it was the housing that came apart revealing a regular iron core, copper coil transformer, a few resistors, and a capacitor. That was the power for my KVM switch, about 5 years of age, and in the area of 1A if memory serves me right. Replaced it with a switched, and selectable output unit, and haven't thought of it since. I'm guessing cheap OEM items may still use a regular transformer, saving development costs but spending on expensive materials as copper instead.
B!

[Eamoe]

Originally Posted by Loedown

leave it plugged into your Amiga for about 1 minute with the Amiga power switch on then it will discharge the main filter cap, after that go mad.

OK thanks. Sounds easier to discharge compared to CRT's...

[Shadowfire]

Back in the day, I did have an A500 power supply that killed off two A500's before I realized the power supply was destroying the machine. I had sent the machine into our local computer shop (a place called Friendly Computers, defunct for years now) for a trade-in on an Amiga 2000. They tested it, said it didn't work, and sent it back to me. I'm the kind of guy that likes to take things apart and see how they are put together, and months previous I had opened up the PSU.

Well, I didn't check the PSU out right away. A friend offered to loan his A500 to me for the exchange while I sent mine out for a board swap, and I hooked it up and it worked for about 2 minutes before dying. At that point I opened the PSU and realized the shop had swapped PSU's while testing machines (heat sink in the PSU was VASTLY different than what was in my original PSU).

Unfortunately I didn't have any leg to stand on to get remuneration, so I was out $1300 (2 board swaps and a new heavy duty PSU). Needless to say they never got any more business from me.

Based on the information you posted here, I don't believe that your PSU will damage the machine. I wouldn't worry about it unless the 5V line was over 5.5V, the +12/-12 volt lines on the OP's power supply are within operational parameters (the only thing they drive in the Amiga is the audio opamps which are +15/-15 rated rails).

[Hewitson]

5.5V? Absolute maximum voltage for TTL is 5.25V.

[Loedown]

Originally Posted by Hewitson

5.5V? Absolute maximum voltage for TTL is 5.25V.

Where do you get this from?

[Hewitson]

Datasheet for any 74 series IC.