Power supply questions (A500/600/1200)

[ptyerman]

Use a Goliath myself, excellent PSU. Hard to find now though. The best "new" options are already mentioned except a ATX converted PSU which can be purchased at most Amiga stores.

[arkpandora]

Thank you for all the help. I understand that a PSU is hard to predict as long as you don't know what it is really made of.

I have read that for the Commodore 64 there is an adapter you connect between the PSU and the computer that protects the computer by cutting the power off in case of excessive voltage. Is there anything similar for the Amiga?

I have found this other informative thread about "killer power supplies".

Can you confirm that I should only test a PSU after adding a dummy load on the +5V and +12V pins, in order to get correct measurements, and protect the PSU from breaking down if it is old or bad quality?

[ptyerman]

Most PSU's (of the newer type especially) require a load on them anyway to get a accurate reading, without a load you would get very inaccurate results.
Some switch mode types won't even give out any voltage without a load on them and turn themselves off, Amiga PSU's don't do this but could give a false reading without a load on them.

[idrougge]

The C64 "saver" box was made because the +5V regulator of the C64 PSU often fails in a spectacular way, killing the C64. This has yet to be documented on the Amiga, which uses totally different designs for its PSUs.

[Daedalus]

Originally Posted by scuzz

On my 1200 tower the power is single blue wire going to the centre of the square connector. Works... but not good.

That's just the -12V for serial and audio - the power to the actual digital parts will be provided elsewhere, probably through the floppy power connector on the motherboard.

[Franchute13]

Hi.
To test a PSU, it must have a load. You can assemble this circuit to generate a load.
http://eab.abime.net/showthread.php?t=70016
When the circuit is connected to the PSU, measure the voltages.
It is advisable to change the electrolytic capacitors of the PSU.

[Pat the Cat]

Originally Posted by arkpandora

Hi!



1. Is there any risk of damaging the Amiga if I use them without checking them first?

Yes. It is very unlikely you will damage the Amiga, but it is possible. One fairly good idea is to test the output of the voltages from the power supply with a multimeter, to make sure that +5V isn't coming out as something dangerous like 24V.

The chances of actually damaging an A500, A600, or A1200 is less than 1 in a 100 though. This becomes zero if you just test the PSU voltages first, with a multimeter.

Originally Posted by arkpandora

2. If I need to check them first, can you direct me to any tutorial on the art of checking an Amiga PSU?

This is a very indepth article.

http://ianstedman.co.uk/Amiga/amiga_..._supplies.html

This is a more general article on using a multimeter to measure voltage.

https://learn.sparkfun.com/tutorials...-a-multimeter/

Here's a video if you are really struggling.

http://r.duckduckgo.com/l/?kh=-1&udd...%3DZBbgiBU96mM

Originally Posted by arkpandora

3. Should the capacitors of an original A500/600/1200 PSU be changed, like those of an Amiga motherboard (I know my Amigas' motherboards will have to be recapped as well)?

Sometimes yes, most often no.

Originally Posted by arkpandora

4. If I want to buy a new PSU that I can use without risk of damaging the Amiga, which type of PSU should I buy? It should be powerful enough to power an expanded Amiga 1200.

Thank you so much for any help!

Already covered (picoPSU) - the issue with consumer Amiga PSUs is the connector. I still haven't come across a supplier for a square pin DIN connector. So you have to use the connector from the original PSU.

This last point I could well be wrong on, maybe I've been looking in the wrong places for 30 years.

Originally Posted by demolition

The shield (outer square) should be connected to Ground.

I usually stick to AC mains earth pin rather than ground for shields.



This isn't always an option, and can produce problems with analog systems like audio, video displays, serial and parallel connections, etc.



9/10 time it makes no difference, I'll admit.

[Daedalus]

Originally Posted by Pat the Cat

One fairly good idea is to test the output of the voltages from the power supply with a multimeter, to make sure that +5V isn't coming out as something dangerous like 24V.

As has been pointed out already a couple of times, if it's a SMPS you might need a dummy load on the output, otherwise it could actually read an incorrect voltage even if there's nothing wrong with it.

Already covered (picoPSU)

A nice alternative that was also mentioned, and doesn't mean also having to find a suitable power brick for it is the Mean-Well PT-65B, which also has the advantage of being cheaper than the equivalent PicoPSU and power brick combo. A number of people have fitted these inside original Amiga PSU cases, which could be an option if the case is the right size.

I usually stick to AC mains earth pin rather than ground for shields.

This isn't always an option, and can produce problems with analog systems like audio, video displays, serial and parallel connections, etc.

... Which is why it's important to keep your grounds, shields and earths separate (and know the difference between them if you change an existing setup).

[wiz12]

Is this the perfect replacement PSU for the Amiga voltage-wise and would there any wiring guide?

Originally Posted by Daedalus

A
A nice alternative that was also mentioned, and doesn't mean also having to find a suitable power brick for it is the Mean-Well PT-65B, which also has the advantage of being cheaper than the equivalent PicoPSU and power brick combo. A number of people have fitted these inside original Amiga PSU cases, which could be an option if the case is the right size.

[pyksy]

Originally Posted by wiz12

Is this the perfect replacement PSU for the Amiga voltage-wise and would there any wiring guide?

It should work. I replaced mine with a Mean Well RPT-60B and a friend of mine replaced his with a Mean Well RT-50B. Both got installed inside the original PSU cases.

There was no need for a wiring guide because all the wires were clearly labeled on the original power brick PCB.

[Daedalus]

Originally Posted by wiz12

Is this the perfect replacement PSU for the Amiga voltage-wise and would there any wiring guide?

The voltages should be perfect, it's just a matter of reusing the existing cabling. You'll need to solder the cables (or find appropriate crimp connectors), but it's pretty trivial, and less involved than adapting a PicoPSU. As was said, just match the labels with the appropriate connections. If the connections aren't labelled in your Amiga PSU, the guide linked to above gives a clear pinout at the Amiga end so you can figure it out with a multimeter.

[arkpandora]

Thank you everyone for your valuable help!

Can you confirm that testing an original Amiga PSU without load will not damage it? Setting up a dummy load is beyond my capabilites at present, so I will begin by measuring without load, even if it produces inaccurate results. Even using a multimeter is new for me!

But according to what I have read, my 20-year-old internal PC PSU modified with an Amiga connector might be old or cheap enough to suffer from being tested without load, so I will probably look for nearby help.

[xArtx]

Here’s an example of measuring a linear supply, in this case a CD32, with a cheap multimeter.
It’s easier than you think. For switchmode, it will be the same once the supply is loaded down.
[ Show youtube player ]
There is some inaccuracy in the meter itself, and due to added cable length, temperature etc. but this is acceptable.
Also for linear supplies the meter should become stable, and not jump around too much (power supply ripple).

Now this Pico PSU thing...
Does anyone have one of these that could tell me the two chips on it?
And verify the yellow parts are just covered ferrite inductors?



Because what I suspect is the two duplicate regulator circuits are more or less example
schematics from the chip manufacturer, and at best they trademark their name,
and have copyright on their PCB artwork, but the patent pending part is a joke.

That kind of thing motivates me to reproduce and publish the full schematic,
and also email it to them from my ISP.

[Daedalus]

I'm not sure what your ultimate goal is there... I have a couple of them, and from memory they are inductors but I'm not going pulling them out of where they are to read the chip numbers. There are already plenty of (poorly executed) clones around if your issue is with the company themselves. Just bear in mind that there are two double-sided PCBs back-to-back, so there are a lot more parts involved than you can see in that image. If you were going to reproduce the schematic, you'd really want to get your hands on one yourself and do a full stripdown to identify all the components used. Most PSUs are based on pretty standard schematics anyway, it makes perfect sense to start with the manufacturer's application notes.

Patent Pending means just that - it's pending. It hasn't been granted yet, and doesn't do anything to affect its chances of being granted. Nothing wrong with that. But if you can get hours of enjoyment from going on a little crusade, then by all means.

[xArtx]

It’s the “Patent Pending” that gets on my nerves if indeed it’s just the regulator chip manufacturer’s example, and no new IP.
That would be close enough tot see from a data sheet.. a half hour job at best, not a crusade.

[idrougge]

Originally Posted by xArtx

That kind of thing motivates me to reproduce and publish the full schematic,
and also email it to them from my ISP.

Why would you want to do that? The Chinese have already done it years ago.

[xArtx]

Because the “Pat Pend” is only there to imply there is some IP to be protected,
which is fair enough if there is, but I seriously doubt it.

For the initial first year patent, little if anything of IP has to be demonstrated,
and it’s also very cheap, but there’s also no reason I see to apply for it at all,
other than to discourage others from making or open sourcing the same thing,
which again, is fair enough if there’s something unique about it, but again, I doubt it.

[Daedalus]

It's getting a little off-topic here, but without having looked at the patent application you can't know what they believe is novel. It probably isn't the use of standard circuit designs that's novel enough to warrant a submission, and if that's what it's about, there's an abundance of prior art so it'll never be granted. I'd suspect it's more about the form factor, which may not have been done before. But that's up to the respective patent offices to decide upon. Companies submit patent applications all the time for almost anything they develop in house, and while they try to maximise their chances of having them granted by making sure their concept is unique, it's accepted that a certain proportion will always be rejected, and rightly so in most cases.

[arkpandora]

Here’s an example of measuring a linear supply

Thanks! As my lack of experience is partly due to my lack of common sense, even using such an extension lead ended in a pair of pliers would not easily have come to my mind.

[wiz12]

Just performed a voltage test on one of my PSUs and I'm getting 9.3v on the +12v rail. Should I be concerned?