A600 391029-01 Power Supply - integrated Voltmeter for +5V and +12V

[solarmon]

This weekend I decided to mod an A600 391029-01 power supply to have an integrated voltmeter that shows either (switchable) +5V or +12V

I have an 391029-02 for my A1200, but this is the 'heavy' type with with potting compound - so very hard to mod :



I might chip away at the potting compound at a later date.

So for this project I acquired a 'light' 391029-01 power supply, which does not have any potting compound in it:




I wanted to be able to show both +5V and +12V. Maybe at a later date I can figure out how to get -12V shown too.

The voltmeter I have is only for one input voltage source, so I needed a way of switching between the two.

The voltmeter has a 5 digit display. This will allow me to see how much the voltage reading fluctuates.

The schematic I have for this setup is:



I won't bore you with the build pictures here, but they see be seen at:

https://imgur.com/a/xNegdzd

Here's the finished build showing +5V and +12V readings, with the switch on the side of the power supply (sorry for the poor lighting):





The +5V reading was very stable, albeit a bit high, and sat at a constant 5.262V (sometimes 5.261V).

The +12V reading was not so stable with the last two digits tending to fluctuate frequently. I don't whether this is typical or not, or a sign of a particular power supply capacitor needing replacing?

[8 Bit Dreams]

I think it's better to get a display where 5v & Amps are shown instread of 12v since this voltage doesn't really matter so much, and is used by floppy drive only where amps are really important, especially, if accelerator card, ide & CF Card are in use...

[solarmon]

Originally Posted by 8 Bit Dreams

I think it's better to get a display where 5v & Amps are shown instread of 12v since this voltage doesn't really matter so much, and is used by floppy drive only where amps are really important, especially, if accelerator card, ide & CF Card are in use...

The voltmeter is all I got at the moment. I do have a separate ammeter display, but I will getting a combined voltmeter and ammeter to replace the existing voltmeter.

I'll still have the 5V/12V switchable for the voltmeter part, but obviously the ammeter part will just have to be for the 5V.

[solarmon]

I added a 70mm 12V fan, but connected it to 5V instead. Ideally it should be temperature controlled, but for now this'll do.

I just hot glued the fan to the grill at the top of the case.

[Daedalus]

Originally Posted by 8 Bit Dreams

I think it's better to get a display where 5v & Amps are shown instread of 12v since this voltage doesn't really matter so much, and is used by floppy drive only where amps are really important, especially, if accelerator card, ide & CF Card are in use...

True, and most newer floppy drives don't even use the 12V at all, so it's only for the audio and serial circuitry, where stability/smoothness is more important than precision.

It has to be mentioned however, that there are a couple of issues with using the same type of panel meter that measures voltage and current:

First, they typically measure current on the ground side, which means that while you can see the voltage on the 5V rail, the current display will be the total current on all three rails. Naturally, the +/-12V rail currents will typically be low, but it's not guaranteed.

And second, current is measured by checking the voltage drop across a resistance. This normally doesn't cause much of a problem, but on the Amiga, the ground and shield are common at the PSU and usually connected via a choke at the Amiga. This means that the voltage drop will in part be negated by current flowing through the shield path, thus giving you an inaccurate low reading on the ammeter. This can be resolved by disconnecting the shield at the PSU end, but I prefer not to do that as it's been designed to be connected for noise and protection reasons.

[solarmon]

Thanks, some really good points have been highlighted for me to get my head around!

Originally Posted by Daedalus

First, they typically measure current on the ground side, which means that while you can see the voltage on the 5V rail, the current display will be the total current on all three rails. Naturally, the +/-12V rail currents will typically be low, but it's not guaranteed.

Can this issue be resolved by having the 5V line going through the ammeter instead, if technically possible?

Originally Posted by Daedalus

And second, current is measured by checking the voltage drop across a resistance. This normally doesn't cause much of a problem, but on the Amiga, the ground and shield are common at the PSU and usually connected via a choke at the Amiga. This means that the voltage drop will in part be negated by current flowing through the shield path, thus giving you an inaccurate low reading on the ammeter. This can be resolved by disconnecting the shield at the PSU end, but I prefer not to do that as it's been designed to be connected for noise and protection reasons.

In this particular power supply itself I am finding that the shield and ground are not connected - there is no continuity between shield and ground pins indicated when using a multimeter. Or is the shield and ground connected at the Amiga motherboard end?

Again, can this issue be resolved by having the 5V line going through the ammeter instead, if technically possible?

[Daedalus]

Originally Posted by solarmon

Can this issue be resolved by having the 5V line going through the ammeter instead, if technically possible?

No, at least not with how these combined meters that run off the same supply work. A completely separate meter that can accept its own *isolated* supply will be able to do it, but they're generally much more expensive.

In this particular power supply itself I am finding that the shield and ground are not connected - there is no continuity between shield and ground pins indicated when using a multimeter. Or is the shield and ground connected at the Amiga motherboard end?

Ah, that's interesting... I've never done much with the potted type PSUs other than salvage the Amiga cable from dead ones, but the non-potted ones I've seen do have the earth and ground connected at the PSU end. In that case, the second issue is moot. Does this PSU have the mains earth connected to anything? I've sent all mine to the recycling so I can't check...

Yep, the shield and ground are connected within the Amiga, via a ferrite choke with close to 0 resistance.

Again, can this issue be resolved by having the 5V line going through the ammeter instead, if technically possible?

Same story as above. If you measure on the +5V rail, it should be fine but you need a meter capable of that. It will work without the shield ground connected, but if it wasn't connected anyway in that PSU then nothing to worry about. This is how it is done on the ready-built replacement PSUs with built-in voltmeters/ammeters. I prefer to have the shield connected personally when such an option is available...

[solarmon]

Ah, that's interesting... I've never done much with the potted type PSUs other than salvage the Amiga cable from dead ones, but the non-potted ones I've seen do have the earth and ground connected at the PSU end. In that case, the second issue is moot. Does this PSU have the mains earth connected to anything? I've sent all mine to the recycling so I can't check...

Sorry, maybe my OP might have been a bit misleading if just looking at the pictures. My mod is on a 391029-01 light and non-potted power supply. The image of the potted one was just to show the potted 391029-02 one that I also had.

No, at least not with how these combined meters that run off the same supply work. A completely separate meter that can accept its own *isolated* supply will be able to do it, but they're generally much more expensive.

The combined voltmeter/ammeter I am getting has the option of using independent or shared power - as can be seen by the diagrams below



If powering it 'independently' (bottom diagram) , albeit with the same 5v line, would this still have the issue you describe?

[Daedalus]

Ah, ok, sorry I didn't catch that. Fair enough, there are many different types of power supply that Commodore used - I haven't seen them all.

The independent supply really needs to be independent - to use the same 5V rail you need to treat it as shared. However, it looks like your meter can't measure on the positive rail in either configuration, so it doesn't really matter. It will end up reading the combined current for all rails. If you don't have any 12V loads other than the motherboard, you can probably say that that's around 0.1A so not a huge error really...

[8 Bit Dreams]

Found this on Ebay: Lot Nr. 264244745026
Donno may it be used or not, its basically the same solarmon suggested

[solarmon]

I got the combined voltmeter and ammeter and installed it:



(It probably needs calibrating - there are a couple of potentiometers on it)

This the +5V voltage and usage when loaded in to ClassicWB via CF, and with an 8MB memory expansion card in 4MB mode, and with an internal Gotek with an 2.42" OLED screen and trackball.

The wiring of this setup is now as such - which is getting quite busy and a bit hard to manage inside the power supply: