What's going on with these PSU measurements?

[exxos]

Might be an idea to take a image of how you have all this wired up..

Also if you move the scope cable around, does the voltage change ?

[Steve]

Update:

For some education I watched this video:



I learned that transformers can act as basically a radio receiver/amplifier, which is where that large 50hz signal was coming from when the PSU was disconnected from AC entirely.

Here are some photos of my setup:

setup.jpg

closeup.jpg

Here are some readings & observations:

Here the Amiga PSU is DISCONNECTED from the AC inlet, we can see the transformer is picking up 50hz 'wirelessly'

NewFile0.jpg

Here all I have done is plug the Amiga PSU into the mains - but NOT turned it on... the large 50hz sine wave has disappeared, although, there is still a small amount left. I assume it's gone because it's connected to common ground now.

NewFile1.jpg

Now I have turned the Amiga PSU ON. I can see that the line has become quite noisy, but it's still oscillating at 50-55hz. The sine wave seems to be going +5v to -5v

NewFile2.jpg

Here I zoom in to a 1ms timebase, we can see the curvature of the waveform, am I right to think that the 5v DC is the 'noisy' aspect of the waveform?

NewFile3.jpg

Here we are zoomed in to 100ns, at this zoom you cannot see it's being carried on a 50hz sine wave. But I can't really see any 'noise' anymore, which is the kind of thing I'd like to measure, like Exxos does for the Atari PSUs.

NewFile4.jpg

Overall, I just can't get my bearings with this. I don't understand why the DC signal is being transported on a sine-wave. So I don't understand how to even begin to measure it for noise.

Edit: perhaps it's right what Exxos said earlier, and the rectification is broken? I've got a bunch of new Schottky diodes on order, when they arrive I'll replace the old diodes and see what happens.

[exxos]

Have you continuity tested or diode tested the diodes just to verify they are not shorted or open circuit ? may be better to remove them though for testing.

I don't see how your scope is picking up mains though. If the power supply is physically unplugged from the wall then there would not ordinarily be anything which could pick up mains.

Maybe you have mains cables running nearby or something ? It may be better do just try a different room or lift the whole setup in the air to rule that out.

[Steve]

Have you continuity tested or diode tested the diodes just to verify they are not shorted or open circuit ? may be better to remove them though for testing.

I don't see how your scope is picking up mains though. If the power supply is physically unplugged from the wall then there would not ordinarily be anything which could pick up mains.

Maybe you have mains cables running nearby or something ? It may be better do just try a different room or lift the whole setup in the air to rule that out.

The scope doesn't pick up anything at all unless it's connected directly to the 5v of the PSU (which is disconnected from mains) In the video he explains that transformers act as an antenna?

[exxos]

The scope doesn't pick up anything at all unless it's connected directly to the 5v of the PSU (which is disconnected from mains) In the video he explains that transformers act as an antenna?

The scope probe could act as an antenna. This could be compounded by when you have it connected to the power supply despite it being disconnected from the mains. Effectively making a "longer antenna".

I don't know where your house it all is, but is there anything which could potentially use a lot of amperage or pumps / motors or something ?

Like I said before though, you can get interference the same way with LED lighting, albeit with a lot high-frequency. I even had a 12V fan running when I was originally doing my power supply tests and it drove me nuts because I ended up with a 400hz pulse showing on the scope all the time. I thought it was being generated on the power supply itself but it turned out to be the fan causing it all :roll:

I cannot really tell where you have the setup. But if it was on the floor for example, and had mains cables running underneath the floorboards, you could potentially pick up interference by your scope probes that way. So I would just suggest repeating your tests in a completely different room away from as much electronics as possible.

The scope itself should not be susceptible to the noise because it is shielded internally. I mean I have my scope balanced on top of my power supply most of the time and don't really see any significant interference on my scope screen. However I do keep the scope leads away from potential sources of noise as much as possible. If the scope need wrapped around the transformer I would assume it would show a similar problem to what you see.

Even so, I would be surprised if it is acting like an antenna that way but currently I cannot think of any other way you could see those signals. I guess you could try removing the power supply and just connecting a long piece of wire and see if that acts the same..

[Steve]

I just measured the 4x large diodes with my multi-function tester:

DT SPR50
Uf=460mV
Ir=0.70uA

And here are the new schottky diodes you recommended I swap them for:

STPS SL40 CHN240
Uf=205mv
Ir=4.4uA

Are those readings looking good / improved over the originals? (measured out of circuit) I know Ir is to do with maximum input current, not sure what Uf means though.

[exxos]

I think it is supposed to be Vf ? Voltage Forward. The voltage drop across the diode. So generally the lower the voltage drop is, less heat is generated in the diode. In terms of switch mode power supplies it can increase efficiency a fair bit as well.

[rubber_jonnie]

If I get a chance I'll hook up my 'scope to my A1200 PSU and see if I pick up anything off 5v.

[Steve]

Ok, cool @exxos
thx @rubber_jonnie

[rubber_jonnie]

Ok, cool @exxos
thx @rubber_jonnie

I just did a quick check, I'll do a more thorough look later, but I don't see anything like your seeing.

One thing I noticed was that your probe GND was connected to Earth in one of your pictures.

Earth and GND are not necessarily the same thing, what happens if you connect your probe GND to the GND for 5V? Do you still see the same weirdness?

Measuring between the shield on the power plug and the earth of the 3 pin plug, I get about 2R, not a lot, but not a straight short, suggesting they may not be the same thing. Be worth giving it a go.