Analysis tools - newbie question

[Bo Deadly]

Can I ask a potentially silly question. Why are you nulling? Is it not sufficient just to FFT the output signal?

The null test was just to test the test rig. But it didn't work so I'm wondering why.

 

[Bo Deadly]

Maybe now is the time to ask "how do I FFT the output signal?"

You're already doing it. That's what the Flux software is doing. FFT is the name of the mathematical formula used to create a frequency spectrum plot or sometimes referred to as a bode plot.

 

[morls]

I don't think it makes any significant difference. What does make difference is how they are connected to "ground". Ground circulation currents must be thoroughly separated from sensitive nodes.
Analyze your lay-out. Consider every connection as a resistor.

Ok, thanks. I've attached a couple of pics of the grounding setup. I have 2 separate connections to chassis - 1 for the PSU, and 1 for the audio. They are on either side of the rear panel. With mains/transformer grounding I included a couple of ground connections from the PSU area of the PCB. On the audio side, there is one ground from the PCB as well as the XLR i/o.

The null test was just to test the test rig. But it didn't work so I'm wondering why.

I'm looking into this, at the moment I'm checking out the Flux Analyser program to see if it's a setup issue. The audio (on DAW) for the tests was analysed by Flux via a plugin on the output bus, and I'm wondering if this might have something to do with it. Flux also works as a standalone analyser with its own signal generator, so I'm working to get this up and running.

I also downloaded REW, which looks very promising. I ran a quick lookback test from an output to an input, here are the results:

Input device: Merging Ravenna ASIO driver
Input: 8: 8
Channel: Right
Output device: Merging Ravenna ASIO driver
Output: 4: 4
Input RMS target: -12.0 dB
Actual RMS at 1 kHz: -12.1 dB
Sample rate: 88200 Hz
Input volume: no control, Sweep level: -12.0 dB
20 Hz .. 20 kHz flatness: +0.0, -0.2 dB
-3 dB points: <2.0 Hz, 38.814 kHz

Cheers
Stephen

 

[abbey road d enfer]

Ok, thanks. I've attached a couple of pics of the grounding setup. I have 2 separate connections to chassis - 1 for the PSU, and 1 for the audio. They are on either side of the rear panel. With mains/transformer grounding I included a couple of ground connections from the PSU area of the PCB. On the audio side, there is one ground from the PCB as well as the XLR i/o.

I don't know where these green wires on the 1st pic go, but I think there are too many connections to the audio ground.

 

[Bo Deadly]

Ok, thanks. I've attached a couple of pics of the grounding setup. I have 2 separate connections to chassis - 1 for the PSU, and 1 for the audio. They are on either side of the rear panel. With mains/transformer grounding I included a couple of ground connections from the PSU area of the PCB. On the audio side, there is one ground from the PCB as well as the XLR i/o.

You need to twist each of the AC pairs together. Not super tight but just enough so that the two are touching and there's no gap in between. This is because current in the wires emits electromagnetic radiation that gets picked up by the analog circuit. If you twist them together, the magnetic fields created by currents running in opposite directions cancel each out. This is super important for the heater wires. So twist the blue and red from mains, the red and black going to the power switch, the purple and white of the PT primary, the green wires of the one secondary and the red wires of the other secondary. Then measure again. I think you'll find your noise drops by quite a bit.

Also, technically we (I) advocate that the chassis be connected to the power supply ground at the filter cap(s) and that's it. So one single fat wire between the chassis and the PCB. Right now you have at least 3 such connections just from looking at your pics. Then the point on the PCB at the filter cap ground (sometimes referred to as "star ground") is used to connect PS ground to the analog circuit ground 0V. In your case the PS and analog circuit are all the same board so you don't have a choice about that. But I would run one fat wire from the chassis bolt next to mains to the "star ground" on the PCB. It looks there are a few solder points with a chassis symbols that are supposed to be the "star ground". Again, ideally this should be next to the filter cap ground but it's got a big ground plane so maybe it's not going to be a big deal if you just pick one of those solder points with the chassis symbol.

But twist the wires first and see what happens to your spectrum. The multiple chassis <-> 0V connections won't matter nearly as much.

The analysis software is not responsible for not getting a good null test. The samples of each channel have to be aligned perfectly and my guess would be that they simply are not. I wouldn't break your back trying to figure that out. Null testing is very useful but it's a little more advanced than you need right now.

 

[warpie]

To me it looks like you've connected the pin1s to the XLR lug, to the chassis and to the audio ground on the PCB. Pin1 should be connected only to the XLR lug. Disconnect that 3-green-wire-lug from that bolt that is attached to the chassis and measure again.

 

[morls]

You need to twist each of the AC pairs together. Not super tight but just enough so that the two are touching and there's no gap in between...then measure again. I think you'll find your noise drops by quite a bit.

Thanks.

Also, technically we (I) advocate that the chassis be connected to the power supply ground at the filter cap(s) and that's it. So one single fat wire between the chassis and the PCB. Right now you have at least 3 such connections just from looking at your pics.

Ah, ok, thanks for that. Up to this point I've been grounding each point of the PCB where marked, instead of just one. I didn't want to run any risk of not grounding properly.

The analysis software is not responsible for not getting a good null test. The samples of each channel have to be aligned perfectly and my guess would be that they simply are not. I wouldn't break your back trying to figure that out. Null testing is very useful but it's a little more advanced than you need right now.

I'm going to try and work this one out, after I do this basic testing on my gear.

I don't know where these green wires on the 1st pic go, but I think there are too many connections to the audio ground.

To me it looks like you've connected the pin1s to the XLR lug, to the chassis and to the audio ground on the PCB. Pin1 should be connected only to the XLR lug. Disconnect that 3-green-wire-lug from that bolt that is attached to the chassis and measure again.

Thanks abbey and warpie, I'll sort this out and re-test.

 

[morls]

I've done some re-wiring, but it hasn't fixed the issue, in fact it's worse! I'm going to start a new thread for this, as it's not about analysis software any more.

Thanks so much for all the help so far, I really appreciate it and have learnt a lot.

Cheers
Stephen