Closed gkasprow closed 6 years ago
@gkasprow I check the DAC supplies in the overheated condition, both +12V and -12V are nominal.
So, AFAICT, the absolute max operating temperature of the DAC is +85C, but the overtempeature cutout only trips at 120C...
so it seems that my DACs died in more spectacular way...
@hartytp @gkasprow I have measured the noise of the Zotino with the filter input grounded, and the DAC removed (to not overload the SMPS). The noise curve is below (previous DAC at 0V data for reference): There still seems to be significant noise at ~100 kHz, even without the dying DAC heavily loading the SMPS
The low frequency stuff is mysterious. Shouldn't that SMPS be at ~1 MHz >> 50 kHz?
Thanks for doing that Chris!
There's obviously something severely broken with this board. IMO there is no point debugging this further, given how much abuse these boards have been through. Better to move onto producing the v1.1 design asap.
@cjbe could you short the output CMC and repeat the measurement? It can pickup some external noise...
@gkasprow I assume you mean short both coils of the CMC in order to effectively remove it? I am testing this with your 'ribbon to BNC' test boards, so there is both a CMC on the Zotino, and a CMC on the test board. I assume I should short both?
Yes.
@gkasprow The broad-band noise also looks orders of magnitude higher than it should be, so I don't think this is the chokes (we always blame those poor things and they've never been the problem yet!)
Can we defer further tests until we've got the v1.1 boards ready, please?
@gkasprow I shorted out the CMC's on a grounded channel - this did not significantly change the ~100 kHz noise, but increased the 50 Hz and harmonics pickup by >50 dB (so the CMC's are doing a good job!)
Looks like (at least part of the problem) is the power supply lines.
If you want to try to track this down further, you could try replacing the SMPS with a bench PSU and see what happens to the noise. Also, that would let you check the current draw on the +-13V lines and check that they match the power budget...
I'd suggest one crazy thing - measurement without the power supply. @cjbe What tools are you using to measure the noise density? I'd like to recreate it in my lab.
@hartytp I updated schematics of Zotino and initial components placement
@gkasprow I am using an R&S UPV Audio Analyzer. I will do some more sanity checks tomorrow to trying and narrow down the source of the noise.
I have access to Agilent SSA, in dedicated faraday cage room. I will check tomorrow what is the lowest frequency it can measure.
Closing this as we're done with tests for Zotino v1.0. v1.1 discussion at #446
Issues:
Tests that can be done without an Artiq driver:
Tests that need an Artiq driver:
Configuration for measurements is: DAC->3m SCSI cable->HD68 to IDC->IDC to BNC->scope/DVM/noise analyser. CMCs and a LPF on the IDC-BNC board as previously discussed. LPF chosen to avoid degrading the DAC noise or bandwidth too much.