Re-running the Sonnet model with a full gerber import and no simplifications to see if the ground plane geometry has anything to do with what I'm seeing.

FFT was definitely the right move, I have much better dynamic range now.

Seems like I have about 10 dB of stopband rejection (not great at all, lol) and Fc right around 9.91 GHz.

@dlharmon is this much extra IL in line with what you'd expect on oshpark from plating and copper roughness?

The design target for this filter was 10.25 GHz Fc and 250 MHz -3 dB BW.

The actual as-fabricated filter seems to be more like 9.92 GHz Fc and 292 MHz -3 dB.

I don't really feel confident making any claims about stopband rejection because of the poor dynamic range of this setup, but the hump around 10.7 GHz is not ideal. I saw some ground plane resonances in the simulation and wasn't sure if they were real, it seems they might be? Once I get more detailed S-parameter data with better dynamic range I'll go back to the simulations and revisit.

It's... definitely a filter. I can say that much.

Not sure what's up with the 0 dB regions at start and end of the sweep, it's clearly a bug (The waveform behavior doesn't match, it's definitely not actually 0 dB)

I'm pushing limits of what I can do as far as dynamic range easily. The Siglent is putting +16 dBm into the doubler (which has an absolute max of +20 so it's safe but not leaving much margin). The doubler is only specced for 10-20 GHz input so 5-10 GHz input and trying to sweep down to 4 GHz on the input means I'm operating out of spec but it seems to mostly work.

The doubler has ~12 dB of conversion loss so I only have about +4 dBm coming out of it, which isn't a ton of stimulus energy when the DUT is this lossy.

I'm not sure why it's coming out as 30 dB IL. My simulations said it would be more like 15, plus copper roughness and ENIG plating losses which I did not expect to add another 15 dB.

I expect the filter rolloff is a lot sharper than what's shown here but the -45 dB loss range is putting me basically at the noise floor of the scope.

I could try to get more dynamic range by adding an amplifier after the doubler, but i'm a bit sketched out about putting big RF PAs on something going into a $10K-to-repair-if-damaged 16 GHz frontend that is rated for +/- 2V p-p.

Probably better to just send the thing to somebody who has a real VNA that can work at these frequencies.

Characterizing my first draft of the 10 GHz bandpass filter.

Be gay, don't admit to crimes on fedi

BRB just daydreaming about being able to buy a JESD204B to PCIe bridge chipset COTS

Who have you spent more time window shopping at?

Right. I forgot about this part.

Out come the flush cutters

Lots of hand soldering to do before it's usable though

Which means I have at least one switch worth of line cards now (this one. plus the one I hand soldered a year or so ago). Good progress.

MCU comes up, it turns on all the power rails, and the MCU is responding over UART.

3V3 reads 3.285V.
2V5 reads 2.472V.
1V0 reads 0.997V
1V0_2 reads 0.993V

I don't have it logging all of the sensor values to UART right now (you need I2C from the main board to do that with the firmware I loaded) but it's alive enough for me to be reasonably confident it won't need massive rework.

So I think I'm good to go put the RJ45s and proper power connector on.

3.3V rail comes up OK, it's pulling 65 mA from the 12V input which seems probably reasonable for everything mostly idle?

MCU shows up on SWD, time to try actually flashing it

It's fiiiine this is totally the intended power feed point

All known problems on this board have now been fixed. Time to power it up and see if the MCU is alive.

I'm going to do something truly unholy here, powering it up from a daisy chain of banana plug -> SMA -> SMB -> U.FL connectors to feed 12V into what is supposed to be a test point.

This is not a great idea and it'll probably run warm, but I won't be actually linking up the PHYs to a link partner - there wont even be any RJ45s soldered on yet.

The goal is to check for obvious shorts or things that will require rework before I hand solder the non-reflowable RJ45 and Mini-Fit connectors.

Put the 0R on R15.

This is the remote sense path for the 3.3V rail. Forgetting *that* would be Bad (tm)

I don't have any 1K ohm resistors at home, they're still on the PNP at my friends shop.

None of the three missing ones are critical. Putting 1.2K on instead.

This is an easy fix.

Time to try reworking and bringing up the switch line cards I made over last weekend.

Number 2 has a short on the 1V0 rail so I'll save it for last, I want to try to localize the short before I blindly rework it.

003 looks like it will be the least work to get going so that's going to be the next focus.

It's missing three 1K and one 0 ohm resistors due to PNP programming errors, and has one misaligned resistor due to a reflow defect. Should be quick to rework.

And then of course it still needs the RJ45s and the DC power connection added but I want to do bringup testing before I install any non-reflowable parts