LeCroy DDA-125 Stuck in standby (I think)

[Converter]

Just do it the same as LeCroy does. This will slightly complicate the disassembly, but there is free space and you can add a new connector to the middle of the conductor. I recently repaired my DDA-120 power supply, I was too lazy to add another connector - I just cut this wire, and then connected it again with soldering.

[DaJMasta]

Hahaha, that's LeCroy's official solution?

While it's perfectly reasonable to do... somehow I feel bad about messing up the original drop-in modular design, even if it ended up not being great in the long run, it seemed like a nice solution 


I may try just tinning the socket connectors first just to see if I can get around jumpering them, but we'll see if it does anything.  As it stands, with maybe a 20-30 pound weight on the top of the power supply, I think it would boot up quite repeatedly, though this may just exacerbate the problem with time.  It's odd though, looking at the contacts on the acquisition board side, it seems like most electrical contact is being made on the side of each blade, in a circular pattern over the fork gap.  I think the contacts on the PSU side have been sort of snapped in over little plastic bumps which make both the contact point and the locking mechanism to hold on the connector, but since the connector is melted, I don't think there's a way to get it off non-destructively anymore aside from desoldering everything from the main board.

[Leo Bodnar]

I think the contacts on the PSU side have been sort of snapped in over little plastic bumps which make both the contact point and the locking mechanism to hold on the connector, but since the connector is melted, I don't think there's a way to get it off non-destructively anymore aside from desoldering everything from the main board.

You'll find that plastic is very brittle and will crack easily when you try removing it. 
And when you get to the PSU connectors you'll find blade terminals just like on acq board with measly clips around them.

I have used DIN 41612 Type H11 female socket connectors for repair. 
Opened them up, extracted the terminals and soldered them to cut off stubs on the PSU.

Doing it right needs complete replacement of both connectors.

Leo

[Converter]

Hahaha, that's LeCroy's official solution?

Do you doubt this? It's normal practice for LeCroy to add a bunch of new wires to their devices
Previously, I showed a photo of my DDA-120. And here you will see the same thing in the photo of the other guy: Click
And also for example Here
Since the added wires on the LeCroy board already have so many, it does not look unusual

[DaJMasta]

Haha, I didn't but I still think it's funny.

Bending over the blades on the bottom terminals for extra strength is a nice touch.  I'll make my attempt this evening, and maybe I'll end up with another example of a nice bodge.

[DaJMasta]

I gave tinning a shot, and it seems to have done the trick.  I ended up removing a bit of solder after this pic, and then did a little extra carving of the damaged housing in the top, as the tolerance for the contacts actually seems to be fairly tight.  Especially on the pins that had seen heat damage, it was tough to get the silver solder I used to stick, but an upgrade to the flux I was using did a decent job.  I've got maybe 8 or so boots without a failure so far, but I think it's good to wait and see a bit, since it's a sort of problem that could worm its way back.



Anyways, hooked it up to the RF gen to check whether it was really running 8GS/s and what the -3dB point really was.  The 8GS/s mode is available on any channel, to my surprise, that special mezzanine board seems to be full-featured.  Not only that, but with the statistics on, I measure just a hair over 3dB of loss with the peak to peak measurement, though the RMS is not accurate down in this range.  In a cursory glance through things, does this scope have a sin(x)/x interpolation mode that can run with the statistics?



Also worth noting that the lower volts per division modes really attenuate the signal (it does say this in the datasheet), having some initial trigger trouble, I zoomed in to get a good lock, but read about a tenth of the peak to peak value of the signal.  Zooming out got it back, but I was a little surprised the first time.  That trigger trouble was actually caused by using the internal timebase, it seems not so stable when you're only getting a handful of samples per period, when I switched in the same external reference as the RF gen, it was just fine.

Happy Birthday, DDA-125!

[Converter]

Congratulations, DaJMasta.
But, I think that you could turn on the sweep mode in your oscillator and the Oscilloscope FFT would show you the frequency response more representative.
I used a noise generator to test my DDA-120. The result is shown in the photo.
Frequency response for modes less than 10 mV/div. in fact, greatly worsens. It seems, this is said in the datasheet. This is true for all digital oscilloscopes.

[DaJMasta]

Getting more familiar with the interface, sin(x)/x interpolation is really easy, I just hadn't really considered the way this scope is setup in sort of functional layers.  Similar settings to above, I fed a +5dBm sweep from 1MHz to 3GHz into the scope and took some FFTs.  You could see some higher harmonic tones well beyond 3GHz, was a bit surprised to see them showing up with an 8 bit digitizer, but they were certainly visible.  Looks like the filtering is setup as advertised, where that first little knee is right about 1.5GHz (though if I'm interpreting it right, it's reading 1.5GHz as closer to a -5dBm point), but you can still get usable input to 2GHz if you're just looking for signal presence.  There is some signal response beyond there, but it would be so small that it would be tricky to visualize normally with the bandwidth restrictions at very low levels - though i suppose an active probe would overcome that.  And remember, this is the response with the 'built-in PP096'.



It's a neat (not so) little beast, I feel like the UI takes a lot of 'keystrokes' to navigate, but things are setup pretty well.  I think the biggest complaint so far is that they had to shorten measurement names to only a few characters, so trying to guess which 5 letters they used for which measurement has been fun 

[Converter]

Thus, we can say that your frequency response DDA-125 does not differ from my DDA-120 .
Perhaps this is due to the presence of your multiplexer, as in DDA-120, which should not be in DDA-125.

[Enginerding]

Working through an intermittent PS problem exactly like DaJMasta's.  W/ an unloaded PS, C93 (ceramic, 390pF, 1kV), gets up to ~140F in about 20 seconds (thermal cam).

Pulled C93 and tested it with the DE-5000 LCR @120Hz = 401pF, D 0.021, Q 59.4, ESR 12K-300K, -89 phase angle.  The ESR and Q ar alllll over the place.

I grabbed a new (caveat: new from Aliexpress) 470pF, 1kv ceramic.  I tested @120hz = 470pF, D 0.003, Q 135,  ESR 20k-50k, -89.5 phase angle.  ESR also all over the place.  This has to be some kind of measurement error, right?

I've pretty comfortable with using ESR to grade electrolytics, but I've never looked at ceramic caps in the pF range.  I'm stumped by this measurement.