Author Topic: Agilent DSO5054A repair attempt (ch1 polluted with 1Ghz sine signal)  (Read 1381 times)

0 Members and 1 Guest are viewing this topic.

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: us
Hello community!
I am trying to repair Agilent DSO5054A.
It has following symptoms
    1. Ch1 polluted with sine signal 1Ghz Vpp 200V and attempts to self cal fail with error "Ch1 digitizer error"
    2. Ch2 sometimes getting this signal too and it is intermittent
    3. Ch3,4 are ok no pollution and appear to measure correct values.

My thought is that reference 1Ghz signal from 1Ghz oscillator somehow gets into Ch1 and Ch2 missing comparators.
I will be posting my findings here.
Any advice warmly appreciated.  :-BROKE
 

Offline SaabFAN

  • Frequent Contributor
  • **
  • Posts: 649
  • Country: de
Do you mean 200mVpp, or is it really showing 200Vpp?

Is the noise identical on all the Volt-Ranges, or does it change? If it's the same over all the ranges, the fault is more likely to be found in the ADC and the adjacent circuitry, if the amount of noise changes across the ranges, its more likely to be found inside the analog frontend that does all the signal-switching.

I would first look for any gunk or potentially conductive material on the board that could couple a 1Ghz signal into the ADC-Inputs.
After that there's the obligatory check for bad caps and the supply-voltage for the ADC - if the driver of the clock is powerful enough I could imagine that it is able to turn on the ESD-Diodes at the Clock-input, modulating the Supply-Voltage that is somewhere else being loaded down and does not reach the specified value. Cause could be a bad ceramic cap sitting at the very edge of going short circuit somewhere between the local regulation for the ADC and the ADC itself.

Also, if you don't have an infrared thermometer, move your hand over everything on the main-board. If there's a hotspot, that area should receive a closer investigation. WARNING: Have proper ESD-Protection in place and be prepared to get burned. The PSU is easily capable of vaporising traces so any failing part is going to be !!!HOT!!!
« Last Edit: May 18, 2016, 10:26:25 am by SaabFAN »
 

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: us
Thank you SaabFAN
Yes measures across all ranges, so it is not going to analog part.
I will follow you suggestions and report what i found.
 

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: us
Well i have put aquisition board under microscome and found that someone did soldering on it and left it with flux spils. I assume repair was unsuccessfull that is why it was left as is. Made a thermal imgae with my flir one so it is not best. I do see warmth of voltage regulator and comparators. Temps are around 40C-60C
Here is images i made, sorry it is made from my phone, maybe not very good quality.
 

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: us
I wonder is Agilent ever published any schematics for their scopes? Googling did not give me any hint.
 

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 442
  • Country: us
Re: Agilent DSO5054A repair attempt (ch1 polluted with 1Ghz sine signal)
« Reply #5 on: June 02, 2016, 12:15:52 pm »
Ok I have just cleared all that flux mess and 1Ghz vent away.
And Ch2 is in good shape. Even Ch1 shows some vitality signs.
1 It shows waveform on all frequency ranges correctly  and correctly measures frequency
2. at some attenuation ranges it shows proper voltage and waveform amplitude. but at some it looks like signal is pulled down, but devider approximately changes accordingly.
Here is table of signal taken from probe compensation (sorry my signal gen somewhere buried in boxes waiting my new lab and bench)
Volts/divVpk-pk
503
202.6
102.6
52.6
relay switches
22.5
11.38
500m690m
200m300m
100m170m
50m88m
20m125m
So as you can see on range 1Volt/div it starts fooling around.
I have to thoughts
1. Attenuator/amplifier (Agilent 1BN7-8453) is baked at least one its part.
2. Some passive component dragging signal down.

If someone have any ideas please?
Next step it to check simple components some resistors or caps around acquisition block relays and amplifier/attenuator
But already looking for  Agilent 1BN7-8453 on sale because Murphy law will get ya  :D
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf