Help ! Strange Issue with Tek TDS3054B

[Thyratron]

I Have a TDS3054B   that has a strange issue, and wondering if anyone on  the forum might have encountered something similar before . It has a low number of  power ups  ( 179)  and  total operating ti,mme  of  322 Hours .  Firmware is  V3.35
For all Y or Vertical Division settings   from 1V / Div upwards  I am getting  bursts of random pulse on the display . At settings of 500 m V   / Div and lower they are not there  ! ! !
This occurs in both 50R  or 1M  input Z setting.
Not sure if this is significant but the repetition rate  seems to be of the order of   12.4 KHz . Perhaps someone might know   if there is  something   in this model “Banging “  away   at this frequency
I have  been able to zoom in to   an individual; pulse  and  it seems to have a rise time of 30 n S   and  fall time of about  120 n S 

Otherwise the  channel is  functioning normally , especially the Attenuator settings which lead’s me to belied that the  Front end Hybrid   is probably OK and the fault is further back .
Incidentally the amplitude of these   pulses seem to reduce in  sympathy with the Attenuator setting

Running any of the built in self checks and Diagnostics doesn’t throw up any errors.
My next step is to get another scope and probe the Diff output from the Hybrid to see if   the artifact is   present here.
I have attached some images showing the issue and which should be self-explanatory

[sicco]

The power supply inside the scope is the first thing to check. Bad caps.
You can bypass the mains power supply by powering with a single 15VDC 3+ Amp external power supply.

Also you can probe (with another scope) if there’s any ripple on the +5, -5, +15, -15 volts around the BNC input connectors.

[Thyratron]

Hi ,
Thanks for the suggestion . This would normally be  the first thing I woul check especially with older T & M Equipment  like an oscilloscope . However in this instance the remaining three channels are working correctly and without issue so  didn't think the PSU would be suspect as I assume the rails are shared by the the signal paths in the other three channels . That being said  I will check  just to elimite  the PSU as a  possible cause

[eb4fbz]

IMHO something goes wrong on the ADC interleaving.

[R-1125F]

I have a TDS3054B with *exactly* the same issue on channel 1 as as Thyratron shows in their post.  It only appears when switched to 1V per division.  All other channels are OK, so it can't be a power supply problem. Anybody got a clue what's wrong?

[BreakingOhmsLaw]

Had a similar issues before. Try disconnecting each and every device in your lab from the mains. In my case, it was the cheap Chinese power brick from my endmill that spewed crap out onto mains line and protective earth. Years before, had an issue that was caused by an older devolo Powerline-LAN adapter.
That would not explain why you see this only on certain V/div settings, but you never know.

[R-1125F]

It is only on one channel.  Others are fine. The waveform is exactly as shown in the original post for this thread.  Dave Jones also has a video showing a similar problem but he never solved it.  I don't think any outside interference would produce an identical waveform.

[BreakingOhmsLaw]

If it's one channel, then you can rule out mostly anything in the internal power supply.
Try disconnecting the inverter of the LCD backlight please.

[R-1125F]

I tried that and no change. 

[BreakingOhmsLaw]

Okay, let's check if the front end. Please connect both inputs to each other with a regular BNC cable and see if that noise transfers to the other channel.

[R-1125F]

Nothing there.  I connected CH1 to CH2 and nothing on CH2.  The next step would be to open the front end and see if there is anything wrong.   
After watching Dave's video, , my thought is to check the relay that is on the channel.  I don't have time to do that now.  It will have to wait for some other time.

[BreakingOhmsLaw]

When you feed a sine wave or a DC voltage into the scope, does the distortion add to the signal? Or does the signal cover the distortion?

EDIT. Just saw the 4th screenshot, so it adds to the signal.
That's not a buggy relay. To quote Seaman Jonesy from The Hunt for Red October: "That's gotta be manmade"
Start probing stuff that could be related (using the good channel). I²C busses, SPI lines, matrix keyboards. Could be anything that as you stated "bangs away" at that frequency. I'd start with the keypad and the screen.

[R-1125F]

I did not have time to do what you suggest but I did find something interesting.  When the coupling is set to ground, the problem goes away.   When the a probe is grounded, the problem is still there.  There are no schematics available, so what the the coupling setting actually doing?  For AC it must switch a capacitor into the circuit.  But how is it grounding?

Just adding:  When coupling is switched to ground, the noise now shows up a settings starting at 50mV and lower.  I have no idea what's happening.

[sicco]

Unclear what it is, but maybe this schematic is of help.

Suggest probing the R1..R15 pins, with other scope or via other channel, and look for something that pulses in sync with what you see on the scope.

Does the problem go & come with the 'Gain' relay being on or off?

The relays are latching type. They only need a short (~40 ms ish) pulse, and they're powered from the + and - 2.5 rails. The ADG361 has digital outputs, and they're set on/off through SPI commands. If after the 'on' time they're not turned off, then quite large current is drained by the relay, and that can kill the digital outputs. Do all relays still click ok? Is any relay getting warm?

'GND' does not have a relay, but is some solid state switch, controlled by one of those R1..R15 pins.

[R-1125F]

Thank you very much for the schematic.  It is a big help.   I could not measure the bottom relays.  The probe won't fit in that spaces.  There is no odd noise on the gain relay.  I can't tell if the relays are actually working. Thermal image of the relays does not shown any heat, so I don't think anything is stuck on.  I did find matching noise on the hybrid going into the ADC.  It looks like two differential pairs and the have the strange signal. This is on the top end of the hybrid, which would be pins (don't confuse with ICs) U2, 3 and U5, 6. 

I found a large signal in hybrid pin R13.  I am not sure what that does.   Any idea?

[sicco]

That ‘R13’ is only connected to something on CH1. The other three channels have it at GND. Not sure what it is for, but i’m speculating it is intended for some fancy advanced premium option, that requires a dedicated plugin module also. Maybe the TDS3SDI or so.

In your case, see what happens if you also tie the CH1 R13 pin to GND. Does that fix the issue?

[R-1125F]

I tried grounding pin 13 and it did reduce the problem but did not make it go away.  It also appears the problem gets worse at it heats up.  I am not sure what the means.  It could be a bad capacitor somewhere.

[R-1125F]

I am giving up for now.  I found this noise in most places in the unit.  I am out of time to work on it.