Oscilloscope fault ... any ideas where to look?

Thread Starter

iwanczuk

Joined May 15, 2018
8
Hi, I have an Aplab 3706C that was given to me with an interesting fault. The fault appears on both channels equally and also when set to dual channel chop or alt modes. The attached photos show the fault with just one of the channels turned on - the lower the trace is on the screen the more skewed to the right the trace is. Also the effect gets worse the further the trace is to the right of the screen. In addition the higher up the screen the trace is the more it droops to the bottom as it moves across the screen from left to right. And there's more ... the further up the screen you move a trace the fainter it gets to the point of disppearing completely about 1cm from the top of the screen.

I am thinking that it might be a power supply problem? All of the components look clean and there is no obvious damage to anything internally. I have no idea what the power supply voltages are and there is no indictaion on the circuit board of what they should be.

I have not been able to find circuits for the unit and have asked Aplab if they can supply them (I doubt they'll give them to me). If it comes to it I'll trace the power supply board to produce a circuit in an attempt to figure out what the voltages should be.

Any advice would be appreciated. Many thanks in advance.

IMG_0672.jpeg

IMG_0671.jpeg
 

AlbertHall

Joined Jun 4, 2014
11,960
I have a nasty suspicion that the electron gun in the tube may be damaged.
Do you have (access to) another oscilloscope to check the waveforms on the deflection plates?
 

Thread Starter

iwanczuk

Joined May 15, 2018
8
I only have 400V probes on my oscilloscope - I do however have an HT voltage probe for my DMM so will check the voltages and if the aren't too high I'll hook up my scope. Can't do this now but will take a look tomorrow.
 

Thread Starter

iwanczuk

Joined May 15, 2018
8
I did check the voltages on the plates and they were within the limits of my scope probes. Here are the waveforms on the X and Y plates. The yellow and blue traces are the voltages on the plates relative to 0V and the red trace is calculated by yellow minus blue. No sign of the non-linearity seen on the actual CRT of the oscilloscope under test. I think this supports the premise that the CRT is faulty. Thank you for your help (and apologies about the undershoot/overshoot on the yellow trace - I have a dodgy scope probe!)

IMG_0673.jpeg

IMG_0674.jpeg
 

MrChips

Joined Oct 2, 2009
25,931
The probe on channel 1 is not dodgy, it’s the user.
The probe just needs adjusting. There is an adjustment trimmer on the probe. Read up on how to properly adjust the scope probe.
 

Thread Starter

iwanczuk

Joined May 15, 2018
8
The probe on channel 1 is not dodgy, it’s the user.
The probe just needs adjusting. There is an adjustment trimmer on the probe. Read up on how to properly adjust the scope probe.
Thnaks for that - it's been adjusted and it makes no difference - the probe is faulty and a new one is on order.
 

KeithWalker

Joined Jul 10, 2017
2,311
The problem may be in the power supplies, not the CRT. The vertical deflection signal may be modulating the timebase waveform. I believe you have a bad smoothing capacitor in the power supplies.
With the square wave connected to the scope vertical input, use another scope on AC coupling to look at the regulated DC supplies for that signal. When you find it, change the smoothing capacitors in that regulated supply.
 
Last edited:

AlbertHall

Joined Jun 4, 2014
11,960
The problem may be in the power supplies, not the CRT. The vertical deflection signal may be modulating the timebase waveform. I believe you have a bad smoothing capacitor in the power supplies.
With the square wave connected to the scope vertical input, use another scope on AC coupling to look at the regulated DC supplies for that signal. When you find it, change the smoothing capacitors in that regulated supply.
Except that the waveforms on the deflection plates are correect.
Also note the dimming of the trace near the top of the screen, probably where the electron beam is hitting some part of the gun.

I used to repair oscilloscopes many years ago and the company had redesigned the tube supports inside the mumetal shield and it turned out that the supports actually magnified the G forces on the tube. They used to come in with the glass rods that supported the gun elements broken.
 

MrChips

Joined Oct 2, 2009
25,931
In post #1 I see two symptoms.
How do you account for the slope in the vertical transitions particularly in the rising transition?
How do you account for the different trace intensities?

Edit: In response to KW's comments.
 

AlbertHall

Joined Jun 4, 2014
11,960
In post #1 I see two symptoms.
How do you account for the slope in the vertical transitions particularly in the rising transition?
How do you account for the different trace intensities?

Edit: In response to KW's comments.
The elements of the electron gun are distorted in some way so, particularly, the Y plates are not producing a deflection orthogonal to the X plates.
If the gun elements are distorted then the beam may hit some part of the gun and hence get partially blocked.
I saw such symptoms in the broken tubes I described in post #9.
 

KeithWalker

Joined Jul 10, 2017
2,311
Except that the waveforms on the deflection plates are correect.
Also note the dimming of the trace near the top of the screen, probably where the electron beam is hitting some part of the gun.

I used to repair oscilloscopes many years ago and the company had redesigned the tube supports inside the mumetal shield and it turned out that the supports actually magnified the G forces on the tube. They used to come in with the glass rods that supported the gun elements broken.
The signal modulating the timebase is very small and fast (rise and fall times of the square wave). It will not show in the waveforms you show.
Take a closer look at the signal on the horizontal deflection plates using AC coupling and higher vertical gain, with the timebase set to 20uS/div. You will be looking for a very small deviations every mSec.
I also repaired many Tektronix, HP and other scopes. I have seen similar problems and they were always associated with defective supplies.
 
Top