开云体育

Hello everybody,
I have followed this forum for a couple of months. Seems to be full of information. I have just re-capped 7603 mainframe successfully. Both vertical slots basically work (tested with 7A18and 7A26). I have also non-working 7A26 and just wondering if this group could give some help or direction.

Symptoms:
Both traces show no signal on screen.
Traces are visible and are following position and identify controls.
Display shows settings correctly.
Supply voltages (+/- 15V) are not loaded (no tantalum caps shorted).
Visually boards look intact and clean.

Are there any typical fault causing these symptoms?

Best Regards,
Jouko

Do you have a service manual for the 7A26? If not, here are PDFs of the manuals

Looking at the images of the 7A26's insides it should be easy to trace an input signal from the attenuators through the amplification stages and then to the backplane, assuming that you can connect the module to the backplane in a way that allows you to access the insides (I assume that there are extension plugins that let you run a plugin outside the main housing, though I can't find reference to such a thing on TekWiki).

I did something similar with a 475 a few weeks ago, and tracing the signal was quite simple to do, given schematics and board layouts, and a second scope to test with. In my case the issue was an unseated channel switch IC, and tracing the signal through channel 1 lead me straight to the problem.

Looking at the images it doesn't look like the 7A26 has the channel switch that the vertical system of a 475 has (I assume that's located in main unit), which means that the common failure must be elsewhere. The obvious places to look are where the signals exit the plugin, and where power enters the plugin, but you really need to see how far the input signal gets through the amplification chain to make any real diagnosis.

On Tue, Nov 24, 2020 at 11:16 PM, Jeff Dutky wrote:

(I assume that there are extension plugins that let you run a plugin outside
the main housing, though I can't find reference to such a thing on TekWiki).

Fixed extender: .
Flexible extender:

Raymond

The extenders do make things a lot more convenient but you can make many measurements just by taking the left cover off the 7A26, putting the 7A26 in the left vertical slot and taking the left cover off the 7603.

The vertical position and identify controls are placed part way down the amplifier chain, after U1450, so the circuitry is at least working at DC from there onward. If you have means of tracing the signal (eg a second scope) then you should take Simon's route and start by looking for a signal at the output of the attenuator board. If you don't have any other equipment then there is a simple check that will show that everything from the atenuator onwards is OK. It does require making a tentative adjustment that affects the calibration so think twice if you are nervous! At the rear of the attenuator card you will see a large-ish transistor (Q150 in the schematic, mine has a small heat sink attached) and above it there is a preset resistor R160 labelled '2X balance' on the schematic. If the amplifier chain on the main circuit board is OK then a tiny adjustment to R160 will move the beam for the corresponding channel vertically, if you take this route be prepared to redo the balance adjustment after fixing the 7A26.

I would guess that the most likely cause of your problem is that the switches in the attenuators and or the input selectors (AC-GND-DC) need cleaning, if you take an input of at least 1V pk-pk from the calibration generator and try all the V/div settings and both AC and DC coupling then you may find that you intermittently see some sort of signal on the screen.

Regards,

Roger

Thanks for the replies,

R160 didn't have any effect on the trace.
I think the problem might be in the attenuator/source follower part
Attenuator in 10mV/div setting. Resistance from BNC to R130 shows 56R. Probe in place and connected to calibrator I can't see any signal in R130.
Input resistance from BNC is only 340k. When I remove Q150 it is 1M as it should be.
Situation is similar in both inputs. Q150 damaged in both inputs???
I suppose the attenuator switches can't make this behaviour?

Regards,

Jouko

If changing R160 didn't have any effect, then are the +15 VDC and -15 VDC present at its two ends? Did adjusting both R160s not do anything on either channel? I suppose it is possible that something drastic has happened to both baluns (T1301 on both channels), if someone accidentally fed in too much voltage on both channels, but I would have thought that Q150A/Q150B would have died first.

A resistance of 56R between the centre of the BNC input and the non-earthy end of R130 looks to be correct - it is R10.

You might have dirty AC/GND/DC switches - these contacts are a bit prone to getting a bit mucky after years of service, but they should only be cleaned with isopropanol and carefully. It is easy to bend the gold-plated leaves if you are too aggressive and the board substrate can be damaged by other solvents. I have cleaned attenuator cam-switches, but not the AC/GND/DC switches. Others may give you advice on that. You may also have dirt on the attenuator switches, but time will tell.

You may well have faulty Q150s, but since they are dual FETs in the two channels, I would have though the failure of all four devices was a little unlikely.

I don't think that dirty attenuator switches would give the symptoms you describe, but I could be wrong, I often am. Dirt on these switch contacts usually has an effect on the frequency response of the affected attenuator, the dirt being capacitative. It's a pity that both channels are exhibiting the same fault, because otherwise you could swap components like the individual attenuator units and the Q150s between the two.

I have had a problem (still do, to some extent) with one of my 7A26s in that one channel was showing a different gain to the other. I think I got near to sorting that out by replacing the faulty U1350, but my memory is a bit hazy - not that that helps with your problem.

I am assuming that you have the appropriate Service Manual (you are quoting part-numbers) and that you have checked all of the voltage rails. That is often a cause of faults.

Good Luck, they are nice amplifiers,
Colin.

Colin and Juoko,

If the 7A26 is arranged anything like the 475 then cleaning the AC/GND/DC coupling switches is almost exactly the same as cleaning the rest of the attenuator switches: soak a narrow strip of 100% cotton bond paper in IPA, slide it between the open switch leaves and the contact plate on the PCB, close the switch trapping the strip of paper, gently pull the paper out, and repeat several times for each contact point. I found the process more difficult for the attenuator switches on the 475 because you have to remove each attenuator block to get access to the switch contacts, and because there are contacts on both sides of the PCB, the ones on the bottom side being significantly more difficult to access. That said, with patience and diligence I was able to clean all the switch contacts, so I think you shouldn't have any trouble.

-- Jeff Dutky

I have just checked input resistances on a working-to-spec 7A26, the results are quite unexpected!. If I use an analogue meter and have the 7A26 powered off then in the 10mV position the DC 'resistance' from each BNC centre pin to its outer is about 350k measured in both polarities. For 20mV and higher the resistance is 1M. If I use a DMM then with either power on or power off the 'resistance' is 1M as it should be. Clearly Q150 needs to be powered in order to have its gate reverse biassed.

Even if Q150 is faulty, its output, together with the output from the 2X balance pot R160 form the two inputs to the differential amplifier U1350, so if U1350 onwards is all OK, the trace should move.

I have just noticed that there is another DC balance pot (R1353) between U1350 and U1450 so that can also be used to narrow down where the fault is. Since the two beams do move in response to the vertical position front panel control it looks like U1550 in each channel is OK so you can swap the good IC from one channel to the suspect position in the other channel and see what changes.

Regards,

Roger

Thanks for the guidance,

I suppose I have to take the cleaning challenge first as there are couple of suggestions for it.
As I didn't get any signal to R130...

I am also getting another unit to facilitate part swapping.

Regards,

Jouko

Hello,

I tried by swapping one Q150 from working unit. That brought input signal to screen.
I acquired two Q150 as both channels were similar and showed only 340k input resistance.
Those are now in place and partial calibration (not HF compensation) is done. The unit seems to be working.
Rise time is roughly 3ns in 7603 frame (Leo Bodnar pulser). The frame itself with 067-0587-01 calibration fixture showed 2ns rise time.
Does that sound logical?

Fortunately I didn't have to take the cleaning challenge...

Regards,
Jouko

Hi Jouko,

Rise times add approximately in root-sum-squared fashion (square each rise time, sum them, then take the square-root of the result). So if you have a 2ns rise time amplifier in cascade with another 2ns rise time amplifier, the combination will have a 2.8ns rise time or thereabouts.

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070

Thanks Tom,

Just repeated my measurements with the 7603 frame.
2.5ns is more correct value (timebase calibration issue with fastest sweep and 10X). This corresponds to 140MHz.
I measured also frame 3dB bandwidth and got 140MHz.

7603+7A26 combination gave 3ns rise time (correct timebase).
7603+7A26 3dB bandwidth was 100-110 MHz. These gave 3ns rise time (correct timebase).

I am happy with the results. Now I have 4ch 100MHz ”old school” scope in my lab.

Regards,
Jouko