Tek 465M dead again.

[andy3055]

This is the first repair thread: https://www.eevblog.com/forum/repair/tektronix-465m-repair-210933/

My scope died again! The dreaded fuse F558 started blowing up. After trial and error I have found that if the CRT anode is disconnected, the fuse will not blow. I know that the EHT multiplier is good as it is generating the 10kV anode voltage-checked with an HV probe.

I got another 465M which had a different problem (some weird display issue) and I suspected the horizontal module in it. So, I swapped the module from the 1st scope above and that ended up having the same issue of blowing the F558 fuse! The "virus" got transferred to the 2nd scope also 

It seems that the horizontal module in my 1st scope had done some sort of damage on the main board which comes in to play only if the anode is connected. I cannot think what would cause such a condition and any help is much appreciated.

Thanks in advance.

[med6753]

Is F558 the HV Fuse? If yes, is there a 47uf tantalum capacitor downstream from it and before the oscillator? If yes, check that capacitor.

While I'm not familiar with the 465M this is a common fail point on 465/465B.

[andy3055]

Hi Med,

Nice to hear from you. I have attached an image of the relevant page of the s/manual. You are right it is the fuse supplying 32V to the HV oscillator and in this case, it is an 8.2uF cap which is already replaced.

The issue is, it does not blow unless I connect the HV anode lead. All voltages are fine till then. I suspected the tube and even with a known good tube, it does the same. It happened in my 1st 645M and by swapping its horizontal module to the 2nd scope, it is happening in that too! To me, it appears that there is something in the horizontal module that has triggered this. Strange thing is that if I remove the HV plug, even with the horizontal module in place, the fuse will not blow! The vertical amp module has no problem at all.

The SM is here: https://w140.com/tekwiki/wiki/465M

If you can spare a few minutes and give me some pointers, I will appreciate it very much.

Kind regards.

Edit: Once this starts happening, the fuse blows when the HV plug is connected, even if the horizontal module is not in. I have ruled out the HV block and the EHT section of the HV block. The U550 in this case has the tripler as well as the rest of the HV components in one block. It is almost as if the Horizontal module has done some damage on the main board that comes in to play only if the HV plug is attached!
Thanks

[floobydust]

A high voltage multiplier can fail where it outputs a large amount of AC, it fails to fully rectify, so the CRT (capacitance) then appears as a low impedance and looks like a heavy load to the inverter. When the CRT is disconnected, there is no AC load and things can look OK. I think it's if a diode shorts at the multiplier's output, contrasted with a diode shorting at the multiplier's input which always overloads the inverter (unless the GND wire is lifted). When you checked the 10kV output, did you look to see if it's all DC or has AC present? Isn't that high with 2kVx3 or +6kV sort of expected there, total 8kV for the CRT.
I've mentioned you can connect a signal generator to the HV module U550 and inject 20VRMS at say 50kHz and scope the output waveform. It should still triple, assuming the multiplier diodes' high Vf is taken into account. Same for the other diodes inside, you can inject a test signal to see what comes out.
I'm not sure what the "horizontal module" is, the entire A11 board?

[jdragoset]

A high voltage multiplier can fail where it outputs a large amount of AC, it fails to fully rectify, so the CRT (capacitance) then appears as a low impedance and looks like a heavy load to the inverter. When the CRT is disconnected, there is no AC load and things can look OK. I think it's if a diode shorts at the multiplier's output, contrasted with a diode shorting at the multiplier's input which always overloads the inverter (unless the GND wire is lifted). When you checked the 10kV output, did you look to see if it's all DC or has AC present? Isn't that high with 2kVx3 or +6kV sort of expected there, total 8kV for the CRT.
I've mentioned you can connect a signal generator to the HV module U550 and inject 20VRMS at say 50kHz and scope the output waveform. It should still triple, assuming the multiplier diodes' high Vf is taken into account. Same for the other diodes inside, you can inject a test signal to see what comes out.
I'm not sure what the "horizontal module" is, the entire A11 board?

Seems this could be  qick-tested with an HV diode from U550 output  to the CRT anode terminal

[med6753]

Check the HV ceramic capacitors. Although rare I have seen one go leaky (9M when measured with a DMM) and that can kill the HV once you put an additional load on the circuit by connecting the CRT. And of course the tripler itself can have a partial short.

I've seen one CRT develop a very high leakage path to ground which in turn would kill the HV. But that was in a 7904.

[andy3055]

Thanks for the answers gentlemen. To rule out the HV tripler section, I rigged up an external tripler with 3 caps and 3 diodes. The HV voltage in the 465M is 10kV. And the rigged up tripler generates it. As for the rest of the caps and diodes in that U550, I had the working 365 (which had a working U550) also develop the same problem after I swapped the time base/horizontal. Yes, it is one big single unit with a couple of daughter boards on it. 465M has 4 major parts as I see it. The vertical, time base, power supply with the HV components on the main board and the tube assembly. So, when I moved the “parts” around, it was always a complete unit in the case of vertical and time base.

Right now, with the vertical and the time base unplugged, the fuse still blows if I connect the HV lead to the tube. I have also tried 3 separate tubes. This is the reason I get the feeling that the problem is in the main board (in addition to the time base that caused it-maybe). Like Med says, I may have to fabricate a new U550 as I did earlier (my previous thread) but I am not sure if that will resolve the problem. I say this because, the 2nd unit had a working U550 and that unit developed the same problem after I swapped the board. Unlike the 465, the 465M has an “integrated” HV module and the last time around it took me a while to rig the thing.

Thanks again.

[floobydust]

It might be a distraction then that with the CRT disconnected things appear to run, the lighter load lets the inverter run in the usual range and not blow fuses. If, as you are saying, the HV multiplier checks out OK.

There might be a problem with regulation. If the -2kV rail is weak or does not come up, the inverter would then stay full tilt with Q544/Q548 full on. I would measure the -2kV rail as it should be regulating. If the HV is too high, maybe damage the +95V rail capacitors or other parts.

The -ve rectifier diode on HV module you can test (pin 5-6) or with the two 10k resistors (pin 6-9). The HV blanking diodes, you can also check (pin 4-8), (pin 1-2 but should not see any HV unless cap failed).
Use a high voltage say 20-30VDC source and 10k series resistor or something to limit current to a few mA, to test the diodes in the blob. The CRT cathode-grid neons are staying off?

[andy3055]

Thanks for the reply floobydust. The Neons are off and they do come on momentarily when the power is switched off.

I have not yet tested the rest as you mentioned. Especially the -2kV. I am still confused as to how the same time base can cause the identical fault in both scopes. But I may be jumping all over the case here. Need to concentrate on one unit so that it is not a confusion for myself!

[andy3055]

Update:

I checked the -2kV supply and that is good and steady.

Meanwhile I was able to find a main board and a time base/horizontal module from the folks at the TEK museum up in Oregon. Cleaned it up and installed both using my old Vertical module and now all seems to work well. The only thing I noticed is that the display blanks for a fraction of a second every now and then. Obviously, that should not happen. Made sure nothing is overheating on the HV side. Next thing to do is to monitor the LV supplies and see if the glitch coincides with any power supply dip as well.

IMO, this exercise confirms that the original time base did some damage to the original main board, which remains a mystery as I do not have any thoughts as to what that could be.

Any ideas are most welcome.

Thanks in advance.