My 7704A has died :(

[JPortici]

Seems like i've found my new weekend project. I was looking at wiggles on my tek 7704A and it shut down by itself.
After a couple of minutes i was able to power it up again, but apart from the initial flash on the CRT no traces, no readout.

(hint n°1: this fail has put the power supply in protection mode, probably a short of some type)

I already made a topic on my mainframe months ago but i never ended up fixing it, because i didn't know how to discharge the CRT.

-- I AM WELL AWARE OF THE RISKS ---
- How do i discharge the CRT in this particular scope?
- How do i safely check the acceleretaion voltage (21 kV)? There is a test point at what i think is the first tap of the diode multiplier, which is still 3kV. I don't have neither a multimeter nor a probe adequate for this measurement. Is there a workaround? I've seen people mentioning voltage dividers with high voltage resistors but...

[MadTux]

I had some similar problem a while ago with my 7904A:
https://www.eevblog.com/forum/testgear/i-have-an-old-tek-7904-that-power-suply-is-tic-tic-think-is-the-multipler/25/

In this case, it was a bad multiplier which I fixed with lots of duct tape. HV multiplier can be tested with modded audio amp (put a 230V=>12V transformer onto audio output in reverse and feed something like 1Vpp/20kHz into the amp and increase volume while measuring the HV multiplier output with normal 1000V/10MOhm DMM). The multiplier has a connector for the tube which can be removed by gently pulling on it. Simply stay away from the hot end and hold it onto the case to discharge the tube. Nothing really wild, because the tube is small and has low capacitance.

I tested the PSU rails ot ticking PSU with HP-3456As which have statistics mode. So you can recall highest/lowest reading. Since all rails usually reach nominal voltage before the PSU goes into ticking mode, it is likely that there is a short on some rail if it won't reach nominal voltage before ticking starts.

[mmagin]

I would point out that this kind of problem on the switching power supply 7000-series scopes can have nothing to do with the anode voltage multiplier output, so I wouldn't even mess with that initially.

My 7904's first failure was a shorted ceramic capacitor: https://www.eevblog.com/forum/repair/tek-7904-tick-tick-tick-caused-by-a-ceramic-capacitor!/

Some of the documentation (at least for the 7904) lists the resistance you should measure from each of the power rails to ground, that was helpful for debugging.

[tautech]

-- I AM WELL AWARE OF THE RISKS ---
- How do i discharge the CRT in this particular scope?
- How do i safely check the acceleretaion voltage (21 kV)? There is a test point at what i think is the first tap of the diode multiplier, which is still 3kV. I don't have neither a multimeter nor a probe adequate for this measurement. Is there a workaround? I've seen people mentioning voltage dividers with high voltage resistors but...

-Use a grounded screwdriver slipped carefully under the anode boot, do so a couple of times.
-Typically you don't, only check the -3KV EHT, one of the old AVO's are handy for this.
Normally you can disconnect the tripler/quadrupler/quintupler and see if the scope powers up without it.

[JPortici]

still

- How do i discharge the CRT in this particular scope?

i think the CRT is the same mounted on the 7904? even though the 7704 can conveniently be separated into display unit and power board/plugins bay area i'd want to start by enhancing security

thanks tautech, the problem is that i don't know where is the anode thing in this CRT

[guido]

Did you check for shorts in the low power supply lines? Last problem with mine was a shortened tantalum cap on one of the amp pcb's. Quick fix.

Other problems i had was a bad mains rectifier (then the fuse really goes bang) and problems with the resistors in the high voltage part (arc over).

[David Hess]

Seems like i've found my new weekend project. I was looking at wiggles on my tek 7704A and it shut down by itself.

Is this a 7704 or 7704A?  They are very different oscilloscopes with the 7704 having the first *discrete* implementation of the 7000 series switching power supply and the 7704A and 7904 which are contemporary having the first *integrated* version of the 7000 series power supply.

After a couple of minutes i was able to power it up again, but apart from the initial flash on the CRT no traces, no readout.

(hint n°1: this fail has put the power supply in protection mode, probably a short of some type)

I already made a topic on my mainframe months ago but i never ended up fixing it, because i didn't know how to discharge the CRT.

-- I AM WELL AWARE OF THE RISKS ---
- How do i discharge the CRT in this particular scope?

The time constant for the CRT PDA (post deflection acceleration) voltage is 10s of seconds so just wait a few minutes.  If you are paranoid about it which is a good idea, then pull the high voltage connector straight apart and touch the exposed pin to the chassis.

- How do i safely check the acceleration voltage (21 kV)? There is a test point at what i think is the first tap of the diode multiplier, which is still 3kV. I don't have neither a multimeter nor a probe adequate for this measurement. Is there a workaround? I've seen people mentioning voltage dividers with high voltage resistors but...

You mentioned the power supply being in protection mode above.  If it is in tick mode, then there will not be any useful CRT voltages to measure.

The CRT will work without the PDA voltage being present; it will be dimmer, less focused, and the deflection will be about 1/2 of what it is suppose to be.

Is the power supply working at all?  Check the low voltage outputs to be sure before moving on to the high voltage sections.

[JPortici]

clarifications:
- 7704A
- I want to keep being paranoid. I get too many beatings from snowboarding already, if i can avoid being zapped...
- Protection.. this is what happened: was working, then it was not. all lights off.
pushing and pulling the switch didn't bring on any lights.
switch pushed in off position, some 10 seconds later removed all plugins, pulled switch. no lights. pushed, removed power cord.
waited five minutes, plugged in power cord, pulled switch. lights come on (graticule illumination, pushbuttons). Green spot on crt at powr on, no more green from CRT.
There was no ticking. i talked of protection because i tought of ATX psus that shuts down for some minutes after an overcurrent

I did notice that left timebase didn't trigger from vertical source (but would from line. didn't use external so don't know) some minutes before the failure

[David Hess]

clarifications:
- 7704A
- I want to keep being paranoid. I get too many beatings from snowboarding already, if i can avoid being zapped...

There is nothing wrong with that.  The CRT PDA time constant is short enough that by the time I get everything setup after turning the power off, the CRT anode is discharged anyway but I still pull the high voltage connector apart and short it to the chassis.

At this point however I do not think you need to mess with the CRT PDA or circuits.  There are other things to check.

- Protection.. this is what happened: was working, then it was not. all lights off.
pushing and pulling the switch didn't bring on any lights.
switch pushed in off position, some 10 seconds later removed all plugins, pulled switch. no lights. pushed, removed power cord.
waited five minutes, plugged in power cord, pulled switch. lights come on (graticule illumination, pushbuttons). Green spot on CRT at power on, no more green from CRT.
There was no ticking. i talked of protection because I thought of ATX psus that shuts down for some minutes after an overcurrent

The mainframe requires an amplifier or timebase plug-in to be installed into a selected horizontal slot to unblank the CRT.

I did notice that left timebase didn't trigger from vertical source (but would from line. didn't use external so don't know) some minutes before the failure.

How familiar are you with operating these oscilloscopes?  The triggering configuration can be tricky.

I would start with the basics.  Remove all plug-ins, apply power, and check the various supply voltages for level and ripple.

Then install a vertical amplifier into one horizontal slot, power it on, select that slot, and see if the beam unblanks.  Then test the other horizontal slot.