HP 1740A on the bench, help please?

[Specmaster]

I've got a bulk lot of transistors as a experimenters kit where you typically get something like 50 of a type, so I probably have about 1,000 small power / signal T0-92 type transistors

[Specmaster]

The BOM shows the HP part number for Q2 as being 1853-0066 (PNP SI PD=200mw) and that comes as being an equivalent to a 2N3906 (I have plenty) which translates to a NTE159 which has a PD of upto 625mw.

Seems to fit the bill according to the data sheet, lets give it a burl shall we?

[bd139]

2n3906 is the part that came to mind looking at the schematic. Go for it.

[Specmaster]

Q2 replaced, no difference. the volts are the same and no oscillator running.

[bd139]

Ok that eliminated then! Is anything happening on the feedback winding of the transformer?

[Specmaster]

In what way do you mean, the voltage I'm reading on the feedback winding before CR1 is 7.9v and after its 7.26 which is also being seen on the base of Q3 and collector of Q1.
Is the feedback circuit shutting down Q3 to stop it oscillating? Looking at circuit, I take it that if the voltage on the base of Q3 that normal service should return?

[Specmaster]

The feedback circuit should only kick in if the 120v rail drops to 100v, but it currently sits at 122v

[bd139]

Is it at 122v when it fails?

Unfortunately had to take SWMBO out this evening so haven’t had time to play or look at this

[Specmaster]

Yes it is at 122v even when it fails (I take you mean, no HT)

Gone back to basics again as still not getting any HT despite the voltages supplying the board are well within their spec and also replacing Q2 to no avail.

Rail voltages on the main power supply, again all are well within the spec.

Started checking voltages on the pre-amplifier and input attenuation board and the supplies to the board are again within spec, but however the +5v rail is playing up. At TP2 I'm getting +5v but at TP6 +0.5v and at TP5 its +2.8v. previously these were all within spec. 

Will re-look at this area tomorrow but surely should have no effect on the HV not working or vice versa? 

[Specmaster]

Pre amp board voltages all look normal, after due consideration it was down to various switch positions and operating these switches restored the readings, so that was a rabbit hole that I followed far too long. However, as I always do as I go around checking, I always ensure that the plugs etc are firmly plugged in and I was noticing that the -15v rail was disappearing, even from the source, the power supply was reading as low -1v. Off I go again and discovered that the A14 bridge board was the culprit, somehow shorting out?

Backing this plug board off a bit from the boards it was connecting to restored the supply again??

Now once again all voltages are back to spec but not the HV, normal supply volts are still within the spec but can't get the oscillator to kick in again??

[Specmaster]

If I pull the HV board from the parts mule and swap them over, there's nothing to be worried about with regard to charged caps etc on the faulty board is there?

Its been disconnected for a couple of hours and beside the oscillator is not running so surely in reality they should discharged by now?

I've never done anything before that was higher than 415v so I'm just being cautious.

[bd139]

It’s worth shorting both ends of each of the Hv caps then each lead to the chassis. The connection between the multiplier and the board is the most likely to bite you.  While it most likely won’t kill you, it’s not something particularly fun

I do this, on tek scopes anyway, with a 4mm banana lead plugged into the earth hole in the front.

[Specmaster]

Ok, I chickened out on that for the time being, instead I have been trying to work out just how this oscillator is supposed to function and what starts it oscillating and what voltages on the base of Q3 enable the HT or disable the HT output and the how the OP amp interacts with it. I have to say regrettably I can't get my head around it, I'm still really sort of learning this again, it was about 50 years ago when I last did electronics so I've forgotten a lot. 

[bd139]

It’s very difficult to understand so I don’t blame you. I tend to extract bits of the circuits and put them in LTspice and see why they do what they do and divide the whole thing into little blocks. This takes hours. If you look in the picture of the bench above you will see about two hours worth of scribbling and thinking trying to work out something. To the left is a pile of paper stuffed under everything (like Bob Pease’s office) which contains scribblings

Oh also beer. Beer helps!

Not much help that but never give up

[Specmaster]

Yes I did notice, I have LTspice but yet to learn that as well  I have to many things on the go at anyone time to find the time but I guess I'm going to have to make some time.  That is how I thought it would be be of assistance to me by drawing up a circuit and then run it and observe voltages and waveforms etc without having to breadboard everything first. Not used my breadboards yet either 

Anyway, some good news late last night I popped the question about HV oscillator on the HP/Agilent group and someone confirmed that the voltages on the transistors is correct apart from the 7.2v on the base of the HV OSC Q1 which should be less then 1v. They reckoned that the socket on that transistor may not be making proper contact, hence the high voltage. I removed it and dosed it with contact cleaner, plugged it back together and unplugged, repeated it a few times to clean the pins and tried it. The oscillator is back running again and the traces are back on the screen. 

So now I can crack on with the calibration process and correct some of the biases that are present on some ranges 

Thank you for your sterling efforts and lovely tips, you never know, you might make an electronic engineer out of me yet 

Its now 2.20am so I'm off to bed. 

[Specmaster]

Trace on this is not correct, the total length is about 2 to 2.5 times the width of the screen on all ranges and if you pump a square wave into it and get it to display what you think is one cycle, move the horizontal position to the right and the left side is twice the length of the right side (when moving trace to the left on the horizontal position control) is there an internal control for this somewhere? Strikes me that HP seem to use a language all of their own, I mean since when has LINE meant power switch or ON/OFF??

I'll post a vid of this later if that helps to describe what I'm talking about, they say a picture says a 1,000 words, lets see how many a vid says shall we? 

[bd139]

Interesting. Possibly horizontal amplifier balance or problems with the 10x magnification. Definitely post a video.

LINE is American terminology. Everything I have is labelled POWER or just ON OFF.

[Specmaster]

Its nothing to do with the x10, on that time base setting all I get is one long half of a square wave if I press the x10. Yeah, the americans do call power, line and live, hot etc so I was wondering if they had labelled any of the internal adjustment trimmers by ambiguous names as well 

After dinner I'll take a vid and post it via a link to my flickr account.

[Specmaster]

Here is the link to the video which shows the odd width of the horizontal trace and the linearity of it and you can also see some of the flyback lines at times. On some of the slower sweeps they can be seen to double back behind the main trace, but this I think is due to the massive length of the trace in the first instance or I have I gotten that wrong??

https://www.flickr.com/photos/158191416@N07/24056629268/in/dateposted/
 

[bd139]

Interesting. Id concentrate on the fly back / blanking first. That might yield some other clues.

Are very slow speed sweeps linear or do they slow up or speed down during the sweep?

[Specmaster]

https://www.flickr.com/photos/158191416@N07/26134122199/in/dateposted/

This shows the slower sweeps and I don't think it varies during a sweep but it does show the flyback very clearly.

I have not replaced a single cap on this as yet, do think it is likely to a be high ESR or something else altogether.

All I have done so far is extensively clean the timebase switches, (stripped them down to parts, cleaned lubricated and reassembled, replaced those 2 burned resistors on the vertical amp output, repaired the +3V rectifier connections and the plug connection as well, replaced the 3 transistors and cleaned up the socket on the HV oscillator transistor.

Checked all caps on the beast for shorts to ground, also tested the ESR of power filter caps, remade and cleaned all connections at the plug / socket joints in case of bad connections, replaced the timebase control shafts and generally cleaned the whole scope up as it was a bit on the grubby side when I got it.

[bd139]

Ok definitely hit blanking next. The time distortion might be an artefact of that. I think blanking is the most common scope breakage so fingers crossed should be relatively ok to debug.

[Specmaster]

Thanks, will do.

Looking at my scopes tonight, its easy to see how the inputs can get toasted as they don't all have the same capabilities, the Hitachi V-252, Rapid 7025A and the Goldstar OS-9100D only go to 5V per division, the Iwatsu SS-5710 goes to 10V and the HP 1740A up to 20V. So anyone used to working with a 1740A at high voltages who then has to jump onto any of the others could end up overloading the input stages by 2 or 4 times. There ought to be a standard for these things?

[Specmaster]

Thanks, will do.

Looking at my scopes tonight, its easy to see how the inputs can get toasted as they don't all have the same capabilities, the Hitachi V-252, Rapid 7025A and the Goldstar OS-9100D only go to 5V per division, the Iwatsu SS-5710 goes to 10V and the HP 1740A up to 20V. So anyone used to working with a 1740A at high voltages who then has to jump onto any of the others could end up overloading the input stages by 2 or 4 times. There ought to be a standard for these things?

[bd139]

It's possible that can happen easily. The Tektronix 465/475 have thin film attenuator modules which get toasted pretty quickly in this circumstance. Also difficult to get replacements for!

I'm slightly scared every time I put my Rigol near anything more than 100v as they are bastards to repair!