HP1715A repair

[WaveyDipole]

I was hoping to be able to solve this one on my own, but I have run into a bit of a puzzle. The scope was purchased from a member recently with a known fault. It does have a trace, but seen only when you press the beam finder, otherwise it seems to sit off the top of the screen. Sometimes if you tunr the scope off and then on again then you can see a soft trace starting from the left hand side and disappearing rapidly off the top edge.

I have checked all LV lines and they are all spot on, except the unregulated 20v line with is reading 22.3v, on the high side but probably normal for an unregulated line. I also checked the HV line and I got a reading of -2900v, just a tad short of the -2950 it should be.

Since all line voltages se3emed OK, I than had a look at the Y amplifier and found and imbalance in the output voltages, which should be +40v either side, but I am getting +30.5v and +49.5v respectively. At the input to the output transistor array (U2) I get 29.2v and 25.8v instead of what I think should be 28v either side. On the input side, there is a 0.5v difference between the voltage on pin 3 (13.5v) and pin 6 (14.0v). All other voltages around the U1 IC seem pretty close to what they should be. Adjusting the balance pot has no effect. The 53.3v line has the correct voltage and the two film resistors read 164.1 and 164.2 ohms so not too far off the 165ohms they are supposed to be.

I've been scratching my head over this imbalance for a while and eventually it occurred to me to take some readings with the CRT disconnected. If I disconnect the +53.3v supply to the plates via the two 165ohm resistors, or the plates themselves from the amplifier output, then I get equilibrium on the output of both U1 and U2 with 19.1v and 19.2v respectively on the amp output. I also get 53.5v and 53.6v respectively on the pins leading to the CRT Y plates. Reconnecting the CRT causes the imbalance again.

So my question now is, what is causing the imbalance? Is this a problem with the CRT Y plates or the output stage U2? Is there any test I can do to rule out one or the other?

[WaveyDipole]

PS, I have checked the resistance between pin 2-7 and 3-6 on U2 in both directions and it is 986Ω across both paths so it doesn't seem like there are any output transistor shorts. Resistance readings between pins 6-5 and 7-5 are inconclusive due to being in circuit, but are 770Ω in one direction and 850Ω in the other, exactly the same for both sides. I also forgot to mention that resistors R24 and R25 read 71.7Ω and 72.0Ω, so a little less than the 75Ω that they should be.

[Radio Wrangler]

without the schematic in front of me I can only generalise.

With care, you should be able to short the differential signals together at some points in the Y amplifier chain, and if the subsequent stages and the crt are good you should get a fairly central trace.

working the other way, with the input grounded and the shift pot central, you should see reasonably balanced voltages down the chain.

Does the beam switch CH1/CH2/Chop/ALT do anything?

David

[WaveyDipole]

Y-amp circuit uploaded although compression seems to have made it less readable than it already was.

Originally Posted by Radio Wrangler

without the schematic in front of me I can only generalise.
With care, you should be able to short the differential signals together at some points in the Y amplifier chain, and if the subsequent stages and the crt are good you should get a fairly central trace.

I had wondered whether it could be tested that way, for example by shorting the two inputs to U1 or into U2, but I didn't want to risk it without a clearer understanding of exactly where to do it. As a quick and dirty test I shorted the inputs to the the Y-amp board since they are close together and easy to do. This does give me an almost spot on central trace, however voltages on the plates are now 40.1v and 44.5v respectively so still not balanced. The trace itself seems to be more intense to the right hand side unless shorter timebase intervals are selected.

Originally Posted by Radio Wrangler

working the other way, with the input grounded and the shift pot central, you should see reasonably balanced voltages down the chain.

Does the beam switch CH1/CH2/Chop/ALT do anything?
David

Unfortunately, Chop/Alt do nothing. CH1/CH2 do make some visible changes on certain settings, but only when the beam finder is pressed so its difficult to know exactly what its doing, otherwise the trace stays off the screen.

[WaveyDipole]

A closer up view of the circuit attached. Hopefully these will be a bit more legible.

[Radio Wrangler]

This scope has 'distributed' Y plates... a lot of little ones distributed along a solenoidal inductor and the amplifier's pull-up resistors terminate the simulated transmission line.

You can disconnect the connections from the amplifier to the input of each plate line on the tube. Leave the load resistors and add resistors to ground from the tube plate inputs. Ohms law and the terminating resistor and supply values will give the resistor values needed. This should allow you to put the correct voltage on each plate group, and should give a central line. You can play with resistor values to move the line around and check the sensitivity and linearity if you're really suspicious of the tube. This should exonerate either the tube or the amp.

David

[WaveyDipole]

Yes, adding resistors to make a voltage divider made sense. My calculaton came out at about 496Ω. Since the nearest I had to hand was 560Ω I used those. This gave me exactly 41.1v on each plate, but the line was 2 minor marks below centre. I substituted one of the resistors with a 1kΩ. This moved the line 1.5 graticule lines up from centre and plate voltage rose to 45.7v. Swapping them around moved the trace about 2.1 graticule lines down but plate voltage was still 45.7v although now of course on the other plate. Even taking into account the 0.2 shift, it seems the trace moved at least 0.2 further downwards than upwards. I measured both 560Ω resistors and both gave exactly the same resistance reading. This is not a huge amount off and I'm not sure whether that would be normal for the age of the CRT or not, but certainly the trace is a little lower than it should be and the deflection does not seem quite linear. I also monitored the current being drawn which started at about 12mA but drops quite rapidly and settles at about 0.08mA once the CRT gets warmed up and the trace appears. Voltage appears to stay constant. The same result for both plates.

I am not sure what this means just yet, but I'm inclined to think that although the trace is slightly low, the plates are otherwise behaving exactly the same regarding current draw and resulting voltages across the resistors so they are probably OK. Since shorting the two inputs to the amp board resulted in an exactly central trace, I'm inclined to think that the amp itself is probably OK although I still don't understand the 5v difference between the two plates. The service manual indicates that there should be 40v both sides.

[Radio Wrangler]

I agree.

The asymmetry in deflection seems a bit worrying. CRTs need to be quite linear to do their job.

Unfortunately the crt and the custom ICs have all been unobtainable for a long time. I'm not sure what to suggest next.

David

[WaveyDipole]

UPDATE - I performed two additional tests. I shorted pins 5 & 7 on U2. Output voltages on U2 were even at 40.4v on the both outputs however the trace was offset below the center line. I removed that link and then shorted the two inputs (i.e. anodes of CR3 & CR4) as I had done quickly last night, but this time I checked the voltages through the amp. The voltages around U1 were spot on and the two channels were perfectly in balance. I then found that it was possible to adjust BAL so that both outputs of U2 have equal voltages, but the trace would be shifted below the center line again. Adjusting BAL back so that the trace was dead center showed the same variance between the plates (about 5v). So it would seem that this 5v differential is required in order to offset the trace so that it is correctly centered. Whether this is within sensible limits I don't know, but certainly comfortably within the adjustment range.

I would say from this that the Y-amp is working normally although as you say, a bit worrying that such a voltage differential on the plates is required in order to center the trace correctly. The additional offset however, that is sending the trace off the screen, must now be down to the input signal and hence tracks back to the Y pre-amp module. Here I think I am going to run into a bit of difficulty because the circuit diagram for this stage is legible only with extreme difficulty and I will need to order a better resolution manual from Artek media, unless perhaps someone can help me out with a decent image of figure 8.6, page 8-19 which would be appreciated.

PS, I noticed that the two input boards are different - similar but with different transistors and diodes on them. A trimmer has just fallen off one of them - it wasn't soldered on very well - looks like probably a modification , so it seems 'its been got at'....

[Radio Wrangler]

Stupid thought.... have you checked for any residual magnetism around the tube?

David

[WaveyDipole]

No, that's one thing I haven't done. What's the best way to do that?

In the meantime I've been looking at the pre-amp section and am slowly working my way through the circuit. There is an IC (shown as 'SUBSTRATE ASSY' on the diagram) the output of which drives two pairs of transistors. One pair drives the Y amp. The pair of transistors marked Q3 and Q4 were showing different base and collector voltages. The collector voltage difference was significant: 7.0v versus 14.1v. The transistors were checked out of circuit and found to be OK. I put them back in but left the bases disconnected and the output voltages are now pretty close and the trace was dead centre.

I checked the output of the IC and found one side gave a reading of 6.7v and the other 8.0v. There are differences in the channel input voltages as well though and there is other supporting circuitry to check so I'm not concluding that its a faulty IC just yet, but the lack of clarity in the diagram will make further work rather difficult.

I also came across a problem in the timebase, which was behaving rather erratically. An electrolytic capacitor had leaked its electrolyte and this had corroded away one of its leads and some of the track on the PCB. The track has been carefully bridged and the capacitor replaced, but the control continues to behave rather erratically. I have tried a little Servisol on the switches but to no avail. They may need to be dismantled, checked and cleaned. The trace disappears altogether on some settings, is very faint on others and looks fine on still others, although without a signal to look at, I can't tell whether the timing intervals are correct.

I think there is still much work to do, but pre-amp fault cannot be fixed then I would be wasting my time fixing anything else so i will be concentrating on that for now. If it turns out to be a faulty IC, then I guess that will be the end of the project!

[Radio Wrangler]

Just search around with a little compass. It will generally attract itelf to ferrour material, so you're looking for anything which seems to be a stronger effect.

David

[WaveyDipole]

I knew I had one somewhere! I dug it out this morning and had a look. The neck of the tube seems OK but as you move the compass towards the front the compass aligns itself with the length of the tube pointing towards the screen just as I move it past the point where the tube starts to widen. From the front, the compass points in the opposite direction towards that same narrowing point. It would seem that there is some magnetism around the deflection plate area, but wouldn't that be expected due to the magnetic field produced by the deflection coils? I did have the scope turned on shortly before I checked it, but I left it turned off for half hour or so and checked again with similar result.

[Radio Wrangler]

Deflection coils?

There is a small solenoudal winding around the whole tube - axis aligned with the axis of the tube. The scope puts a little current into this (adjustable from a little one way through zero to a little the other way) to rotate the deflection of the beam to make the X and Y deflections agree with the graticule engraved inside the glass. It's called 'Trace rotation' (Tek amused me by having a tweaker accessible from outside labelled 'Trace Rot')

David

[WaveyDipole]

Apologies, I'm probably thinking in terms of a TV CRT, although the 'Trace Rot' control was funny! Whatever the source of the magnetism is, the CRT has been left turned off for a couple of hours and its still there.

Does anyone know whether there is anywhere to look up the part numbers on some components in this scope?

There seems to be two types. The first is in the form of two groups of digits separated by a hypen, i.e. 9999-9999 which I take to be HP numbers. The second starts with a 4, or less commonly with a 3 and then has two groups of digits separated by a hypen. Sometimes there are 3 and sometimes 4 digits in a group, i.e. like 4-999-999 or 3-9999-999. I'm not sure what these references are so I'm not sure what kind of reference guide to look for. I have found a couple of partial HP references but they do not seem to be terribly helpful as they are PDF files of scans with lists that are not in any particular order and they cannot be searched.

One particular part I am looking for is 1826-0553, which I know is an op amp and looks like a metal can 741 in disguise... but I would like to confirm. I am also looking for a transistor marked 4-546-924.

[Radio Wrangler]

4digit dash 4digit numbers are HP stock numbers for purchased parts

5digit dash 5 digit numbers are HP fabricated parts

1826-4 digits is a linear IC, purchased the second lot of 4 digits are chronological with no other meaning.

A transistor with 4-546 will be shortened from 1854-0546 and the 924 is likely to be someone's datecode, whoever made it

1854- is Si NPN
1853- is Si PNP
1855- is Si FET
1820- is digital IC
1901- is diode
1902- is zener
0122- is varactor

01715- will be fabricated parts or documents done for that product.

DAvid

[PJL]

Is the trigger in auto and does it appear to be triggered.
Enable Channel A only. does Position A control change the plate voltages?
Switch to B only and try the Position B control.

I must have the same document as you as it is near impossible to read. The 'A1 SUBSTRATE ASSEMBLY' chip performs the vertical switching. This is when you need a working scope to fix a scope!

Logically, we should be able to ignore everything before this chip as the fault exists on both channels. It might be worth testing that the channel A input does change when it is set to DC and you put a DC voltage on the input. You could also check the offset input voltage changes when you adjust the pot.

There are two connections at the bottom of the schematic that enable the channels. Try measuring these when you switch the A and B display on/off.

The problem with a scope is it only needs to have the correct plate voltages on it when it is displaying the line which can mean the average voltages are meaningless and another scope is your only option.

[Humber888]

1826-0553 crosses to a PMI OP08GJ if I read the slightly fuzzy x-ref documents correctly. I can't post it here as it's 22MB but is available on the web somewhere.

[WaveyDipole]

David, thank you for the useful information on those HP part numbers.

Originally Posted by PJL

Is the trigger in auto and does it appear to be triggered.
Enable Channel A only. does Position A control change the plate voltages?
Switch to B only and try the Position B control.

Yes, auto is on. I had to disconnect the output from the pre-amp (bases of Q3 and Q4) to get the trace to stay in the centre of the display - otherwise the trace is completely off the display and cannot be adjusted back to center. Since no signal is evident - only a line - unfortunately I cannot say whether there is any triggering. Presumably though, if the Y amp is not being driven properly, wouldn't I get only a straight line?

Originally Posted by PJL

Logically, we should be able to ignore everything before this chip as the fault exists on both channels. It might be worth testing that the channel A input does change when it is set to DC and you put a DC voltage on the input. You could also check the offset input voltage changes when you adjust the pot.

I tend to agree and nothing happens on any input, AC or DC. Just to be sure, I disconnected the 3 wires from each attenuator board from the pre-amp board so they are no longer in the picture but the problem persists. The pots do seem to operate correctly and the offset voltage on the wiper changes when I adjust the pot. Both change from 15v to -15v. I also checked the series resistors between wiper and IC and both are OK.

Originally Posted by PJL

There are two connections at the bottom of the schematic that enable the channels. Try measuring these when you switch the A and B display on/off.

I believe I understand which ones you mean - marked 14 and 11, but I can't determine where the switches connect to the pre-amp board. What I can say is that I can see voltage changes on some pins of the logic ICs on the switching board - 3.40v to 4.56v. Is this sufficient to indicate a state change?

BTW, I do have another scope available and the next step will be to see how far I can follow the signal. The only thing I have determined so far is that both channel input/attenuator boards seem to be OK and are passing a signal to the pre-amp board. Also, that 'A1 SUBSTRATE ASSEMBLY' chip is running VERY hot. I can keep my finger on the Y-amp output IC which I expect to run hot, but not on this one. I would not have expected anything to run that hot in the pre-amp stage?

[PJL]

It will contain differential amplifiers, analog switches and a mixer.

The 13/14 links go to a couple of transistors on the Display Trigger Switches circuit a few pages down. That implies that 6.5V is needed to turn it on but I am mostly guessing. There are some pull-up resistors drawn next to the chip (200 ohms?) but I can't read what they are connected to but I guess it should be 6.5V.

If you have a scope, I would have a look at the 13/14 control lines.