Gould OS300 blurred trace

[TonyPhs]

Hi guys.

I recently bought a Gould OS300 oscilloscope off someone and it worked for two days before developing an intermittent fault where the trace totally blurred out. The trace is there but the colour is basically spread across the entire screen. Occasionally the trace would come back and then go again.

I've tried tapping a few components but that has no effect. The time that the trace comes on now is basically zero.

I've whipped off the CRT connector and had a probe with my high voltage probe and found the cathode and associated voltages are fine at -1800ish.

The one thing I'm unsure of, and this is because I don't really know much about the operation of scopes, is that the 1 X plate measures 50v while the other measures about 200v.

The Y plates seem to both measure around 112v. Rail for the plates is +210v I believe.... Edit: I just noticed that when I twiddle the vertical adjustment the Y voltages vary between roughly 60v and 180v, as I adjust one up the other comes down.

Now I'm guessing there needs to be a difference between the plates to actually have an effect on an electron but is that difference normally small or large?

Any help is much appreciated.

That's two working scopes that seem to have broken on me now. Learning a lot but these things are way more complicated than anything I've tried to understand before.

Edit2: The xplate voltage hardly varies at all. It's like it is fixed at the two voltages. Would this cause the blur I am seeing?

[maninashed]

Hi Tony, what I think you are seeing when you say the trace is totally blurred out, is that the trace is hard off screen to the left or right, and the illumination is a reflection off the side of the tube. From your description, it sounds like you have a problem with the X amp. Normally, when the spot is at the centre of the screen, the X plates are at a potential of about 50v. The X amp is basically a push-pull amplifier, you have an imbalance, so check that you have correct voltages from the psu, (there will be + and - rails) check you have all correct, check the o/p transistors, and their collector resistors for o/c.
You need a circuit diagram to work on scopes, make that a priority if you don't have one.
Bill

[TonyPhs]

Thanks Maninashed.

I do have the manual and schematic of the scope so that's not a problem. I just needed an idea of where to start my troubleshooting. I'm still very inexperienced with this sort of thing, despite having spent many hours trying to fix, and failing so far, an EHT section in another Gould scope.

What you are saying makes sense though with the reflection and the voltages I saw.

First thing I will do is replace the reservoir caps as they look pretty old. It makes some sense to me that the caps may have been idle for years with the previous owner and since being used this weekend have warmed up and now have high resistance. I may not have the exact capacitance 22uF in a 300v cap but I do have a 47uF 400v cap. The low voltage one is a 470uF 50v which should be easily replaceable.

I did check the rails this morning and they all seemed to be within the tolerance laid down in the manual, all apart from the +7v which measured 0.05v outside the upper tolerance. Maybe the capacitor change would help that.

I also checked some of the points on the ICs and they were good.

Obviously these tests are done with a DMM so there may be nasty ripple. I do have a dirt cheap USB scope but I only got limited success getting it to work.

[cmjones01]

I'm pretty sure that replacing the power supply capacitors isn't going to help. It sounds like the spot is simply off to one side of the screen, as Bill said.

Does turning the horizontal position control make any difference? What about if you select X-Y mode? What about if you press the 'Bright Line' button, which should force the horizontal sweep to do something?

A look at the circuit diagram will help - I'm sure we can track down what's wrong.

Chris

[TonyPhs]

Hi Chris,

I'm sure it is off to the side now that this has been mentioned but it was the intermittency of the trace that lasted for a few hours before it went completely that made me think of the capacitors.

The attached is the relevant section of the schematic, without values of course as that's elsewhere.

A few things I noticed yesterday while doing my 'looking for glaringly obvious things' check.

Terminal/connection to the board at the X out location, where the two wires go off to the CRT connector appeared corroded and had flux deposits.

With the bowing in the case body the high voltage transistors at the x output, TR513 and TR514 (NSD459) were perilously close to the case. The tab of these carries a good voltage so maybe one or both actually shorted to the case and this gradually broke down the transistor. They aren't showing a short though.

Turning the horizontal makes very little difference. Approx 4-5v from full left to full right. You can't see the blur change at all like it does when adjusting the vertical.

I've tried triggering and using bright line and that seems to make no difference to the blur, although I didn't measure the voltages yet.

I also slowed the timebase right down so the dot would travel really slowly but I noticed no change in blur, again I didn't check the voltages while doing this. I ran out of time last night.

[dragonser]

Hi,
I may be wrong, but I think that the clue to the fault may be in the difference between the voltages on the x plates. I would suggest carefully measuring the voltages at the various points on the X amplifier circuitry and seeing if either of the output transistors may have failed.

regardsPeter B

[TonyPhs]

Thanks Peter,

Just read some of the manual while at work and it's beginning to make some sense to me. I read this bit last night and I had no clue what it meant, assuming it was something I could only measure on another scope.

X output should be +118v at centre screen.
Then 12v/cm each side.


So given those figures at the edge of the screen voltage would be 58v and 178v. I measured with a DIY high voltage probe which isn't super accurate but I had it at 200v give or take so that's 2 cm outside the screen.

So I will jump right on those output transistors TR513 and TR514 and then down to the TR512 and TR511 which have 500mV/cm when I get half a chance.

[TonyPhs]

Right here's some figures.

TR513 collector 207.5v
TR514 collector 41v

No change with any horizontal movement.

TR511 emitter
(should be +3.7v at centre, so ramp from 0.7v - 6.7v at 500mv/cm)

Actual readings are 3v to 9.6v. So ramp is ok but is offset by 2 to 3v.

TR514 base - -3.0v reqd, actual -0.297v

TR512 collector (+0.65v centre, 500mv/cm)

Actual -1.67v to 3.07v. So is moving but isn't going quite as far as the manual says it should. Centre voltage is within a few degrees of knob centre.

Now I'm a noob here but if the pre-amplifiers, if that's what they are, are changing output voltage then the output transistors should be changing voltage as well. Shouldn't they?

[WME_bill]

OS300. Measuring voltages. Could I just add a word about measuring voltages around a troublesome scope.

You must stop the Time Base working, as otherwise voltages measured on the X plates, X amplifier, Beam bright up, and the Time Base itself will give various readings depending on the sweep speed etc etc.

So stop the TB, by switching to X input, or Triggered-not-Auto, or Stability- fully-CCW, or TB off, or External Trig, or Bright-line-off, or X-Y, depending on the style and age of the scope. Otherwise any voltage readings are almost meaningless.

Can you give the X Amplifier output voltage readings with the spot centered and no TB. The X output stages TR513 and TR514 should be the same. The X Shift should change the voltages -+60v and move the spot to edge of the screen and both collectors change, opposite ways. That apparently isn't happening. So look all round the output pair.
Exactly as PeterB -Dragonser has suggested. I would also agree with Bill in his Shed.
Then read the manual about TR515 which is just an extra to help pull the spot back more quickly on flyback.
The previous stage is rather unusual, TR511 TR509, but the manual does explain it quite well. Just check those diodes also.
Only when you have got these voltages about right can you start the Time Base to get an line on the X axis. It sound as though the sweep itself is alright.

There is another current thread about Telequipment D43 with bright-up trouble, where voltages are being quoted apparently with the TB still running (on Auto with Stability various), and so creating the same problems.

It is difficult to give any guidance with lots of meaningless readings being offered. It is amazing what can be done on a faulty scope with an old fashioned voltmeter, provided the TB is off, and no signal input. The X and Y shift controls are quite enough to check out the Display voltages. wme_bill

[TonyPhs]

Thanks Bill.

I put it in x-y mode and moved the dials to the centre position and measured again.

The output pair remain unchanged with an off screen voltage as before.

The preamplifiers all track voltages like before but in a non linear fashion.

I will check the components near the output as there seems to be something preventing the voltage changes coming through.

[WME_bill]

OS300. I am glad to see you are concentrating on a single isolated aspect, and can now work stage by stage back.
Check voltages around TR513 TR514 first. The common emitters should move if the base of TR513 moves.
Then go back to the start of the chain by the X shift control. What happens to TR506 emitter when you move the X shift. Then trace this movement through TR509 base, TR509 collector, TR511 emitter to drive TR513 base. As you follow that sequence, there is a voltage which doesn't move. Also check the zener diode D521, they often fail. The output transistors TR513, TR514 look like an Advance special. I suspect a common high voltage video output like BF258 will do. wme_bill.

[TonyPhs]

Good morning,

I won't have chance to do any component checks until later or tomorrow but before I set off for work this morning I went in and checked some of those voltages.

I didn't check the emitter of TR513/514 for some reason but did check the base and that's static with only a very slight drift from full left to right deflection.

I went all the way back to TR506 emitter and there's 0.288 to 0.454v

TR509 base shows 0.596 to 0.676v
TR509 collector 11 - 0.75 - 0v left-ctr-right

TR511 emitter 10.67 - 140mv - 1--mv left-ctr-right


I looked at D521, a 6v2 zener when I had the voltage offset yesterday and thought I might start there. I don't really understand what it actually does in the circuit though, having not really ever used or studied zeners.

[cmjones01]

From your voltage readings, it looks like there's little or no current flowing in TR513. I suspect it's open-circuit. My reasoning for that is that the voltage on TR511 emitter is a few volts higher than expected, which is consistent with little or no current flowing into TR513's base, and TR513's collector is way higher than expected. That in itself is enough to explain why everything round TR514 is out of whack too - it's a long-tailed pair, so I'd expect that.

My next move would be to take TR513 out and test it out-of-circuit. Just seeing whether the B-C and B-E junctions look like diodes would be enough. If you have a suitable replacement transistor around (as Bill says, any common-or-garden video output transistor should do at least for a test), try fitting that. If you haven't, try swapping TR514 and TR513 over. See what the spot does then.

Chris

[TonyPhs]

Thanks Chris.

I did already check TR513 and TR514 for a short across C-E as that is easy to do while in circuit but was waiting until I was more sure of where the failure was before I started taking stuff off the PCB or cutting legs.

I very much doubt if I have any video output transistors to hand as I've never harvested components from a device that has a CRT output before and I certainly haven't bought any. So I will likely have to buy something, possibly those suggested by Bill if I can find some, or their slightly higher voltage friends the BF259. I would obviously buy a pair.

I'm a little confused how this failure was at first intermittent though.

Tony

[TonyPhs]

Took both transistors off the board.

TR513 was indeed blown. One junction had open circuit as you said, I think it was the B-E.

TR514 remains with intact with diode drops across B-E and B-C.

I tried swapping them over and the attached image shows the result, in Ch2 mode of course. In X-Y I get a spot off to the left of the screen, but at least on it.

The new now working TR513 collector voltage now tracks from 50ish to 110 when I twiddle the horizontal. The new TR514 stays all still. I left it running for a few minutes so there's nothing massively wrong.

So should I troubleshoot further, the case was bowed on top and close to the heatsink collector. Transistors don't normally fail for no reason do they?

I will order some of those BF259 or BF258 right away anyway so at least they can be on the way and this thing can be up and running again so I can continue to troubleshoot my other broken scope.

I did actually spot the discolouration to the PCB around the transistor as noted before but of course that could have been there before.

[cmjones01]

Great! It looks like we've got somewhere. That's exactly what I'd expect with the transistors swapped over. The spot is still to one edge of the screen (opposite where it used to be), but a bit more responsive, since one half of the longtailed pair is now working and thus one X plate is now doing something rather than neither of them doing anything.

I strongly suspect that once you replace the duff transistor, you'll be back in action. I would advise replacing both transistors with the same type since it's useful for their properties to be as similar as possible in a longtailed pair. It will help to keep the X position stable as the scope warms up.

Transistors do sometimes fail for no good reason, yes. They're not perfect. It's possible that if the case did hit the transistor's collector, it could do some damage, but not hugely likely. Whatever happens, when you put the new transistor in, make sure everything is tidily assembled and insulated in the right places. If there's a chance of the metal case contacting that area, put some insulation (even tape is better than nothing) on the inside of the case.

Discoloration on the PCB around transistors like these which run fairly hot in normal use is not unusual, and isn't usually something to worry about.

What package are TR513 and TR514 in? Is it a metal can TO39 like the BF259 replacement?

Chris

[TonyPhs]

Hi Chris,

The existing TR513 and TR514 are in a TO-202 package rather than the TO-39 package the BF259 replacements are in. Fortunately the pins are in the same order so I don't have bend legs in creative ways. I've ordered 5 anyway as the price was not that different from just buying 2 singles. £3.60 vs £4.

The heat dissipation of the two form factors is clearly different though so I'm thinking about attaching a ready made or perhaps creating a heatsink for them as they are quite close physically on the PCB. Maybe a small stick on VGA card RAM heatsink would do. I can't go touching them to assess the temperature and don't have access to a non contact temperature gauge, although I do have something which we use in work to check the temperature of solder pots so that will do I suppose to give ballpark figures.

[maninashed]

Glad you've found the culprit Tony. What transistors were specified on the original component listk? The BF259, (BF305) variant is very common as X or Y output transistors in scopes of this era, and normally they are very reliable, but they do run hot. They are usually heatsinked in some way. If someone has fitted TO220 types they may well have discarded the original heatsinks. Be very careful what you clamp to chassis for heatsinking as the case is usually connected to the collector on BF259.
Bill

[cmjones01]

Hmm, those transistors look quite a lot beefier than the BF259, so be prepared for the BF259s to get rather toasty. If heat dissipation is a problem, I'm sure we can track down some other video output transistors - colour TVs used plenty of them in bigger packages. I can't think of any type numbers off-hand now though.

Chris

[TonyPhs]

The NSD459 were the original specified in the schematic and were those that are still on there, although they could have been like for like replacements I suppose.

The original transistors did not have any additional heatsinking beyond the long blade of the package so I'm hoping the TO39 will be sufficient with something stuck on it to give it a bit of a helping hand.

The attached image shows some available on ebay but I will have to check the size to ensure they don't touch each other as that would not be good.

Funnily enough my other scope, the larger and older Gould OS3000, which has very low EHT, has BF258 on its horizontal output so provided that isn't too hard to get to I might use them in lieu of delivery of the replacements.