Hameg HM307-3 Cleaning and Repair

[WaveyDipole]

Originally Posted by OldTechFan96

The main problem at the moment is the lack brightness from the CRT. I'll start checking the CRT circuit. Especially around the blanking circuit. What does this do?

As you will no doubt be aware, the trace sweeps the CRT from left to right. The sweep is driven by a signal from a ramp generator. Once the trace reaches the right hand side, the beam needs to be switched back to the left so that it can start the next sweep cycle. Although this switching back to the left happens very rapidly, if the beam were left turned on while this was taking place, it would leave an unwanted visible line across the CRT. The blanking circuit cuts off the beam as it is switched back from right to left effectively hiding this unwanted artefact.

[OldTechFan96]

Thanks for that, Wavey Dipole. I think I understand. Does it remove the faint line shown in the attached image?

What do you think about the BF440? With it having the same voltage on each junction I think it may be faulty. I have checked the resistors around the transistors and they are good.

[WaveyDipole]

Originally Posted by OldTechFan96

Thanks for that, Wavey Dipole. I think I understand. Does it remove the faint line shown in the attached image?

Yes, that's a fine example of what happens when the return sweep is not blanked. You get a visible trace of the electron beam being swept back from right to left, but since this happens much faster than it takes to sweep the signal trace from left to right, it appears fainter.

Originally Posted by OldTechFan96

What do you think about the BF440? With it having the same voltage on each junction I think it may be faulty. I have checked the resistors around the transistors and they are good.

After replacing C352 you now have a 33v difference between the -900v and -867v rails (-909-(-874) = 33) which is as it should be. There should also be a 100v difference between the -900v and the -1kv rail. Since you have -909v on the -900v rail, it may be simply that the -1kv rail is at -1009v which is not a major problem. However, the zener should still be maintaining a 100v difference between those two rails. If you measure 100v across C350 then all is well and good. If you are seeing less than that, then this might affect the maximum achievable brightness and there are only two components that are likely to affect this - C350 and the zener itself. Disconnect the capacitor. If the voltage across the zener now rises to 100v replace the capacitor, otherwise replace the zener.

Regarding the BF440, I am not entirely sure, but the reading does look odd. If you have a 2N2907, 2N3906 or 2N4404 to hand then you could try one of these. Even a BC327 might do at a push. It has a lower transition frequency, but it should be fine for a test. Just be careful about the pinout of any substitute. For example, the 2N4404 is EBC vs CBE.

Since we are dealing with pulsed signals it is difficult to understand what is going on at the output using a multimeter alone. The output pulse has an amplitude of +33v relative to the -900v point, but since the switching is so rapid, a multimeter would just average this out, with the result being affected by the duty cycle of the pulse. The voltage drop across the transistor alone would be fairly insignificant (approximately 0.6v) while it is turned on, however, since the blanked state duration is but a fraction of the unblanked state, I would expect the measured reading to "favour" the -900v unblanked state. The fact that the measurement seems to be actually at the blanked state voltage may suggest that T352 is hard on, which may be either because it is leaky, or D351 is faulty or both. Your strange test result does seem to suggest that all may not be well with the transistor, but I would check both transistor and diode anyway.

[OldTechFan96]

There is 100V across C350.

I've subbed in a BC327 in place of the BF440 and the junction voltages remain -875V. Perhaps this is correct?

D351 had the same voltage on each side. It tested good out of circuit. There doesn't seem to be any voltage across it.

Would I be able to use my good scope to try and probe the blanking and unblanking pulses?

[WME_bill]

HM307.
Your picture showing the return trace on posting#82 shows the classic fault condition for failed blanking.
These transistors seem to fail frequently. T352 is failed from your voltage tests. Any switching PNP will do, rated >50v. The 40v rating of your BF440 does not leave much margin for surges.
If your junk box doesn't provide and you have to get a replacement, I would replace both the PNP BF440 and NPN BF199. These are german types.You might find cheaper generic types like 2N3906/2N3904. Your BC327 is fine for the PNP. Try changing the NPN as well. Then look to the Opto coupler.

As WaveyDipole says, you can always try out almost any generally suitable transistors, on the basis that if it gets the thing working, then you can go on to any other faults. You can always go back later to optimise performance.
In this circuit, a replacement transistor which is not switching fast enough would give a smear at the beginning of the return trace on the highest TB speed. You can often live with minor points like that anyway.

As I suggested a long time ago, stop the timebase operating to measure voltages.
The trace will then be blanked. Then switch to External X and the trace will change to full brightness. That should be self evident.
If the optical coupler is failed, IC304, look for a change of 3v at the input R363 when you move the switch. If OK, go to the high voltage points at output of the IC p4 and then the collector of the driver transistor T353.
If you leave the timebase running, with blanking that is a switched pulse, the voltage you read on a meter will depend upon the mark-space ratio of the timebase sweep to flyback, so could be anything.
At worst, use the slowest TB speed, as then an analogue meter will have time to respond and will judder between the two voltages. A digital meter will get confused and just flicker away.

I agree with WaveyDipole over the 100V. May well be the tolerance on the zener Z351. Wind the two set brilliance range presets R356 and R357 to furthest apart, so that the range on the Intesnsity control is greatest.
You have about 100v across Z351 zener. so OK, and the -875V looks right enough, but should change. Reading of 875 suggests T352 driver is hard on. Either the diode D351 has failed, or there is no blanking pulse coming from the Opto coupler.
wme_bill

[WaveyDipole]

Originally Posted by OldTechFan96

There is 100V across C350.

Ok, that's fine. So the zener is OK and the -1kv rail is just a little out. This is not unexpected. The adjustment would be a little coarse and it would be difficult to get it spot on. Its close enough.

Originally Posted by OldTechFan96

I've subbed in a BC327 in place of the BF440 and the junction voltages remain -875V. Perhaps this is correct?

D351 had the same voltage on each side. It tested good out of circuit. There doesn't seem to be any voltage across it.

Would I be able to use my good scope to try and probe the blanking and unblanking pulses?

NO! Not unless you have a x100 high voltage probe. You do not want those kinds of voltages anywhere near the input of your good scope! The only signal you can probe safely is the 3.4v input signal I mentioned that is found at the bottom of the diagram.

[OldTechFan96]

I do have some more BF199s so I'll try changing one later. I'll remember to take measurements with the timebase stopped.

Voltages on the input of R363:

Hor. ext. in: 3.38V

Hor. ext. out: 3.22V

IC pin 4: -871V

C of T353: -872V

I might have some opto couplers knocking around. CN-17 rings a bell. Comments about diode D351 are noted. Another possible failure only at HV?

I'll not do any probing with the good scope. I checked the output of the square generator last week and it was fine.

Attached are some photographs of trace I am getting. One is with the hor. ext in and the other out.

[WaveyDipole]

Originally Posted by OldTechFan96

I do have some more BF199s so I'll try changing one later. I'll remember to take measurements with the timebase stopped.

Voltages on the input of R363:

Hor. ext. in: 3.38V

Hor. ext. out: 3.22V

IC pin 4: -871V

C of T353: -872V

Let us know how you got on with swapping the BF199.

Thanks also for those readings. However, if, in addition to the above if you could provide the voltage readings for those same points with the timebase stopped as suggested by wme_bill, then that should be helpful.

[OldTechFan96]

I should have included last night that the voltage measurements at pin 4 of the IC and the C of T353 were the same, regardless of the position of the timebase switch.

[OldTechFan96]

I replaced the BF199 and nothing changed.

More voltage measurements (timebase off):

IC304:

Pin 1: +1V
Pin 4: -873V
Pin 5: -873V
Pin 6: 0V

T352:

E: -867V
C: -868V
B: -870V

T353:

E: -901V
C: -866V
B: -901V

I do have some CNY17-3 opto couplers. I may have a suitable diode to replace 1N4149, I'll have a look.

[WaveyDipole]

Thanks for these. Is this with Hor Ext in or out? Need to comparison readings for the opposite case please, but still with timer off please if you can.

[OldTechFan96]

The measurements in post 90 were done with the hor. ext. pushed in. Which I think defeats the timebase.

I can redo these measurements with the hor. ext. button in the out position for the sake of comparison.

[OldTechFan96]

More voltage measurements (timebase on, hor. ext. out):

IC304:

Pin 1: +1V
Pin 4: -870V
Pin 5: -873V
Pin 6: 0V

T352:

E: -866V
C: -866V
B: -866V

T353:

E: -899V
C: -865V
B: -899V

[WaveyDipole]

Sorry, the confusion is mine.

Originally Posted by WME_bill

As I suggested a long time ago, stop the timebase operating to measure voltages.
The trace will then be blanked. Then switch to External X and the trace will change to full brightness. That should be self evident.

I was trying to follow WME_bills excellent idea and had wrongly assumed that there was an X-Y mode position on the timebase dial that stops the timebase and then we needed measurements with the Hor Ext in the out and the in position. However, I can see from the pictures and the manual that the Hor Ext switch serves this purpose.

Ok, so we have current at the input of the LED in the optocoupler. The light sensitive transistor appears to be turned on because it is pulling E towards to -867v line. However, T353 is turned off and the circuit is still in the blanked state. Your measurements show the base of T353 at -899v when it should be pulled up to -877v. You already tried substituting T353, so is R362 o/c?

BTW, maybe its the effect of warm up by you show -873v on pin 5 and then -866v at the collector of T352, but both are on the same -867v line?

[OldTechFan96]

R362 measures good.

I measured the voltages on pin 5 and the C of T352 after three minutes:

Pin 5: -867V

C of T352: -865V

[WME_bill]

HM307
Hang on chaps. Bright-up circuit.
Pin 5 collector of Opto-coupler should be same as emitter of T352, not the collector. The collector voltage from your measurements is 2v positive (less negative), which suggests the transistor is hard on. The emitter will be somewhere between 0.6v and 2v positive of the 867V rail.

W/D comments that the voltage at pin 4 of IC304 moves when you switch the blanking input by changing from Ext X to Sweep on, not triggered. But T353 appears dormant. Have you changed it?
My reasoning is that Ext X will give full brightness. Sweep on, not triggered wil give sweep blanked. That means Nornal trigger, not Auto. no signal input, or Trigger Level set way off.

The sweep blanking signal from the timebase into the Opto-coupler is at 0v during the sweep and +3.4V when blanked and TB waiting to be triggered.

Measuring small voltage differences when floating at high levels is neither comfortable, nor likely to be very accurate, unless you have a 5/6 digit voltmeter, to sensibly read 0.6v on 1000v scale, which your measurements of yesterday afternoon imply.
wme_bill

[WaveyDipole]

OldTechFan, sorry, yes, I am confusing things again. I am not very well at the moment so its probably tiredness on my part. I said collector when I meant emitter of T352. When I was looking at from your readings in #93 there is a difference between Pin 5 of the optocoupler (-873v) and the E of T352 (-866v) of 7v which shouldn't be there. You did mention earlier that the -867v line read -874v. In post #84 you also mentioned that all pins of T352 were at -875v, readings that seem reasonably consistent, but now we seem to have -866v at the emitter of T352. Similarly, you had previously reported the -900v line at -909v. I therefore wondered whether you had tweaked the -1kv line at some point between readings or whether the -1kv source was shifting. I suspect the latter due to warming but it does confuse the readings a little.

Getting back to the blanking circuit, yes, agreed T353 is not getting switched on and thanks for the clarification regarding manual trigger mode. I could see that Ext X will give full unblanked brightness, but wasn't sure how to get a steady blanked signal. All the readings so far seem to show an input voltage present, either 3.38v or 3.22v at the input to R363, or 1v at pin 4, which would be about right with the stated 3.2v input. But why is there a voltage present when Ext X is pushed in? I was therefore going to suggest lifting the input end of R363 and grounding it to get a 0v input and then measuring again.

[OldTechFan96]

I've done some more measurements:

Pin 5 of opto coupler: -870V

E of T352: -870V

C T352: -868V

B T352: -869V

T353 has been changed.

The voltage at R366 is -867V and is correct for the blanked state. It stays the same no matter how the controls are manipulated.

The HV pot has not been touched since the -1000V rail was fixed.

-867V rail measures -870V.

-900V rail measures -903V

I've lifted the input side of R363, grounded it and took measurements at pin 4 of the OC IC and the voltage measured is -872V. This does not change nomatter where the hor. ext. button is pressed in or not.

[WaveyDipole]

With that end of R363 grounded, nothing will change with the position of the buttons because we have fixed the state input state to 0v. There seems to be no effect regardless of whether there is a voltage present at the input of R363 or not. Since you have tried substituting the transistors and confirmed that the diode is OK, I think we must be looking at a faulty OC. You mentioned you might have a spare so I suggest you try replacing it it.

[OldTechFan96]

IC304 changed to a CNY17-3. No change in brightness.

New voltages T/B off, hor. ext in:

Pin 1: +1V
Pin 4: -894V
Pin 5: -870V
Pin 6: -897V

T352:

E: -902V
C: -868V
B: -868V

T353:

E: -900V
C: -900V
B: -900V

EDIT:

With the auto triggering switched off, the voltages at R366 change when the hor. ext. button is pushed.

Out: -864V blanked

In: -900V unblanked