Hameg oscilloscope horizontal deflection issue

[tlm283]

Hello everybody,

my Hameg HM 205-2 Oscilloscope has a problem with it's horizontal deflection.
After being on for like 10 Minutes it suddenly stopped drawing the trace properly. To the left it's drawn like normal, and when moving the X-Pos knob it behaves like expected.
The right half of the screen behaves like it is compressed to half the size. When put in X-Y mode the dot can't be moved to the right as well, and when moving the trace with the X-Pos completely to the right, it only covers like 2/3rds of the right side of the screen.
The same happens when putting it into 10x mag, the trace gets cut off on the right side. I attached a picture of the scope, the X-Pos knob was in the center position, and the trace should have covered the whole screen.

With the problem persisting even in the component tester mode, I am very confident that the problem has to be in the final X amplifier. I did the usual things, checking voltages and obviously dead components, but all looks normal. I then looked further into the schematic, realizing the X amplifier is more complex than I thought it would be, and I am clueless what to check next.

The manual can be found here: http://www.sm5cbw.se/hameg/hmosc/hm205-2-man.pdf
The X amplifier is at page 47, but I have a screenshot of the schematic attached.

EDIT: I forgot to mention that D2 and D1 are the output to the plates of the CRT, and that D2 is the plate that does not deflect properly.
D1 has at full deflection a voltage of about 120V, and at no deflection at all 6V.
D2 only has about 73V at full deflection, and about 20V with no deflection.

[tautech]

Sweep linearity is possible the problem and possibly right back at the sweep generator in the timebase.

So it should generate a simple ramp waveform and there should be examples of the waveform and where to check it in the service manual.
The ramp rising voltage when amplified is what's responsible for driving the trace across the display by creating a electrostatic disparity between the 2 horizontal plates.

[tlm283]

Sweep linearity is possible the problem and possibly right back at the sweep generator in the timebase.

So it should generate a simple ramp waveform and there should be examples of the waveform and where to check it in the service manual.
The ramp rising voltage when amplified is what's responsible for driving the trace across the display by creating a electrostatic disparity between the 2 horizontal plates.

I thought that was the problem as well, but the sweep generator is not used when the scope is in component tester or in X-Y Mode.
Also, in X-Y Mode I used to be able to move the dot completely from left to right with both inputs grounded only with the X-Pos knob.
The dot also moves relatively linear on the right side, it just does not move as far.

I think I am going to try to simplify the schematic a bit, I just can't quite figure out how exactly the amplifier works

[shakalnokturn]

I agree with that, if you can't get where you want in X-Y it is doubtful the sawtooth linearity has anything to do with this.
Start by checking the X amp output transistors are tight enough on the rear plate, clean and replace thermal compound if necessary. Any transistor that was very loose is a suspect.
Switch to X-Y, set X position control to midpoint (adjustment, not on CRT) compare voltages on the good and bad sides of the X amp.

I really wouldn't be surprised if it was a dead output transistor (not dead short of course) but on some occasions I've seen BF199's cause funky failures too.

[tautech]

Sweep linearity is possible the problem and possibly right back at the sweep generator in the timebase.

So it should generate a simple ramp waveform and there should be examples of the waveform and where to check it in the service manual.
The ramp rising voltage when amplified is what's responsible for driving the trace across the display by creating a electrostatic disparity between the 2 horizontal plates.

I thought that was the problem as well, but the sweep generator is not used when the scope is in component tester or in X-Y Mode.
Also, in X-Y Mode I used to be able to move the dot completely from left to right with both inputs grounded only with the X-Pos knob.
The dot also moves relatively linear on the right side, it just does not move as far.

I think I am going to try to simplify the schematic a bit, I just can't quite figure out how exactly the amplifier works

OK then we look a bit harder but it's not too hard. 

As we have ~150V do drive the plates with in X-Y mode with zero signal the dot should be centered on the display and an equal voltage on each plate, somewhere ~70V.
Use that as a starting point to check voltages from the plate outputs back down the signal path.

Really we need the service manual that tells how the scope needs be set to check all these things.

[sundance]

https://www.sky-messtechnik.de/reparaturdienst-service/downloads

HM205-2 Service Manual: https://www.sky-messtechnik.de/images/PDFs/4S-0205-02D0.pdf
(unfortunately in German only)

edit: and in English: http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Hameg/hm205-2.pdf

Sorry, didn't see the link in the 1st post....

[tlm283]

With the X-Pos control at the middle, the dot is centered, and both plates have ~70V on them.
With the control moved all the way to the left D2 has ~23V on it, and D1 125V.
With the control all the way to the right, D2 has about 73V on it, and D1 ~29V.

Now if I understand it correctly, T2209 and T2210 are the transistors that are actually controlling the voltage on the plate.
The base of T2210 is a constant -5V and does not vary with the X-Pos control or with a voltage applied to the horizontal input, the base of T2209 changes between -5V and -7V.

I'm not quite sure what to make of that. I expected that the base voltage of both transistors would change, depending on X-Pos and/or the XF input. Does this mean that the base of T2209 does not get negative enough?
Oh, and no transistor in the X amplifier is connected to a heatsink.

[tautech]

Have you checked all the resistor dividers to see if any had drifted high ?

[tlm283]

Have you checked all the resistor dividers to see if any had drifted high ?

Yes I already checked the ones I would think are critical, just checked some other ones as well, but everything seems to be fine.

[tautech]

Hmm, shame our Hameg specialist oldway is no longer with us. 

Have a search through the forum for his many posts on Hameg repairs.

[Chris56000]

Hi!

Both the X plates of the crt are symmetrically driven by identical cascode directly coupled DC amplifiers, so it shouldn't be too difficult to localise this sort of fault!

The simplest method is as follows:–

Switch the oscilloscope to it's SLOWEST timebase speed and connect your d.v.m. to the base of T2206, and switch the triggering onto auto so you get the spot free–running at it's slowest speed, turning down the intensity to make it just visible! As the spot runs from left to right, note the minimum and maximum voltage recorded on the dvm (if your dvm has min/max recording this is a great help) – if not, try and note down the figures as accurately as you can.

Once you've got a pair of voltage figures for the maximum and minimum base voltage on T2206, transfer your meter prod to T2207, this time rotating the "X Pos" control from one end to the other – obviously this is much easier because "X pos" is a manual control.

Once you've got a pair of minimum and maximum voltage figures for the bases of T2206 and T2207, subtract the lower from the higher figure, the difference is the voltage swing at each base needed to deflect the spot across the screen horizontally.

The most important thing that matters is that the voltage changes on T2206 and T2207 bases should be approximately the same, as you're deflecting one crt and it's D1 and D2 plates will have the same sensitivity!

If you measure approximately the same minimum–to–maximum voltage changes at both T2206 and T2207 bases by following the above procedure, then any trace distortion or non–linearity will be down to the final X amplifier stages (T2206, T2208 & T2210 for the D1 plate, and T2207, T2209 & T2211 for the D2 plate) and straightforward component checks on the resistors & transistors should bring any defects to light.

If the voltage swings at T2206 and T2207 bases are NOT approximately equal by using the procedure I suggested, then you need to make a further quick check of the output circuits, by lifting one end of D2204 and D2205, then using short pieces of wire to connect the anode end of D2204 to the anode point of D2205 on the printed circuit board and vice–versa – this changes the D1 and D2 plate amplifier connections to the X amp input and the X pos control over, and if both sets of amplifiers are
working correctly, the trace should show the fault you're seeing on the left hand side of the screen if the fault is before the X final amplifier.

It sounds a fairly complicated procedure, but it's the only quick way of differentiating between X amplfier and timebase failure!

Please try the above if you can and come back if you're still stuck!

Chris Williams

[tlm283]

Thanks for that detailed explanation!

I checked the base of both T2206 and T2207, with the lowest time base available. The voltage on T2206 varies between -4.2V and -5.4V. The voltage on T2207 on the other hand stays at a constant -4.3V, even with changing the X Position.

I then swapped both anodes of D2204 and D2205. The X-Pos control was as expected reversed(when moving the knob to the left the trace moved to the right, like intended), but the trace still looks the same, with the right side "compressed".
That means that everything to the right of D2204 and D2205 should be working properly, right?

So I checked the emitters of both T2202 and T2203, again using the slowest time base/X-Pos control.
The emitter voltage of T2202 changed between -4.2V and -6.6V, and the emitter voltage of T2203 changed between -4V and -4.35V when changing the X-Pos control, and changing +/-200mV when using the timebase.
The base voltage of T2203 changes with the X-Pos knob between 0.6V and ~5.2V.

I don't want to jump to conclusions, but for me a change of a few 100mV seems a little bit low, especially considering that the change on the base of T2202(about 7V) results in a change of a couple of volts, but a change of 5V on the base of T2203 only causes a change of a few 100mV on the emitter. Because the circuit looks so symmetrical I would have expected that both transistors behave somewhat equal, especially when the base voltages are about the same.

Thoughts?

[StillTrying]

but the trace still looks the same, with the right side "compressed". That means that everything to the right of D2204 and D2205 should be working properly, right?

No, as the fault didn't swap screen sides, everything to the right of the diodes is suspect, which is what Chris56000 says.
I don't know much else!

[tlm283]

Yeah you are right, I got that completely wrong.

[StillTrying]

T2206, T2207, and T2212 are just emitter followers so it should be possible to eliminate them with DC voltage measurements with the X-trace stopped. That would be close to leaving just 2 transistors and their resistors.

Have you measured the values of the resistors in circuit while switched off!, some of the high values will read low in-circuit but at least you've the same on the other side to compare.

[tlm283]

I went through all resistors, and they all seem to be ok.
I also removed the XY-Board to inspect the solder joints (as the X-Amp is conveniently directly under the CRT and i had to remove all boards to get the board out) but they all looked fine, I went and reflowed some of the joints but no  improvement.

I also disconnected both plates, maybe one was somehow shorted, but again, no change.
I'm slowly thinking that maybe the X-Amp was functioning all along, but it just does not make any sense with the component tester/X-Pos not working.
Really weird.

[StillTrying]

The base of T2210 is a constant -5V and does not vary with the X-Pos control or with a voltage applied to the horizontal input, the base of T2209 changes between -5V and -7V.

It's strange that one base voltage varies quite a bit but the other doesn't while both emitters are connected.
Have you checked that each base-emitter is always ~0.6V.

[tlm283]

Have you checked that each base-emitter is always ~0.6V.

Both T2209 and T2210 have a base-emitter of ~0.9V, but I think the transistors are fine and it is just because of the supporting circuitry.
I think I will take the XY-Board out again tomorrow and measure them out of circuit.

[tlm283]

I took the board out of the scope again to remove the transistors and checked T2206, T2207, T2208, T2209, T2210, T2211, and T2212, and all of them seem to be ok.
I also swapped around T2208 with T2211, T2209 with T2210 and T2206 with T2207, hoping that if one of these transistors had a weird fault it would be on the other plate, but again, no luck.

[shakalnokturn]

I'd really doubt it is the problem but you can rule-out the CRT by crossing the D1 / D2 outputs to the plates.

Other than that I'd be re-checking the 4 diodes and linked resistors.

And R2236, R2237, R2231, R2234.

[tlm283]

Checked the plates, resistors, diodes, no change.
Should I check the traces again? Maybe there was a broken trace i did not catch.

[shakalnokturn]

You've handled a lot of the troubleshooting the same way I would have, I was expecting the problem to be found when you switched transistors over from 1/2 to the other.

Checking traces, why not, it's not a very common problem but when out of ideas might as well...

A few thoughts on the testing components did you do most of it in circuit? Without getting fooled?
Comparing both 1/2's? Measuring two parallel 100k resistors in circuit you may read something around 50k and accept that as OK, but if you measure the same setup elsewhere without referring to the schematic you may read around 100k and admit it is OK looking at the colour code when in fact one of the resistors has gone open. I've been fooled once before in a similar case.

You're left with T2202 and T2203 that haven't been swapped around yet. At this point I'd be removing R2201 and T2204 and trying to drive T2202, T2203 externally (bench supply or scope's supply rails and a couple of potentiometers). Not saying it is the wisest thing to do, sometimes it just helps to split things up to test and understand.

[StillTrying]

Do the 3 emitters connected together at the center of the output stay at about -4V, while the X position is moved from side to side.

[tlm283]

A few thoughts on the testing components did you do most of it in circuit? Without getting fooled?
Comparing both 1/2's? Measuring two parallel 100k resistors in circuit you may read something around 50k and accept that as OK, but if you measure the same setup elsewhere without referring to the schematic you may read around 100k and admit it is OK looking at the colour code when in fact one of the resistors has gone open. I've been fooled once before in a similar case.

I was aware that measuring in circuit is a bit risky, but I always compared both values to the ones from the other plate, and they were always pretty much the same.
But considering that the transistors are definitely working and there is pretty much nothing else to fail it has to be a resistor.
I'm tempted to desolder all resistors right of D2204 and D2205 to check them out of circuit.

Do the 3 emitters connected together at the center of the output stay at about -4V, while the X position is moved from side to side.

They are a constant -5.6V when moving the X-Pos.

Thanks for all the help you guys provided, really appreciated.

[tautech]

I'm tempted to desolder all resistors right of D2204 and D2205 to check them out of circuit.

Sometimes you just need to pull a couple of components and this can isolate many for accurate tests.
A study of the schematic can identify the ones you need to pull. 
I certainly wouldn't be pulling them all unless absolutely necessary.