Tektronix 2213 trace starts in middle of screen

[moose]

Hi
i have a Tek2213 that is used ocassionally as a 2nd scope.
The current time that i wanted to use it i noticed the following...the trace started near the middle of the screen. If i tried to reposition (with horiz position control) the trace to the left, it would simply bunch up at the point the trace started.
Moving the trace line right functioned as per normal.


i swopped left & right crt deflector leads & that merely reproduced the issue, but now on the right.

As per service manual, i set it to trigger-auto, chan 1-gnd.
Looking at the waveforms (24/25) on the main scope both waveforms mirror images of each other except the left values were 10V higher.

i then looked at the waveforms of the sweep generator...those also looked fine.

is it possible the horiz position control pot itself could be faulty? what should the resistances at both extremes be?

any suggestions as to where to focus next?
[edit] i did check all PSU voltages & they are all fine.

thanks in advance

[neillnz]

The manual says the horizontal postion pot is 5 k Ohm. It shows it as a Voltage divider for the +8.6 V and -8.6 V rails, see the fold out 7 which is page 187/224 of the PDF from Tek: https://w140.com/tektronix_2213_service.pdf 

What is test point 23 doing? Does it swing +1 V to -1.3 V?

I Used to fix these things for the NZ distributor in the early nineties.  Cannot recall seeing this particular fault, it was almost always power supply issues.  One day an engineer from Beverton was visting and he asked what was a typical fault, I told him that's easy it's almost always PSU and he said that's odd because they rarely had PSU issues, I always wondered if it was an issue with using a primarily 110 V design at 230 V. 

[moose]

Hi Neil
Thanks for that info (& history - saved for future ref!).

as for TP23: it shows the right shaped waveform.
          With the Horiz controller positioned
                     max left    +2.2 to +0.28
                     centre       +0.8 to -1.46
                     max right   -0.18 to -2.56
I'm uncertain whether this is good, bad, or indifferent..

TP22 also had right waveform swinging from +11.4 to +0.2 (regardless of horiz controller position).

cheers

[bdunham7]

You can put the scope into XY mode with no inputs (and both set to GND) and then you'll have static voltages on all of the horizontal amplifier sections.  Try to center the dot and move it around to see how each half of those amplifier sections responds to the horizontal position control.   

[BlownUpCapacitor]

If I were to make a wild guess, one of the supply rails for the horizontal amp is bad. Check supply voltages near the horz amp.

Also do some signal tracing to figure out whether the horz amp is really the problem. Perhaps it's only the sweep not going to the negative (or positive maybe it's inverted) when that happens, but considering the horizontal position is experiencing the same effect, I would bet supply voltages in the horz amp.

[moose]

thanks for the suggestions...
XY mode - whilst the dot was centered i could only move it vertically. And when i connected a diode & a cap to view the waveform it was basically a kinky line. So i must assume that XY-mode is non functional (something else to investigate ).

i had a look at all the voltages as per the circuit diag 5 (sweep generator & logic) and circuit diag 7 (xy amp/horiz output) ....everything appears to be close to what its supposed to be & the pwr rails inputs are spot on. Admiteddly i ignored the XY section.
Checking the horiz position knob output voltages, they varied correctly so i can exclude that.

[bdunham7]

Can you pull out the 10x horizontal magnification knob on the timebase knob?  Do that with a normal trace showing in auto trigger mode, see if the trace expands to cover the screen.  If so, I think your problem may be Q730 or very near it.  You can just measure the voltages (referenced to ground) at all 3 pins of Q730 and Q731 and report back.  Do that with the horizontal knob centered and in XY mode with both inputs set to GND, then with the horizontal knob turned a bit to the left and then a bit to the right. 

[moose]

the 10x mag merely magnified the signal over the same portion of the right side of the screen......a few days ago when i 1st started with this it did go right across the screen.
Q730 -0.32 +0.4 -7.68
Q731 -0.32 +0.4 -7.64
(pain to find them & then get to them!)

however, something happened in the last few hours...the right side of the trace is now chopped before the extreme right of the screen...and...the waveform (at R788 & R778) going to the L & R horiz deflection plates is not as clean as it was.
is it possible that transistors Q789/785 and Q779/775 are faulty??

Also, i noticed today that pin 5 & 12 of Q603 (diag 5 F4) remains at 3V...i'm sure that yesterday there was a square wave on it...so does that mean Q603 is faulty

[bdunham7]

It looks like the issue is further down the line.  You can try probing the voltages on all of those transistors in the same way without worrying too much about exactly what they are doing--just look for about 0.7V from base to emitter.  Include Q736 just to be complete, but anywhere between point 23 and 24/25 is game.  Don't overlook the beam find circuit and Q747 as a malfunction there would certainly cause the symptoms you are seeing.  Maybe even check Q747 first.

[moose]

i am really at a loss now...i checked voltages at the following:
Q736, Q763, Q753, Q765, Q770 and Q747 - none of the voltages were much out of what was written on the schematics (and all base/emitter were around 0.6-0.8 V.

[bdunham7]

i am really at a loss now...i checked voltages at the following:
Q736, Q763, Q753, Q765, Q770 and Q747 - none of the voltages were much out of what was written on the schematics (and all base/emitter were around 0.6-0.8 V.

What about Q775/779/785/789 and Q780?

Where is the dot on the screen when you are doing this?  If it is approximately where it is supposed to be then perhaps the voltages won't be off.  Try moving the horizontal position all the way towards the side that the dot won't go to and try again.  And check points 24 and 25 each time as well. 

[moose]

ahh, my bad...all the above were with chan 1 GND, sweep 0.1mS
i will redo with chan1 & 2 at GND, and sweep at XY and horiz pot at full left.

[moose]

ok, done..
values are 1st horiz full left & 2nd horiz full right, a single value indicates no change left/right
                   B               E                 C
Q747        -5.1            -5.7          -2.8/-4.2   
Q763     -1.6/+1.2   +0.8/-2.1     +4.6/+4.8     base/emitter high.....faulty??
Q753           0             -0.8            +5
Q765        +6.8          +6.2            +8.5
Q770         +5            +6.2            -0.5            base/emitter high.....faulty??
Q779         +96           +97            +50/+63     base/emitter high.....faulty??
Q775           0               0              +50/+63
Q789         +96           +96            +44/+33    surely this & Q779 should match??
Q785          0               0               +45/+33    surely this & Q775 should match??
Q736         -7.7           -8.3            +2/-1
Q730         -0.3          +0.3            -7.7
Q731         -0.3          +0.3            -7.7           
Q720         -0.4             0              -1.5           

also, i noticed that when the horiz control was full right, the dot would wonder left to right over the width of the sweep which is about 1".

[bdunham7]

You missed Q780?

I'll sleep on this, perhaps something will come to me.  The output portion of the horizontal amp is not really balanced or symmetrical in design but the output should shake out to be pretty balanced--and yours obviously isn't.  It also lacks range.  Can you also check the voltage across VR781 and also confirm that your -8.6V supply is good?  Measure that at R771 and R881, one leg of each should be at -8.6V.

[moose]

yes..the voltages at R771 & R781 both spot on -8.6, voltage across VR781 14.95 (and at each end -13.95 & -0.6)

[bdunham7]

I'll agree that Q763 looks problematic if you look at it in isolation.  If I'm thinking clearly (no guarantees) it looks like it is reversed biased in one case and should be completely off but it still allowing current through and in the other you have +3.3V_BE and it appears to not be fully on.  I'd still like to see what Q780 is doing.  As for the other two transistors that you marked as possible problems, I'd leave Q779 alone for now (but keep it in mind) and IDK what to say about Q770.  That can be replaced with a 2N3906 if you have one and want to do a quick substitution.  Q763 is designated as a "select" MPS918 which may be a bit of an issue as the p/n is obsolete so you may not have an easy time getting one let alone an assortment from which to select one--if we even knew the selection criteria.  Central Semiconductor has the PN918 which appears to be a drop-in replacement for the MPS918, but the "select" part IDK about.

Looking at the circuit as a whole, it also appears that your signal (difference when you change the knob position) appears at the base of Q763 as expected and really never appears beyond that.  If you have the time and inclination--as in you prefer precision diagnosis to shopping for a few parts--you might try another set of measurements, one with the horizontal position knob set so that you get exactly 0.00 volts at the base of Q763 (get Q780 readings this time) and another with the same settings but using a scope with a 1kHz 1V_p-p signal on CH1 (X) with that channel set to 0.5V/div.  With the scope probe (10X at least, 100X 100M if you have it) find the signal at the base of Q763 and then start tracking it through the circuit.

[moose]

thanks for all that....it may take a few days to get all done.

[moose]

oops...i somehow skipped over Q780 from my results sheet with the rest of the values from yesterday.
                   B               E                 C
Q780        +5.2            +4.4         +14.0 
Q763     -1.6/+1.2   +0.8/-2.1     +4.6/+4.8

Q780 is spot on the expected values.

i'll get on to your suggestions...

[moose]

i thought i would relook at U603 (sweep generator & logic waveforms, sheet 7)
Putting the scope as per their settings, the waveform of TP3 is as theirs, however, the waveform of TP5 & TP6 is NOT!
With TP5, the hi time is fine, but the low time is 1/2 a cycle of the TP3 wave.
Similarly with TP6, its low for the correct time, but its high time is also 1/2 a TP3 cycle.
And with TP14, the high lasts also 1/2 a TP3 cycle.
Looking at the input signals (S,R,C) for the output (Q) at pin9, pin10 (S) is constant H..is THAT supposed to vary H/L?
And pin 13 (R) is low for 5 cycles (of TP3), but comes hi for only 1/2 a TP3 cycle...is that supposed to be like that?

So perhaps the issue is somewhere in the sweep circuit?

[BlownUpCapacitor]

If X-Y is not working, that means that the sweep circuit is likely fine and that the fault it else where.. I suggest putting it to X-Y mode again, but this time feed a sine wave into CH1 and use another oscilloscope to follow the signal from WFM22, check your way through the the A12 board circuit making sure to follow the signal, and then go to WFM23. Then check the base of Q780.

If possible, please post photos of the waveforms. There should be no huge voltage offsets.

The fault is likely not in the trigger circuit nor before it because if it was, the sweep should work fine. Having both X-Y and normal sweep be faulty indicates a problem in schematic 7, the XY AMPLIFIER/HORZ OUTPUT circuit.

This is assuming there is only one fault though, not two. But two faults reproducing the same error seems unlikely.

[bdunham7]

i thought i would relook at U603 (sweep generator & logic waveforms, sheet 7)

The manual I'm looking at has the sweep generator and logic on sheet 5.  Do you have a link for yours or can you post it?

So perhaps the issue is somewhere in the sweep circuit?

I think that pulling out the 10X horizontal mag and not seeing the trace expand acorss the screen pretty much rules out the issue being at any point prior to that area.  Also, you're seeing a significant shift at the base of Q763 as you change the horizontal position in XY mode.  At this point I'd recommend obtaining a replacement transistor for Q763 and also getting a few 2N3906, if you don't have them on hand, to replace Q770.

[moose]

appologies....yes, sheet 5 not 7 (it would appear dont know my 5s from my 7s    )

i'll order some of them 2N3906 chips...on that note, trying to establish equivalent parts for obsolete parts i find rather problematic.
what would be valid replacements for the following (or are my choices valid?):
Q789 SPS6700 -> MPSA92 (?)
Q780 S032077 -> ?
Q763 SPS8002     -> MPS918 -> BC237/2N5209
Q731 SPS8223 -> MPSH69 -> 2N4557
Q736 SPS8224 -> MPSH11 ->BC847

being so far away, shipping costs are more of an issue than component costs so i tend to order from aliexpress (2 weeks shipping).

[bdunham7]

Is that SP8002 number for Q763 from the manual or is that what is on the transistor body?  I've no datasheet for that, but the MPS918 that the manual lists has a relatively low gain (min=20) and a 600MHz f_t.  So I don't think your listed subsitutes will work properly.  But I suppose you could just put a 2N3904 in there for testing, you should then be able to determine if that was the problem or not.  You should stock 2N3904/2N3906 if you're going to have electronics as your hobby!

It's probably going to be hard enough to readjust and calibrate the scope after replacing Q763 (there's a reason they are "select" and I'm guessing it is gain related) so once you verify that's the issue as best you can, I'd recommend getting as close a match as possible even if you pay some shipping.  The Central Semi PN918 would probably be your best bet.

AliExpress and eBay aren't totally terrible places to do some semi shopping, but the risks are pretty high and I'd be reluctant to buy anything that I don't have the capability to verify the specs on myself.

Edit:  Element 14 Australia has MMBT918LT1G in a SOT-23 package with a minimum order of 5.  If you have good fine soldering ability you could make that work temporarily, although I think the power dissipation is less than the TO-92 version.

[moose]

sheesh....typo...SP8002 should have read as SPS8802-1

i'll stock up on the 2Nxx & get a few PN918 as you suggested.

element14/mouser/rs-comp/digikey all have a shipping fee of NZ$19 or thereabouts...but if i get all the components at once it at least brings down the per item cost.

[moose]

the components arrived today...so i can give an update.
I removed & checked the following:
Q763 hFE=6 so replaced it with SS9018 hFE=154
Q780 hFE=154 so kept it
Q785 hFE=104 so replaced with 2N551 hFE=157
Q789 hFE=84 so replaced it with MPSA92 hFE=162
Q779 hFE=162 so kept it
Q775 hFE=101 so replaced it with 2N551 hFE=161

net result was a marginal improvement so the problem is not with that circuit.
i didnt check Q770 or Q765 as i dont see them being the cause.

Could the caps in the left part of the horiz circuit cause this issue at all??

On the other hand could it be that the trace is simply starting too late?
and if so wouldn't that indicate a blanking timing issue or a sweep issue?