Author Topic: Tektronix 2246 flickering screen, ch2 not working. (Ch2 fixed) New help needed! (Read 9220 times)

grusus · « **on:** May 25, 2020, 03:02:45 pm »

Hi.
This is my first post on this forum so hello to everyone!
I apologize in advance for my bad english. (I am from Poland)
My friend gave me this scope for free few days ago. (It was thrown out of company where he works). Don't know its history.
I am trying to troubleshoot it. I am newbie to scopes repair. I have experience in tube amp build and repair, also in electronics in general.
Everything is on video.

https://drive.google.com/file/d/1GUUY_F4MY3gDyiwNQZAO2gofrW0KoW46/view?usp=sharing

What I've done so far:
Voltages within specs. 50mv ripple on +130V, others less than 5mv. (tested with voltmeter and another scope)
Scope is responsive. All things seems to work fine but it burns my eyes with this flickering. Everything is ripped and cuted. Traces are not continous.
Also I cannot see trigger indicator on screen (I was able to see it once with full intensity. I don't remember if it was readout intensity or A or B).

About dead channel 2:
Trace is somewhere but very weak and very distorted. Position seems to do nothing. No trace visible. (It can be spotted on beam find)
Trace (I input 1kHz square) is ok at input of AT127. At input of U122 there is very distorted triangle, also on output. (on ch1 there is nice square).

Many thanks for any hint where to start with.

I use this service manual: https://www.tek.com/manual/2246-service-manual

Mine is very early model: 1Y 2246 B045340

grusus · « **Reply #1 on:** May 30, 2020, 10:24:53 pm »

Anyone?

grusus · « **Reply #2 on:** June 18, 2020, 07:28:24 pm »

I have changed all electrolitics on all boards, and checked other caps against shorts. This helped a little.

I was checking all test points with another scope and found something interesting.
Schematic on page 323, Waveforms on page 319. - I will refer to them in my post.

TP 6A and 6B looks fine. (pins 3&5 of U802)
Pin 1 is readout input. Looks like 6A but with lots of garbage. Its shaking also. I think it should looks like 6A
On the output of U802 Pins 12&13 +8,8V's are spot on. However 6D is square with lots of garbage but very low amplitude (many times less than it should be) and 6E is more like square also but higher amplitude than 6D.
6H&6I -58V and -88V. And also like square wave. They Don't even looks like they should.

Any sugestions where to look next rather than changing U802? ( tek234-0401-21)
Also does anybody know where I can get it from along with at127? Maybe some direct replacement's?

BreakingOhmsLaw · « **Reply #3 on:** June 18, 2020, 08:40:14 pm »

I'd start checking around the circuitry for the Z-Input, itlooks like it is blanking. However, the text should never be affected, Z-Inut is gated to the traces only.
The high voltage seems ok, or you would see the image pumping in size as well, but it just blanks.
As the backlight of the buttons are flickering as well, you do have a power supply problem somewhere. If all rails measure ok, a bad ground connection might be the trouble.
Check the ground points.

grusus · « **Reply #4 on:** June 18, 2020, 10:52:49 pm »

Yes flickering of backlight of the buttons stopped after recap. Now only cuted traces with blank spots is the problem and flickering of the readout(when readout is turned ccw flicker of traces and ghosts on crt dissappears). I suspect horizontal circuit as I mentioned in previous post.
Q809&Q810 and surrounding resistors are ok. Voltages around also. I will check test points for z input tomorrow, as i ended checking on horizontal amplifier because i found issue there.
Image in size is rock steady.

grusus · « **Reply #5 on:** June 20, 2020, 09:21:14 pm »

I made some photos of traces.
V/div and time base like in service manual.
6A: https://photos.app.goo.gl/3JPDuHeFoFpadMsQA
6B: https://photos.app.goo.gl/Qw58roiYCzeScTVQ7
When I go from A sweep to B sweep I see this:
https://photos.app.goo.gl/yyLausR5frKo5FMv5
Trace is on left half of screen.(that was from time I get the scope)
When I turn sec/div 1 position ccw i get this:
https://photos.app.goo.gl/aiXmWaZn2VDtXKoc9
6B then is: https://photos.app.goo.gl/nkmv5ZFTY3fGdKAP9
Turn sec/div back one position cw I get this:
https://photos.app.goo.gl/UUfgEAHQD5jTt4tUA
6B Looks this: https://photos.app.goo.gl/JAdj8BHENuPWcrm26
I think that is not a normal behaviour.

6C: https://photos.app.goo.gl/zTbj4Fm3LGL17TYu7
6D: https://photos.app.goo.gl/UNecEuXAGtQXzjSR7
6E: https://photos.app.goo.gl/yiykE79w3CnfjKah9
6F: https://photos.app.goo.gl/4Z1engjbXPfp5hzM8
6G: https://photos.app.goo.gl/RjmTHAermev4fa9r6
6H: https://photos.app.goo.gl/XJtnovY3V2vqp9TV7
6I: https://photos.app.goo.gl/NaW8gZc32HAAKFuX8

Scope behaves like this (previous video was before recap):
https://photos.app.goo.gl/GMHjh4wK8cWLfcCc9

I am wondering why holdoff is influenciing traces the way they cuts, and also readouts flickers more or less with different position of holdoff. Also scale ilumination has influence on overall flicker.
Other thing I noticed: There should be grid with dots all over crt when readout is turned on (seen on others video). Here they are present but they are jumping/floating around from left to right very quick.

Other thing is Pin1 of U 802 is Readout horizontal input and looks like this:
https://photos.app.goo.gl/TYZogpEe5RccrkTXA

Output of U801A Pin1 is:
https://photos.app.goo.gl/tDQGWUuNRZYTp7jMA
https://photos.app.goo.gl/aWBX1pYvkmcbiNAq8

I am wondering if there is some timing issue, considering strange behavior in b sweep and no stable matrix of dots with readout. Maybe it is influencing also a sweep and overall performance. Where to look for it?
As I said, I am noob in scope reapirs, trying to learn from people who are much better than me. This is for my personal use. Any help like "look in baseline stabilizer" or similar will be apreciated.
Many can say buy a working scope. There are two things: Money (good scopes cost a lot) and It is nice to use equipement that you have reapired yourself.:-)

Also scope behaves very odd when it is cold. It must run for 5 minutes and is operational as on video in this post. For first 5 minutes uncal are blinking, brightness of crt varies. and eventualy it gets as bright as it shold be. There are random brights up during first 5 minutes.

BreakingOhmsLaw · « **Reply #6 on:** June 21, 2020, 07:11:03 am »

Some Questions:
- Does that behaviour change when switching between alt and chop mode?
- Does the Z-Input work?( If you don't have a signal generator, just connect a x1 Probe to the Z-Input and connect it to the probe calibration signal output.)

Seeing your latest video, that still looks like something is messing with the Z-modulation.

grusus · « **Reply #7 on:** June 21, 2020, 03:02:22 pm »

First few minutes of operation at cold start up(left off overnight).:
https://photos.app.goo.gl/P1hn33ym8zmrWHf27

Quote

- Does that behaviour change when switching between alt and chop mode?

See video:
https://photos.app.goo.gl/3y8rhsYRvHtdtoLF7

Quote

- Does the Z-Input work?( If you don't have a signal generator, just connect a x1 Probe to the Z-Input and connect it to the probe calibration signal output.)

BreakingOhmsLaw · « **Reply #8 on:** June 21, 2020, 08:39:27 pm »

Okay, we can learn a couple of things here:
The blanking changes when we switch between alt and chop.
Quick theory ofvoperation recap:
In chop mode, the scope switches between the two inputs during the trace, instead of drawing the traces after each other.

Chop is used to display the two traces time synced. When it changes to the other trace, the scope briefly blanks the trace to avoid drawing lines between the traces whenever it chops to the other trace. Generally, the same circuit that does the blanking in chop mode, also serves the Z-Input. I am on mobile now, so i can't dig into the service manual of this particular scope. When you switch to alt, the blanking stops (or just sees the blanking from Z-input, as I don't know if the Z-Input was active in your first video).
So you need to look into the part of the circuit that creates the chopper frequency and does the trace blanking. Read the circuit descriptions, those are usually very good in Tek service manuals. Don't measure any test point for now - first, really wrap your head around the circuit and understand how it works.

As those gaps in the trace are very wide, i wonder whether the chop frequency is simply too low. Do the gap "pauses" change with different timebase settings? Locate the source of the chop frequency for starters and check it.

grusus · « **Reply #9 on:** June 21, 2020, 09:38:09 pm »

External Z-axis input is active only on those two videos:

Quote

Quote
- Does the Z-Input work?( If you don't have a signal generator, just connect a x1 Probe to the Z-Input and connect it to the probe calibration signal output.)

My signal generator is 0-3,8V square at 10000Hz. Multimeter shows ac on output of generator.
See video:
Holdoff full cw:
https://photos.app.goo.gl/JbHSfhqot4DuVBZc7
Holdoff full ccw:
https://photos.app.goo.gl/beh9krctcym6mdWu5

Different timebase settings makes chops wider or more narrow. To make them pause I must adjust the holdoff. Even if the do not pause, they are there but scrolling very fast throu whole waveform.(Don't know if you can see this on video).

I will check service manual for that and reply to you when i will get into some thoughts.
I am not in a hurry with this scope, so if you will manage to locate yourself the source of chop frequency before me please reply.
I have surgery in hospital from june 29, and will be off in hospital for 5-7 days. So no panic, no hurry :-)

grusus · « **Reply #10 on:** June 21, 2020, 10:31:21 pm »

Manual p. 53 explains it as I think.

As I think:
Clock for chop blanking is generated at U600 pin 18, then goes to U602. It is test point 4F. Clock source is 10MHZ crystal oscillator that goes to pin 1 (TC) of U600
There is also another clock input to U600 pin 2 (LFC), from probe adjust circuit
Holdoff is also connected to U600.

If what I am thinking is right, I should test 4F (blanking generated by U600), 4G (10MHZ crystal input to U600), 4H (the other clock input to U600).

I will do that tomorrow. :-)

Also I found something in Troubleshooting scopes by tek on page 40 (46 according to pdf reader):
https://www.sphere.bc.ca/test/tek-parts/troubleshooting-scopes.pdf

From what I have noticed all neons are propably good - they flash during turning off the scope.
HV caps tested with ESR meter and capacitance meter - good. I do not have HV leakage tester that can go almost 3kV.
Heather to cathode leakage in unblanking circuit? Where to look for it? They do not mean crt heater to cathode leakage? But when there will be leakage from cathode to heater of crt neons should light during normal operation? I am confused.

BreakingOhmsLaw · « **Reply #11 on:** June 21, 2020, 10:50:00 pm »

I don't think it is leakage in the CRT, the blanking looks too digital.

Also, check the selection circuit. The 2246 has separate pots for trace A, B and the readout, which is nice.
Depending on what the scope is drawing, the right intensity signal must be fed into the blanking circuit, test point 7B.
U602 must select *one*, and only *one* intensity from the opamp U1001. It will switch one of the transistors Q1001 through 1004 on. Q1005 is controlled by the Z-input.
Check:
Q1001 to Q1005 - if failed short, two signals get mixed.
CR1001 and CR1002, these mix Z-Blanking into Intensitiy A and B. If they fail short, A and B get mixed.

Edit: The chopper frequency is 1,25MHz with 200ns blanking, so in chop mode you should see a signal 600ns ON, 200ns OFF at 7B.
As your second cope is anaolog, you will be facing some trouble triggering that - the scope will update the readout in between traces, so disable that. But looking at your screenshots, you already figured that out yourself.

You will still get an additional blank during flyback of the trace to the left, just select a low timebase setting on the dead scope and you'll be fine.

rf+tech · « **Reply #12 on:** June 22, 2020, 01:20:20 am »

The 2246 has an adjustment (R724) for Readout Jitter, the procedure is paragraph 7 of Section 5 of the service manual.

I think BreakingOhmsLaw is on the right track and adjusting R724 is not likely to correct the problem, but may reveal some additional clues. It would be wise to mark R724 before altering its setting.

RF+ Tech

grusus · « **Reply #13 on:** June 22, 2020, 02:14:53 pm »

Mine doesn't have R724. It is very early model.(look at the end of first post) R724 was my first thought after reading service manual :-)
https://photos.app.goo.gl/Q6k2vWTNtEyj51gD6

I was monitoring pins 17-20 of U600 while changing A alt and B + regulating RO intensity.

Only A:
Signal is present on pin 17. other three are very low with some garbage. RO makes signal on pin 20 to come up, also is added to pin 17.

Alt:
Signal is present on pins 17,18,19. RO intensity makes signal on pin 20 to come up + also added to 17,18,19

B:
Signal is present on pin 18. other three are very low with some garbage. RO makes signal on pin 20 to come up, also is added to pin 18.

LM324(u1001) is making problems? I have some on hand.

EDIT:
Maybe there is some kind of instability on U1001 making hf oscilation on 17-20pins of U600? I think this "garbage" as i said in upper section seems to be low level hf oscillation. I will check that and post photos in next post.
/Edit

Q1001-Q1005 seems fine. No shorts. Tested with diode tester in circuit and voltage drops seems fine all junctions.
CR1001 and CR1002 also looks fine.

7B in alt mode(.2V 1ms):
https://photos.app.goo.gl/QhGybyn3vXNy9oet5
7B in chop mode(.2V 1ms):
https://photos.app.goo.gl/S3BoaWg3VY6CK5yv5
7B in chop mode zoomed(.2V 1us):
https://photos.app.goo.gl/Gtb8qJDQS4gA4yCz5

Yeah readout intensity is always off :-) almost:-) during measurements.
I have also DSO chinese small digital 1 ch scope if you preffer :-)

About diagram 4 test points.
Checked all just in case.

4A: https://photos.app.goo.gl/k5QuiAPzVqiz13wz7

4B: depends on holdoff:
https://photos.app.goo.gl/PkyR6uxqxPU9NBDQ6
https://photos.app.goo.gl/VTEujTk2Zsu86k6x8

4C: depend on holdoff:
https://photos.app.goo.gl/49W4HVX14epxr1Zh8
https://photos.app.goo.gl/Kdji3TBwa1NLCTmw8

4D: https://photos.app.goo.gl/YTAzB3Hs9viSXXMx9

4I: https://photos.app.goo.gl/cNPhUh2RZvzVorETA

4E: https://photos.app.goo.gl/6JjX9oXBrAuS5SoH6

4F:
RO int at 0:
https://photos.app.goo.gl/ns33sALJKYZi2Exu9
RO int turned even slightly on:
https://photos.app.goo.gl/eGSgmAS7FCeLDQ9J8

4G: 1v 0.2us
https://photos.app.goo.gl/1ACwZaQpRGkPbM2v6

4H: https://photos.app.goo.gl/1jLUK3Bt47wggyqb7

I think I covered everything :-)

BreakingOhmsLaw · « **Reply #14 on:** June 22, 2020, 06:31:03 pm »

Okay, 7B looks exactly what is is supposed to look like. Now we have this ruled out, lets take a look at the trigger circuit.
I have the sneaking suspicion that it does not wait for hold-off timer to finish before starting the next trace.
Let me dig into that circuit tonight and i will come back to you.

Edit: Please select single beam, slowest timebase setting and verify that the holdoff oscillator is only triggered/running when the sweep has finished and not while sweep is running. TP 4I looks okay, but i want to double check if that is correlated to the sweep actually ending.

grusus · « **Reply #15 on:** June 22, 2020, 09:21:07 pm »

4I: I think that this is not what you expected :-( Seems fine.
https://photos.app.goo.gl/QTXHWLj9EJf8VA166

U602 18,19,20 when A sweep is chosen without readout. 50mv@5ms
https://photos.app.goo.gl/xZKDYhFNxnyJy1x79

EDIT
Here are at: 50mv 5us.
https://photos.app.goo.gl/ydj7GjzREKoZAaqo6
And direct output of U1001 pin 8 or 7 (same traces) 50mv@5us
https://photos.app.goo.gl/5R73sjkcScU56Av58
/Edit

While pin 17 of U602 is:
https://photos.app.goo.gl/BMTQmJCUKCG5b6dM8

Also I catched light ups on cold start up: (sorry for background music)
https://photos.app.goo.gl/LufcFkhNoHkAHf9Y8
Also trigger led is blinking.

Another thing on cold startup. Even if i turn readout full ccw - they are still there.:
https://photos.app.goo.gl/AoCZNrB3nLEbc57o9

I am andjusting RO brightness only when they fully appear on screen on video. When the pattern is seen with barely visible readout, it is full ccw. (and yeah I have cleaned this pot and tested it many times - it is good)

rf+tech · « **Reply #16 on:** June 23, 2020, 12:32:38 am »

Quote

Another thing on cold startup. Even if i turn readout full ccw - they are still there

During the cold startup when the RO intensity is full ccw, is there a signal present on U602 pin 20??

BTW, if you have not searched the forum, another 2246 owner posted last November, with Z axis problems. Although the OP never responded and the outcome is unknown, the same circuitry is involved:

https://www.eevblog.com/forum/repair/tektronix-2246-z-axis-issues/

The OP in that thread provided some DC measurements.

grusus · « **Reply #17 on:** June 23, 2020, 10:24:09 pm »

Quote

During the cold startup when the RO intensity is full ccw, is there a signal present on U602 pin 20??

Yes.
https://photos.app.goo.gl/adxJmUETUFK9Zim1A

Maybe you will not see it, but on cold start up pin 20 is almost the same as after some time. Thing is that when scope warms-up i must turn RO intensity up and then down and then pin 20 is quiet, and ghosting effect from readout on crt then dissapears with RO intensity turned full ccw. When sope is cold sometimes it do the trick, sometimes doesn't. But ghosting is always at turn on cold. Probably signal on pin20 of U602 either.

I will check voltages in Z -axis tomorrow - run out of time today. I have searched forum and seen that topic, but as you wrote outcome from that topic is unknown.

rf+tech · « **Reply #18 on:** June 23, 2020, 11:51:40 pm »

Preset RO intensity control to mechanical center. At cold start, note if ghosting is present, U602 pin 20 noisy.
Check U1001C pin 8 DC voltage and voltage at R506 RO control wiper side.
Both should be close to zero Vdc, with pin 8 1.53 * RO wiper Vdc.

Turn the RO intensity up slowly, check that U1001 pin 8 smoothly goes to +3.8 Vdc.
Turn the RO intensity down slowly, check that U1001 pin 8 smoothly goes to –3.8 Vdc.
Is there any sudden change at U1001 pin 8, when U602 pin 20 becomes quiet?

Does cold spray on U602 or U1001 have any effect? Hot air?

Following with much interest

grusus · « **Reply #19 on:** June 24, 2020, 03:33:33 pm »

I am back with some results :-)
At the end is scan with measured voltages.

I have noticed something today that I overlooked. No matter if scope is cold or warmed up, when I power on the scope with RO intensity turned full ccw this gosting effect is present till I turn RO Int slightly cw and then back - then it dissappear. Difference is that on cold startup i doesn't do the trick. Ghosting is present till scope warms up and I Use RO intensity the way I explained before.

See video (actually i captured one of very rare situations when after restart with ro intensity turned ccw, there was no ghosting effect (usually it is almost always like on first start on this vid.)

. Those was restarts after warm up.
https://photos.app.goo.gl/oeVBXWVrw1xtBbCL6

In case we missunderstood of ghosting effect.
Normally on this scope when RO intensity is turned up there is a matrix of dots generated with readout and readouts are displayed sharp with no blurr around and no flickering. In my case matrix of dots shakes and are blurry + they are running from lef to right of crt and readouts have blurr also, and are slightly dancing around + flickering.

Voltages goes smoothly from 3,8 to -3,8 and also from 2,5 to -2,5.
No sudden change on pin 8
At mechanical center of RO int voltages are near zero V.

I run out of cold spray - must order. Heating with hotair shorten time needed to make ghosting dissappear. U1001 during operation i s normally cold. U602 gets warm approx +50 degrees celcius (122 F).
U600 is also cold as cucumber :-) Maybe slightly warm.

I am slowly suspecting U602 is going bad. What should I check also to rule out or confirm U602 fault?

Edit:
Look what I found today:
https://photos.app.goo.gl/RPQQkWBhBpVKKWdGA

DS2704 is glowing. I am pretty shure it didn't before. When I turn focus up it is glowing brighter. When down it dimms.
New problem or indication of problem that already was there? :-) This scope will kill me. :-)

rf+tech · « **Reply #20 on:** June 24, 2020, 10:11:01 pm »

Okay, the DC voltages all look reasonable, so I think we can rule out the analog portion of RO intensity.

I agree that U602 looks suspect, given that extra heat shortens the time for ghosting to clear. One can try ice in a plastic bag, to cool U602.

How to proceed. Following W2304 (top left) pins 10.11.12 to sheet 11 shows each line is essentially not connected. The right end of R505, R506 and R507 each go to U506, a CD4051 multiplexer. There is a remote chance U506 may be defective and some crosstalk from another channel is interfering with the RO intensity signal. CD4051 is inexpensive and easier to change, than U602.

Looking at U602 bottom left, address lines A0 and A1 provide control over which of the four intensity gate outputs are active at a given moment. If the address decoding inside U602 is knackered, then one or more of the other three intensity gates could be turned on while RO is active. Try removing Q1001, or just lift the emitter, to eliminate the /A intensity gate signal and see what effect, if any, this has. Then do the same with Q1003 and Q1004. What happens with all three intensity gate signals disconnected?

DS2704 glowing, Channel 2 not working. This is mild.

grusus · « **Reply #21 on:** June 25, 2020, 01:32:30 pm »

Hi I was doing my weekly order from part shop and I ordered some cold spray and CD4051 + some other goods.

Here are results. (removed all of mentioned components, not only lifting emitter.

Q1001 removed only:
https://photos.app.goo.gl/CwJFG4ygYeditFMV6
That trace which i seen on crt: I was on A sweep and tried to move trace up and down and it was moving.

Q1002 removed only (1001 back in place):
https://photos.app.goo.gl/os2GupcJk6iPafud8
Ghosting still present.

Q1003 removed only:
https://photos.app.goo.gl/gaJ9RwNB5SaXr91d9

Q1001-1003 removed:
https://photos.app.goo.gl/rrd5L59zoodTusMu5

Q1004 removed only:
https://photos.app.goo.gl/XWp3KixkNR3sSV6Z6

Q1001,Q1003,Q1004 removed:
https://photos.app.goo.gl/zL5ZL5GcCPRtqWCc8

Q1001-Q1004 removed:
https://photos.app.goo.gl/92KqESBrx1bUhvpW6

Still you can see traces from A and B sweep. Even with A,B, RO intensity turned fully CCW.

Checked Q1001-Q1004 out of circuit. Seems ok on transistor tester.

Also checked CR1001&CR1002 out of circuit. Also seems ok. Tested on multimeter in diode mode and with T7 multi function tester.

It is getting into suicide mission to make this scope working again

Tube amps, even those nasty ones are much easy to fix :-) Even laptop motherboards never made to me this much of trouble. Maybe that's because I know how to fix them

BreakingOhmsLaw · « **Reply #22 on:** June 25, 2020, 02:52:13 pm »

Perhaps the fault is in the blanking circuit itself. Are you bable to measure up to 3000V?

Try:
- single trace mode A, GND
- Apply 3V to the Z-Input
- hide readout

The screen should be completely blank. If it isnt, the blanking circuit itself (past 7B) may be flawed. The circuit will generate a voltage up to 2700V, you need a HV probe to test.
You can get a good one for around 150zł on ebay, they are no longer used much since CRT TVs are gone.

grusus · « **Reply #23 on:** June 25, 2020, 03:25:49 pm »

Here are the results:
https://photos.app.goo.gl/x45VZydsNB8Pvrwm7

Oops I bet that I have adjusted grid bias before.

Here is after adjustment.
https://photos.app.goo.gl/8HLGoy7qP2BrKCr39

Can you pm me some links to good HV probes? My multimeter goes to 1000V DC (Brymen BM817)

rf+tech · « **Reply #24 on:** June 25, 2020, 11:26:40 pm »

Grid bias is relative to the cathode, just the same as in a tube amplifier. The absolute -HV is of somewhat secondary importance. As long as the trace can be focused sharply, the -HV is close enough to not be a concern.

Simply adjusting the bias control to extinguish the beam, with 3 Vdc on the Z-axis input, will get closer to the optimum setting.

Best would be to follow the Grid Bias adjustment procedure outlined in section 5-2. This procedure does not require a HV probe.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Tektronix 2246 flickering screen, ch2 not working. (Ch2 fixed) New help needed! (Read 9220 times)

Share me