Tektronix 7623 Issue with Readout

[Kidon]

Hi,

I have this 7623 3-Slot mainframe.  It was working fine this morning.
I moved it to another room, and when I powered it up, I noticed unusual characters where the readout normally is.

Has anyone ever seen something like that before?  Can anyone help diagnose and fix the problem?

Here are pics of this strange phenomenon...

Thanks for your help.

[wn1fju]

I have seen the similar thing many times with my Tek 7600-series mainframes.  It usually can be fixed by powering the scope off, removing all the plug-ins, taking a Q-tip with alcohol and rubbing the edge connectors of the plug-ins, then insert and remove the plug-in a few times to clean the connectors.  Let the alcohol dry for a few minutes and see if that helps.

[Kidon]

I have seen the similar thing many times with my Tek 7600-series mainframes.  It usually can be fixed by powering the scope off, removing all the plug-ins, taking a Q-tip with alcohol and rubbing the edge connectors of the plug-ins, then insert and remove the plug-in a few times to clean the connectors.  Let the alcohol dry for a few minutes and see if that helps.

Thanks, I’ll try that.  Although it does the same thing with any plugin I use.  I have several, and whatever combination, it’s the same.

[Kidon]

I have seen the similar thing many times with my Tek 7600-series mainframes.  It usually can be fixed by powering the scope off, removing all the plug-ins, taking a Q-tip with alcohol and rubbing the edge connectors of the plug-ins, then insert and remove the plug-in a few times to clean the connectors.  Let the alcohol dry for a few minutes and see if that helps.

Well, I did and there is no change at all.

Can anyone be of some help??  I didn’t do anything before or when it happened,  But maybe something got disconnected inside when I moved it from one room to the next???

[MadTux]

Since the < and I symbols are close to correct 1 and 2 readout and since the problem happens across different plugins, there has probably something drifted away inside the mainframe.
Character skip is also close in range, so fits the symptoms as well.
There probably is a potentiometer on readout board or so, to adjust for correct character display.

Tekwiki about Tek7000 readout system:
http://w140.com/tekwiki/wiki/7000_series_readout_system
Table with current to character map:
http://w140.com/tekwiki/wiki/7000_series_readout_system#/media/File:Tek7000-readout-charset.jpg

[Kidon]

Since the < and I symbols are close to correct 1 and 2 readout and since the problem happens across different plugins, there has probably something drifted away inside the mainframe.
Character skip is also close in range, so fits the symptoms as well.
There probably is a potentiometer on readout board or so, to adjust for correct character display.

Thanks.

All the potentiometers are documented, and unfortunately none fit that description...
Maybe something else drifted, but I haven’t the slightest idea where to even start looking...

Also, the 2nd input of one of the 7A26’s is stuck on « IDENTIFY » all the time, and the trace barely moves when I press the button.  But that was also the case when the readout worked properly.
Any clés on that will also be more than welcome.

Regards

Tekwiki about Tek7000 readout system:
http://w140.com/tekwiki/wiki/7000_series_readout_system
Table with current to character map:
http://w140.com/tekwiki/wiki/7000_series_readout_system#/media/File:Tek7000-readout-charset.jpg

[MadTux]

Have you checked the row/column match pots in the schematic below? That's where I would start fiddling.

Otherwise, do you have a second 7xxx mainframe?
If nothing else helps, I usually start swap boards with a known good instrument, to narrow the problem down to a board, to keep my sanity, like if the problem is caused by leakage/noise/bad contacts....

There likely is something misadjusted with the row decoder, decoder thinks it's in row2/0.1mA mode while it should be in row1/0mA.

[Kidon]

Have you checked the row/column match pots in the schematic below? That's where I would start fiddling.

Otherwise, do you have a second 7xxx mainframe?
If nothing else helps, I usually start swap boards with a known good instrument, to narrow the problem down to a board, to keep my sanity, like if the problem is caused by leakage/noise/bad contacts....

There likely is something misadjusted with the row decoder, decoder thinks it's in row2/0.1mA mode while it should be in row1/0mA.

No I don’t have another 7xxx mainframe.
Row/Column match pots?  I haven’t seen that yet.  In the manual?

[MadTux]

Don't know if it's written somewhere in the manual, but in the schematic, it's R2184 and R2243, in the bottom left, C/D-3/4 coordinates on schematic.

I guess these pots are there to adjust the current offset which the character decoder gets. And since the characters are decoded at an offset, I would just fiddle a bit on those pots and see if it improves anything, those trim pots on the readout board don't affect calibration after all and are rather easy to adjust back again.

[Kidon]

Have you checked the row/column match pots in the schematic below? That's where I would start fiddling.

Otherwise, do you have a second 7xxx mainframe?
If nothing else helps, I usually start swap boards with a known good instrument, to narrow the problem down to a board, to keep my sanity, like if the problem is caused by leakage/noise/bad contacts....

There likely is something misadjusted with the row decoder, decoder thinks it's in row2/0.1mA mode while it should be in row1/0mA.

Ok, I see what you’re talking about.
I’ve fiddled with these two pots.  The column one has some visible effect, but not much, and not in the right area.  It changes the “F” into a “J” and the “T” into a “R”.  The Row pot has absolutely no visible effect whatsoever.

[Kidon]

Don't know if it's written somewhere in the manual, but in the schematic, it's R2184 and R2243, in the bottom left, C/D-3/4 coordinates on schematic.

I guess these pots are there to adjust the current offset which the character decoder gets. And since the characters are decoded at an offset, I would just fiddle a bit on those pots and see if it improves anything, those trim pots on the readout board don't affect calibration after all and are rather easy to adjust back again.

Don't know if it's written somewhere in the manual, but in the schematic, it's R2184 and R2243, in the bottom left, C/D-3/4 coordinates on schematic.

I guess these pots are there to adjust the current offset which the character decoder gets. And since the characters are decoded at an offset, I would just fiddle a bit on those pots and see if it improves anything, those trim pots on the readout board don't affect calibration after all and are rather easy to adjust back again.

I don’t have these references on the readout board I have.

This is what I have:

R2214 Column Match
R2273 Character Height
R2183 Row Match
R2291 Vertical Separation
R2128 Character Scan

I’ve played with all of them, but still nothing.

[MadTux]

That distorted IDENTIFY is interesting though. If I understand Tek7000 readout function decription correctly, IDENTIFY is done on timeslot TS1 and there is not supposed to be any more characters after the IDENTIFY command.

The "Zeros and Logic IC" then triggers something in the row/column decoder, so that they generate currents so that IDENTIFY appears on screen. As far as I know, there shouldn't be any more characters after the IDENTIFY command, all other timeslots should be 0mA on row/column.

Those 2 dots on 2th and 3th character/timeslot TS2/TS3 means that something goes wrong during those 2 timeslots. First thing I would check if all the timeslot lines have valid pulses on them, maybe a timeslot line has double pulse on it or something which causes the row/column decoder to go crazy. Should be easy to measure with a second scope, since they all appear in sequence.

Other interesting thing to do would be to set a single readout plungin into 100µV or 100mV mode (with only that readout on screen), probe testpoint 8/10 with another scope and see if those voltages look similar as the example from Tek 7904A manual. (TS1:R3/C2: "add 2 zeros", TS2: R2/C1: "inverted", TS3 with R2/C10 for ">" uncal symbol, TS4: R1/C2: "1", TS8: R4/C2: "µ", TS9: R5/C2: "V")

[Kidon]

That distorted IDENTIFY is interesting though. If I understand Tek7000 readout function decription correctly, IDENTIFY is done on timeslot TS1 and there is not supposed to be any more characters after the IDENTIFY command.

The "Zeros and Logic IC" then triggers something in the row/column decoder, so that they generate currents so that IDENTIFY appears on screen. As far as I know, there shouldn't be any more characters after the IDENTIFY command, all other timeslots should be 0mA on row/column.

Those 2 dots on 2th and 3th character/timeslot TS2/TS3 means that something goes wrong during those 2 timeslots. First thing I would check if all the timeslot lines have valid pulses on them, maybe a timeslot line has double pulse on it or something which causes the row/column decoder to go crazy. Should be easy to measure with a second scope, since they all appear in sequence.

Other interesting thing to do would be to set a single readout plungin into 100µV or 100mV mode (with only that readout on screen), probe testpoint 8/10 with another scope and see if those voltages look similar as the example from Tek 7904A manual. (TS1:R3/C2: "add 2 zeros", TS2: R2/C1: "inverted", TS3 with R2/C10 for ">" uncal symbol, TS4: R1/C2: "1", TS8: R4/C2: "µ", TS9: R5/C2: "V")

I do have several other scopes I could use to check that.  I just don’t know how to do it (yet).
It’s a learning curve at the moment.

On a side note, I did remove the readout board to check the grid bias and other voltages (some were a bit off) since I got this scope a few days ago with 7 plugins, along with a 466 at a very attractive price (for Europe, I guess...).  They were both quite filthy, and in need of a calibration.  So I’m also trying to follow all the procedures, starting with this one.

But I agree, it is interesting.

[jxjbsd]

Interesting topic, my 7623 oscilloscope also has a problem of trajectory interruption.  It is necessary to understand the principle of the analog character generator, so that we can really understand where the problem is.

[doktor pyta]

In my case the problem was an IC that stores characters (PROM) on readout module.
First of all look at the manual and try to locate which IC generates characters that should be displayed but are wrong.
Often this guides to one specific IC.
Then locate the IC on PCB and try to swap with IC known to be good.

[David Hess]

The lines between the readout board and plug-ins are multiplexed placing the outputs from the plug-ins all in parallel.  So check the readout with only one plug-in installed, and move it between slots, because one malfunctioning plug-in can cause the readout from the other slots to malfunction.

[Kidon]

The lines between the readout board and plug-ins are multiplexed placing the outputs from the plug-ins all in parallel.  So check the readout with only one plug-in installed, and move it between slots, because one malfunctioning plug-in can cause the readout from the other slots to malfunction.

I have tried all the 7 plugins I have individually, as you suggested, and the result is the same.   

[Kidon]

Can anyone who has a 7623 tell me if something is connected to that location and to what/where?
See pict.

Although I don’t see any wires up in the air anywhere, maybe I missed something...

Maybe it disconnected while I was moving the scope....?

[MadTux]

Not connected, see pics:
https://amplifier.cd/Test_Equipment/Tektronix/Tektronix_7000_series_mainframe/7623A/images/7623_rechts.jpg
http://w140.com/tekwiki/images/f/f4/USA_Tektronix_7623A_Right1.jpg

If you have too much trouble repairing that board, you can also send it over to me for a little cash, I'll fix it.

[Kidon]

Not connected, see pics:
https://amplifier.cd/Test_Equipment/Tektronix/Tektronix_7000_series_mainframe/7623A/images/7623_rechts.jpg
http://w140.com/tekwiki/images/f/f4/USA_Tektronix_7623A_Right1.jpg

If you have too much trouble repairing that board, you can also send it over to me for a little cash, I'll fix it.

Unless you’re located 1 mile from my place in Paris, you’re probably way too far.   

[Kidon]

Not connected, see pics:
https://amplifier.cd/Test_Equipment/Tektronix/Tektronix_7000_series_mainframe/7623A/images/7623_rechts.jpg
http://w140.com/tekwiki/images/f/f4/USA_Tektronix_7623A_Right1.jpg

If you have too much trouble repairing that board, you can also send it over to me for a little cash, I'll fix it.

Link returns a 403 error

[David Hess]

I have tried all the 7 plugins I have individually, as you suggested, and the result is the same.   

That is actually good because it narrows the possibilities.

Reseat the connectors and ICs on the readout board.  Check the connections between the readout board and interface board which is where the plug-ins attach.

[Kidon]

I have tried all the 7 plugins I have individually, as you suggested, and the result is the same.   

That is actually good because it narrows the possibilities.

Reseat the connectors and ICs on the readout board.  Check the connections between the readout board and interface board which is where the plug-ins attach.

Thanks.  Will do and report back.

[Kidon]

I have tried all the 7 plugins I have individually, as you suggested, and the result is the same.   

That is actually good because it narrows the possibilities.

Reseat the connectors and ICs on the readout board.  Check the connections between the readout board and interface board which is where the plug-ins attach.

While checking for possible disconnect wires, I noticed this.
Both transistors have very bad and cold solder joints, and the one on The left side is actually moving.  Could very well be related.  It might have moved during transport from one room to the next and cause this behavior.
It’s not gonna be easy to resolder.  Not the most accessible area...

Could that be the problem...?   

[jxjbsd]

The 7000 and 400 series are the transistors in the Tektronix oscilloscope, which are basically installed in this way, usually without too many problems.  You can try to compress it.