Topics

7A26 CH2 Readout


Stephen
 

One of my 7A26’s has a readout problem on CH2. It displays « INDENTIFY » all the time, and never displays the voltage range. Also, the push button has no effect on what is displayed, and only move the trace up just very little compared to the other channels. This happens whatever slots it’s in. So it must be something to do with the plugin itself, not the readout board of the mainframe. All other 7A26 I have, don’t have that issue.
I looked at the schematic, and guessed that something must be wrong with either Q620, or CR621.
PS: The serial # of this particular unit is below B149999.
I replaced Q620, and now it displays the voltage range, but not « IDENTIFY » anymore when the button is pressed. But the trace moves a tiny bit, like it was before. Replacing CR621 didn’t help either.
I don’t quite understand how this circuit works, but I assume that something around that Darlington must be bad.
BTW, the schematic calls for a 2n5305, and I replaced it with what I had, a BC517. The pinout is different, but I’m pretty sure I got it right.

Any help is welcome.

Thank you.


Eric
 

I had one that was doing the same thing on channel 2 it ended up being a bad BNC connector on the front panel. There is a low Ohms resister from the BNC scene ring to the push button switch if you lift the resister and it drops out of identify check the resistance to ground from the pin on the jack on a healthy jack it should show open mine was 39K Ohms.

Eric

On 8/20/2020 4:01 PM, Stephen wrote:
One of my 7A26’s has a readout problem on CH2. It displays « INDENTIFY » all the time, and never displays the voltage range. Also, the push button has no effect on what is displayed, and only move the trace up just very little compared to the other channels. This happens whatever slots it’s in. So it must be something to do with the plugin itself, not the readout board of the mainframe. All other 7A26 I have, don’t have that issue.
I looked at the schematic, and guessed that something must be wrong with either Q620, or CR621.
PS: The serial # of this particular unit is below B149999.
I replaced Q620, and now it displays the voltage range, but not « IDENTIFY » anymore when the button is pressed. But the trace moves a tiny bit, like it was before. Replacing CR621 didn’t help either.
I don’t quite understand how this circuit works, but I assume that something around that Darlington must be bad.
BTW, the schematic calls for a 2n5305, and I replaced it with what I had, a BC517. The pinout is different, but I’m pretty sure I got it right.

Any help is welcome.

Thank you.


 

- Pushing Identify should move the trace up at least one minor division. Not sure if that happens, since you're writing "a tiny bit".

- If you don't have an extender, you may consider putting the 7A26 without its right side panel into the rightmost slot (yes, that's a "horizontal" slot but this a 7000 system, use the orthogonal architecture to your advantage), remove the right side panel of the 'scope and measure voltages on the readout boards live.

I guess you already checked the following:
- Without depressing the Identify button, the voltage at the Identify ring around the BNC input connector should be at about -15V. Depressing it should change that to about 0V. See Dwg 3.
- The connection between CR621 (anode), R620 and R621 should behave about the same (voltages and level shift). If it doesn't, something 's probably (still) wrong with / around Q620.

- Did you measure voltages around the darlington on the CH1 readout board and compare? You may consider swapping the darlington from CH1 in.

Raymond


Stephen
 

On Thu, Aug 20, 2020 at 09:33 AM, Eric wrote:


I had one that was doing the same thing on channel 2 it ended up being a
bad BNC connector on the front panel. There is a low Ohms resister from
the BNC scene ring to the push button switch if you lift the resister
and it drops out of identify check the resistance to ground from the pin
on the jack on a healthy jack it should show open mine was 39K Ohms.

Eric
You were right. The ring was shorted to ground. Removed the little shield at the back of the BNC, desoldered the 100R resistor from the ring, and the ceramic cap from the ground, inspected the BNC from the back. I couldn’t really see anything except for maybe a tiny « something » stuck in between the ring and the shield. I’m not very sure what it was. I started twisting very slightly the BNC from the front, less then 1/4 of a turn, and sure enough continuity from the ring to ground disappeared. Now all is working perfectly......
Except.... BACKWARDS. 🤷‍♂️

It’s always on « IDENTIFY », and when I press the button, the trace moves exactly as it should, and the display shows the voltage range... Backwards....

Hmmm.. What did I do wrong? Did I put the new Darlington equivalent (Q620) backwards? That wouldn’t do that, would it?


Stephen
 

On Thu, Aug 20, 2020 at 10:04 AM, Raymond Domp Frank wrote:

Interleaved.


- Pushing Identify should move the trace up at least one minor division. Not
sure if that happens, since you're writing "a tiny bit".
What I meant was about a 1/10 of a division.... virtually nothing. Barely visible.

- If you don't have an extender, you may consider putting the 7A26 without its
right side panel into the rightmost slot (yes, that's a "horizontal" slot but
this a 7000 system, use the orthogonal architecture to your advantage), remove
the right side panel of the 'scope and measure voltages on the readout boards
live.

I guess you already checked the following:
No, I don’t have an extender. And I haven’t checked the voltages yet, precisely because of that.
I didn’t know you could do that to check voltages. But I guess that makes sens. Thanks for the tip, it may come in handy.

- Without depressing the Identify button, the voltage at the Identify ring
around the BNC input connector should be at about -15V. Depressing it should
change that to about 0V. See Dwg 3.
- The connection between CR621 (anode), R620 and R621 should behave about the
same (voltages and level shift). If it doesn't, something 's probably (still)
wrong with / around Q620.

- Did you measure voltages around the darlington on the CH1 readout board and
compare? You may consider swapping the darlington from CH1 in.
I haven’t checked any voltage, because of the lack of an extender (previous answer)

Thank You Raymond.


 

On Thu, Aug 20, 2020 at 10:01 PM, Stephen wrote:


I replaced Q620, and now it displays the voltage range, but not
« IDENTIFY » anymore when the button is pressed. But the trace moves a
tiny bit, like it was before.
That seems incompatible with this:

On Thu, Aug 20, 2020 at 10:01 PM, Stephen wrote:

Also, the push button has no effect on what is displayed, and only move the trace up just very little
compared to the other channels
and this:

On Thu, Aug 20, 2020 at 11:48 PM, Stephen wrote:


You were right. The ring was shorted to ground.

It’s always on « IDENTIFY », and when I press the button, the trace moves exactly as it should, and
the display shows the voltage range... Backwards....

Hmmm.. What did I do wrong? Did I put the new Darlington equivalent (Q620) backwards?
That wouldn’t do that, would it?
So that's different behavior from what I understood earlier.
It looks like the darlington isn't functioning as it should.

The simplest thing to do probably is putting the 7A26 in the rightmost slot as I described earlier and do live (comparative) voltage measurements.
Creating a convenient working environment often makes up 50% of the job.

Raymond


 

Your latest message arrived just after I sent mine. Please take that into account when reading it.

Raymond


Stephen
 

On Thu, Aug 20, 2020 at 11:08 AM, Raymond Domp Frank wrote:


On Thu, Aug 20, 2020 at 10:01 PM, Stephen wrote:


I replaced Q620, and now it displays the voltage range, but not
« IDENTIFY » anymore when the button is pressed. But the trace moves a
tiny bit, like it was before.
That seems incompatible with this:

On Thu, Aug 20, 2020 at 10:01 PM, Stephen wrote:

Also, the push button has no effect on what is displayed, and only move the
trace up just very little
compared to the other channels
and this:

On Thu, Aug 20, 2020 at 11:48 PM, Stephen wrote:


You were right. The ring was shorted to ground.

It’s always on « IDENTIFY », and when I press the button, the trace
moves exactly as it should, and
the display shows the voltage range... Backwards....

Hmmm.. What did I do wrong? Did I put the new Darlington equivalent (Q620)
backwards?
That wouldn’t do that, would it?
So that's different behavior from what I understood earlier.
It looks like the darlington isn't functioning as it should.

The simplest thing to do probably is putting the 7A26 in the rightmost slot as
I described earlier and do live (comparative) voltage measurements.
Creating a convenient working environment often makes up 50% of the job.

Raymond
It seems « incompatible » because you’re mixing excerpts from different time lines. As I mentioned, things have happened in between.


 

On Fri, Aug 21, 2020 at 12:20 AM, Stephen wrote:


It seems « incompatible » because you’re mixing excerpts from different
time lines. As I mentioned, things have happened in between.
I thought as much but couldn't match it. Actual situation is you have a reversed operation of Identify after replacing Q620 and removing a short between the Identify ring, right?

Raymond


 

On Fri, Aug 21, 2020 at 12:25 AM, Raymond Domp Frank wrote:


between the Identify ring, right?
...between the Identify ring and ground, right?

Raymond


Stephen
 

On Thu, Aug 20, 2020 at 11:25 AM, Raymond Domp Frank wrote:


On Fri, Aug 21, 2020 at 12:20 AM, Stephen wrote:


It seems « incompatible » because you’re mixing excerpts from
different
time lines. As I mentioned, things have happened in between.
I thought as much but couldn't match it. Actual situation is you have a
reversed operation of Identify after replacing Q620 and removing a short
between the Identify ring, right?

Raymond
Correct. I replaced Q620 with an equivalent that has a different pinout. I think I did put it the right way.
Now, after I did this, I didn’t have « IDENTIFY » showing up all the time, but the trace was still not moving upwards the way it should. And instead of « IDENTIFY », I had the voltage range displayed. Which is the normal behavior.

Then I checked what Eric suggested, and sure enough, the ring was shorted to ground, the same way his was.
When I removed the short, the trace was moving about 1 division BUT, « IDENTIFY » was the default readout, and the voltage range was showing when the button was pushed. In other words, it’s behaving as it should, meaning the button does have an effect on the readout and the trace, but backwards. Am I making sense?


Stephen
 

On Thu, Aug 20, 2020 at 11:27 AM, Raymond Domp Frank wrote:


On Fri, Aug 21, 2020 at 12:25 AM, Raymond Domp Frank wrote:


between the Identify ring, right?
...between the Identify ring and ground, right?

Raymond

Yes


 

On Fri, Aug 21, 2020 at 12:36 AM, Stephen wrote:


Then I checked what Eric suggested, and sure enough, the ring was shorted to
ground, the same way his was.
When I removed the short, the trace was moving about 1 division BUT,
« IDENTIFY » was the default readout, and the voltage range was showing
when the button was pushed. In other words, it’s behaving as it should,
meaning the button does have an effect on the readout and the trace, but
backwards. Am I making sense?

Yes, as a description of the actual situation, It seems to make sense.
However, the BNC Identfy ring (which was shorted to gnd) and the Identify button are switched in parallel, see dwg 3 (if I'm looking at the right schematic version). That means, if *either* is connected to gnd, the other doesn't matter and has no further influence, unless... the ring-short wasn't low-impedance enough. That would explain the slight movement when pressing Identify.
The trace jump doesn't need and doesn't work in parallel with the readout showing either IDENTIFY or voltage scale. That's the darlington circuit's area. The trace jump circuit is completely on dwg 3.

Again, I advise setting up and doing the live measurements. Being able to look at CH2 and where necessary comparing CH1 and CH2 physically and logically next to each other should really make this a lot easier.

Do have a look at the junction of CR621, R620, R621 and C621 in CH2 and compare with CH1.
This is such a simple circuit that finding the fault shouldn't be that difficult. This remark is in no way meant to criticize you; there's no reason, looking at what you've achieved with this group recently. Just make life a bit easier, especially if you don't understand the circuit completely, as you wrote earlier.

Just one bit of information that you may not be aware of: The whole circuit of Q620 and what is drawn to its right in dwg 5 is meant to set currents that the analog readout control uses for generating its information to be displayed: Resistors are switched in and out and Q620 is just one of the devices that with its series resistor R622, controls a current, like the other mechanical switches to the right of it with their series resistors like R630, R631, R647, R648, etc. Read the interesting description of how the readout works here:
https://w140.com/tekwiki/wiki/7000_series_readout_system
or in one of the 7000-series mainframe manuals. Many (not all) go into great detail on this. It's a brilliant circuit, devised by a Brit called Barrie Gilbert.

Don't forget (in general): It's not completely unthinkable that a previous owner modified the circuit...

One last question: Are the voltage scale indications (V/div) correct?

Raymond


Stephen
 

On Thu, Aug 20, 2020 at 12:29 PM, Raymond Domp Frank wrote:

Interleaved.

Yes, as a description of the actual situation, It seems to make sense.
However, the BNC Identfy ring (which was shorted to gnd) and the Identify
button are switched in parallel, see dwg 3 (if I'm looking at the right
schematic version).
I’m sure you are.

That means, if *either* is connected to gnd, the other
doesn't matter and has no further influence, unless... the ring-short wasn't
low-impedance enough. That would explain the slight movement when pressing
Identify.
I suppose it was very low imperfections. It was just, I think, what seemed to be, some kind metal part, less than a 1/10 of a mm long. I was very hard to tell, even with the magnifying glass I used.

The trace jump doesn't need and doesn't work in parallel with the readout
showing either IDENTIFY or voltage scale. That's the darlington circuit's
area.
That’s what I figured, without understanding how it actually works.

The trace jump circuit is completely on dwg 3.
Okay.

Again, I advise setting up and doing the live measurements. Being able to look
at CH2 and where necessary comparing CH1 and CH2 physically and logically next
to each other should really make this a lot easier.
I will definitely do that, or try to, next time I run into an issue. And since I bought quite a few things recently that need restoring or minor adjustments, it shouldn’t be long before I do.

Do have a look at the junction of CR621, R620, R621 and C621 in CH2 and
compare with CH1.
This is such a simple circuit that finding the fault shouldn't be that
difficult. This remark is in no way meant to criticize you; there's no reason,
looking at what you've achieved with this group recently.
Thank you. I’m doing my best, in my spare time, to educate myself and try to understand all this, little by little.
Since I had the readout board unscrewed to take those measurements, I figured I’d put back the original Q620.
And guess what. Everything is now back to normal, and working properly. I haven’t tried, but I must have mixed E and C on the equivalent I used. Could that reverse the order of things like I experienced?

Just make life a bit
easier, especially if you don't understand the circuit completely, as you
wrote earlier.
That is correct, I don’t.

Just one bit of information that you may not be aware of: The whole circuit of
Q620 and what is drawn to its right in dwg 5 is meant to set currents that the
analog readout control uses for generating its information to be displayed:
Resistors are switched in and out and Q620 is just one of the devices that
with its series resistor R622, controls a current, like the other mechanical
switches to the right of it with their series resistors like R630, R631, R647,
R648, etc. Read the interesting description of how the readout works here:
https://w140.com/tekwiki/wiki/7000_series_readout_system
or in one of the 7000-series mainframe manuals. Many (not all) go into great
detail on this. It's a brilliant circuit, devised by a Brit called Barrie
Gilbert.
Will definitely read it. Thank you 😊

Don't forget (in general): It's not completely unthinkable that a previous
owner modified the circuit...
Yes, nobody knows what previous owners did. But in this particular case, the original issue Was just the shorted BNC/Identifying Ring like Eric had experienced.

One last question: Are the voltage scale indications (V/div) correct?
Yes. Even when the order of operation was reversed, this was correct.


 

On Fri, Aug 21, 2020 at 02:43 AM, Stephen wrote:


I was very hard to tell, even with the magnifying glass I used.
This shows how easy it is to misinterpret. Actually, a resistance problem with BNC rings isn't uncommon but because you mentioned some movement, I thought it couldn't be a simple short. I didn't think that the movement was as minuscule as it was, and when I suggested its minimum, your reply didn't make me "switch back". I have learned a thing in this thread as well...
Of course, it only takes some small resistance in the button/ring combination to create the very small movement that you did see. BTW, Eric wrote 39KOhm, he must have meant 39 Ohm.
As it is, different resistance values between ring and gnd are used to indicate the attenuation factor of a probe. As an example, 10X is indicated by a resistance of about 10...11KOhm ISTR. Some vertical amplifiers recognize other attenuations.

Raymond


Stephen
 

On Thu, Aug 20, 2020 at 02:25 PM, Raymond Domp Frank wrote:

Interleaved.

This shows how easy it is to misinterpret. Actually, a resistance problem with
BNC rings isn't uncommon but because you mentioned some movement, I thought it
couldn't be a simple short.
I don’t exactly know what it was. It was extremely hard to tell. And I’m not even 100% I saw
it right. The only sure thing is that turning the actual front end of the BNC, a bit less than 1/4 of a turn - in my case CW - removed the short. BTW, I was actually trying to unscrew the thing, not trying to fix the short. But, in between little twists, I decided to recheck for that short anyway. And bingo, it was gone.
Maybe that’ll help someone, someday.... Needless to say, UNSOLDER EVERYTHING FROM THE BNC BEFORE YOU DO THAT. Otherwise, you’re going to twist everything that may cause more damage than what you initially started with.

I didn't think that the movement was as minuscule
as it was, and when I suggested its minimum, your reply didn't make me "switch
back". I have learned a thing in this thread as well...
😉

Of course, it only takes some small resistance in the button/ring combination
to create the very small movement that you did see. BTW, Eric wrote 39KOhm, he
must have meant 39 Ohm.
As it is, different resistance values between ring and gnd are used to
indicate the attenuation factor of a probe. As an example, 10X is indicated by
a resistance of about 10...11KOhm ISTR. Some vertical amplifiers recognize
other attenuations.
From memory, it a 100R resistor in this case, I think.


Stephen
 

Do you know if inverting C&E on Q620 would cause « IDENTITY » and the « VOLTAGE RANGE » to invert?


Colin Herbert
 

If I remember correctly, those BNC inputs are fastened into "D"-shaped holes. I think I recall trying to do some work on one of my 400-sries portable scopes where I had problems with the readout ring. I found it just about impossible to remove the BNC because the fastening-nut, which is behind the panel, is pretty much obscured to any tool that might be able to loosen it. If you have attempted to turn the BNC from outside the plug-in, I wouldn’t be surprised if you have done a little physical damage to something. Behind the front-panel and connected to the readout ring by a lug on the BNC, there is a pi-network of two capacitors (C13 and C15) and one resistor (R13). You may well have disturbed a joint between one or other of these or perhaps even broken a component.
Incidentally, the way I solved (read "bodged") the repair to my readout ring was to drill a very small hole (just big enough for a piece of 24-gauge tinned-copper wire to go through). I then connected the inner end to wherever it was supposed to go inside the scope and soldered it to the outside of the readout ring as closely to the ring as possible - as flush as I could make it. OK, on the 400-series this readout ring only serves to switch the illumination between the "X1" and "X10" positions, but I think the principle would be the same. You just have to remember when plugging something into that BNC that there is a bit of wire soldered to the readout ring and be careful.

Good Luck.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Stephen
Sent: 21 August 2020 09:23
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7A26 CH2 Readout

On Thu, Aug 20, 2020 at 02:25 PM, Raymond Domp Frank wrote:

Interleaved.

This shows how easy it is to misinterpret. Actually, a resistance problem with
BNC rings isn't uncommon but because you mentioned some movement, I thought it
couldn't be a simple short.
I don’t exactly know what it was. It was extremely hard to tell. And I’m not even 100% I saw
it right. The only sure thing is that turning the actual front end of the BNC, a bit less than 1/4 of a turn - in my case CW - removed the short. BTW, I was actually trying to unscrew the thing, not trying to fix the short. But, in between little twists, I decided to recheck for that short anyway. And bingo, it was gone.
Maybe that’ll help someone, someday.... Needless to say, UNSOLDER EVERYTHING FROM THE BNC BEFORE YOU DO THAT. Otherwise, you’re going to twist everything that may cause more damage than what you initially started with.

I didn't think that the movement was as minuscule
as it was, and when I suggested its minimum, your reply didn't make me "switch
back". I have learned a thing in this thread as well...
😉

Of course, it only takes some small resistance in the button/ring combination
to create the very small movement that you did see. BTW, Eric wrote 39KOhm, he
must have meant 39 Ohm.
As it is, different resistance values between ring and gnd are used to
indicate the attenuation factor of a probe. As an example, 10X is indicated by
a resistance of about 10...11KOhm ISTR. Some vertical amplifiers recognize
other attenuations.
From memory, it a 100R resistor in this case, I think.


Stephen
 

On Thu, Aug 20, 2020 at 11:12 PM, Colin Herbert wrote:


If I remember correctly, those BNC inputs are fastened into "D"-shaped holes.
I think I recall trying to do some work on one of my 400-sries portable scopes
where I had problems with the readout ring. I found it just about impossible
to remove the BNC because the fastening-nut, which is behind the panel, is
pretty much obscured to any tool that might be able to loosen it. If you have
attempted to turn the BNC from outside the plug-in, I wouldn’t be surprised
if you have done a little physical damage to something. Behind the front-panel
and connected to the readout ring by a lug on the BNC, there is a pi-network
of two capacitors (C13 and C15) and one resistor (R13). You may well have
disturbed a joint between one or other of these or perhaps even broken a
component.
Incidentally, the way I solved (read "bodged") the repair to my readout ring
was to drill a very small hole (just big enough for a piece of 24-gauge
tinned-copper wire to go through). I then connected the inner end to wherever
it was supposed to go inside the scope and soldered it to the outside of the
readout ring as closely to the ring as possible - as flush as I could make it.
OK, on the 400-series this readout ring only serves to switch the illumination
between the "X1" and "X10" positions, but I think the principle would be the
same. You just have to remember when plugging something into that BNC that
there is a bit of wire soldered to the readout ring and be careful.

Good Luck.
Nothing seems to be broken or damaged in any way. But the important thing is that it now works perfectly as expected. So I doubt anything is broken or damaged.
But thank you for the heads up on the 400 series. I will definitely keep that in mind if I ever need to mess with the BNC’s on the ones I have.


Colin Herbert
 

I have just had a look at the BNC inputs of one of my 7A26s and they are certainly a lot more accessible than the inputs on the 400-series scopes. One thing that I did notice, though, was that the connections to the back of the POSITION/IDENTIFY switches and the associated components were very, very close to earthed metalwork around the input BNC and attenuator housings and could easily short out. Something for us all to be wary of, I think.
Colin.

PS: I have been meaning to remark on your ready access to an electronics component shop in Paris. Apart from Cricklewood Electronics (North London), there aren't many component shops any more, if any at all. It is pretty much always online ordering in the UK these days. Having said that, I order from Cricklewood and their service is quick. Added to that, there are several places that we in the UK can order parts from, so it's no big problem except when you would probably like to look before you buy. C'est la Vie.

Nothing seems to be broken or damaged in any way. But the important thing is that it now works perfectly as expected. So I doubt anything is broken or damaged.
But thank you for the heads up on the 400 series. I will definitely keep that in mind if I ever need to mess with the BNC’s on the ones I have.