Topics

519 -- broken rate generator; stuck at a very high frequency?


@0culus
 

Hi all,

I recently became the proud owner of this gigantic hunk of atomic history. Scope is in excellent shape and runs mostly fine but there are three problems that I have found and would like to try to diagnose and fix. The most mysterious one (and the subject of this thread) is that the rate generator is stuck at ~250 kHz and does not respond to the controls, except ever so slightly in the x1000 multiplier position. This implies to me that the multivibrator is running, but there is something wrong with the circuitry that controls the multivibrator, namely, C920(A,B,C,D,E), R925 (maybe this has drifted high?), and R923 (which seems to be healthy). Looking at the schematic, there seems to be very little to even go wrong. I did some contact cleaning as well as tried other 6DJ8s to no avail.

Anyone else with 519 experience encounter this failure mode?

Thanks,

Sean


Tom Lee
 

Never owned that scope, but as you say, there's very little that can go wrong. Troubleshooting should not take a lot of time unless the components are inaccessible.

R925 is a 1% resistor, so I doubt that it is a carbon comp, but if it is, it certainly could have drifted high, but I don't see how that would cause the symptoms you are seeing. I would first suspect the switches, though. If they are making poor contact to the timing caps, that could easily produce what is happening. Did your contact cleaning include the timing switch?

Otherwise, do theĀ  normal things: Perform a careful visual inspection of the circuit in question. Any toasted resistors? Sad-looking caps (check C916; old electrolytics always deserve scrutiny)?

Measure voltages.

Check resistances.

There are so few components and nodes that you should be able to run through that list in about the same time as it takes to read this email.

If you suspect something's up with the switches or timing caps, try tack-soldering a 0.01uF or 0.1uF cap across the switch terminals and see if you are able to affect the frequency.

Enjoy that Nuclear Nugget. It's definitely a piece of history!

--Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 10/17/2020 21:39, sdturne@q.com wrote:
Hi all,

I recently became the proud owner of this gigantic hunk of atomic history. Scope is in excellent shape and runs mostly fine but there are three problems that I have found and would like to try to diagnose and fix. The most mysterious one (and the subject of this thread) is that the rate generator is stuck at ~250 kHz and does not respond to the controls, except ever so slightly in the x1000 multiplier position. This implies to me that the multivibrator is running, but there is something wrong with the circuitry that controls the multivibrator, namely, C920(A,B,C,D,E), R925 (maybe this has drifted high?), and R923 (which seems to be healthy). Looking at the schematic, there seems to be very little to even go wrong. I did some contact cleaning as well as tried other 6DJ8s to no avail.

Anyone else with 519 experience encounter this failure mode?

Thanks,

Sean




@0culus
 

Hey Tom,

Thanks for the brainstorming. I do need to turn the scope on it's side so I can get to the bottom panel and look under there. That, however, is a tomorrow problem. :o) I did deoxit the range switch, which is what is switching the timing caps in and out. No change there. I did note that one set of contacts makes a tiny arc when they move past each other.

I definitely will enjoy it! It's a really neat scope, and it's pretty cool to see it next my 7104.

Sean

On Sat, Oct 17, 2020 at 10:05 PM, Tom Lee wrote:


Never owned that scope, but as you say, there's very little that can go wrong.
Troubleshooting should not take a lot of time unless the components are
inaccessible.

R925 is a 1% resistor, so I doubt that it is a carbon comp, but if it is, it
certainly could have drifted high, but I don't see how that would cause the
symptoms you are seeing. I would first suspect the switches, though. If they
are making poor contact to the timing caps, that could easily produce what is
happening. Did your contact cleaning include the timing switch?

Otherwise, do theĀ  normal things: Perform a careful visual inspection of the
circuit in question. Any toasted resistors? Sad-looking caps (check C916; old
electrolytics always deserve scrutiny)?

Measure voltages.

Check resistances.

There are so few components and nodes that you should be able to run through
that list in about the same time as it takes to read this email.

If you suspect something's up with the switches or timing caps, try
tack-soldering a 0.01uF or 0.1uF cap across the switch terminals and see if
you are able to affect the frequency.

Enjoy that Nuclear Nugget. It's definitely a piece of history!

--Tom


@0culus
 

OK, I got the scope on it's side. That wasn't as hard as I feared, because the 519 is very much bottom heavy. Looking at the part of the rate generator circuit that is on the bottom side, I can see nothing obvious, so I'm going to do some measurements of the passives and see if anything sticks out.

Sean


@0culus
 

Couldn't find anything amiss...all the passives appear to be fine. I scoped the multivibrator as shown in the schematic, and it is works fine. It responds to the controls as it should. I reckon the problem is most likely the avalanche transistor. Luckily, I also have a 577D1 now...

Sean


@0culus
 

Well, no dice there either. The transistor seems fine. I'm officially stumped. The contacts for the multiplier switch are working fine, establishes by the fact that the multivibrator is both running nominally (which also implies the two 6DJ8s are fine) and controllable by the multiplier switch and the cycles/sec pot. There's really not much else to this circuit...nothing could really go wrong with the charge line right? It's just a coil of coax.

Sean

On Sun, Oct 18, 2020 at 03:34 PM, @0culus wrote:


Couldn't find anything amiss...all the passives appear to be fine. I scoped
the multivibrator as shown in the schematic, and it is works fine. It responds
to the controls as it should. I reckon the problem is most likely the
avalanche transistor. Luckily, I also have a 577D1 now...

Sean


Chuck Harris
 

Hi Sean,

I know nothing about the 519, but I do know that tektronix
had a habit of making all sorts of configurations available
on similar scopes. The folks that would buy a 519 often
had divergent interests.

Look carefully for binding posts that should have a strap
connecting them. They could be on the front panel, or the
back.

There also could be other forms of strapping inside of the
scope. It should show up in the schematics.

-Chuck Harris

sdturne@q.com wrote:

Well, no dice there either. The transistor seems fine. I'm officially stumped. The contacts for the multiplier switch are working fine, establishes by the fact that the multivibrator is both running nominally (which also implies the two 6DJ8s are fine) and controllable by the multiplier switch and the cycles/sec pot. There's really not much else to this circuit...nothing could really go wrong with the charge line right? It's just a coil of coax.

Sean

On Sun, Oct 18, 2020 at 03:34 PM, @0culus wrote:


Couldn't find anything amiss...all the passives appear to be fine. I scoped
the multivibrator as shown in the schematic, and it is works fine. It responds
to the controls as it should. I reckon the problem is most likely the
avalanche transistor. Luckily, I also have a 577D1 now...

Sean





@0culus
 

Hi Chuck,

I haven't seen anything like that inside or outside the case. I will keep my eyes peeled though.

Sean

On Sun, Oct 18, 2020 at 05:04 PM, Chuck Harris wrote:


Hi Sean,

I know nothing about the 519, but I do know that tektronix
had a habit of making all sorts of configurations available
on similar scopes. The folks that would buy a 519 often
had divergent interests.

Look carefully for binding posts that should have a strap
connecting them. They could be on the front panel, or the
back.

There also could be other forms of strapping inside of the
scope. It should show up in the schematics.

-Chuck Harris


Tom Lee
 

So, the multivibrator is producing the correct waveforms with controllable frequencies, and the only problem is that the final output is running at 250kHz, independently of what the multivibrator is doing? Have you verified that the swing out of the multivibrator is correct? And is your rate generator an early or late one?

As you said, there isn't much left to check, so by definition you're close!

--Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 10/18/2020 16:10, sdturne@q.com wrote:
Well, no dice there either. The transistor seems fine. I'm officially stumped. The contacts for the multiplier switch are working fine, establishes by the fact that the multivibrator is both running nominally (which also implies the two 6DJ8s are fine) and controllable by the multiplier switch and the cycles/sec pot. There's really not much else to this circuit...nothing could really go wrong with the charge line right? It's just a coil of coax.

Sean

On Sun, Oct 18, 2020 at 03:34 PM, @0culus wrote:

Couldn't find anything amiss...all the passives appear to be fine. I scoped
the multivibrator as shown in the schematic, and it is works fine. It responds
to the controls as it should. I reckon the problem is most likely the
avalanche transistor. Luckily, I also have a 577D1 now...

Sean


@0culus
 

Tom,

Yes, that is correct. The output seems to not be correlated whatsoever with the multivibrator. The transistor is avalanching...the waveform that comes out is consistent with what it should be I think. It's just stuck at one frequency that is much higher than anything that the settings of the controls would suggest.

"Swing out" is a term I am not familiar with...perhaps I know it by another name?

Based on serial number, the rate generator should be the later revision.

Sean

On Mon, Oct 19, 2020 at 12:18 AM, Tom Lee wrote:


So, the multivibrator is producing the correct waveforms with controllable
frequencies, and the only problem is that the final output is running at
250kHz, independently of what the multivibrator is doing? Have you verified
that the swing out of the multivibrator is correct? And is your rate generator
an early or late one?

As you said, there isn't much left to check, so by definition you're close!

--Tom


Tom Lee
 

Hi Sean,

Sorry for the truncated, uncpunctuated verbiage -- compression has obscured the message. Here's a less cryptic version: Have you verifed that the swing of the signal coming out of the multivibrator is correct (both in amplitude and dc level)?

Either way, your avalancher is triggering on its own. The problem could be as simple as the collector supply being set too high (there's a pot for that), or possibly having a lot of noise on it. The way the avalancher is supposed to work here is that the dc supply should be set just below the point where the transistor will avalanche. The signal from the multivibrator kicks it over the top of the hill, causing the transistor to avalanche in synchrony with the multivibrator's triggering signal.

I don't have the schematic in front of me, but you should be able to (mis)adjust the pot to prevent the avalancher from avalanching at all. Try taking it to that point, and then gradually adjusting the pot to increase the collector voltage until the transistor starts to avalanche reliably, and under the control of the multivibrator.

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 10/19/2020 15:03, sdturne@q.com wrote:
Tom,

Yes, that is correct. The output seems to not be correlated whatsoever with the multivibrator. The transistor is avalanching...the waveform that comes out is consistent with what it should be I think. It's just stuck at one frequency that is much higher than anything that the settings of the controls would suggest.

"Swing out" is a term I am not familiar with...perhaps I know it by another name?

Based on serial number, the rate generator should be the later revision.

Sean

On Mon, Oct 19, 2020 at 12:18 AM, Tom Lee wrote:

So, the multivibrator is producing the correct waveforms with controllable
frequencies, and the only problem is that the final output is running at
250kHz, independently of what the multivibrator is doing? Have you verified
that the swing out of the multivibrator is correct? And is your rate generator
an early or late one?

As you said, there isn't much left to check, so by definition you're close!

--Tom


Tom Lee
 

Also, you say that the avalancher seems to be doing the right thing. Would you happen to have a screenshot or something to show what that output looks like? Avalanchers can exhibit several pathologies (which is why not all transistors -- even from the same production run --
avalanche similarly). Avalanching is very sensitive to surface effects, for example, and not all surfaces age gracefully. A transistor can look otherwise fine as a transistor, but stop avalanching well.

Just a thought.

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 10/19/2020 15:14, Tom Lee wrote:
Hi Sean,

Sorry for the truncated, uncpunctuated verbiage -- compression has obscured the message. Here's a less cryptic version: Have you verifed that the swing of the signal coming out of the multivibrator is correct (both in amplitude and dc level)?

Either way, your avalancher is triggering on its own. The problem could be as simple as the collector supply being set too high (there's a pot for that), or possibly having a lot of noise on it. The way the avalancher is supposed to work here is that the dc supply should be set just below the point where the transistor will avalanche. The signal from the multivibrator kicks it over the top of the hill, causing the transistor to avalanche in synchrony with the multivibrator's triggering signal.

I don't have the schematic in front of me, but you should be able to (mis)adjust the pot to prevent the avalancher from avalanching at all. Try taking it to that point, and then gradually adjusting the pot to increase the collector voltage until the transistor starts to avalanche reliably, and under the control of the multivibrator.

-- Cheers,
Tom


Chuck Harris
 

There is one other way, and that is motorboating.
The multivibrator multivibrates, causing a glitch
in the power supply, which triggers the multivibrator.

Just a thought...

-Chuck Harris

Tom Lee wrote:

Hi Sean,

Sorry for the truncated, uncpunctuated verbiage -- compression has obscured the
message. Here's a less cryptic version: Have you verifed that the swing of the signal
coming out of the multivibrator is correct (both in amplitude and dc level)?

Either way, your avalancher is triggering on its own. The problem could be as simple
as the collector supply being set too high (there's a pot for that), or possibly
having a lot of noise on it. The way the avalancher is supposed to work here is that
the dc supply should be set just below the point where the transistor will avalanche.
The signal from the multivibrator kicks it over the top of the hill, causing the
transistor to avalanche in synchrony with the multivibrator's triggering signal.

I don't have the schematic in front of me, but you should be able to (mis)adjust the
pot to prevent the avalancher from avalanching at all. Try taking it to that point,
and then gradually adjusting the pot to increase the collector voltage until the
transistor starts to avalanche reliably, and under the control of the multivibrator.

-- Cheers,
Tom


@0culus
 

Tom,

I put some photos in the album here: https://groups.io/g/TekScopes/album?id=255396

The trace shot is the output of the rate generator, at 10 ns/cm. The crt on this particular scoped is stated as 0.29 ns risetime and 9.1 V/cm.

On Mon, Oct 19, 2020 at 03:21 PM, Tom Lee wrote:


Also, you say that the avalancher seems to be doing the right thing. Would you
happen to have a screenshot or something to show what that output looks like?
Avalanchers can exhibit several pathologies (which is why not all transistors
-- even from the same production run --
avalanche similarly). Avalanching is very sensitive to surface effects, for
example, and not all surfaces age gracefully. A transistor can look otherwise
fine as a transistor, but stop avalanching well.

Just a thought.


@0culus
 

Tom,

It's all good; thanks for the explanation. I have not checked for dc level coming out of the multivibrator.

I turned R931 (avalanche control) some and observed the effects. There is, as expected, a threshold where no avalanche occurs. Beyond, this, turning the pot affects the frequency of the output considerably before it dies away again and the other end of the pot. Still no control with the multivibrator though. Maybe there's an open circuit hiding somewhere?

Sean

On Mon, Oct 19, 2020 at 03:14 PM, Tom Lee wrote:


Hi Sean,

Sorry for the truncated, uncpunctuated verbiage -- compression has obscured
the message. Here's a less cryptic version: Have you verifed that the swing of
the signal coming out of the multivibrator is correct (both in amplitude and
dc level)?

Either way, your avalancher is triggering on its own. The problem could be as
simple as the collector supply being set too high (there's a pot for that), or
possibly having a lot of noise on it. The way the avalancher is supposed to
work here is that the dc supply should be set just below the point where the
transistor will avalanche. The signal from the multivibrator kicks it over the
top of the hill, causing the transistor to avalanche in synchrony with the
multivibrator's triggering signal.

I don't have the schematic in front of me, but you should be able to
(mis)adjust the pot to prevent the avalancher from avalanching at all. Try
taking it to that point, and then gradually adjusting the pot to increase the
collector voltage until the transistor starts to avalanche reliably, and under
the control of the multivibrator.

-- Cheers,
Tom


@0culus
 

I tried another thing, just to be sure...I substituted a 2N2369 (I have a lot of them on hand for building avalanche pulsers) that seemed to have largely similar curves as the existing transistor...this changed nothing as to the lack of control over the frequency of the output...but it did avalanche nicely when I adjusted the avalanche control. :o)

At this point, I'm honestly thinking it could be a cold solder joint. I'm going to need to get the scope on it's side up on a suitable work surface....this working on the ground is for the birds!

Sean


Tom Lee
 

I agree that looking for an open circuit somewhere, whether within or between components, makes sense.

One clue is that the free-running frequency of the output is too high. Injection-locking needs the free-running frequency to be below the locked frequency. If the avalanche stage is running too fast, then the RC that controls the free-run rate isn't doing its job.

You are very close to finding it, I'm sure.

Good luck!

-- Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 10/19/2020 19:49, sdturne@q.com wrote:
I tried another thing, just to be sure...I substituted a 2N2369 (I have a lot of them on hand for building avalanche pulsers) that seemed to have largely similar curves as the existing transistor...this changed nothing as to the lack of control over the frequency of the output...but it did avalanche nicely when I adjusted the avalanche control. :o)

At this point, I'm honestly thinking it could be a cold solder joint. I'm going to need to get the scope on it's side up on a suitable work surface....this working on the ground is for the birds!

Sean




@0culus
 

Tom,

I think I found it. Page 5-15 in the manual describes it:

"76. Set the MULTIPLIER switch to OFF and check to see that the waveform disappears from the screen of the Type 519.

If Abnormal...

The Avalanche stage is free running. Adjust the AVALANCHE SET control to the point where the generator just stops free running. If the adjustment cannot be made, check R933, R932, R931 and 0934."

I was able to make this adjustment with the help of another oscilloscope that offers slower time base settings, and it now appears to be under control of the rate generator controls!

Sean

On Mon, Oct 19, 2020 at 08:06 PM, Tom Lee wrote:


I agree that looking for an open circuit somewhere, whether within or between
components, makes sense.

One clue is that the free-running frequency of the output is too high.
Injection-locking needs the free-running frequency to be below the locked
frequency. If the avalanche stage is running too fast, then the RC that
controls the free-run rate isn't doing its job.

You are very close to finding it, I'm sure.

Good luck!

-- Tom


Tom Lee
 

I'm a bit confused, because I thought you had already tried this, but if it's all ok now, it's all ok now!

Congratulations!

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 10/19/2020 20:20, sdturne@q.com wrote:
Tom,

I think I found it. Page 5-15 in the manual describes it:

"76. Set the MULTIPLIER switch to OFF and check to see that the waveform disappears from the screen of the Type 519.

If Abnormal...

The Avalanche stage is free running. Adjust the AVALANCHE SET control to the point where the generator just stops free running. If the adjustment cannot be made, check R933, R932, R931 and 0934."

I was able to make this adjustment with the help of another oscilloscope that offers slower time base settings, and it now appears to be under control of the rate generator controls!

Sean

On Mon, Oct 19, 2020 at 08:06 PM, Tom Lee wrote:

I agree that looking for an open circuit somewhere, whether within or between
components, makes sense.

One clue is that the free-running frequency of the output is too high.
Injection-locking needs the free-running frequency to be below the locked
frequency. If the avalanche stage is running too fast, then the RC that
controls the free-run rate isn't doing its job.

You are very close to finding it, I'm sure.

Good luck!

-- Tom


@0culus
 

I hadn't tried this adjustment while looking at the signal with another scope...I may have stumbled upon this without knowing since the slowest sweep on the 519 is rather fast compared to general purpose scopes.

Now to move onto two other problems...namely the reed relay appears to be dead and the sweep is kinda messed up at 2ns/cm, mainly on the right side of the crt. The trace sort of coils into a spiral at the end of the sweep...it's weird. Maybe an artifact of the rf power tube that drives the time base?

Sean

On Mon, Oct 19, 2020 at 08:23 PM, Tom Lee wrote:


I'm a bit confused, because I thought you had already tried this, but if it's
all ok now, it's all ok now!

Congratulations!

-- Cheers,
Tom