Topics

7623A Storage problem


Colin Herbert
 

My 7623A scope has recently developed a problem when asked to store anything. It doesn't show a trace at all. I can see the time-base triggering, but there is no trace at all to be seen on the CRT, other than the fact that the screen erasing also seems to be working as it should. Neither trace nor readout is shown on the CRT. If I return to non-store, the trace is still not there, nor is the readout. If I power down and then power up, after switching to "non Store", normality is returned. If I power down and power up while the scope is still in "Store" mode, it does a single storage cycle only.
I fully admit that I have not examined my Service Manual so far, as storage CRTs confuse me. It has been a while since I last used storage on this scope, but it was working a while ago. Is this a simple fault easily rectified? Any help gratefully received.
TIA, Colin.


Roger Evans
 

Colin,

I don't claim to be an expert on storage CRTs but I spent a surprisingly pleasant couple of weeks tracking down the several problems on a 7623A which not only had some real faults but also had the storage board wrongly wired up after removal by the previous owner. The performance of the 7623A was so good that when I spotted a 'spares or repair' 7633 at a bargain price I thought I would have another go. I did get it working but the storage tube was well used and its writing speed was not up to the (good) 7623A.

The methodology I used was to set up a repetitive store and erase cycle and follow the waveforms that the manual shows for each of the storage electrodes and each storage mode. The waveforms are complex and rely on some logic and 74123 monostables to set up the timing and then the logic switches different voltages into the op-amps that drive the relatively high voltages on the storage electrodes. Between the two scopes I must have replaced four of the 74123s and one or two 74xx logic ICs. The other failure common to both boards was (one or more of) the set of four transistors that control the 600V line (Q1728 ..Q1743) and resistors around these had drifted significantly in value. The 7633 also had internal tracking of the PCB traces (four layer board) in the vicinity of the high voltage area which I finally managed to fix by drilling and filing until I had cleared the leakage and adding jumpers for the tracks that had been removed in the process.

Having a digital storage scope to follow the erase and store cycle made life a lot easier than trying to remember what had just flashed on the screen and wouldn't return for a few seconds. A wild guess based on your symptoms would be that one of the 74xx logic ICs has failed but finding the culprit may take time.

Regards,

Roger


Colin Herbert
 

I have had some further observations, not good, I'm afraid.

I thought that I would leave the scope in "store" mode and see if any improvement happened. It didn't. When I powered down, left it a while and then powered up in "non-store" mode, I was hoping to see the scope working as it had been before, but the screen was totally blank; no beam, no beam-finder, no readout. I have checked all of the low-power voltages and they are all within spec. Also, I can see the saw-tooth output on the back of the scope, so some circuits are working. However, the -1475 V high voltage is absent. I assume this will have the effect of losing the beam, but what could have caused it's failure?

Another anomaly is that the voltage at the "collector" test-point on the storage board is -1.4 V, when it should be +150 V.

I would certainly like to get the "non-store" mode working again, possibly by repairing the -1475 V on the CRT, then assessing why the storage mode wasn't working properly.

TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 26 August 2020 15:38
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I don't claim to be an expert on storage CRTs but I spent a surprisingly pleasant couple of weeks tracking down the several problems on a 7623A which not only had some real faults but also had the storage board wrongly wired up after removal by the previous owner. The performance of the 7623A was so good that when I spotted a 'spares or repair' 7633 at a bargain price I thought I would have another go. I did get it working but the storage tube was well used and its writing speed was not up to the (good) 7623A.

The methodology I used was to set up a repetitive store and erase cycle and follow the waveforms that the manual shows for each of the storage electrodes and each storage mode. The waveforms are complex and rely on some logic and 74123 monostables to set up the timing and then the logic switches different voltages into the op-amps that drive the relatively high voltages on the storage electrodes. Between the two scopes I must have replaced four of the 74123s and one or two 74xx logic ICs. The other failure common to both boards was (one or more of) the set of four transistors that control the 600V line (Q1728 ..Q1743) and resistors around these had drifted significantly in value. The 7633 also had internal tracking of the PCB traces (four layer board) in the vicinity of the high voltage area which I finally managed to fix by drilling and filing until I had cleared the leakage and adding jumpers for the tracks that had been removed in the process.

Having a digital storage scope to follow the erase and store cycle made life a lot easier than trying to remember what had just flashed on the screen and wouldn't return for a few seconds. A wild guess based on your symptoms would be that one of the 74xx logic ICs has failed but finding the culprit may take time.

Regards,

Roger


Roger Evans
 

Colin,

The voltage on the 'collector' together with the voltage on the front mesh of the storage tube depend on a +600V supply derived from a separate winding on the HV transformer. So the lack of +150V on the collector traces back to the lack of -1475V on the CRT.

The usual suspects for failure of the -1475V are the heat-sinked drive transistor Q1195 and the power supply decoupling components C1198 and L1198. Q1190 is also worth checking and C1224 (is it one of the failure prone tantalums?) will kill the drive to Q1195 if it (ie C1224) goes short circuit. There is a fuse in the unregulated supply to Q1195, (F814) which is on the low voltage regulator board.

There are of course other potential failures on the HV side of the -1475V supply.

Regards,

Roger


Roger Evans
 

Apologies - I skipped the very important 'check all the low voltage DC supplies', there are so many inter-dependencies in these scopes.

Roger


Colin Herbert
 

Hi Roger,

Thanks for all the ideas. I am a bit careful with voltages over 1 KV, but in the past, when I was still at school, I caught a belt at 2 KV on a homebrew scope that I was trying to build. This was from a mains-transformer supply with a voltage doubler and so was low-impedance and capable of quite a whack. It came about from a leaky capacitor which was supposed to be keeping the X-modulation input safe. More fool me!

With regard to the 7623A, I think I mentioned that I had checked all of the voltages from the low-voltage DC supply and found that they were all in spec. I might also have mentioned that I could see the sweep output on the BNC at the back of the scope, so a good few voltages are obviously ok.

I suppose I should gird my loins and try to gain safe access to the high-voltage board and that heat-sunk transistor Q1195.

Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 11:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Apologies - I skipped the very important 'check all the low voltage DC supplies', there are so many inter-dependencies in these scopes.

Roger


Roger Evans
 

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger


Colin Herbert
 

I have found that F814 (2A) on the LV power supply was blown and continues to do so, possibly because of a failure of Q1195 or C1198 (47uF). My question is how can I get to either of these components? I can see from the Manual that C1198 is on the HV board and that I might have to get in there to look at it. But Q1195 isn't on the HV board because it is heatsinked, but where is it located? I have a transistor checker and can thus check it out, if I can locate it.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 14:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger


Miguel Work
 

I remember that is mounted in the HV metallic shield/box



-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Colin Herbert via groups.io
Enviado el: viernes, 25 de septiembre de 2020 18:47
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] 7623A Storage problem

I have found that F814 (2A) on the LV power supply was blown and continues to do so, possibly because of a failure of Q1195 or C1198 (47uF). My question is how can I get to either of these components? I can see from the Manual that C1198 is on the HV board and that I might have to get in there to look at it. But Q1195 isn't on the HV board because it is heatsinked, but where is it located? I have a transistor checker and can thus check it out, if I can locate it.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 14:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger


Colin Herbert
 

This may be a little premature, but if the problem is a failed Q1195, it looks as if it a Tektronix-made 2N3055-family NPN power transistor. Tek show its part-number as 151-0140-00. The question I have is, can I just put any 2N3055 in there? I recall there being some differences between old-style and new-style 2N3055s, in terms of gain, if I remember correctly. How might this affect my choice of replacement?
TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Miguel Work
Sent: 25 September 2020 21:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

I remember that is mounted in the HV metallic shield/box



-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Colin Herbert via groups.io
Enviado el: viernes, 25 de septiembre de 2020 18:47
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] 7623A Storage problem

I have found that F814 (2A) on the LV power supply was blown and continues to do so, possibly because of a failure of Q1195 or C1198 (47uF). My question is how can I get to either of these components? I can see from the Manual that C1198 is on the HV board and that I might have to get in there to look at it. But Q1195 isn't on the HV board because it is heatsinked, but where is it located? I have a transistor checker and can thus check it out, if I can locate it.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 14:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger


 

You want an old slow (low Ft) non-epitaxial 2N3055 for the HT inverter.

David

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Colin Herbert via groups.io
Sent: 04 October 2020 11:28
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

This may be a little premature, but if the problem is a failed Q1195, it looks as if it a Tektronix-made 2N3055-family NPN power transistor. Tek show its part-number as 151-0140-00. The question I have is, can I just put any 2N3055 in there? I recall there being some differences between old-style and new-style 2N3055s, in terms of gain, if I remember correctly. How might this affect my choice of replacement?
TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Miguel Work
Sent: 25 September 2020 21:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

I remember that is mounted in the HV metallic shield/box



-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Colin Herbert via groups.io
Enviado el: viernes, 25 de septiembre de 2020 18:47
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] 7623A Storage problem

I have found that F814 (2A) on the LV power supply was blown and continues to do so, possibly because of a failure of Q1195 or C1198 (47uF). My question is how can I get to either of these components? I can see from the Manual that C1198 is on the HV board and that I might have to get in there to look at it. But Q1195 isn't on the HV board because it is heatsinked, but where is it located? I have a transistor checker and can thus check it out, if I can locate it.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 14:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger


Tom Lee
 

As David Partridge says, it's risky to pop in a modern replacement. It's not just a matter of gain (I assume you meant beta). The original 2N3055 was made by a now-extinct "hometaxial" fabrication process, which produced surprisingly rugged, inexpensive, and very slow devices. Modern planar versions of the 3055 are much, much faster, but also less rugged. The higher speed is not always a virtue -- it can provoke wild and not infrequently destructive oscillations in circuits that rely upon the slowness of original 3055s for stability. It is often possible to add capacitances to make modern devices slower in the right ways, but it's an ugly circuit-specific kludge that doesn't fully reproduce the glorious sluggishness of the original.

-- 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 10/4/2020 03:28, Colin Herbert via groups.io wrote:
This may be a little premature, but if the problem is a failed Q1195, it looks as if it a Tektronix-made 2N3055-family NPN power transistor. Tek show its part-number as 151-0140-00. The question I have is, can I just put any 2N3055 in there? I recall there being some differences between old-style and new-style 2N3055s, in terms of gain, if I remember correctly. How might this affect my choice of replacement?
TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Miguel Work
Sent: 25 September 2020 21:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

I remember that is mounted in the HV metallic shield/box



-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Colin Herbert via groups.io
Enviado el: viernes, 25 de septiembre de 2020 18:47
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] 7623A Storage problem

I have found that F814 (2A) on the LV power supply was blown and continues to do so, possibly because of a failure of Q1195 or C1198 (47uF). My question is how can I get to either of these components? I can see from the Manual that C1198 is on the HV board and that I might have to get in there to look at it. But Q1195 isn't on the HV board because it is heatsinked, but where is it located? I have a transistor checker and can thus check it out, if I can locate it.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 14:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger





















Colin Herbert
 

Thanks for that, guys. I have three old 2N3055 in my "junk box". One doesn't have a manufacturer's name but is also labelled 3055H and has (using my MK-328 transistor tester) what I assume is a beta ("B") of 30, another labelled ITT has a beta of 34 and one marked RS has beta of 12. Which of these would be appropriate, if any? I also have a newer one with a much higher beta, which I assume is not good in this application. Please forgive my ignorance on this.
TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Tom Lee
Sent: 04 October 2020 12:12
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

As David Partridge says, it's risky to pop in a modern replacement. It's
not just a matter of gain (I assume you meant beta). The original 2N3055
was made by a now-extinct "hometaxial" fabrication process, which
produced surprisingly rugged, inexpensive, and very slow devices. Modern
planar versions of the 3055 are much, much faster, but also less rugged.
The higher speed is not always a virtue -- it can provoke wild and not
infrequently destructive oscillations in circuits that rely upon the
slowness of original 3055s for stability. It is often possible to add
capacitances to make modern devices slower in the right ways, but it's
an ugly circuit-specific kludge that doesn't fully reproduce the
glorious sluggishness of the original.

-- 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 10/4/2020 03:28, Colin Herbert via groups.io wrote:
This may be a little premature, but if the problem is a failed Q1195, it looks as if it a Tektronix-made 2N3055-family NPN power transistor. Tek show its part-number as 151-0140-00. The question I have is, can I just put any 2N3055 in there? I recall there being some differences between old-style and new-style 2N3055s, in terms of gain, if I remember correctly. How might this affect my choice of replacement?
TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Miguel Work
Sent: 25 September 2020 21:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

I remember that is mounted in the HV metallic shield/box



-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Colin Herbert via groups.io
Enviado el: viernes, 25 de septiembre de 2020 18:47
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] 7623A Storage problem

I have found that F814 (2A) on the LV power supply was blown and continues to do so, possibly because of a failure of Q1195 or C1198 (47uF). My question is how can I get to either of these components? I can see from the Manual that C1198 is on the HV board and that I might have to get in there to look at it. But Q1195 isn't on the HV board because it is heatsinked, but where is it located? I have a transistor checker and can thus check it out, if I can locate it.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Roger Evans via groups.io
Sent: 23 September 2020 14:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

Colin,

I had the afterthought about low voltages since the regulation of the power transistor Q1195 relies on +5V, +15V and +130V being OK. The components I mentioned are all on the low voltage side of the transformer T1199 and can be checked with the power off. Q1195 is attached by a plug and socket so it can effectively be tested out of circuit, Q1190 can be checked as a pair of diodes for open circuit / short circuit. The two capacitors are likely to be short circuit if they have failed and L1198 will be open circuit if the fuse didn't do its job in time.

At least you don't have to wait 30 minutes for the high voltage to leak away!

Regards,

Roger






















Thomas Garson
 

If your 2N3055s are in aluminum packages, they are likely Motorola or Fairchild fab, and will be epitaxial (high gain). If they are in steel alloy cans, they could be RCA, which would be hometaxial, or slower and lower gain. There were a lot of smaller fabs making 2N3055 because that part number had become ubiquitous. Most, perhaps all, used hometaxial process, and steel cans. Possibly those foundries had acquired obsolete lines which had been surplussed out by the bigger operations when they changed over to epitaxial.

You didn't specify what collector current your tester operates at. I suspect its in the low ma range. At 10 ma, hometaxial parts would show hFE close to your 30 & 34. hFE maxes out at about 70 at 3a. I wouldn't use the part with hFE of only 12, or the high gain part.

I think that when RCA was killed by GE, their power devices were still primarily (exclusively?) hometaxial. The RCA 40436 speced identical to the 3N3055. I used a number of them in one off power amps and linear supplies during the early 1970s. Another RCA option would be 2N3772.

I agree that it is not a good idea to install newer epitaxial devices in a circuit whose design hails from before about 1975. There is a good chance the part will not behave nicely. As the unwashed might say: "Magic smoke may appear".

Somewhere, I've got a bag of old, but mostly unused, 2N3055s, should it become an issue. All the ones I just found in a Google search were planar epitaxial, which work fine in newer designs that expect the higher speed.

Anecdote: 1972. To get cheap high power for small time rock bands, I would take 2 Peavey PA120, which was a 2 ohm capable 1ch power map entirely built on a 5u rack panel. PA120 used house branded 2N3055 equivalent devices. Circuit probably copied from RCA Transistor Manual. I put two of them in one cabinet, back to back with end to end fan added, configured for bridged mono operation: Easy 500w.r.m.s into 4 ohms. If one blew up (not easy), I used RCA 2N3055s as replacement devices. They were very reliable.

Thomas Garson
Aural Technology, Ashland, OR
By my calculation, the dynamic range of the universe is roughly 679dB,
which is approximately 225 bits, collected at a rate 1.714287514x10^23 sps.

On 10/4/20 8:26 AM, Colin Herbert via groups.io wrote:
Thanks for that, guys. I have three old 2N3055 in my "junk box". One doesn't have a manufacturer's name but is also labelled 3055H and has (using my MK-328 transistor tester) what I assume is a beta ("B") of 30, another labelled ITT has a beta of 34 and one marked RS has beta of 12. Which of these would be appropriate, if any? I also have a newer one with a much higher beta, which I assume is not good in this application. Please forgive my ignorance on this.
TIA, Colin.
-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Tom Lee
Sent: 04 October 2020 12:12
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem
As David Partridge says, it's risky to pop in a modern replacement. It's
not just a matter of gain (I assume you meant beta). The original 2N3055
was made by a now-extinct "hometaxial" fabrication process, which
produced surprisingly rugged, inexpensive, and very slow devices. Modern
planar versions of the 3055 are much, much faster, but also less rugged.
The higher speed is not always a virtue -- it can provoke wild and not
infrequently destructive oscillations in circuits that rely upon the
slowness of original 3055s for stability. It is often possible to add
capacitances to make modern devices slower in the right ways, but it's
an ugly circuit-specific kludge that doesn't fully reproduce the
glorious sluggishness of the original.
-- Tom


Brad Thompson
 

Thomas Garson wrote on 10/4/2020 3:09 PM:

If your 2N3055s are in aluminum packages, they are likely Motorola or Fairchild fab, and will be epitaxial (high gain). If they are in steel alloy cans, they could be RCA, which would be hometaxial, or slower and lower gain. There were a lot of smaller fabs making 2N3055 because that part number had become ubiquitous. Most, perhaps all, used hometaxial process, and steel cans. Possibly those foundries had acquired obsolete lines which had been surplussed out by the bigger operations when they changed over to epitaxial.
<snip>

Hello--

In addition to deliberate process changes, some manufacturers reportedly labeled
their parametric rejects as 2N3055s. The rejects met or exceeded the 2N3055's
registered specs which were relatively relaxed to begin with.
I haven't searched the web, but a basic and easily duplicated fT measurement
setup would be helpful for smoking out the too-fast "2N3055"s.

73--

Brad  AA1IP


Tom Lee
 

There are also notorious Chinese knock-offs which contain low-power (2N2222?) dice in what appear to be period-authentic packages. They'll pass the standard multimeter diode test, but don't survive actual use. The ft test that Brad T. suggests is a very good way to identify non-genuine 3055s.

--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 10/4/2020 12:42, Brad Thompson wrote:
Thomas Garson wrote on 10/4/2020 3:09 PM:

If your 2N3055s are in aluminum packages, they are likely Motorola or Fairchild fab, and will be epitaxial (high gain). If they are in steel alloy cans, they could be RCA, which would be hometaxial, or slower and lower gain. There were a lot of smaller fabs making 2N3055 because that part number had become ubiquitous. Most, perhaps all, used hometaxial process, and steel cans. Possibly those foundries had acquired obsolete lines which had been surplussed out by the bigger operations when they changed over to epitaxial.
<snip>

Hello--

In addition to deliberate process changes, some manufacturers reportedly labeled
their parametric rejects as 2N3055s. The rejects met or exceeded the 2N3055's
registered specs which were relatively relaxed to begin with.
I haven't searched the web, but a basic and easily duplicated fT measurement
setup would be helpful for smoking out the too-fast "2N3055"s.

73--

Brad  AA1IP




SCMenasian
 

Colin,

2N3055H transistors are hometaxial and should be OK. The 'H' designation was added some time after the epitaxial process was introduced. Of course, all earlier 2N3055s were also hometaxial and one should be careful about 2N3055Hs made by some unknown FAB in China.

It should be a fairly simple matter to make a test fixture which will be an oscillator with a modern 2N3055 and only amplify with an original one.

Stephen


Thomas Garson
 

The junctions of that type of counterfeit power transistor will have a higher forward voltage than a genuine part. Usually they read as much as .1 to .15 volts higher on a typical DMMs diode test range. They will also have different junction capacitance. I can't give numbers as I'm not where I have proper tools. However a data sheet for the McCoy will provide at least some of that information.

A curve tracer will reveal a lot. Small signal transistors have very different family of curves than most any power transistor. While the Tek curve tracers are the tool of choice, I've gotten a lot of good mileage from the Heathkit variety. The Heathkit curve tracer, which does not have a display, will work with any DC coupled 'scope and can sometimes be found on eBay for a low price that belies its usefulness. Even if broken, they are relatively simple and easy for a moderately knowledgeable tech to fix.

Thomas Garson
Aural Technology, Ashland, OR
By my calculation, the dynamic range of the universe is roughly 679dB,
which is approximately 225 bits, collected at a rate 1.714287514x10^23 sps.

On 10/4/20 12:59 PM, Tom Lee wrote:
There are also notorious Chinese knock-offs which contain low-power (2N2222?) dice in what appear to be period-authentic packages. They'll pass the standard multimeter diode test, but don't survive actual use. The ft test that Brad T. suggests is a very good way to identify non-genuine 3055s.
--Tom


NigelP
 

Another good source was the old SGS or SGS-Ates 2N3055H, possibly also SGS-Thomson as it later became known; can't remember when we stopped doing hometaxial devices, it was a long time ago!

Regards

Nigel G8AYM (ex STMicroelectronics, SGS-Thomson, SGS-Semiconductor and SGS-Ates)


Colin Herbert
 

I have just checked all of my 2N3055 stash using a magnet and they all have
steel casings.

I am afraid my electronics knowledge is too limited for me to be able to
knock-up any test-rigs to do electrical checks on them. Any hints?

TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Thomas
Garson
Sent: 04 October 2020 20:10
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

If your 2N3055s are in aluminum packages, they are likely Motorola or
Fairchild fab, and will be epitaxial (high gain). If they are in steel
alloy cans, they could be RCA, which would be hometaxial, or slower and
lower gain. There were a lot of smaller fabs making 2N3055 because that
part number had become ubiquitous. Most, perhaps all, used hometaxial
process, and steel cans. Possibly those foundries had acquired obsolete
lines which had been surplussed out by the bigger operations when they
changed over to epitaxial.

You didn't specify what collector current your tester operates at. I
suspect its in the low ma range. At 10 ma, hometaxial parts would show
hFE close to your 30 & 34. hFE maxes out at about 70 at 3a. I wouldn't
use the part with hFE of only 12, or the high gain part.

I think that when RCA was killed by GE, their power devices were still
primarily (exclusively?) hometaxial. The RCA 40436 speced identical to
the 3N3055. I used a number of them in one off power amps and linear
supplies during the early 1970s. Another RCA option would be 2N3772.

I agree that it is not a good idea to install newer epitaxial devices in
a circuit whose design hails from before about 1975. There is a good
chance the part will not behave nicely. As the unwashed might say:
"Magic smoke may appear".

Somewhere, I've got a bag of old, but mostly unused, 2N3055s, should it
become an issue. All the ones I just found in a Google search were
planar epitaxial, which work fine in newer designs that expect the
higher speed.

Anecdote: 1972. To get cheap high power for small time rock bands, I
would take 2 Peavey PA120, which was a 2 ohm capable 1ch power map
entirely built on a 5u rack panel. PA120 used house branded 2N3055
equivalent devices. Circuit probably copied from RCA Transistor Manual.
I put two of them in one cabinet, back to back with end to end fan
added, configured for bridged mono operation: Easy 500w.r.m.s into 4
ohms. If one blew up (not easy), I used RCA 2N3055s as replacement
devices. They were very reliable.

Thomas Garson
Aural Technology, Ashland, OR
By my calculation, the dynamic range of the universe is roughly 679dB,
which is approximately 225 bits, collected at a rate 1.714287514x10^23 sps.

On 10/4/20 8:26 AM, Colin Herbert via groups.io wrote:
Thanks for that, guys. I have three old 2N3055 in my "junk box". One
doesn't have a manufacturer's name but is also labelled 3055H and has (using
my MK-328 transistor tester) what I assume is a beta ("B") of 30, another
labelled ITT has a beta of 34 and one marked RS has beta of 12. Which of
these would be appropriate, if any? I also have a newer one with a much
higher beta, which I assume is not good in this application. Please forgive
my ignorance on this.
TIA, Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Tom
Lee
Sent: 04 October 2020 12:12
To: TekScopes@groups.io
Subject: Re: [TekScopes] 7623A Storage problem

As David Partridge says, it's risky to pop in a modern replacement. It's
not just a matter of gain (I assume you meant beta). The original 2N3055
was made by a now-extinct "hometaxial" fabrication process, which
produced surprisingly rugged, inexpensive, and very slow devices. Modern
planar versions of the 3055 are much, much faster, but also less rugged.
The higher speed is not always a virtue -- it can provoke wild and not
infrequently destructive oscillations in circuits that rely upon the
slowness of original 3055s for stability. It is often possible to add
capacitances to make modern devices slower in the right ways, but it's
an ugly circuit-specific kludge that doesn't fully reproduce the
glorious sluggishness of the original.

-- Tom