2nd life for a 575 after resting for 30+ years in the attic #photo-notice


Joe
 

Hi,
four months ago really by chance I was able to add a Tek 575 to my little collection. As it has got serial # 1818 it is a pretty early one and the cabinet has a definitely different texture than later instruments. It had been resting for more than 30 years now in an attic after serving a Siemens Munich lab engineer.
So I asked here and there for advice how to bring it back to life safely (so many thanks Dennis T. !) and I learned that my method to reform electrolytics - practised on many tube radios before - seemed to be more than rude. I promised to give it much more time!
Another thing that I learned is that selenium rectifiers do have a very limited life expectation, "the question is not if but when they will fail". So I decided to do the Mod M5272 to upgrade to silicon rectifiers first of all.
And well, recapping is a must. I selected all of the HV caps and these "bumble bee" called caps and some "Good-all" caps to be replaced. Most of them proved to be more or less leaky!
Today was the day: After several hours of gradually reforming electrolytics I dared to try if I could get a trace on the screen. Finally rising the variac to 90% of normal line voltage rewarded me not only with a bright and sharp trace but with a family of curves when a BF 259 transistor was to act as a guinea pig! I am happy now that I took the long way to save that instrument.
Now it is time for some tuning, checking the tubes (most of all original but some German "Siemens" tubes), cleaning the faceplate and doing longer tests before recalibration will be attempted.
I have just started a photo album "Tek 575". Again many thanks for all of the advice!

Joe


 

My understanding of the paper tubes over the metal can capacitors is that this was done when the can was not at ground potential. The cardboard tube is meant to insulate the metal can, either to prevent shorting to nearby components, or to prevent people from touching the non-grounded cans and getting a shock.

-- Jeff Dutky


 

Hi Joe,

this tale reminds me of my 575, that was sitting together with a 576 for a number of years.... in an attic.

I took down the 576 and it behaved like the proverbial old lady... bad HV transformer, recently a bunch of bad transistors...

Eventually I took a look at the 575. Having the habit of first trying out what happens when rising the variac, thermal cam at hand, well... it worked flawlessly!! Did some realignment, not much was needed. It still has all its original caps, bumble bees, selenium stuff and whatnot that might fail one day. Now its sitting in a (dry) basement, on a shelf connected to the power line where I can switch it on from time to time. So fare no issues.

On the other hand, the old lady... the 576 is sitting on my bench now, in regular use, and the step generator begins acting up (again), doing steps only if well warmed up ;-)
Probably the IC that enables/disables the step counter has an issue...

cheers
Martin


John
 

Joe:
congratulations, and I hope you enjoy using this wonderful piece of test equipment. I had a very simlar experience with a 575 which I bought "blind" on an auction site a couple of years ago (not Ebay). I've managed to retain the HV electrolytics, but changed all the black jobbies. On my unit, there were quite a few tired valves/tubes: in particular the triodes used as HT regulators. Oh, and the fan bearings were sticky with old green oil, so the fan would not run until things got hot. It took some effort to completely clean the oilite bearings.
John


Joe
 

Hello,

Today I let the unit running for an hour at regular mains voltage and did comparison tests on a pair of BC 238 C transistors. What can I say? I am deeply impressed by the bright, sharp and stable display. That is great! Nevertheless I registered a minor problem that has to be solved: The flyback traces of the beam are "visible, too bright".

@ John: Yes I forgot to mention these sticky fan bearings. Well, a little "Ballistol" oil did the job quickly.
@Jeff: Thanks, your explanation must be right. That tells us the Tek designers did a very good job. I know a lot of other instruments where several components are tied to higher voltage levels and not any provision is done to prevent shorting or so.

BTW: All of my BC 238 C transistors produced very similar curve families, so they seem to have close tolerances. I think I will try some older OC 26 now and see what happens.
Joe


Mlynch001
 

If my memory is serving me correctly, the 575, like the 576 curve tracer does not “blank” the fly back retrace, so there is always a slight retrace visible.

--
Michael Lynch
Dardanelle, AR


Chuck Harris <cfharris@...>
 

On, I think all, of the tektronix curve tracers,
the forward trace, and the backward trace are equivalent
as there is no z-axis control, save for the intensity control.

So, unless you are planning to add a whole bunch of twidgety
circuitry that senses when the sine wave hump that traces the
horizontal trace reaches its peak, and blanks as it returns to
its minimum, I would suggest that you live with it.

The degree of looping that you are seeing can be increased by
stray capacitance, and by thermal effects in the transistor that
cause it to warm on the forward trace, and being warm, retrace
a different path on the retrace. The knowledge it presents is
also useful. You should be able to calculate all sorts of thermal
effects from the width of the loop.

Enjoy! It is a good curve tracer.

-Chuck Harris


On Tue, 30 Mar 2021 13:41:57 -0700 "Joe"
<The-Lohrbach-Family@t-online.de> wrote:
Hello,

Today I let the unit running for an hour at regular mains voltage and
did comparison tests on a pair of BC 238 C transistors. What can I
say? I am deeply impressed by the bright, sharp and stable display.
That is great! Nevertheless I registered a minor problem that has to
be solved: The flyback traces of the beam are "visible, too bright".

@ John: Yes I forgot to mention these sticky fan bearings. Well, a
little "Ballistol" oil did the job quickly. @Jeff: Thanks, your
explanation must be right. That tells us the Tek designers did a very
good job. I know a lot of other instruments where several components
are tied to higher voltage levels and not any provision is done to
prevent shorting or so.

BTW: All of my BC 238 C transistors produced very similar curve
families, so they seem to have close tolerances. I think I will try
some older OC 26 now and see what happens. Joe





 

Hi Joe,

Michael and Chuck are right - there is no such thing as a retrace on the (analogue) CRT curve tracers. A retrace is something you have to adresse when you display a raster.

The HV in these units (as well as in scopes) is generated with a flyback transformer circuit, but that has no connection to the horizontal deflection circuits.

cheers
Martin


Sean Turner
 

Joe,

As others have pointed out, there is no blanking on the 575. However, it is worth studying the manual and understanding what the different positions of the "Steps/Sec" control do for you. The retrace line from the middle and bottom positions can be useful, as it can be used as a load line!

Sean

On Tue, Mar 30, 2021 at 01:41 PM, Joe wrote:


Hello,

Today I let the unit running for an hour at regular mains voltage and did
comparison tests on a pair of BC 238 C transistors. What can I say? I am
deeply impressed by the bright, sharp and stable display. That is great!
Nevertheless I registered a minor problem that has to be solved: The flyback
traces of the beam are "visible, too bright".

@ John: Yes I forgot to mention these sticky fan bearings. Well, a little
"Ballistol" oil did the job quickly.
@Jeff: Thanks, your explanation must be right. That tells us the Tek designers
did a very good job. I know a lot of other instruments where several
components are tied to higher voltage levels and not any provision is done to
prevent shorting or so.

BTW: All of my BC 238 C transistors produced very similar curve families, so
they seem to have close tolerances. I think I will try some older OC 26 now
and see what happens.
Joe


Keith
 

Love this story! This is so typical of how comparatively well tube gear ages. I’ve seen this over and over with various tube vs. solid state hi-if and studio gear. The tube stuff - which we were told by industry experts was “unreliable” compared to the fancy new transistor stuff they were pushing on us - has in practice outlived and out-valued the transistor stuff. I love the irony of it all.

You just have to have hung on to the transistor stuff long enough to see the failures happen.

For example, I have on display at our production facility a circa 1961 Roberts 770x (all tube) and a circa 1962 Roberts 771x reel to reel. Both are virgin mint condition pieces that have been babied their whole life. Guess which one still works? 🤓

Of course I know in my heart that Tek could never have done the cool stuff we love and depend upon these days without modern transistor designs.
But I also know from some decades of personal experience that my really old tube stuff tends to come on and work no matter the age, while my really old transistor stuff does not so much exhibit this same robustness.


Joe
 

Keith,

the same thoughts apply to "musical instruments". In my living room we have got a Rock Ola 434 "Concerto" juke box circa 1969 tube with tube amplifier (4x 6BQ5 PP). Once you have listened to the old 7" vinyls on this one you are sure nothing else compares to it!

Well I have found out in that Tek 575 there is at least one weak emission 6AU6 tube in the sweep amplifier. I am going to check all the tubes on a tube tester, maybe I will have to put in some new ones. Up to now I haven't seen one that I would classify as "impossible to find" as NOS. I would not want to use Chinese or Russian substitute tubes, just for the reason that this was not original.

Joe


Bert Haskins
 

On 3/31/2021 3:38 PM, Keith wrote:
Love this story! This is so typical of how comparatively well tube gear ages. I’ve seen this over and over with various tube vs. solid state hi-if and studio gear. The tube stuff - which we were told by industry experts was “unreliable” compared to the fancy new transistor stuff they were pushing on us - has in practice outlived and out-valued the transistor stuff. I love the irony of it all.

You just have to have hung on to the transistor stuff long enough to see the failures happen.

For example, I have on display at our production facility a circa 1961 Roberts 770x (all tube) and a circa 1962 Roberts 771x reel to reel. Both are virgin mint condition pieces that have been babied their whole life. Guess which one still works? 🤓

Of course I know in my heart that Tek could never have done the cool stuff we love and depend upon these days without modern transistor designs.
But I also know from some decades of personal experience that my really old tube stuff tends to come on and work no matter the age, while my really old transistor stuff does not so much exhibit this same robustness.
Ya pays your money, ya takes your choice.

Both my really old tube stuff and my really old transistor stuff are all going strong.
I have moved into my eighth decade of working with electronics and the only tubes that I feel any affection for are CRTs.

What is it in the 771x reel to reel that can't be fixed?




 

On 31. Mar 2021, at 21:38, Keith <coolblueglow@gmail.com> wrote:

Love this story! This is so typical of how comparatively well tube gear ages...
Transistors vs. tubes.

I know another story: mechanical vs. electromechanical vs. computers... in railway signalling!

Here in Germany, but actually all over the world, you still see many mechanical interlockings in operation (in english its called signal boxes, sometimes). In the 30ies up to the 70ies came relay technology, and computer technology beginning in the 80ies.

Guess what: the mechanical ones live forever (i.e. since more than 100 years), as long as there is someone with a hammer, a lime and an oil can. The relay ones continue their service, weakest point I know of is the cabling between relay groups where wires with red or brown isolation become brittle. Relays are still in production, so they can be repaired without problems. Yet the first computer interlockings turned out to be a pain in the ass to keep running when the semiconductor industry decided to no longer make an 8085 processor. You have to redesign the hardware and parts of the software with all the safety certification stuff, which is coming close to a complete redevelopment. Today this problem is addressed by an "obsolescence department". Yeah!!

Safety-wise computer and relay interlockings are comparable. A bit better than mechanical interlockings as they ease workload from the signaller. Nevertheless today, every manager in railway industry seems totally drunk when boasting a new generation of "digital interlockings". They forgot that mechanical interlockings were already fully digital...

cheers
Martin


Harvey White
 

In 1961, I'd call tube technology "mature".  Barring new designs (nuvistor, etc), Tubes were well understood.  Transistors were not, and certainly the manufacturing process was still young.

Harvey

On 3/31/2021 3:38 PM, Keith wrote:
Love this story! This is so typical of how comparatively well tube gear ages. I’ve seen this over and over with various tube vs. solid state hi-if and studio gear. The tube stuff - which we were told by industry experts was “unreliable” compared to the fancy new transistor stuff they were pushing on us - has in practice outlived and out-valued the transistor stuff. I love the irony of it all.

You just have to have hung on to the transistor stuff long enough to see the failures happen.

For example, I have on display at our production facility a circa 1961 Roberts 770x (all tube) and a circa 1962 Roberts 771x reel to reel. Both are virgin mint condition pieces that have been babied their whole life. Guess which one still works? 🤓

Of course I know in my heart that Tek could never have done the cool stuff we love and depend upon these days without modern transistor designs.
But I also know from some decades of personal experience that my really old tube stuff tends to come on and work no matter the age, while my really old transistor stuff does not so much exhibit this same robustness.





Brenda
 

Hello Joe,

First off, I just want to say that's great that you got that 575 up and running! The only thing that I would say is not to rely on your tube tester for testing tubes. Tektronix says it over and over in the manuals for the scopes that I have, I don't know the quote right off-hand, but manuals do say not to rely on tube testers, replace only when that tube is actually causing a problem, otherwise leave it alone. I have several tubes in my Tektronix 535A and both of my 561A's that have issues, but I use them because they are not causing any problems. With tubes getting harder to find NOS, in my personal opinion, use that tube until it does create a problem.

Brenda


Michael W. Lynch
 

On Wed, Mar 31, 2021 at 04:28 PM, Brenda wrote:


The only thing that I would say is not to rely on your tube tester for testing
tubes. Tektronix says it over and over in the manuals for the scopes that I
have, I don't know the quote right off-hand, but manuals do say not to rely on
tube testers, replace only when that tube is actually causing a problem,
otherwise leave it alone.
In contrast to most audio gear, most tubes in TEK instruments do not need to be at the pinnacle of their operational function to work as designed. Tektronix made sure that a tube that might test "weak" in a tester will still operate acceptably in their instruments. Laughably, those people who scavenge tubes from this old gear probably do not realize that many of these tubes are not in the best condition. I guess that they think since the tubes are in that pretty TEK Blue Cabinet, that this imparts "magical powers" to said tubes.

--
Michael Lynch
Dardanelle, AR


Renée
 

I have always believed in the wisdom told to me eons ago- the best test bed is the equipment itself, if the circuit works leave it alone. most circuits were designed to work just fine at 50% tube functionality...so far that has held true..there are always exceptions...but then the circuit was not working.....
Renée

On 3/31/21 4:37 PM, Michael W. Lynch via groups.io wrote:
On Wed, Mar 31, 2021 at 04:28 PM, Brenda wrote:

The only thing that I would say is not to rely on your tube tester for testing
tubes. Tektronix says it over and over in the manuals for the scopes that I
have, I don't know the quote right off-hand, but manuals do say not to rely on
tube testers, replace only when that tube is actually causing a problem,
otherwise leave it alone.
In contrast to most audio gear, most tubes in TEK instruments do not need to be at the pinnacle of their operational function to work as designed. Tektronix made sure that a tube that might test "weak" in a tester will still operate acceptably in their instruments. Laughably, those people who scavenge tubes from this old gear probably do not realize that many of these tubes are not in the best condition. I guess that they think since the tubes are in that pretty TEK Blue Cabinet, that this imparts "magical powers" to said tubes.


 

On 1. Apr 2021, at 01:37, Michael W. Lynch via groups.io <mlynch003=yahoo.com@groups.io> wrote:
.... In contrast to most audio gear, most tubes in TEK instruments do not need to be at the pinnacle of their operational function to work as designed. Tektronix made sure that a tube that might test "weak" in a tester will still operate acceptably in their instruments...
Very interesting remark. I think Tek has been obliged to make a very tolerant desigg. When you sell a scope with 100+ tubes inside, making it work with even "weak" tubes is not a nicety but really a necessity. Otherwise you would end up with a doorstop.

Makes me think of the ENIAC and its 18000 tubes. They say they always had to replace a handful of tubes when they switch it on (or so). The design must have been very tolerant, too, otherwise it would not have been usable at all. But I don't think they had to swap 18000 tubes to find out which one created a problem. Anybody knows how they detected the tubes that made problems within a reasonable amount of time?

cheers
Martin


 

Martin,

I expect that finding faults in the ENIAC was pretty much the same as finding faults in any digital computer: unlike a tube radio (or oscilloscope) a digital computer is composed of a bunch of relatively loosely coupled subassemblies, and faults can be easily isolated to the subassembly before you need to start doing detailed tests. Assuming a parallel organization (not a very good assumption for machines of the vintage of ENIAC) the memory system would be composed of a set of identical amplifiers and flip-flops (and the storage devices, maybe a mercury delay line). If a tube failed in one of the amplifiers it would be immediately obvious because you would have a single bit stuck on or off in every word fetched from memory. Similarly for a failure in the logic or control assemblies.

I also expect that they had test routines designed to detect and isolate all sorts of specific failures (I don't know about the ENIAC, but these sorts of test routines were commonplace by the 1960s), so you would be able to narrow down the source of a malfunction pretty quickly using the test routines, and then only have to test several dozen tubes in the failed subassembly.

-- Jeff Dutky


Keith
 

Interestingly, tube manufacturers produced a whole series of special tubes for computers. Most of them were twin triodes, the later ones loosely based on the same general physical design as the familiar 12Ax/Ay/Au/7.

The uniform features touted by the manufacturer for these so-called “special quality” computer equipment tubes was generally something like this

“Specially designed to operate for long periods in cutoff” etc.

Here’s an example - the 7044 from about 1969

https://frank.pocnet.net/sheets/127/7/7044.pdf

Here’s another one - a true SQ computer tube built for 10,000 hours...the EC180 - circa 1968

https://frank.pocnet.net/sheets/009/e/E180CC.pdf

(The “regular” version of this tube was the 7062)

In these computer tube designs, they were ruggedized with an emphasis on functioning as an on-off switch rather than a linear amplifier. Not a surprising design quality, given the application.

Much to the chagrin of the audiophooles, these design features mean they generally don’t make very good choices for Hi-Fi. Of course this doesn’t stop them from using them on occasion and applying the usual audio-fool language. “Broad soundstage! Z-axis Sonic dimensionality is breathtaking, blah blah blah” 🤣