Tektronix Type 575 curve tracer : repair and restoration

[med6753]

Following too.

[factory]

Yeah Vince,

The Selenium smells bad, but is highly toxic as well. The only remnant of my 555 is the cart I had for it, now the 575 sits on it. If I just had room to roll it around!!! 

I'm a little surprised the Tek diode modification doesn't use any dropper resistors, particularly for the +100V supply, as the pass tube will be dropping the excess voltage from the more efficient silicon diodes, then again they were probably adequately rated (other TE such as PSU's may need further modification, to avoid exceeding maximum voltage ratings of pass components).

I would have thought TE manufacturers would have used parts with decent ratings (unlike TV/radios where lower cost/rated parts might be used), plus the Tek 575 seems to have decent forced air cooling for them too (unlike those used in industrial, crane* & lift/elevator etc. control boxes which might only be passively cooled), the low voltage diodes also have fuses fitted before them, which should prevent overheating from a shorted smoothing capacitor.

*I remember maintenance bringing a rectifier for a crane brake to me several times over a few years, to replace the shorted silicon bridge, to me it looked to have originally had a metal or selenium rectifier.

Selenium rectifiers & suppressors are still available from GD rectifiers, they can't be the only supplier.
https://www.gdrectifiers.co.uk/products/category/power_assemblies/selenium_rectifiers/
https://www.gdrectifiers.co.uk/products/category/power_assemblies/selenium_suppressors/

Personally I have lots of test gear with Selenium diodes and no plans to replace them unless they fail, which could be due to overload from shorted capacitors, or the voltage drop increasing with age, YMMV.

It would be interesting to hear about any Selenium diodes in old test gear (not TV or radio) that have failed with smoke being released, just can't seem to find anything with a quick search.

David

[Vince]

Thanks chaps ! 


@Robert : power supplies all good.

-150V (reference rail) : -149,0V ripple 2mV on AVG DMM (so 6mVpp or so ?)
+100V : +98.8V  ripple 3mV AVG, <10mVpp
+300V : +298,9V ripple 4mV AVG so 12mVpp or something I guess.

Have not dared looking at the ripple on my scope (no diff probe here nor isolation transformer beefy enough), even though when you think of it, the Tek having a linear supply, its transformer is actually isolating the scope, doesn't it... and its earth pin isn't even connected..

Still, I am paranoid so I didn't scope it.

It's worth mentioning how thoughtful Tek was wrt to testing the power rails : on a 500 scope there are no test points. You need to tilt the scope on its side, remove the bottom cover, and probe directly on ceramic strips, on fully exposed circuitry... better watch your fingers and not slip your probe, or it could ruin your day and/or that of the scope !

But... on the 575 it's much easier and safer : you can leave the instrument upright in its normal position. Just pop the RHS cover and you have 3 test points right there, laid on the side/edge of the lower deck, conveniently labeled as well. Test points are well insulated / safe.. it's hard to screw up when doing the measurements ! A delight....

@ Factory : Wow incredible that one can still get these old fashioned Selenium rectifiers, good to know ! Who knows maybe one day some tiny fab in china will start making obsolete Germanium trannies as well ! 

BUT .... BUT !!   What do you know... I checked the service manual and this Selenium looking unit is apparently GERMANIUM not Selenium !  6 cells, rated 0.5A / 300V each. So maybe they are more reliable and don't smell crap and intoxicate you with their fumes.  I will therefore leave them well alone since they work fine right now. Also... they are NOT used to rectify the main power supply.... no. They use it to rectify the Collector voltage !


- The two crusty power Germanium transistors : looked them up in the schematics (see attached). They have an easy life, not going to burst into flames any time soon ! I mean.. they have nothing to do with the power stuff associated with the collector voltage/current....nope. These two Germanium trannies, as beefy as they look, are tied together in a Darlington kind of way, and see only 28Volts, and current wise they only supply the BASE current for the DUT !
So not too worried about them, and even if they do go the instrument will not catch fire...

- Good news : I managed to FIX something !     When I removed the RHS cover panel, I looked more closely at the compartment at the top, wrt to the door being stuck closed. I noticed the hinge (a simple metal rod) was sticking out quite a bit on the visible side, like 8mm or so. Looked a little suspicious to me. I tried to push on it with my finger to see if it would budge... it did ! It went back into place... then all by magic the door alignment corrected itself, and now the door can be closed and opened just fine, every time, never misses  a beat ! One problem fixed !

I updated my original introductory post in the thread to reflect that... all the problems are listed there quite tidily, so I think it's quite practical to just go back there to update it anytime I manage to fix something. Makes it easy to see how the restoration is progressing.

This compartment is also MUCH deeper/bigger than its counterparts in 500 scopes (where it's super shallow and you can not fit anything more than the booklet it's intended to host), because well... there is less stuff to cram in the instrument.
It's actually large enough to host a set of probes I think ! Too bad it's not a scope, doesn't need probes ! 
Oh well, I guess you can use it to store some bacon to keep it warm... 


Spent a few hours playing with the thing to work on the trace instability issue. I think I got somewhere. Didn't find the root cause of the problem so no fix, but I gathered lots of interesting clues which is a good start...

1) Took a wooden stick, and tapped on all of the tubes, every component on every ceramic strip, tapped on all the power components on the LHS where the collector stuff is. Did the same underneath the instrument as well (flipped it on its side). Even tapped on the CRT where the deflection plates terminals exit the glass tube. Also tapped on the big connector at the back of the CRT... then wiggled the wiring harness connected to it... then its parent harness...

Result : NADA ! Not a sausage !   

2) Played with the controls / knobs, to see if there were dirty even though I had already given them a good shot of Deoxit the other day... still nothing. Not a dirty switch issue.

3) Gave the top and bottom decks a good shag... nothing.

The instability remains there, perfectly unchanged, no matter how I torture the instrument.

Kinda disappointed.... no easy fix for Vince !

So then tried the next approach : tried using the instrument, playing with all the knobs, trying different things to see if I could make the instability go away, or at least discern a pattern at some point.

That's where I made progress ! Again no fix just yet, but still some progress I think !    I will let you judge of that...

In short :

1) The instability remains NO MATTER WHAT

2) It is present on both X and Y axis.

3) Its amplitude remains unchanged REGARDLESS of X and Y amplifier setting/sensitivity. The  instability spans about 2 small graticule divisions on the X axis, and a bit more, almost 3 divisions on the Y axis.

4) Is present whether I test a transistor or a diode. That is, whether you make use of the Base Step Generator or not, hence one can rule out the entire Base Step Generator subsystem, I think...

5) It is present with or withOUT a DUT present ! That is, you connect nothing, you get just a flat line on the screen, zero current to be measured, hence should have zero "crap"/instability on the Y axis... yet you STILL get crap on the Y axis...somehow.


Then I managed to find a pattern at last : I tested a Zener diode, so that I can display a sharp "knee" on the screen. This gives me a clean horizontal and vertical trace on the screen, that I can use as a " test pattern " to observe what the display is doing exactly, on each axis.
That was a good idea... I made a short video illustrating this :




What this shows is that the "span"/width of the unstable area of the trace, varies depending on WHERE it is located on the CRT !
Weird.
It does so similarly on both axis, it's symmetrical left / right and top / bottom.

Basically, it is most unstable on the outer part of the CRT, that's where you get the 2 or 3 division span. Then as the trace gets closer to the center of the CRT, the extra/unwanted/random crap/traces, progressively get closer to one another, and toward the center of the CRT they kinda "merge" and give you the illusion of a solid (hence stable) line, albeit a very thick line.

The video illustrates this :

1) I start with the trace centered on screen : the vertical part of the trace is centered, so is the horizontal part. You can see the trace looks thick but stable.

2) Then I move the trace vertically, downward. You can see the trace become more and more unstable as I get closer and closer to the bottom of the CRT.

3) Then I move the trace to the top of the CRT and you can see it acts similarly, it's symmetrical.

4) Then I center the trace and this time move it horizontally, left to right, and you can see a similar behaviour.


So ... it's weird, but at least I have now a pattern that's perfectly reproducible, like Swiss clock work, so I have something to chew !
So looks like it's to do with the high-level stuff so to speak. Does not matter what you test or even don't test with the instrument, does not matter how you twiddle the amplifier knobs or collector voltage/current settings, or base step generator.... it's above all that, it's common to all of these things...

That rings " power supply issue "... but they are fine.

So something else... The CRT itself looks fine...  so something in between... final drivers for the CRT plates... but it's weird that it would affect both X and Y at the same time... so it should be something common to both X and Y display... pfffffff..... I really don't know.  Need to study the service manual, theory of operation, overall block diagram, schematics, hoping to get a bright idea some day...

[Vince]

Forgot... I turned  the thing off... then an hour later went to turn it back on... and it would not ! Dead as a dodo !

Blown fuse.... looked really crusty, and was not the appropriate one ! Read 1.6A fast blow. Should be 2A slow blow...   all the big 500 Scopes have a similar fuse, so I had already a 10 pack of those in stock, yeah ! Popped a new fuse, back to life ! 

So the fuse was not the right one, this was yet another problem I had not listed... but now FIXED ! 
One less problem on this bad boy ! Will get there one small step at a time, but I will get there

[xrunner]

Forgot... I turned  the thing off... then an hour later went to turn it back on... and it would not ! Dead as a dodo !

Blown fuse.... looked really crusty, and was not the appropriate one ! Read 1.6A fast blow. Should be 2A slow blow...   all the big 500 Scopes have a similar fuse, so I had already a 10 pack of those in stock, yeah ! Popped a new fuse, back to life ! 

So the fuse was not the right one, this was yet another problem I had not listed... but now FIXED ! 
One less problem on this bad boy ! Will get there one small step at a time, but I will get there

That's one small step for Vince - One giant leap for Tektronixkind. 

[Vince]

OK, already another update ! Could not sleep so been burning the midnight oil working on the thing, investigating further this trace instability issue.

To be 100% sure it was not the CRT subsystem at fault, I decided to do a quick test to rule it out, as there is something one can do easily... At the CRT glass, I unplugged the 4 wires for the deflections plates, so I can drive them myself, manually so to speak.
This way I could see if the instability would go away or not.

I connected one set of plates to my sig gen so as to be able to sweep the beam, and draw a straight line across the display. Then I could use my lab power supply to drive the vertical plates to make the line go up and down, and see whether the trace would become unstable as I move away from the center of the CRT. That was the plan...

At first I connected the sig gen to the horizontal plates so as to draw a horizontal line, seems more natural to me, somehow. However even when cranking my sig gen full blast, i.e. 20Vpp... that would produce a deflection / trace length of only 1.5 DIV or so !     Not very usable for my purpose here. So instead I connected the sig gen to the vertical plates, as it turns out they are a bit more sensitive. They would produce a deflection of 2.5 DIV or so.. still not great, but much better none the less.... so had to make do with that.
Then I connected my lab DC power supply to the horizontal plates. Twin floating 30V outputs. Put them in series so I can do 0 to 60Volts. Even 60V would give me only a 4 DIV deflection ! I need a high voltage adjustable supply I think...

The trace was extremely blurry, buried into a huge blob of noise. But, I could still discern the trace in all that noise, and could conduct my little experiment. The result was not what I had hoped.. at all.... the trace would STILL exhibit the same instability issue !!   

So... it means it does come from the CRT subsystem itself ! That explains everything, it's the easiest and most logical explanation I thought... but.... fuck me if the CRT is bad all is lost and it's the end of the restoration !! 

OR... I thought with some "luck", it could be due to the  CRT supporting circuitry, rather than CRT itself . I mean there is focus, astigmatism, geometry settings in the CRT gun, and all these things interact with each other and can produce weird and wonderful things on the CRT target...

So, I figured maybe a bad cap somewhere in the CRT circuitry... but for now, while I am here, let's just clean all the contacts as a first measure ! Yes it's the only contacts I had not yet cleaned, I admit... 
So I cleaned the pins and wire connectors  for the deflection plates. They look dark grey, as if they were oxidized... but not really. It kinda looks like it's their normal finish... still, just in case, I put some Deoxit in there and used a tiny/pen style wire brush.

Then I moved to the back of the CRT neck. Pulled the big round connector, Deoxit in there and on the pins on the CRT side. Brushed the pins. Exercised the connector a few times on the CRT pins to help the Deoxit work its way.

While I was in there with this area apart, I did some cleaning. Got rid of the ton of crusty debris trapped around the neck, bracket, nylon clamp etc... looked horrible. Used the tips of my fucked up cheap tweezers as a pick tool to dislodge every crusty bit in  every nook and cranny. Then a bit of cleaning with hot water/detergent in a spray bottle, a brush, kitchen towels..
Looks 10 times better than before. Much cleaner, no crusty debris all around the neck/clamp.. now looks much more appealing. Not as off-putting as it did.

See previous / original pics posted on the first page, to appreciate the change ! 

With all the crust removed you can better see all the parts of the mechanism so you better see what's going on and needs doing... clearly as I initially thought, the part that embraces the neck and rotates inside the metal bracket... is indeed not there, or not anymore. There just isn't anything in there. The neck is floating a quarter inch away from the metal bracket !   

So I am starting to think that maybe, since the seller did Frankenstein this unit from 2 units... maybe he just had no good part in either unit, so simply didn't put one in, simple as that !    Picture of the missing part, exploded view from the service manual.

So, once I had "refurbished" the CRT, I powered the unit and gave it a test ride : PROBLEM FIXED !!!!   

Stable trace now, left to right and top to bottom ! 

But after having worked on the CRT, of course the alignment was completely off, like... off by 45° or so ! Needless to say the picture looked a bit weird.... so before taking pics of my success... I grabbed a pair of pliers to grab the neck and rotate it while powered up. A bit stressful I must say... Got it almost perfectly lined up with the graticule... a bit more... a little bit more... allllmost there.... OH NO, trace is unstable again, and massively so, it's all over the shop, NOOOOOOOOOO !!!!!!! 

But I was not too worried, it was just proving the point. The problem was now most certainly due to a bad contact in the big neck connector. So I just moved the CRT a tiny bit and hey presto, picture perfectly stable again !    Played with the thing for 30 minutes, rock solid    PHEW !!!   

So the mystery is solved then.... I am glad it was that simple, no weird electrical bug to chase in the amplifiers or god knows what other part of the instrument !

A few pics to prove my success....

However I still have this problem with the squiggly lines towards the right end of the curves, that "jump" from one curve to the next. But it's a minor issue. Does need to be solved of course, but it's minor.  The instability issue was the real show stopper so I wanted that fixed or at least diagnosed, before I go any further with the restoration.

The squiggly lines however, whatever it is, I am sure is bound to be a minor problem which can be fixed any time in the restoration process, either before or after the full nuts and bolts cosmetic restoration.

For now, in the interest of making nice clean pictures, I made the squiggly lines go away by playing with the X amplifier controls, so as to make the squiggly lines go far away off screen, far away to the right.

Plus, I may have some clue about it. I noticed something interesting : the " STEPS / SEC " toggle switch. It has 3 positions :

120 / 240 / 120.

I.e 2 or 4 times the line frequency. Except of course in France / Europe it's 50Hz not 60Hz like in the USA, so I would get 100/200  refresh rates, not 120/240 !  Yet, the face plate of the instrument does say 120/240 not 100/200, even though it was not manufactured in the USA. Serial number does start with a 'zero' digit, not a 'B' (Beaverton / US plant). So I find this kinda lame from Tektronix ! 

Also, I don't understand why there are TWO positions marked " 120" ! There is nothing on the face plate that tells you what the difference between the two is ! Well maybe if I RTFM I would learn about it...

Anyway, the interesting thing is : the squiggly lines show up only in TWO of these 3 positions ! The 240 position, and only ONE of the two " 120 " positions !  Yes... somehow they don't behave 100% the same.
So that's great. I can now study the schematics and see how this working "120" setting differs from the other 120 and the 240.

So, I am making progress I think ! 

At any rate, now that the instability issue has been diagnosed and kinda fixed it seems, the unit now works well enough electrically, for me to be happy moving on with the restoration !!! 

[Vince]

Forgot to add : I looked a bit into that Selenium rectifier, and the silicon rectifier mod Tek offered.

I looked the latter up at the end of the service manual. My unit already has this mod implemented....

The big "Selenium" rectifier "stack" we see in the collector sweep area of the instrument, is nothing to do with the rectifier mod listed in the manual.
Indeed as the parts list pretends, although it does look like your typical Selenium rectifier, it is not. It is indeed Germanium not Selenium.

I looked it up. The part number printed on it is " GE4JA211CX278 " quite a mouthful eh ?!

The first two characters " GE " sound good eh ? Sounds like ... GE-rmanium ! 

Wanted to be 100% sure so googled it. ZERO hit ! However if you remove the last 5 characters and only consider " 4JA211 ".... then I found ONE hit.

OK to cut it short,  " GE " stands for " General Electrics " not " GE-rmanium ". Google found me the relevant page of a period GE product catalogue. They advertise this very rectifier.  "4JA211 " was the name of a series of rectifiers they made. Catalogue clearly states it's Germanium not Selenium, and it even shows a photograph of the rectifier, which matches 100% what's in my 575...

So there you go ! 

So no Silicon mod for me, it's already done. And no discarding the crappy / dangerous / ticking time bomb Selenium rectifier... as I don't have any in this unit !    So I will leave this Germanium rectifier stack well alone...

[Robert763]

Given the location in the circuit I would suggest that the Germainium rectifier was chosen because of the lower forward voltage drop compared to other technology available when the instrument was designed. A modern schottky diode might e siutable as a replacement. But.
"If it ain't broke, don't fix it"

[xrunner]

So, I figured maybe a bad cap somewhere in the CRT circuitry... but for now, while I am here, let's just clean all the contacts as a first measure ! Yes it's the only contacts I had not yet cleaned, I admit... 

Well I'm glad it's fixed but your killing me with this kind of suspense "Yes it's the only contacts I had not yet cleaned, I admit... "

 

But about that supporting collar. What is the temperature right there where it sits. I imagine it's hot but how hot? It can be designed for a 3D printer job but the material will have to be selected based on the temperature there,. i.e. PLA probably wouldn't work but there are others available ...

[Vince]

But about that supporting collar. What is the temperature right there where it sits. I imagine it's hot but how hot? It can be designed for a 3D printer job but the material will have to be selected based on the temperature there,. i.e. PLA probably wouldn't work but there are others available ...

Well I don't know I didn't put a temp probe in the thing while it was running !   Though that would be interesting I must say..... need to buy a temp sensor for my DMM first ! 

However no need to bother reinventing the wheel Martin.M  said a guy in his local forum, in Germany then, actually MAKES this very part already !  So I can just contact him.  However I don't even need to do that, no need spend any money : As I had hoped, I was able to salvage this part from my favorite donor, my crusty type 531A scope ! 
The part is made of hard plastic, that used to be clear plastic I assume... which is now heavily yellowed.... but it's still in perfect shape !
Maybe I could try " retro-brighting " it ?!  Worth a shot I suppose...

So that's what I have been doing today. Taking this part off of the 531A, which is a bit of a pain, then putting it on the 575. I was almost done then hit a brick wall : before I can put the CRT back in place, I need to tighten a bolt that secures the back of the CRT shield, to the metal bracket that holds the neck in place. But... that bolt goes through a little hollow spacer (see picture).. which I LOST !    When I took things apart, I heard it fall inside the scope, sure of that... it fell inside the scope not on the ground... searched on the ground anyway, no joy. Nowhere to be found in the scope either... I don't understand.... 

Maybe I will find it on the bench somewhere... some day. I can't pinch that spacer from the 531A because it is designed differently. Instead of having one single long screw going throughout a hollow spacer, what they did was use a threaded spacer, on both sides, with two tiny screws, one at the top one at the bottom of the spacer. So no good for me.

But luckily I think I can make a new spacer easily. My local hardware store sells anodized aluminium tubes that look exactly like what Tek used :

https://www.leroymerlin.fr/produits/quincaillerie/corde-chaine-cable-tole-fer-et-profile/fer-et-profile/profile-aluminium/tube-rond-aluminium-anodise-brillant-l-1-m-x-l-0-8-cm-x-h-0-8-cm-x-diam-8-mm-80108686.html

Well that's just an example. They have several sizes/diameters available so need to take measurements so I know what to buy once I get there... hopefully tomorrow (22H30 here, too late...).

While assembling the mechanism, I noticed that the white nylon clamp on the 575 is broken   
Luckily not in a show-stopping way, as you can see on the picture. The clamp part itself is intact, so it can grad the CRT neck just fine, that's the most important.
However as you can see the bottom part of the clamp, which holds the big nut you use to rotate the clamp... is cracked. Luckily this area is still strong enough to operate properly even with such a massive crack, phew !
I might try to repair it though....
I can't use the clamp from the 531A because it's broken too, but in a different place. It's the clamp part itself that's broken in two, so it's 100% unusable, scrap.

However... I could use this useless clamp to help me repair the one on my 575 ? I mean I could scrape some plastic material off of it to try to "weld"/fuse the crack in the 575 clamp ? Could also use it to practice welding on it, before I attempt a repair on the 575 clamp....
This way I could practice with the hot air station to figure out the optimum temperature to melt it. Hot enough to get a good bond, but not too hot that it deteriorates the plastic.

Yeah that could be a little project 

But for now it's gonna stay as it is, since it works... I will just make a new spacer real quick as it's really mandatory... but I will leave the eventual repair for the later stages of the restoration. For now, I am doing the first pass only : trying to figure out how I can fix things, try to see what parts I can salvage from others scopes... solve the big issues first. The fine tuning and polishing, that will be done later...

Anyway, I am happy I have this CRT assembly sorted out ! I am glad I have this 531A scope donor, would be screwed without it !  So next time someone asks me why I keep buying all these scopes.... well I need freaking PARTS at hand so I can restore those that can be saved and deserve to !
For example I have FIVE type 317 scopes ! Yet I have already restored one of them... so I don't need to keep the other four. Will I sell these 4 then ? NO WAY !!!     They are like little gold mines for parts of all kinds to save my bacon when restoring other scopes !

[med6753]

Why do you think I have a closet full of parts scopes ranging from a 2465 DMS to a Type 310A and several in between? Plus several 560/500/7000 series dead plug-in's. And at least 4 boxes/totes of additional parts. 

[Vince]

Good to know, if I need something I will PM you ! 

[Vince]

Worked a few hours on the thing again today. Making progress.


CRT
- Tried "retro-brighting" the plastic clamp for the neck, the part that I was missing. Didn't work at all, so I will have to get used to its massive yellowing.

- Lost spacer : went to the H/W store, bought a 8mm OD anodized aluminium tube and cut a new spacer, 8mm long. Used my tiny precision files made in China, to deburr the thing and make it look nice. Super happy with the result, looks like I bought one ! 

- put everything back together, reinstalled the CRT and gave it a test. I tested a Zener diode so as to display a test pattern that would give me a horizontal and vertical line as a reference. I played with the red knob of the clamp, to rotate the CRT, and it worked just fine. Te CRT is perfectly held in place, no wobbling around anymore, and I can easily fine tune the rotation. Had no trouble at all lining up the Zener diode pattern exactly to the graticule !   See pics.

KNOB REPLACEMENT
Big knob used to select the series resistor for the base of the transistor under test. Seller put an incorrect knob in there. Grey instead of black, and also with a big hole in the center, probably destined to host a secondary small red knob, I presume.

Once I had removed this horrible knob, I was in for a nasty surprise, see pic. Underneath the knob, 2 HUGE scratches, VERY deep!!!    Probably the guy didn't have the correct imperial size 1/16" Allen key so attacked the knob from under using a flat screw driver or something, and went crazy with it...     Problem is that even with the knob in place, you can still see a little bit of the scratches, it can't hide it all 

I replaced it with the proper knob, courtesy of my 531A donor as per usual.

Now come the good surprise : before, with the old knob I found it was behaving strangely: you could switch from the last position of the switch (highest base resistance, 22k ) to the first lowest resistance, 1 ohm) !! No stop ! I fond this crazy as one could easily by mistake blow the transistor under test in the blink of an eye ! I was bitching at Tektronix !
BUT... when I went to put the new knob in place, I had forgotten what position the old knob was in, so I reset the shat to put at one end. Screwed the new knob and.. hey presto, now you can't blow your tranny !  You can't switch carelessly from the highest to the lowest anymore ! Tek was not faulty, the old knob was imply not adjusted properly that's all ! Problem solved   
So the seller once again demonstrated his "talent"... destroys the face plate while taking the old knob out. Then can't even be bothered to put the correct knob and to top it all can't even be bothered to align the knob properly, meaning basically he had no clue what base resistor he was using as the knob was displaying garbage ! 
this guy blows my mind, really does !!! And he teaches electronics to engineers he said !! 

TEST PANEL
Well I call that the "Test panel" for lack of a better term... it's the "pod" at the bottom of the instrument that holds the sockets to fit the DUT. Well the bottom right corner of said panel is bent out of shape, it's sticking out. This also causes a binding post that so happens to live there, to consequently follow suit : it's not upright anymore, it's leaning a bit to the left, hence it no longer aligned with the other two binding posts above it. So I went to pull the panel out of the pod, hoping / thinking it would be easy and I could try to straighten the panel at my (mechanical) work bench.

Well it was not so easy...
There are four screw one can see from the outside, one in each corner.. I though OK let's just remove them and I can pull the panel. Nope. The top 2 screws you can remove. The bottom two you can loosen them but that's all, then they spin round, no luck. There is a nut behind the panel but it's not captive. Tilted the 575 on its side... oh there is a bottom cover one can remove. 3 screws. Then you get to see what's inside the pod, and you can remove  the two nuts. But the panel still won't move at all.
For that you need to free the switch : removed its big retaining nut from the outside. 16mm ( 5/8"), but need a deep socket so as to be able to clear the bat of the toggle switch. Standard socket no joy. So how do you do ? Well lucky me, I had a deep socket for the spark plugs on my car... it so happens to also be 16mm, how cool  ! 

So ow you can free the panel but... you can only lift it by a couple inches tops ! Because the bat of switch won't clear the panel, almost but not quite ! So the switch is keeping you prisoner !!!   So to pull the panel for good, one would have to desolder the very many wires going to the switch, which was way more than I had planned on doing, and way more than I could do this evening. So I tried to straighten the panel in situ ! Put a couple layers of masking tape on it to avoid marking it, then grabbed a pair of pliers and had a go at it, crossing fingers and proceeding with caution as you can imagine.
Well it went well. I managed to bend the corner of the panel into shape without making it look worse than it already was. It's spotless too, tape did its job.

Well to be honest I should have bent it a tiny little bit more... The corner of the panel looks good now, not sticking out any more, but the binding post is still sliggggggggghtly misaligned, just a hair. So I think later on I will probably have another at it. But I am splitting hair really. It looks fine. If you don't know what to look for, you would never see it I bet !
So overall very happy with the result.

ELECTRICAL
I just noticed that if you test a transistor, and if you look very closely at the CRT for the horizontal parts of the curves... then you see that lines are not "solid"... they show like a tiny ripple all over them ! OK  you say need to recap the power supply... maybe but as I said earlier, I have already tested the 3 main rails, -150/+100/+300 and they are all fine with very low ripple, it's within spec.

It does not do it when I test a diode... only with transistors.

I think I found the problem... think it's indeed a power supply, but not an obvious one.
As I had the 575 on i's side looking at the bottom of the test pod... my eyes inadvertently glimpsed at some canned caps, and noticed that two of them had a suspicious looking terminal ! See pic.  The positive terminal is covered in ... I don't kn,ow what. Looks like these caps leaked. Looked them up in the schematic, and then it all made sense : they are not part of the 3 main supplies I had already tested good. No. These 2 caps are part of a gang of 4 identical smoothing caps all in parallel : x4 2000uF caps, that filter... the power supply that supplies the base current for the transistor under test ! So yeah, I can understand how these faulty caps could forward their ripple to the screen, and why it would not affect diodes, since in this case, a dipole, you are not using the Base Step Generator so not using this power supply !

So... that's cool, looks like I may well be soon fixing yet another problem, and make the display on this puppy even more betterer as some would say ! 

I will replace all 4 of these caps I think, not just the 2 faulty ones. Not going to put crapacitors so I need to order online, so will cost an arm in shipping/minimum order, for 2 Euros worth of caps, pfff.... what wouldn't I do for my little 575...
Will have to buy some more stuff to meet the minimum order... but no more than that.

BTW, I pulled the condom off of one of these  cans, and it says 30th week of 1963. So mine is 58 years old !
Will probably be 60 by the time I am done restoring it....
Caps all rated at 20V.

CLEANING
Now that I fixed the test panel and knob.. thought it was about time to finish the job by giving the front of the instrument a good cleaning. Basic stuff, 30 minutes, no taking apart of anything, just some detergent with hot water in a spray bottle, a stiff tooth brush to clean all the crap on the knobs and every where... and off you go. 30 minutes later it now looks much better. Test panel is now so much more pleasing to look at ! Just compare before and after pics...

Now that it's clean I can see what defects / imperfections are left. Answer : all of them. That poor face plate is riddles with a myriad o very deep scratches, worn out spots, stains... it's hopeless, will have to leave with that. Again the poor 575 was in such a poor overall shape inside out the it was clear from day one that I could not do miracle and turn it into a concourse winner no matter how hard I try... it's worn out, can't fight that ! 
So the goal has alwyas been to make it look as good as possible, do everything that's reasonable, but that's about it.. I am no magician ! 

Anyway, I enjoy it much better already with what I fixed so far and this first pass at cleaning it, and anticipated the future improvement I will make to it's appearance as I go. So I think I can make it look quite decent and that will somewhat balance things out....

Test panel : as you can see, the paint is worn out / faded toward the left. I think because that's where people put their greasy fingers whenever they stick a tranny in the test socket. However, all is not lost ! I noticed that the artwork is NOT printed on the panel, as it is on the 500 scopes. No... look like Tek wanted to innovate with the curve tracer... all the artwork is actually engraved !    ALL of it... text, lines, symbols, absolutely everything. So basically I could fix the paintwork by just adding some more paint in the  "grooves", then wiping the excess and that's it ? Will need to experiment and ask around what kind of paint is best suited for that application. Need to have it custom mixed too, to get the colour right... so not something to improvise, but sounds doable eh ?!

Then I looked at the face plate.... you won't believe it.... they went wild and also engraved the face plate too ! ALL OF IT ! There is not a single thing that is printed on the face plate ! Even the Tektronix logo and accompanying text, it's all engraved !  wtf:
SO !!! That means that ONE DAY, when I am retired and have lots of time, and some money to spend... I COULD fix this 575 after all ! I could just sand down allll the scratches and blemishes and ever defect, make it 100% perfect. Then I would have to send it to some specialized company so they can re-anodize it just like Tek did. Then I can go wild with black/red/blue paint and redo all the artwork, then a couple coasts of lacquer to protect it all, again like Tek did... and hey presto, a brand new face plate !! 

I could also do the same all the chassis parts, most notable the lower deck, which are pitted/ corroded beyond recovery. Cleaning it will never make them look shiny and spotless again, forget it 
So here too, I could just sand them to removed every bit of pitting, then anodizing, then I will have to redo all the printing on it (vacuum tubes types and part number mostly), then lacquer. Brand new chassis.

But that's just a dream, it's well beyond the scope/ambition and budget of this restoration. But that's OK. Everything I do now, is something I won't have to do later on if I decide to have another go at the restoration one day.



Well overall it was quite a day for my 575 I think. Made quite some progress. fixed a few things, diagnosed some others, gave it a good clean, now looking much better.... yes it's a good day for him !


A few pics and off to bed, it's 7AM FFS ! 

[Vince]

..

[Vince]

Attached the schematic of the power supply, showing where are my two bad caps. So again it's the power supply that supplies current to drive the base of the transistor under test. It's not regulated, as simple as it can get : a few rectifiers, then followed by those 4 big 2000uF can caps in parallel. My 2 bad caps are two of these four.

[tautech]

TBH you'd be nuts not to replace the few extra PSU caps whilst you're in there. 

[med6753]

Agreed, take no prisoners. Replace them all, including the cans in the rest of the PSU.   

[xrunner]

Looking good Vince - long way to go.

[Vince]

No shotgunning here !  I am not in the US? I have no admiration for big weapons and mass shootings. I prefer finesse !   

The caps in the other power rials do their job just fine as is attested by the ripple that's within spec. Their job is to contain ripple. They do their job just fine, so by very definition, they are GOOD.  Have been good for 60 years, and you would like me to replace them for modern crap (no matter brand it is) that will last just a few years at best ? and might even die before the old one would have, had I let them in place ?! Silly ! 

These caps are going nowhere. I will replace them if and when necessary.

The only shot gunning I am will to do as I already is said, is to replace all 4 caps in the faulty power rail, rather than just the 2 bad ones.. because they are all in parallel, it's one entity. If 2 failed, maybe the other will soon, or are on their way already, don't know haven't checked. So it does make sense to replace all 4 at once, I fully agree with that.

But no more.

Get over it you mass shooters ! 


Trying to think of what I could do next on the 575, might have an idea or two. Stay tuned 


Oh BTW I checked all my old Tek scopes, and go figure, many of them DO have an engraved face plate too !  and I never noticed it thus far !!!

I am not sure why some are engraved and others are printed. It's not a price point problem I think, because some of my cheap/modest type 317 are engraved, yet the fancy super expensive type 547 and 549 are printed instead.
hell, I have even two scopes of the same type, two 515A... coming both from the sample plant ( code/prefix '1' ) one is engraved, the other is printed ! 

My best guess is that it's technological problem ? Maybe they sued engraving in the early years, then when printing became available they moved to that as it was probably cheaper / faster to produce ? If so, the S/N of my two 515A ought to reflect that. The S/N of the engraved one ought to older than that of the printed one... gonna check right now...

[med6753]

If you want to tempt Murphy and his capacitor goons then be my guest. Better to start with a "clean slate" rather than going back in again at a later date.

Kill them all. 

[Vince]

There is no reason to replace them early, no problem going back there later because in this 575 access to all htese caps is easy. I do not have to take anything apart to get to them. Whether I replace them this week or in 3 years makes absolutely no difference.

I do agree with your point though... you have to balance Murphy with the physical accessibility of the caps. In my 317 or a very crowded 547 or 549 or 545, it could well be that need to tear half the scope apart to get to the caps, in which case yes better to do it while you have the scope apart during its restoration.

But it's not the case in this 575. I just need to pop the covers off and the caps are all right there at my disposal.

Another good example would be a portable scope from the '70s. It's so tight in there, such a pain to take apart to get access to anything, that yes, once you have ti apart, you REALLY don't want to ever have to do it again ! 
So yes for this type of scopes, I too would replace all electrolytics, aluminium or dipped tantalums alike.

[Vince]

Just checked serial numbers on my two 515A. Seems to comfort my theory : the oldest of the two is the one that's engraved, newest is printed.

I might even be able to tell pretty accurately WHEN did this change occur, because it so happens that my two scopes have VERY close S/N !

100719 and 100800 ! 

So if I looked at date codes on the can caps, that would narrow it down pretty well.

[tautech]

If you want to tempt Murphy and his capacitor goons then be my guest. Better to start with a "clean slate" rather than going back in again at a later date.

Kill them all. 

Agree although Vince would be wise to at least measure the remaining 10 PSU caps to not only preempt their failure but also to protect the now days valuable glowing glass.
I do understand his resistance to yank those few big caps out and lose the originality of the design but there is always the option to repack the cans with new caps.

[Vince]

I am all for testing Tautech, as soon as I get my hands on an ESR tester ! 
So many things to buy, so little money !

[bd139]

Make one. I did  . Based on this excellent design: https://www.eevblog.com/forum/projects/5-transistor-esr-meter-design/

My attempt. Not pretty but it works: https://www.eevblog.com/forum/projects/5-transistor-esr-meter-design/msg2327025/#msg2327025

There's also an ESR adapter for DMMs here as well. I might build this as well as I have dropped my analogue one about 20 times recently and it's looking a bit bashed up now 

https://www.eevblog.com/forum/projects/esr-meter-adapter-design-and-construction/