Author Topic: Tektronix 453 Oscilloscope no display  (Read 2396 times)

0 Members and 1 Guest are viewing this topic.

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Tektronix 453 Oscilloscope no display
« on: July 27, 2018, 03:11:29 am »
Hey all! I'm going to preface this by saying I've got very limited knowledge of electronics. I'm much more savvy on the software side of things. I have a very basic understanding of how things work but that's it. I'm a pretty quick learner but you may have to go into more detail than you usually do in order to get me to really grasp a concept but I'm definitely excited to jump into this project!  :-/O

Wall of text warning! TL;DR at the bottom!

Bit of backstory: My dad has worked for IBM since the 1960s repairing the System/360 Model 30s and continued to work for IBM up into the 1990s. In 1996, when IBM was liquidating most of their old equipment, a coworker of his had stumbled upon this Tektronix 453 oscilloscope that IBM was determined to get rid of. He asked my dad if he wanted the poor thing, to which he of course said yes. Throughout the years it sat in the closet collecting dust, only to be brought out occasionally when our projector television was on the fritz and we needed some test equipment. Eventually I stumbled across the poor neglected thing and, while not knowing a lot about oscilloscopes, I was really attracted to the aesthetic and vintage value it had. I was also into sound reproduction systems so I had it set up for some time as a waveform viewer for my stack's audio output. This was probably about 8 years ago now, and back then it worked fantastically. The only problem I ever had was that the fan was quite noisy (which I've found is really common with these old 453's) and some of the functions didn't seem to work (though that could very easily have been user error, I basically messed with the thing and taught myself how to use it instead of trying to find a manual). Eventually it found it's way out of my system and retired itself to the closet again while things in the house got messy.

Fast forward a few years and I cleaned some stuff up and found some space to set it back up as a waveform viewer for another sound system. Got everything cable-managed, and flipped the switch. The front panel lit up, the scale illuminated but it was eerily quiet.... the fan had stopped working while it was sitting in the closet I guess... I took note of it and made sure not to run it for more than a few minutes at time (so I can assure you it hasn't been overheated). While continuing to fool around with it, I couldn't get any trace or sweep to show up on the scope. Thinking okay, maybe a knob or two got moved and the trace is off the display I hit the trace finder. If I remember correctly, if you hold down the button the the trace mark should persist on the screen but it only flickered for what seemed like a millisecond, mostly in the middle of the screen, then immediately disappeared. I'm like damn, this has more wrong with it than just a fan. I messed with the A and B vertical position knobs which didn't affect the location of the trace whatsoever.

I'm aware there are some pretty dangerous voltages (about 4kv to my understanding) and, for the most part, I've figured out where they are and I know the standard practices of being careful around them. I've created a grounding tool and I make sure to only use one hand at a time with this thing so that if I do get zapped, at least it isn't going over my chest. With those precautions taken care of, I started exploring around this thing. Firstly I made it a point to try to find all of the fuses. I've found five: 2 on the side of the case accessible from the outside, and three more in the back accessible with the removal of the rear cover. All of these fuses look good, so I decided to go looking online. I found this forum post on AntiqueRadios: http://antiqueradios.com/forums/viewtopic.php?t=143928 which describes a similar yet vague issue where the poster was given some good troubleshooting advice concerning the high voltage supply for the CRT. Unfortunately I don't think I have any meters that can handle 4kv so I dare not try to measure voltage at the tube. Following the advice found on that thread, I discovered I'm getting 22v at the fuse, capacitor, and transistor. I'm using the aluminum case as a ground and I was getting 22v at the transistor Q930 from both the outer shell of the transistor, and the inner post. I assume that's normal? He said I should be getting 18.7v so I'm a little high, not sure if that's a problem or not.

I decided to check out the HV rectifiers anyway and found something interesting. The rectifiers this model came with are glass tube 5642's. Mine are black "plastic" with a part number 152-0408-00 for the smaller 2 and 152-0192-00   for the larger one. I found out that these are Tektronix part numbers but I'm guessing the scope was serviced at some point in it's life and that they were replaced. Not being sure if the same diagnostic techniques could be applied, I checked for resistance (after grounding) at all three and none of them were open. According to http://www.sphere.bc.ca/test/tekequiv.html however, your standard ohmmeter can't be used on the larger one for reasons I still don't understand. I'm also using an IBM 8060A/AA multimeter by the way.

While I was this far in, I also disconnected the fan from the case, only to confirm that it's not socketed or clipped, but rather soldered in place, making removal a bit of a task, especially considering the placement of the solder points. I did, however, play around with the fan while the scope was powered, found that in one direction it resisted my movement, and in the other it encouraged it but at no point was I able to give it enough momentum for it to continue moving on it's own. I also noticed that the motor casing got quite warm. So we know power is getting to the motor so I suppose we're either underpowering it, or the motor is just toasted.

At the end of all this I don't feel like I'm any closer to figuring this out and was hoping some of you wizards could help point me in the right direction, give me some troubleshooting/probing guidance and help me restore this poor thing. It's been on this planet for 50 years and I want to see if I can give it at least another few decades if not longer. If you feel it would be beneficial for me post high resolution videos or images of anything, let me know, I'd be more than happy to oblige. If you've gotten this far I applaud you for just that  :-DD

TL;DR: Tektronix 453 Oscilloscope, early model (S/N 10562) with the Nuvistor tubes. Used to work, now it doesn't anymore. Symptoms are a non-functioning fan (but the motor gets hot), no trace or sweep, trace finder doesn't persist on screen and the A and B channel vertical adjustments don't affect it's position, which is horizontally precise but flickers vertically. Found 5 fuses, all of which are closed and not broken. Getting 22v at transistor Q930 and none of the three HV rectifiers in the HV box are open. Help!
 

Offline PaulAm

  • Frequent Contributor
  • **
  • Posts: 684
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #1 on: July 27, 2018, 12:25:16 pm »
If you get any kind of trace on the CRT, even briefly, the HV is working fine.  If it's working, don't mess with it.  If you need to adjust it you will need a meter and HV probe that can safely handle 1500V, SO DON"T MESS WITH IT.  ;D

The first order of business is to check all of the low voltage power supplies to make sure they're working properly.  If you haven't already, get the service manual.  Free copies are available on the web.  If the power supplies are not right, nothing works!  The manual will show you where the test points are and what values you should see.  Do not do anything else until you've done this.  Do not pass GO, do not collect $200.

Once that's correct, the transistors are socketed and it's possible oxidation is causing a bad connection.  You may need to go through and reseat all of the transistors.  Most of the time you can pull them up a bit and push down again.  You could also pull them out, clean the leads with Deoxit and then put them back.  Make sure all the leads are back where they came from in the socket.

There have been some threads on rebuilding the fan on those scopes; you might want to check out tekscopes at groups.io

Tektronix also put out a publication on troubleshooting oscilloscopes.  You should find that and read it.
 

Offline particleman

  • Regular Contributor
  • *
  • Posts: 115
Re: Tektronix 453 Oscilloscope no display
« Reply #2 on: July 27, 2018, 01:08:05 pm »
453s are great scopes. Very much so worth repairing. Like PaulAm said you need the manual which will provide all the test points and wave forms to check. Also take a look at c937 while you are there anyway. Get the manual and report back!
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #3 on: July 27, 2018, 05:36:58 pm »
If you get any kind of trace on the CRT, even briefly, the HV is working fine.  If it's working, don't mess with it.  If you need to adjust it you will need a meter and HV probe that can safely handle 1500V, SO DON"T MESS WITH IT.  ;D

Brief is an understatement but it's definitely bright for the flash it gives, I'll heed your warning and not mess with it!  :P

The first order of business is to check all of the low voltage power supplies to make sure they're working properly.  If you haven't already, get the service manual.  Free copies are available on the web.  If the power supplies are not right, nothing works!  The manual will show you where the test points are and what values you should see.  Do not do anything else until you've done this.  Do not pass GO, do not collect $200.

I did download three manuals, including a user manual and a technical manual from the internet. I had skimmed them and they're pretty intimidating at a first glance. After reading your message I opened them back up and found the power supply testing sections and the textual sections are actually pretty user-friendly if you can get around the jargon! I'll definitely be following those instructions and returning with my findings once I get back to the scope (work sucks!)

Quote
453s are great scopes. Very much so worth repairing. Like PaulAm said you need the manual which will provide all the test points and wave forms to check. Also take a look at c937 while you are there anyway. Get the manual and report back!

I did check C937's input side and got the 22v that I was getting at the fuse, which I suspect is slightly high but I don't think it's out of spec. I guess you want me to check it with an ohmmeter? From what I understand I'm supposed to see 0 impedance and then it should slowly creep to infinite impedance over time?
 

Offline particleman

  • Regular Contributor
  • *
  • Posts: 115
Re: Tektronix 453 Oscilloscope no display
« Reply #4 on: July 27, 2018, 06:31:25 pm »
I did check C937's input side and got the 22v that I was getting at the fuse, which I suspect is slightly high but I don't think it's out of spec. I guess you want me to check it with an ohmmeter? From what I understand I'm supposed to see 0 impedance and then it should slowly creep to infinite impedance over time?
[/quote]


I would at least check to make sure C937 is not shorted. Ideally I would replace it if I had a good replacment high ripple low esr cap. That cap takes a pretty good beating.
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #5 on: July 28, 2018, 02:29:32 am »
So after some digging it turns out I have a capacitance meter. I know the capacitor should be removed from the circuit but I tested it in-circuit anyway and got a reading of about 65µF, occasionally climbing to 67µF then dropping to 55µF and climbing back to ~65. I can't read the rating on the capacitor that's in there because of the way it's facing but the circuit diagram on Figure F0-26 shows it should be 47µF. Is it completely necessary to remove the capacitor to check this or can I assume it's bad from here?

I also went ahead and tested the low voltage test points from Figure 6-4 and the results were really not good. Looks like you guys were spot on about bad voltages on the low side. I'm getting -3.8v at the -12v test point, 7.8v at the +12v test point, and 62.9v at the +75v test point. I'm still reading the manual, but where do you guys think I should head next? I guess I have to remove the capacitor and test it anyway, don't I?

Edit: Since I had it opened up, I decided to get some readings for the malfunctioning fan, since it should only be about 115VAC I figured I would be safe. I read in the manual that the windings are split and that the fan is only connected to one half of the primaries so that it always has 115v no matter if your source voltage is 230 or not. I'm in the US so I'm running 120vac and I have the range switch set to high, though low makes no difference as 120 is within both ranges. So I went ahead and pulled the fan and prodded around a bit as you can see in the figure below. There are 5 wires attached to the fan. Two pairs of wires go to what looks to be the posts to half of the transformer's (T1101) primaries, which means I should be getting ~115vac to them. I'm not sure what the orange(? I'm colorblind) is as I couldn't trace it back far enough to see where it terminated but I'm assuming it's the ground. So I tested the two posts (circled in blue and yellow) using the chassis as a ground. The nearest post (yellow) I was getting 122.9vac out of, which seems healthy if a bit high. The other post, however, was only 28.5vac!
« Last Edit: July 28, 2018, 03:26:26 am by UltimateNoob »
 

Offline ferdieCX

  • Regular Contributor
  • *
  • Posts: 152
  • Country: uy
Re: Tektronix 453 Oscilloscope no display
« Reply #6 on: July 28, 2018, 09:07:29 am »
The 453 came with different types of fans, you will find a discussion about 453's noisy fans here :

http://www.antiqueradios.com/forums/viewtopic.php?f=8&t=188277&start=80

My own 453 (FET type) uses a motor that looks like the one in your picture. It was very noisy when I got it, so I changed the ball bearings and now it sounds quite normal.
Strangely, when I repaired computers with another 453 in the early 80s, the fan was hardly noticeable. Of course, this was in a normal working environment and not in a private home.  :)

To repair the low voltage PS is your first task.
Anything connected to the capacitor under measure will badly alter your readings.
 

Offline JohnPen

  • Regular Contributor
  • *
  • Posts: 228
  • Country: gb
Re: Tektronix 453 Oscilloscope no display
« Reply #7 on: July 28, 2018, 09:14:24 am »
The -12v rail is the reference for the other voltage rails so it is the most important one to correct initially.  Check the main smoothing cap 1500 ufd c1112 by paralleling a known good cap of similar value and see if the -12v improves.  Also check the rectifier diodes D1112 are OK.  Try adjusting R1122 10 degrees each way and seeing if that has an impact on the -12v rail.  If this doesn't work then it gets more complicated as the load on the -12v might be pulling the rail down.  This can be due to a failing decoupling tantalum cap on one or more modules.  This is hearsay for me as my 453 only had main capacitors and a diode giving problems.  The manual as you will have seen is very comprehensive so working your way through from the power supply area should give you a start. 
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #8 on: July 28, 2018, 09:57:07 pm »
So I un-soldered C937 and tested it on my capacitance meter and got 65µF still. I tested a few other capacitors I had laying around and they were all high so I'm starting to wonder about my meter... I'll probably replace the cap anyway since it's out. As far as C1112, I assume you're talking about the one of the big guys in my attached image? The manual doesn't appear to be real clear on where exactly C1112 and D1112 are, but I'm guessing they're there on the board that those big caps are attached to. What's the best way of getting to those? They're pretty smooshed in there and it seems like I can either go from above or below but I'd rather not remove more than absolutely necessary. Thanks guys!
 

Offline particleman

  • Regular Contributor
  • *
  • Posts: 115
Re: Tektronix 453 Oscilloscope no display
« Reply #9 on: July 28, 2018, 10:29:16 pm »
So I un-soldered C937 and tested it on my capacitance meter and got 65µF still. I tested a few other capacitors I had laying around and they were all high so I'm starting to wonder about my meter... I'll probably replace the cap anyway since it's out. As far as C1112, I assume you're talking about the one of the big guys in my attached image? The manual doesn't appear to be real clear on where exactly C1112 and D1112 are, but I'm guessing they're there on the board that those big caps are attached to. What's the best way of getting to those? They're pretty smooshed in there and it seems like I can either go from above or below but I'd rather not remove more than absolutely necessary. Thanks guys!

I have two 453s and one parts unit with a bad CRT  none of them have had bad filter caps. They have had bad transistors,bad wet tantalums. I would find a way to measure ripple before I removed them.
« Last Edit: July 29, 2018, 02:38:35 am by particleman »
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #10 on: July 29, 2018, 12:58:10 am »
So you're saying I need a scope to repair my scope?  :-DD

To be fair, I happen to have a 1500µF/30VDC cap laying around that tests okay (again high on my tester but just within 10%) so if I can get to and identify C1112 it wouldn't be very hard to parallel this cap in with it. Again though, not sure on the best way to get to it.
« Last Edit: July 29, 2018, 01:20:03 am by UltimateNoob »
 

Offline particleman

  • Regular Contributor
  • *
  • Posts: 115
Re: Tektronix 453 Oscilloscope no display
« Reply #11 on: July 29, 2018, 02:38:15 am »
So you're saying I need a scope to repair my scope?  :-DD

To be fair, I happen to have a 1500µF/30VDC cap laying around that tests okay (again high on my tester but just within 10%) so if I can get to and identify C1112 it wouldn't be very hard to parallel this cap in with it. Again though, not sure on the best way to get to it.

A scope would be awesome but you could use a voltmeter set to ac voltage?
 

Offline ferdieCX

  • Regular Contributor
  • *
  • Posts: 152
  • Country: uy
Re: Tektronix 453 Oscilloscope no display
« Reply #12 on: July 29, 2018, 07:58:40 am »
If you measure the DC voltage across C1112, you should measure something between 18 V and 21 V .
As already said, you can measure the ripple setting your DMM to AC V. Even if you meter is not True RMS, you will get an idea about it.
If your DMM can measure frequency, it is a good idea also to measure the frequency of the ripple across C1112, it should be 120 Hz.
If you get 60 Hz, one of the rectifiers D1112 is not working.
« Last Edit: July 29, 2018, 11:09:01 am by ferdieCX »
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #13 on: July 30, 2018, 12:28:04 am »
Holy crap Tektronix really had to pack this thing full of stuff in order to get it to fit under an airline seat didn't they :wtf:  These diodes were not easy to get to and the capacitor is even worse!! As you can see in my attached image, I've unsoldered the bottom four rectifiers, tested them, and did an awful job resoldering them back in but they were all good (allowed flow in one direction but not the other). The capacitor C1112 is underneath the far ceramic bridges. This is going to be a total bitch to get to while the thing is running so I'm going to try to get some clips on it and mostly reassemble (because I've got one board almost completely detached to get here) I just have to make sure there's not chance of shorting from my clips to anything nearby which is a challenge considering how packed this is... I'm really hoping at this point that the cap isn't bad because changing that thing out looks like it would be the worst...
 

Offline PaulAm

  • Frequent Contributor
  • **
  • Posts: 684
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #14 on: July 30, 2018, 12:56:35 am »
Did you measure the ripple with your DMM set to an AC range?  I forget the spec, but you should only see a few mV if the cap is good.  You're likely to see a couple volts from the symptom description.

It is tough getting in there.  Pulling boards on that scope is a bit tedious.
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #15 on: July 31, 2018, 04:10:51 am »
Update:

So I checked the voltage across C1112 and I got 21.44V with a ripple of 330mV. Seems the cap is good according to that. I took a closer look at the circuit diagram with my father and it seems if I were to check voltage on the individual posts on C1112 I should get about 5.95V and about -12v ± the voltage drop across R1129 is, which at 0.3 shouldn't be very much? I'm getting +17.4v on one post and -3.9 on the other which seems very off. I'm starting to wonder both if I checked the correct rectifiers and also if my problem lies before the cap. What I'm about to do is locate C1111 and test the AC voltage across it which I believe I should see 16v (my DMM is an IBM 8060A/AA which is listed as a true RMS meter). If I see 16vac at C1111 then I must have checked the wrong rectifiers or something and I'll have to revisit them, if I'm not getting 16vac well... My issue lies further back I suppose. Does this make sense? Am I following the circuit diagram correctly, gurus?   :-DMM
« Last Edit: July 31, 2018, 05:40:46 am by UltimateNoob »
 

Offline ferdieCX

  • Regular Contributor
  • *
  • Posts: 152
  • Country: uy
Re: Tektronix 453 Oscilloscope no display
« Reply #16 on: July 31, 2018, 07:39:11 am »
The unregulated part (C1112 and D1112) is working properly. I suspect that the regulator itself is not working. Another possibility is that the -12 V rail is consuming too much current, but 330 mV ripple seems too low for that.
Check the voltage across R1129. If it is about 0.6 V or higher, the current limiting transistor Q1129 is acting. In this case, you have a short in the -12 V rail in one of the boards, probably a capacitor.
If the current limiting is not acting, check with the ohmmeter Q1114, Q1124, Q1133 and Q1137. Also, if they are socketed, there could just be a bad contact.
« Last Edit: July 31, 2018, 07:50:45 am by ferdieCX »
 

Offline UltimateNoob

  • Contributor
  • Posts: 8
  • Country: us
Re: Tektronix 453 Oscilloscope no display
« Reply #17 on: July 31, 2018, 05:48:46 pm »
The unregulated part (C1112 and D1112) is working properly. I suspect that the regulator itself is not working. Another possibility is that the -12 V rail is consuming too much current, but 330 mV ripple seems too low for that.
Check the voltage across R1129. If it is about 0.6 V or higher, the current limiting transistor Q1129 is acting. In this case, you have a short in the -12 V rail in one of the boards, probably a capacitor.
If the current limiting is not acting, check with the ohmmeter Q1114, Q1124, Q1133 and Q1137. Also, if they are socketed, there could just be a bad contact.

I'm oddly relieved that the mystery continues... changing that capacitor out would have been an absolute nightmare. What's nice is everything after the cap (with the exception of R1112) is extremely easy to access and I should be able to fully reassemble the scope to test all of those! I'm at work right now and have plans later but late tonight if I get some time or definitely by tomorrow I'll test all of those components and reseat the socketed transistors just in case. Thanks for the help, I'll update once done!
 

Offline PA4TIM

  • Super Contributor
  • ***
  • Posts: 1122
  • Country: nl
  • instruments are like rabbits, they multiply fast
    • PA4TIMs shelter for orphan measurement stuff
Re: Tektronix 453 Oscilloscope no display
« Reply #18 on: July 31, 2018, 07:24:38 pm »
I also have a very early one, mine is dated 1966, the scope is introduced in 1967. But for me it is more a collectors item, I have more than one scope   ;) I sometimes use it for fun

Work systematic and think logical, just like programming.
For instance, your fan problem. The thing gets hot. It can be coincidence but there is also a chance that it is somehow related. In that case, just test without the fan. Just an example. 
Often just one thing goes bad so do not start swapping random things.

Caps that go bad still can have a good capacitance. Always measure them outsite the circuit. A good LCR meter ( measures C and DF) is usable. C measurement of a DMM or a "China special" is only to trust on new caps. But the best test is if they function well, this is not a switcher so they are not very critical. Just measure the ripple. Best with a scope but a DMM is better as nothing.
Caps are trouble makers in cheap consumer stuff, not in test equipment. However they can go bad. Tantalums can be a bigger risc if they are used "on the edge" but the nice thing is, most times they short so you can find them easy or they let you now they are dead by doing boooooom  :P

Always First: optical test, reseat all transistors, this can be a huge source of problems. Clean switches, especial pushbuttons. I "Repaired" a lot of Tek gear just by cleaning the swithes and transistors sockets. Use a cleaner/lubricator like Kontact K61, there are much more brands but I use this a long time. Be careful with aggressive contact cleaners. They often need cleaning after application or you get big problems over time

Second: measure the powersupply. ALWAYS. The number one source of problems and it influences everything. Check All voltages, under conditions and on the location the manual states. If ripple is more important as "normal" they state it.
Do one thing at a time.

Read the theory of operation, they are really good.

Some tips:
You can measure current by measuring the voltage over a resistor (fi to check is some current flows) Current is the stuff things work on. If a transistor conducts there (most times) is a current flow.
Also use things like the 0,6-0.8v voltdrop (silicium) over the base-emitter junction or over a diode in a DC circuit, etc. This tells you if a transistor is short or open in situ.
If you have a thermocouple or IR temp meter you can check if things get to hot (but only if you have an idea of how hot is to hot.

There are very high voltages in a scope so be careful. You need a HV probe to measuring them (safety and loading) Otherwise, do not mess with it.

Be not afraid of "to much work" or to difficult. Often it looks much harder as it is to get things apart. Make lots of pictures and notes. Look good before you do things, do not use force. Often you waist more time in thinking how to do it an other way as it takes time to take it apart right away. I had to take a spectrum analyser complete apart (including removing the CRT) to replace a dead rifa cap and although it was a lot of work, it was nearly as hard as it looked. Sometimes you need to desolder lots of wires but in the end it is only a few seconds a wire and if you make pictures you solder them back as easy.

Do not start turning trimmers etc. Mark their position. Sometimes you need to turn them to see if there is a response so you can set them back but only if you are sure you need to do that. Aligning a scope is a lot of work, really a lot, and you need a lot of gear and it can be complicated. Often a lot of things depend on each other.

Important: almost all service manuals do have faults, besides that, often there are several versions of a scope, probably there are in between versions. And you do not know the history. It even can be a mix from several scopes, people can have made modifications etc. But if you are not sure and do not have enough knowledge, believe the manual. 99% of the time it is right, for that 1% the forum probably can help you.

Use your rubber duck  8)


« Last Edit: July 31, 2018, 07:29:09 pm by PA4TIM »
www.pa4tim.nl my collection measurement gear and experiments Also lots of info about network analyse
www.schneiderelectronicsrepair.nl  repair of test and calibration equipment
https://www.youtube.com/user/pa4tim my youtube channel
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf