Taking care of a Tektronix 465

[Martini]

A very generous someone gave me his old scope (which he replaced with a new Rigol) with a "non working channel 2".

It's true but you will also see that the channel 1 has a weird behaviour. I've had to replace the cord (the Swiss one for a French one), so it's possible I damaged something while removing the cover (it has some dents, it may have "unsocketed" a transistor or two while I was sliding it) although I took the most care I could.


I tried to read the service manual but that's far out of my league. Could you suggest me a series of tests to be ran to check what works and what doesn't ?
After that, I'll need your help for the diagnosis too



The video to give you an idea of the faults I've identified (Ch1, Ch2, timebase, volts/div, trigger):




(I'm currently reading the TEK's troubleshooting guide. Interesting but not all that easy to digest for me)

[Fank1]

While it may not be all of your problems, you really need to clean all of the switch contacts.

[rockets4kids]

A really great resource for the older Tek scopes is the TekScopes mailing list on Yahoo.  You will find lots of people there who know the common Tek scopes, and even a few who can answer questions about the real oddballs, too.

[tombi]

The weirdest problems are often multiple problems. Trick is where to start.

With the weird angular stuff at the end of the trace - it's odd this doesn't happen at higher sweep rates.

Could it be that the probe calibration output is the thing that is broken? Do you have another signal source you could try? Like a signal generator or even a 555 circuit. The bit where adjusting the trigger moved the trace makes sense as by lowering the trigger you are then triggering on the bit at the right of the screen. Also at around 16s in you could see short sections of square wave and then the signal fade off to the bottom.

The fact that you get no sweep at some sweep rates could be switch contacts in the time/div selector.

Tom

[Martini]

Thanks for the help guys!

I'll check the yahoo mailing list but I fear I won't find the interactivity there is on a forum.

Could it be that the probe calibration output is the thing that is broken? Do you have another signal source you could try? Like a signal generator or even a 555 circuit.

An Arduino's PWM:

The bit where adjusting the trigger moved the trace makes sense as by lowering the trigger you are then triggering on the bit at the right of the screen.

I'm not  familiar with the concept.


Sometimes, I loose the trace completely. As if it couldn't trigger. But if I switch it off and on again (I'm a Windows user), it reappears.

Also, in Auto mode, it should display a trace no matter what, shouldn't it?

[tautech]

A few hints to help you on your way. 
The PSU MUST be to spec, check this FIRST. Use a DMM on AC mV for a quick ripple check.
The Ch 2 problem, attenuator contacts, attenuation/divider resistors.
You may need a second scope to follow the signal path to find faults, but not always. Keep reading.

Waveform falling away at end of sweep.
Might be vertical output stage or sweep linearity, too early to tell for sure.

You'll notice DC voltages on the schematics, they can be very helpfull in the absence of more advanced test gear than your DMM.
For these voltages to have any meaning the scope usually must be set to specified settings. RTFM. 

Print the schematic pages and overlay/tape them together if needed.

Sort ONE problem at a time, step by step as they say.

[Martini]

The PSU MUST be to spec, check this FIRST.

It isn't on the +5V rail:


I can't check the high voltage since my multimeters don't go beyond 200V DC.

Use a DMM on AC mV for a quick ripple check.

One of my multimeters shows 0.7V on the +5V rail, 0 or nearly on the others.
The other one (multimeter) goes completely bonkers when I check the AC. Showing 230V AC on the 110V test point



Also, I sprayed the "trigger view" switches ModemHead spoke about with contact cleaner. They didn't look bad beforehand and this made no difference.




(My subscription to the yahoo group is pending)

[tautech]

Now you have some study and investigations to do. 
IMO the +5V rail is overloaded and is likely effecting the linearity of the sweep. That is reproduced in the fall-off of the waveform on the RH side of the screen.

This appears to be the only problem and you must find what is causing the +5V rail overloading.
Just one overloaded rail can effect all others so don't adjust anything until you find the cause.

A2AT in this Tek 468 tread: https://www.eevblog.com/forum/repair/tektronix-468-repairrefurb/
had many bad MLCC caps that was overloading the PSU in his scope.

PCB's can be disconnected for fault finding as outlined in the manual.
Read his thread and study the manual.

[Martini]

Now you have some study and investigations to do. 

I know. But I need some contextual guidance.

I have basic skills to use a scope and I can connect my mailbox to the internet but that's it, I have a very hard time understanding what I read in the manual or other topics.

you must find what is causing the +5V rail overloading.

Yep.
Well, how do I doo that?

A2AT in this Tek 468 tread: https://www.eevblog.com/forum/repair/tektronix-468-repairrefurb/
had many bad MLCC caps that was overloading the PSU in his scope.

That's a very good example of an out-of-my-league thread 

[tautech]

Now we likely know where to look, can you seek help from the person that kindly gave you the scope?

Also do some Googling on CRO operation principles, this will help you understand the Service manual and the schematics better.
Again I urge you to RTFM more, seek out how to isolate PCB's to find where the PSU +5V excess current draw is happening.

[Martini]

Now we likely know where to look, can you seek help from the person that kindly gave you the scope?

Unfortunately, he's hundreds ok kilometers away.
I don't know how good he would be at this either.

[PaulAm]

That 0.7V ac on the 5V rail probably indicates a bad filter cap.  They fail open all of the time.  If you had an ESR meter you could check those pretty quickly.  Hard to say what your other meter is doing, but other filter caps could be bad as well.

I would not recommend a scope as a first trouble shooting experience.  It's possible if you persevere, but you're going to have to put a lot of effort into it.  You're going to have to keep reading the service manual until it makes sense.  Look on this as an opportunity to gain an education 

[Martini]

Hard to say what your other meter is doing, but other filter caps could be bad as well.

I would not recommend a scope as a first trouble shooting experience.  It's possible if you persevere, but you're going to have to put a lot of effort into it.

I'm not really choosing 

[tautech]

Study the manual and find the waveform for the sweep generator. It is a triangle/ramp waveform.
If the +5v is used to produce the sweep, it is not attaining it's full amplitude ie. not a full triange.
This is responsible for the screen waveform falling away IMO.

Use of another scope will confirm ripple and ripple frequency and the corrupt sweep waveform.

All this happens because the sweep controls the voltages applied to the horizontal plates that electrostatically drive the electron beam horizontally across the display from left to right. If the sweep is not linear, you have a display like you have posted.

The real cause is the low +5V rail.
Find what's pulling it down and you will have a good scope.

[Grapsus]

Hi,

I think you need to isolate the 5V PSU (disconnect it from any load). There is a procedure in the service manual for doing this.

Once isolated, if the reading becomes correct it means the 5V rail was overloaded and the cause is somewhere else, if the reading is still too low it would suggest that the PSU itself has a problem.

I found the 465 to be very hard to repair as a beginner. Moreover the different boards are pretty hard to take apart and don't even dream about removing the main board unless you're ready to reassemble a 5000 piece Lego kit. There is a lot of interaction between different blocks. They're all made of discrete transistors, sometimes using weird configurations and frequency dependent feedbacks. In order to find faults you don't necessarily need to understand every circuit in detail, but a general understanding of the blocks and their interconnection in important.

What they don't tell you in the manual is that the voltages in blue and the reference waveforms are not obtained with the same settings ! I found that very disturbing. The voltages in blue are basically bias values, they're taken when the scope does the least amount of work possible, sweep not running, screen blanked etc. It's very important to put the scope in this state and measure those voltages first. These measurements will quickly identify obviously failed components. Then, the waveforms are drawn for the most complex configuration that mobilizes as many circuit as possible. For example all the Z-axis waveforms show the scope in "A intensified" mode. If you don't use the same settings, the waveforms can be very different and they don't tell you which settings they used, so the only way is to try different combinations are understand what it does to the signal. The waveforms will really help in more subtle cases.

Good luck,

[Martini]

Merci Grapsus. Pas simple de savoir qui fait quoi quand on est un débutant complet :-(



So, to do the things in the right order, I shall:

• Resocket every resockettable component with contact cleaner
• Clean the switches
• Read more about the general principle of the scope and which board does what
• Isolate the PSU and hope the fault lies within this board, which will make for an easy fix afterwards

I found a sorry looking puppy:


It's on the main board (bottom of the scope), on the same line as the X10 MAG button
Does someone know what it does?

[Addicted2AnalogTek]

I picked up a 465 a while back that "had a non-working channel 2" and multiple "issues" with channel 1.

I had to clean it multiple times over to get it to properly function, but ALL of the problems went away after doing so.  I cleaned every one of the gold finger-contacts and pads for the switches with Isopropyl Alcohol and a camel hair brush (can find at almost any hardware store that has painting supplies - or art stores).  Or you can use pieces of paper that are soaked in alcohol and close the contact over it, then gently pull it out, and repeat.  I also wiggled all of the socketed transistors and made sure the couple of "IC's" (transistor arrays) were fully seated.   I did have to take multiple boards off in order to properly clean everything.

I also had similar AC ripple on my 5V rail, and 15V rail.  Adjusting the 55V rail back to dead-on eliminated that. It was only 120mV high, too. 

You can also check the PSU rails to see if any of them are folded by using a DMM.  Connect the negative lead to the emitter of the pass transistor and the positive lead to the base. CAREFULLY (don't short anything or you'll be replacing them and the zeners! ... ahem. )   

But anyway, you seem to be on the right track and with good help, so I'm sure you'll get it sorted out.  I just thought I'd share my experience with you since I had similar issues. It also appears that your scope sat for quite some time as mine did. The moisture and disuse allowed oxidation to build.

Since I have a working 465 (SN is above B250000)  I can check voltages on it for you to compare against if you need.

[Addicted2AnalogTek]

I found a sorry looking puppy:


It's on the main board (bottom of the scope), on the same line as the X10 MAG button
Does someone know what it does?

This cap is in the Horizontal Amplifier circuit, and it is a decoupling cap off of the -8V rail. 
Part Number C1220  -  1000uf 20% 10VDC 109D tantalum.  You could probably get away with replacing it with any electrolytic between 800uf and 1200uf, 10V-25V for now (if you have something on hand) and then order the actual part. I'd try my best to get a wet tantalum replacement at the correct rating.
This could very well be part of your nasty looking horizontal sweep problem.

I have the paper manual if you need any info from it. 
You can download one from here:
http://www.ko4bb.com/manuals/index.php?dir=Tektronix/Tektronix_-_465_Oscilloscope

This is probably going to be the best one for download. It's got the largest file size.

[NilByMouth]

Use a DMM on AC mV for a quick ripple check.

One of my multimeters shows 0.7V on the +5V rail, 0 or nearly on the others.
The other one (multimeter) goes completely bonkers when I check the AC. Showing 230V AC on the 110V test point

My cheapo multimeter does the same thing. Did you try swapping the leads round (red to ground)? Mine also measures double the voltage of batteries too with it set to AC voltage range. I bought a new meter!

[Martini]

I'll try lead-swapping next time I open it up.


It's feeling a bit better now that the guys from the yahoo! group advised me to replace the 5V rail caps :

[Addicted2AnalogTek]

It looks like you still have a bit of noise on the horizontal of your trace.  That 1000uf 10V cap you noticed on the main board (horizontal amp circuit) appears to be the original 109D tantalum.  Hopefully that is still good, but in all 5 of the 400 series scopes I have purchased and repaired, that cap has been replaced with an ITT or UCC type SL (low ESR) electrolytic cap.  (good luck finding a 1000uf 10-20V wet tantalum cap for less than $40 -- or at all!!)   It could just be dirty timebase switch contacts, which a good blast of Deoxit Gold should fix.  When it comes to these old scopes with the gold plated finger switches and old pots, I've found both Deoxit Gold and D5 to be a cant-do-without item.

[Martini]

I also have a ripple on the 110V line I didn't have before replacing the two filtering caps.
Maybe I should replace the three power supply caps remaining and then adjust the 55V rail.


The project is on hold for now (I don't have enough space on my desk to work on the scope for the moment  ) but I'll keep all that in mind.

I didn't even know a company could make different contact-cleaning sprays...and that you would actually need several.

[PaulAm]

The Tek recommended method for cleaning those tiny gold contact fingers is to soak a small strip of copier paper in isopropyl alcohol and drag it gently through with the contact closed.

Personally, I would not use any type of contact cleaner on those contacts.  Maybe on the push button switches, if nothing else worked, maybe.

[Addicted2AnalogTek]

I was just suggesting a method that didn't involve taking the scope apart entirely to access the switches for the timebase.  I personally would rather take them apart and use IPA with paper or a camel hair brush, but I have also blasted them with deoxit gold and it works quite well.  I haven't had any adverse effects from doing so after a couple of months, either.

[Addicted2AnalogTek]

Also I think that it is worth mentioning:   Nearly all of the old wet tantalum caps seem to be plated in silver or some similar tarnishing metal.  That is why many of the 109D or other wet tantalum caps you will find in these older pieces of test equipment have black-looking end caps and sometimes the leads as well.  That is nothing to worry about.  You will also notice this on any component with silver plated leads, such as old style zener diodes.  I have even found 109D caps that appear to have 'corrosion' or some kind of crusty buildup at the negative terminal which ended up being nothing. It may have just been from moisture adding to the oxidation.   

Point being that you can't always go by the way a wet tantalum cap looks. Just like any other cap, your best bet is going to be some in-circuit or out-of-circuit ESR testing.  I just use a scope and function gen to test in circuit.   W2AEW's youtube page has a video with instructions to do it.  I find it extremely useful for preliminary cap checks/replacements before even powering on an old scope.