Tektronix 465B - help needed -> fixed!

[nomad86]

Hi,

First up, a disclaimer, I am an electronics beginner so please don't assume too much!

I have recently acquired a Tektronix 465B and HP 8011A pulse generator as a package deal. All seemed to be working well and I was playing around with the scope learning how it all worked. All the functions that a new, inexperienced in electronics, user would be expected to find seemed to be fine. Then I forgot to turn off the signal generator before I turned the scope back on while it was hooked up. What I observed is as follows:
1. The power light comes on
2. The lights behind the volts/div, indicating the selected setting, also came on for a short period before immediately going out
3. The screen then proceeded to light up slowly with a very diffuse glow on the right hand side of the screen

The fan -may- have also stopped although this could just be due to the auto start up sequence... although as this is analog I expect the fan is based of a thermistor of some sort? so as it had been running and was already warm it should have kept going?

Following this tragedy the last step is all that happens on a restart with no input. The focus does seem to have some effect (very minimal) and other lights come on when the corresponding buttons are pushed.

No magic smoke was seen/smelt escaping.

I have attached two images below showing the current state of the scope (you -may- be able to see that it is faintly glowing on the rhs of the screen) and the settings for the pulse generator when it all went south.





As I said before, I am a novice, although the reason I got hold of this scope is to learn electronics and this may be a good (or bad) way to dive into the deep end straight up. I am very willing to try fix it myself with a little advice from the gurus here.

Essentially I am asking very humbly if those of you with experience in these matters think that I have a chance of fixing this and if so how do I go about it?

[PaulAm]

Forgetting to turn off the signal generator has nothing to do with your scope problems.

The 465s have a few common failure points.  Two of the most common are filter caps that fail open and bridge rectifiers that fail.  The very first thing to do is to check the power supply voltages.  You can get a service manual free off the net and there are test points located on the interface board (the large board on the bottom).  Check the +55V, +110,  +15, +5 and -8, they should all be very close (within 0.1V).  If you can check it, ripple should be in the mv range.  If the +55 is off, nothing else will be correct since everything uses that as a reference.  If the voltages are out of spec, or zero, you get all manner of weird behavior.  Fix the PS before doing anything else.

If all the voltages are correct, there's likely a problem in the horizontal amp somewhere.  Unfortunately, it's helpful to have another scope when you need to fix a scope.  It's fairly straightforward to track back until you find the bad section.  If you don't have another scope, check the wires on the horizontal deflection plates; it's possible one might have come off.  Otherwise, you might get an idea what's going on using a DMM on the horizontal output.  You might have a bad transistor there somewhere.  I think they may have started soldering in the transistors on the 465b; on the 465 they were in sockets and it was trivial to pull one to test it (although the sockets are a big failure point as well).

[kfitch42]

What happens when you hit "Beam Find" ?

Do any other knobs (besides focus) make a difference in the 'glow'? In particular horizontal or vertical position. What about intensity?

Knowing which controls seem to effect the output can help in narrowing down where to look.

[Paul Price]

I am quite a fan of Tek scopes.

I think your fan should start from the get-go. Power supply failure?

You've already  probably got past these simple settings  below but.....

Have you set the Timer Base to something abut 1 mSec/div (Is the Horiz possibly set to Ext. Inp.??)
Have you set the Vertical controls and Horiz. Pos controls to middle of their range?
Have you set the Trigger to "Auto"

[nomad86]

Thank you all for your replies, they are greatly appreciated.

Before I take the case off and start poking around I will try to answer the operational questions.

I have the scope set to the following settings:
- all position knobs in the middle of their range
- both volts/div = 1V
- time/div = 1ms
- trig mode = auto
- trig source = norm
- ch1 & ch2 buttons depressed
- Horiz display A button depressed

(hope I haven't missed something important)

Q: What happens when you hit "Beam Find" ?
A: Nothing... (this was working before)

Q: Do any other knobs (besides focus) make a difference in the 'glow'? In particular horizontal or vertical position. What about intensity?
A: All position knobs have no visible effect. Intensity causes the glow to change in intensity very slightly but only in certain parts of the knob sweep and no change is visible between these little 'flickers'.

---

Something else that I forgot to add previously is that after the scope stopped working properly and I was fruitlessly poking buttons and turning dials the following occurred. When rotating the time/div knob a single very bright blob appeared at the far left of the screen, like the entire output of the electron beam was loosely focused just off the screen. With further rotation of the same knob this blob moved from left to right sometimes with the standard sweeping movement of a trace and other times it just stopped in one place. Meanwhile the indicator lights on the volts/div knobs occasionally lit up as I fiddled with time/div. This occured for about two or three back and forths of the time/div knob before it reverted to the 'diffuse glow on the right'. I have not been able to replicate this since.

---

PaulAm, thank you for your detailed instructions. I am also grateful to hear that the signal generator is not the cause, I can now tell the wife that I didn't kill it, it just 'happened'!

I have gotten hold of a service manual and will attempt to trace the fault. I will report back soon!

Many thanks once again.

[nomad86]

I have checked the regulated voltages and they have values as follows:
(checked at location marked with circle in image below, ground used is the one to the left of the circle)

+55 = +46.6
+110 = +101.4
+15 = 0.0
-8 = +0.6

So it would seem there is definitely something wrong with the PS.

Following the power supply isolation flow diagram in the manual I am now looking to check the un-regulated +55V line. Unfortunately I can't find anything in the manual that specifically states where this is on the board. I am assuming that it is the point marked in a square in the below image. I just want to make sure that this is the correct test point as this point is in an area marked "DANGEROUS VOLTAGES" and I like being alive.

Now, maybe getting head of myself, if this point is low then capacitor C4429 and C4439 need to be checked. The layout in the manual is a bit confusing so I thought I would ask here. In my picture I have labelled two double solder points with what I think is on the under side of the board, am I right with this?



Looking forward to more good advice.

[ivan747]

A good way to measure things around high voltage equipment is to clip the negative lead of your multimeter to a ground point on the scope and only use the positive lead. Keep a hand out of the place (maybe in your pocket or on your back) and keep the instrument disconnected.

[Anson]

Check these two caps. The plus signs I drew are the positive leads.

[nomad86]

Anson,

forgive the complete novice question here, but when you say "check these two caps" I interpret this as check the voltage from ground to the positive leads you have marked, or should I be checking the capacitance somehow?

Thanks for your input.

UPDATE: I went ahead and tested the voltage from ground to the centre point of all groups of 3 solder blobs and got the following voltages:

-8 = +15.4
+15 = +25.2
+55 = +80.3
+110 = +163.5
+5 = +10.5

hope this helps...

[Paul Price]

Check the caps means see if they open or shorted.  You can bridge a cap with a similar voltage rating and >= capacity. If you have a multimeter you can see the cap charge on the ohms scale if it is disconnected from the circuit.
If the circuit is power, you can use the AC scale on your MM and read the AC voltage. The voltage across an electrolytic cap should be close to zero, if it isn't you've probably found the problems.

[Rerouter]

-8 = +15.4

if you meant a negative there, that would imply the caps are atleast not shorted or open circuit, could still be cruddy from age, but points the finger towards the regulation stage,

[nomad86]

-8 = +15.4

if you meant a negative there, that would imply the caps are atleast not shorted or open circuit, could still be cruddy from age, but points the finger towards the regulation stage,

Nope, I get +15 as shown in pic. It should be some negative value right? Am I using incorrect measuring technique? That is why I posted a pic so you can tell me if I am doing it all wrong.



I have checked the voltages from the centre post to the outer post, with the positive (red) DMM lead on the centre post and get the following readings when using the AC scale:

-8 = +31.8
+55 = +176
+110 = +184.2
+15 = +54.3
+5 = +22.2

and with the DC scale (just for completeness)

-8 = +14.7
+55 = +79.6
+110 = +83.0
+15 = +24.9
+5 = +10.5

[nomad86]

Ok, so I have placed an order for a FX-888 solder station and UNI-T UT61E multimeter so I can de-solder and test some of those caps. In the meantime I thought I would test a few voltages on the +15V rail as it shows an approximately correct unregulated voltage (test = +25.2V, spec = +21.4V) but I get 0.0V on the regulated test point. The following pic is what I have measured with my $20 DMM.



Now, my knowledge of circuit theory comes from a (10yrs ago) 1st year general electronics uni. course and so does not quite cut it for interpreting what these values really mean. However, I am suspicious of U4411B and Q4401, as the resistors seem to check out and so do the removable transistors. I tested the transistors with the diode function on my DMM according to http://www.wikihow.com/Test-a-Transistor. The op-amp is a plug in unit and so should be testable once I have other supplies I am waiting on (breadboard, assorted resistors, caps, etc). The transistor, Q4401, on the other hand is soldered in so I have to wait for my soldering iron. Also, I am really puzzled by the change in voltage from R4403 to C4331 from 0.01V to 5mV, or is this probably just the accuracy of my DMM?

My question is, am I barking up the wrong tree? Also, should I be disconnecting all the other boards from the interface board according to the isolation procedure in the manual? I have gotten the impression from forum posts that this is not always necessary and I would like to leave as much alone as possible.

I am trying not to be a "please spoon feed me" type so if I am coming across that way please tell me!

EDIT: If this topic should be in the "Beginner" section then please feel free to move it!

[Anson]

What voltage do you have on the 8 pin of the opamp U4411? I suspect you have low voltage into the amp.

[SeanB]

C4431 or something on the output is dead shorted. Q4401 should be getting rather warm in this condition. Check the resistance across C4431 with the power off, anything other than a reading below 100R is OK.

[nomad86]

Thanks very much guys.

Anson:
The 8pin shows a voltage of +6.4V so is definitely low. Should be +22V if I am reading the schematic right. Looks like we are pulling a bit too much current through R4412 (off the +55V line) to get this voltage drop. I am assuming this is related to the problem with C4331 discussed below.

SeanB:
If you were meaning to say C4331 then I think you may have something. The resistance across C4331 is 0.7R which is exactly what I get if I simply cross the probes on my DMM. Q4401 also heats up quite quickly. I'm not not sure how hot it is supposed to get but I think it is too hot, good thing it is mounted directly to the chassis.

It looks like you have guided me into finding at least one fault, greatly appreciated!

[SeanB]

So now disconnect the output from that rail and measure again. If still a short C4431 is shorted, otherwise it is fine and the fault lies on something powered from the supply rail, probably another tantalum capacitor.

[nomad86]

Ok, disconnected everything specified in the isolation procedure (P4061, P4060, P4057, P4058, P4398, W7292, W4481, W4440) and I see the same values as before. So it is pretty much definite that C4331 is shorted. I have on hand a 47uf +16V tantalum cap from jaycar but the one on the board has written on it K47 +20V. Should this be a problem? It should only be seeing +15V so I am assuming the +20V spec on the existing cap is for a safety factor? They also look somewhat dissimilar with the one on the board being about twice the size as the jaycar one. Maybe I should just order another cap (maybe http://au.element14.com/kemet/t350k476k020at/capacitor-tant-47uf-20v-1-2-ohm/dp/1171967) and be patient...

[Rerouter]

higher voltages only, a 1V margin with a tant on a supply rail is not a nice mix,

[Anson]

Couldn't he use a mica or ceramic cap here? Haven't studied up on capacitor differences yet. I know some are polarized like electrolytic's and some are not like ceramics. Are tantalum's polarized? I haven't done anything with them yet I here they are horrible caps.

[SeanB]

Replace it with a 100uF 35V 105C low ESR capacitor instead. Do the same to the others in the power supply of the same value. Simpler than looking for a 47u 35V unit, and probably cheaper as well.

[nomad86]

So I should try to stay away from the tantalum caps?

I looked up 100uF 35V caps on element14 (hoping for quick postage in Aus) and found the following:

CAP, ALU ELECT, 100UF, 35V, CAN

(there were others similar to the above link but it was the smallest)

I am assuming that the polarization due to being an electrolytic (thanks Anson) shouldn't be an issue as long as I align it correctly on the board?

Otherwise I am back to the tantalum caps:

element14 tantalum caps

From what I have gathered from reading around is that the values of the capacitors in the power supplies of these old scopes are not too important i.e. as long as it is equal or higher then it should be good. So I should be looking for a relatively cheap option that is slightly over-specced compared to the original part?

Many thanks for this advice. I am on a real learning curve at the moment and this help is greatly appreciated.

[chulomex3]

Did you get this fixed?  let me know if you need anything as far as documentation.

[Anson]

So I should try to stay away from the tantalum caps?

Totally up to you. Some people say they are great and so do some books and papers on the net. Others say they are horrible and have a tendency to short and burst into flames. Never seen that  personally yet and so far in the last few weeks as I've been working on my own tek scope I haven't had a problem with them.

I am assuming that the polarization due to being an electrolytic (thanks Anson) shouldn't be an issue as long as I align it correctly on the board?

Most definitely.

From what I have gathered from reading around is that the values of the capacitors in the power supplies of these old scopes are not too important i.e. as long as it is equal or higher then it should be good. So I should be looking for a relatively cheap option that is slightly over-specced compared to the original part?

I think as close as you can get to the original is best. Be sure you only do this in the power supply part though. If you have issues on the vertical or horizontal sections you will want to try and stick to the originals and even stay with the same types. Different types of caps react differently to different frequency oscillations.

[nomad86]

Well, a very big thank you for all those who have given me input! I can report a working Tektronix 465B thanks to eevblog.com and its members.

I couldn't hold out any longer and decided to temporarily solder in my replacement of C4331 with my tantalum cap from Jaycar which is only rated for 16V. I have another on order that is rated for 35V but the 16V one was burning a hole in the desk just looking at me. Soldered it in and success! Immediately had the single trace that you get with no input, ch1 & ch2 and a few buttons didn't work but I was very happy regardless.

Then I realised I hadn't reattached the plugs on the interface board that I had disconnected as part of the power isolation routine. Plug those in and even better success, everything works!! Checked the regulated voltage test points and they are all within approx 0.2V with most within about 0.1V.

At first I thought there must be other things wrong as the trace was very 'wiggly' until I realised my fluoro desk lamp was polluting the signal something chronic.

Here is a pic proving that it works. If it looks a bit weird that is because the CRO is upside down on the desk and I have rotated the photo.



I am planning on replacing the 16V cap asap with the 35V one on order. Overall I am very surprised that I only needed to replace the one cap. I am half expecting something else to release the magic smoke any time now.

Once again, many many thanks to all those who have supplied advice and comments, it is very greatly appreciated.

Now for a more mundane, although related, question. One of the volts/div knob seems to give me a noisy trace when I have it on 0.2. I am assuming this is due to cruddy contacts in the selector switch mechanism. The manual says to only use isopropyl alcohol and that is about all they say. I am wondering if there are any experienced hands out there who can give me an idea of procedure for cleaning these switch assemblies?