开云体育

Hello everyone,

A couple of months ago, I purchase a Tektronix 317 scope, serial 3507, and an assortment of 10x probes from a NH Craigslist ad. This is my first scope. I intend to use it for helping diagnose my vintage tube audio equipment. This old war horse has seen its share of battles- the front bezel is all nicked up, the carrying handle was replaced with heavy-duty rope, and there isn't a single straight panel left! However, the inside is in excellent shape and was all original except for some replacement tubes. The scope has the manual too. It was last calibrated by Tek in Boston on 10/27/1977 according to a sticker on the front.

Before I fired it up for the first time, I replaced the half-dead 6080 tube and replaced the following paper caps with film caps of the same capacitance and same or greater voltage: C806 and C815 in the HV circuit, C171 in the preamp section, and C603, C614, C638, C668 in the LV power supply. All the other tubes checked out OK on my tester. While warming up, the trace comes in nice and bright. The horizontal and vertical positioning are just about dead-on.

However, as the 317 warms up to full temp, the trace drifts off the screen. After fiddling with some knobs, I was able get it back by cranking the horizontal knob fully clockwise and turning the intensity up a bit. Vertical is more or less the same. The trace shows up only as a dash at .1 and greater V/div, but appears as a loop at .05V and lower. No amount of adjustment of the time base knobs brings back the nice, straight line. I can get the trace to fire off a sweep sometimes by turning the red Trigger Selector, V/Div, and Time/Div knobs a certain way, and the speed of the "sweep" roughly matches the Time/Div setting, but it only shoots once. The traces wiggle when I touch the probe, so it is picking up, it's just not displaying correctly. In the horizontal amp circuit there were three 6BQ7A/6BZ7s, which are notorious for cathode-to-plate shorts. They all tested good with no shorts on the tester, but to be sure it wasn't those I replaced them with three good 6DJ8s I had on hand- no dice, same behavior.

If this is something that can be fixed with simple calibration, let me know, but I see nothing that relates to this odd behavior in the manual. I see no arcing or sparking with the covers off. Any other tips I could follow to get the trace back to normal?

No specific experience with a 317, but did you replace electrolytics as
well? Its old enough to give it consideration.

Id also check power supplies (voltage and ripple) while its “bad”

Bad caps would be my first thought for something of that age

But beyond that I cant make much more in the way of useful suggestions

David

On Thu, Mar 15, 2018 at 12:02 PM CZ4A via Groups.Io <artofruin1285=
[email protected]> wrote:

Hello everyone,

A couple of months ago, I purchase a Tektronix 317 scope, serial 3507, and
an assortment of 10x probes from a NH Craigslist ad. This is my first
scope. I intend to use it for helping diagnose my vintage tube audio
equipment. This old war horse has seen its share of battles- the front
bezel is all nicked up, the carrying handle was replaced with heavy-duty
rope, and there isn't a single straight panel left! However, the inside is
in excellent shape and was all original except for some replacement tubes.
The scope has the manual too. It was last calibrated by Tek in Boston on
10/27/1977 according to a sticker on the front.

Before I fired it up for the first time, I replaced the half-dead 6080
tube and replaced the following paper caps with film caps of the same
capacitance and same or greater voltage: C806 and C815 in the HV circuit,
C171 in the preamp section, and C603, C614, C638, C668 in the LV power
supply. All the other tubes checked out OK on my tester. While warming up,
the trace comes in nice and bright. The horizontal and vertical positioning
are just about dead-on.

However, as the 317 warms up to full temp, the trace drifts off the
screen. After fiddling with some knobs, I was able get it back by cranking
the horizontal knob fully clockwise and turning the intensity up a bit.
Vertical is more or less the same. The trace shows up only as a dash at .1
and greater V/div, but appears as a loop at .05V and lower. No amount of
adjustment of the time base knobs brings back the nice, straight line. I
can get the trace to fire off a sweep sometimes by turning the red Trigger
Selector, V/Div, and Time/Div knobs a certain way, and the speed of the
"sweep" roughly matches the Time/Div setting, but it only shoots once. The
traces wiggle when I touch the probe, so it is picking up, it's just not
displaying correctly. In the horizontal amp circuit there were three
6BQ7A/6BZ7s, which are notorious for cathode-to-plate shorts. They all
tested good with no shorts on the tester, but to be sure it wasn't those I
replaced them with three good 6DJ8s I had on hand- no dice, same behavior.

If this is something that can be fixed with simple calibration, let me
know, but I see nothing that relates to this odd behavior in the manual. I
see no arcing or sparking with the covers off. Any other tips I could
follow to get the trace back to normal?



--

Sent from Mobile Annoyance thingy, please pardon any typos

Bleh, I was afraid you'd say that! Two of the can caps look hard-to-access. They all look to be original, and even if one or more were replaced during the last calibration in '77 you are right that they are old enough for replacement to be considered.

For the smaller-value caps I can likely get away with mounting pencil caps under the chassis and keep the cans in place for appearance's sake.

Hello anonymous ! :-) I guess "CZ4A" is not your real name ;-)

Well I am hardly an expert with tube stuff, as I only have one of these beasts.... yes, a Tek 317 just like yours.
Got it last summer, fixed it a few months ago, learned some stuff along the way... thanks to the people on here obviously.

My first thoughts :

- Don't lose sleep on the big electrolytics... it so happens that they were top quality and more often than not, they are still good today ! Mine certainly are, and my unit is even older than yours (SN# 2369). All power rails were spot on, and ripple on all f them was negligible.

- You say that the trace brightness is good at power up, but dims a bit once fully warmed up (an indication of how much time that actually means, might be useful : 2 minutes, 30minutes ?). I guess that can only mean that there is still some issue in the CRT/ HV section that needs sorting out. You already replaced the two paper caps in the CRT section (I did too, that fixed the dimming trace issue on my scope, would dissapear after only 60 seconds), but I gather that another common problem on these scope is the HV transformer which becomes lossy once, precisely... warmed up). So maybe you need a new tranny. The issue is so common, that there is someone on here who used to wire them by the dozen to help others fix their scope...

- As I understand you, you already have narrowed it down to the horizontal section of the scope... so well, just keep trouble-shooting it further, thanks to the comprehensive service manual, and your existing experience working on tube stuff... shouldn't take you too long I would think ! :-)


I understand this is your first scope... however it's gonna be difficult checking ripple and waveforms at various points, without... a second scope. Just pick the cheapest/crapest basic old analog scope you can find locally. Even if just single channel, even if just 1MHz B/W... that would still be night and day to assist you in trouble shooting your 317. Obviously you will need to use x10 probes, given that voltages can go up to, IIRC, 400+ Volts (for the CRT circuit). A HV probe would be nice too, to check what voltage you actually get on the CRT anode... I picked a decent 40kV one for 75 Euros shipped. You can find cheaper since here you only need a 10kV probe.

But at first, you can start by checking all DC voltage levels indicated on the schematics (in the area of interest), using a simple digital multimeter...
Of course start by checking all the voltage rails. You can use your DMM in AC mode to get a first taste for ripple levels, short of a scope at the moment.

Keep us posted on your progress :-)


Regards,

Vincent Trouilliez

Forgot : the caps you actually need to worry about are not the electrolytics (at least not at first), but rather all the paper caps, the black and red "Sprague" ones. You already have replaced the two from the CRT/HV section... maybe there is one causing trouble in the horizontal section...

CZ4A is most likely an amateur radio call sign (Canadian).

Thanks,
Barry - N4BUQ

Thanks for the info.. I am no radio amateur so it didn't make any sense to me I must admit...

I guess I am old fashioned but I prefer to call human beings via their real name, feels more natural and fitting to me ?!
It's also easier to remember a real name than a call sign, at least for the average human being that I am.

Anyway, looks like he is not far from a diagnosis on his Tek 317, though would be nice to have a few piccies showing the problem, and/or uploading a little video on Youtube maybe ?


Vincent Trouilliez - VT7760... err no, just made that up, just Vince, that will do fine :-)

Hi Vincent,

I've read your thread on eevblog, which was helpful. The reason I focused on the horizontal amp is that unlike the situation with your 317, the traces are still there after warm-up, just in the wrong position and malformed. The caps in the horizontal amp section are all dogbone-shaped ceramic or mica. It's possible they've gone bad, but IIRC ceramic and mica caps last longer than paper or old electrolytic.

All the paper caps in this serial # are Good-All rather than Sprague. I've replaced the paper caps on the chassis, but still need to get to the paper caps on the rotary switches.

I hope the flyback transformer isn't damaged! Besides the two paper caps I replaced in the HV circuit there is an electrolytic (C807, 8uf, 450V). I've never seen a single-lead cap like it before. It's held to the chassis with a clamp. Looks to be body-negative. Another one is C154 (500uf, 6V) in the preamp. Are these still available outside of fleaBay, or will it suffice if I get modern two-lead caps and route the negative leads to the chassis?

I have an Eico 950B cap bridge I can test the caps with. I do need to get pictures and measurements from the LV rails, but right now I don't have the scope in front of me. I'll be sure to post back here with my findings.


Hi Barry,

It's actually the chassis code for my favorite car, the Mitsubishi Lancer Evolution 10.

Regards,
John - CZ4A

Hi John,

Oh, so you found my topic on EEVblog... well, not much I can add then...

I think however you should not waste time and money replacing all these tube and caps (other than the obvious paper caps of course), unless proven faulty.

The scope has to be diagnosed first, then once the fault is pinpointed, you just replaced what actually is faulty.

Hell, even the paper caps weren't that bad on my unit. I replaced them for good measure (though like you, I still haven't found the motivation to get to the ones hidden in the rotary switch assemblies... a real pain in the butt aren't they !), but really it didn't make any difference. Only the two in the HV / CRT section were bad and causing my trace going dim and disappearing off scale/drifting after a minute. But all the other paper caps still aren't bad to the point of creating any noticeable symptom...

You have one good thing going for you : your fault is not intermittent ! You can reproduce it at will and therefore do all the necessary measurements.. so eventually you will narrow it down.

Again, even if you don't have second scope to check waveform, I would think that once the fault has manifested itself, and is "stable", you should be able to find something wrong somewhere, just by simply checking DC levels with a DMM, using the schematics. This is easy and cheap to do... as you probably have a DMM I would think (?), and if not, given that you only need very basic DMM for this particular task, then even a good brand/quality DMM, bought used, should be cheap. I don't know, an old Fluke 75 or something, whatever. Oh, just checked your Eico 950B... so you really are into tube stuff... if you are "against" modern digital multimeter, then you can still get a high-impedance voltmeter, I saw that in the tube era there existed high-impedance meters, that used a tube on their input stage, as buffer/high impedance votlage follower.. so you could use that. But then you would have to find one, and spend time and money restoring it... would be easier, short term, to help you in diagnosing that 317, to just get a cheap DMM for the time being...

As from the single-leaded clamped electrolytic, yes the can is negative,just like the big cans in the LV power supply, and just like... I think any metal can cap ?!... Even on my 1980 vintage counter (the Enertec 2618, also on EEVBlog), hence much more modern than the 317, the filter caps in the powers supply section were all of the can type, with the negative connected to the can, and a single lead in the center.
Sure enough you can replace those with a modern/two lead cap, which I did. It's just a cap... no magic there.
As for buying a brand new canned cap with a center lead... I don't know if you can get them, other than some specialized sites selling recap kits for particular devices that are popular. so most people who want to retain the looks of the old can, just reuse/restuff it. I am not aware of a cap manufacturer that mass produces this type cap... though would be great of course, would male restoration easier...

CZ4A, ROTFL ! Every time you see such a signature it's a call sign, for once it's not, loved it ! :-)

Vince - L48Y08 then

Btu again, I would not waste and effort doign anything to them

Once you have the DMM, check all voltage

Oops ! The last two lines of my previous message, after the signature.. what are they doing here... just ignore them.

Unfortunately unlike forum posts, you can't edit/correct e-mails... once it's sent, it's sent...will proof read better next time, sorry :-/


Vince

Hi John,

That's what I get for guessing... :)

Thanks,
Barry - N4BUQ <- Amateur Radio Callsign

I've tested the -150V rail with my DMM and it's reading at -204V. o_0 That's a problem. It's throwing off the +100V and +300V rails as well. Turning the R617 (-150V Adjust) pot doesn't seem to do much. Focusing on the -150V part of the LV supply, I've replaced both paper caps with film already and replaced the half-bad 6080 tube with a new one. The 6AN8 tube in that section tests good. That leaves C610 (2-section metal can cap, 160uf and 10uf 450V) and the 5651 tube. The 5651 glows like normal. So I believe my next move will be to remove the can caps and test them with my Eico 950B.

Regarding the Eico, it's not like I'm opposed to using modern instruments. It's more that a) I'm cheap and the Eico and 317 were cheap, and b) the 950B is useful for vintage tube electronics. Some DMMs have a cap testing function, but they don't supply enough voltage to effectively test HV caps like one would find in a vintage hi-fi amp. The 950B isn't lab grade but it's pretty accurate and just as important goes up to 500VDC for leakage testing. That also makes it useful for reforming electrolytics that have been sitting around for a while.

I took some pictures of the strange traces. I'll get them uploaded once I figure out how.

-John

Here are some photos I took of the trace behavior.

/g/TekScopes/photo/38177/0?p=Name,,,20,1,0,0

While warming up.

/g/TekScopes/photo/38177/2?p=Name,,,20,1,0,0

Warmed up, ≥0.1 V/div

/g/TekScopes/photo/38177/1?p=Name,,,20,1,0,0

Warmed up, ≤0.05 V/div

Thanks for the pictures, it looks funny indeed.

OK, so I guess there is a reason why we keep saying to check voltage rails before anything else ! ;-)
Such a massive deviation over the nominal value will obviously cause weird and wonderfull things all over the scope, so there is no need to spend even one second trying to make sense/trouble shoot your trace problem... the only thing you must concentrate on, is to fix the LV power supply, hoping that it being faulty for God knows how long, did not cause permanent damage to some sections of the scope :-/

Of course the good news is that the LV power supply is very simple in design, and physically easy to access. So trouble shooting it should be quick and probably won't require a scope. Your DMM should probably suffice.

So just work your way from left to right, in a logical fashion... using your DMM while trying to make your measurements as quickly as possible to avoid exposing the scope to such out of spec voltage rails for an unnecessary length of time :-/


1) Check AC voltage at all the secondary windings. I gather this particular transformer is as reliable as the earth going round the sun if not abused, but assuming things is just the only thing not to do when you are trying to trouble shoot anything...Murphy is always there watching every one of us ;-)

2) If the all the secondary windings read OK, then let's now concentrate only on the -150V reference rail : with the scope turned off, check all 4 diodes of the full wave rectifier. Just unscrew the plastic transparent guard plate and check them with your DMM... well might be easier to just check all of the diodes in there, rather than try to identify which ones are actually used in the -150V rail.. only takes a few seconds after all.

3) If all test good, then turn the scope on again, and check the voltage (referenced to ground) at the right side of the 10 ohms resistor R601. As the schematic says, should be +72V

4) Then check that the all important voltage regulator tube, the heart of all this, does give you the expected -85V


This should only take a few minutes and will give you a strong starting point...

Do it and report back :-)


--
Vince

1) All transformer windings were in range. I should note input from mains is 121V and the windings are set for 117V.

2) All diodes tested good. I removed D152A and D152B since the cathodes are facing each other and throwing off my DMM's continuity tester, but those were OK once taken out.

3) This was about +70.

4) Didn't test. Should have, but didn't.

I removed C601 and C660 and tested them on the cap bridge. They weren't too hard to get to. Both sections of C601 are fine, but C660 appears to be dead. It's not registering any capacitance at all. I've got a line on a modern 160uf 450V cap that should work as a replacement.

OK great, a dead cap, hopefully that will fix it.

As soon as it's replaced, measure the output of the voltage regulator tube, it's the most important thing, this thing is the reference ;-)

Vince