Yet another Tektronix 468 repair

[JacquesBBB]

Maybe inspired by Addicted2AnalogTek
https://www.eevblog.com/forum/repair/tektronix-468-repairrefurb/,

I just got a Tektronix 468 from my favorite dumpster. Today I have  plug it and start to check
the main power supply.

The good news is that I have a  beam, and it is in much better shape than what I could have expected
from a thrown away device, but there are certainly some  fix to do in order to have it in proper condition.
It dates from 1988.

[JacquesBBB]

The reading of the PSU were surprisingly good

Nominal    Measured   After little tuning
(V)           (no tuning)

+55         54.9             55.00
+15         14.97           15.00
 +5           4.99             5.00
-8            -7.98            -8.00
+110     111.0            111.6

The measures were made with a Keithley 197.

Then I looked to the ripple with my brand new Rigol DS1054Z that I just got out of the box.
The ripple p-p is overall larger than  the nominal values from Table 4-11 of the service manual.

     
(V)          nom. ripple    measured ripple
               p-p

+55          4mV            27 mV
+15          2mV            20 mV
 +5           2mV           20mV
-8            2mV            23mV
+110      20mV           50mV

I put  the screenshot of the 110 V  ripple (8.jpg)  and of the 55V ripple (6.jpg).
All the others look like the 55V ripple.

Questions :

I assume that  these values are too large, and may lead to more fuzzy trace. I assume also that there are due to
one or several defective capacitors.
Are there some obvious culprits ?

Is this level of ripple really annoying ?

Thanks

[JacquesBBB]

Now the first problems.

One of the channel seems to be OK,  at least when testing with the internal square wave.
But for ch1,  sometimes its OK, and then becomes erratic, with
a shape that is no longer square.
This depends on the selection of the vertical scale.

Do you have any hint ?

[tautech]

Ripple.
Your screenshots seem wrong, frequency wrong.
For a sanity check, use your DMM in AC mV.
Initial voltage measurements indicate little wrong, just maybe some drift in value since last Cal.

I'd say you need more practice with a DSO. 

[android]

I was initially confused by the trace on the bottom pic until I realised that the scope was upside down.
Hmmm...even your avatar is upside down....I mean, thanks for trying to make it easy for us Australians, but it's really not necessary! 

[PaulAm]

Your attenuator switch may be dirty.

Inside the attenuator are little gold forks that get moved by a cam on the range switch.  Soak a small piece of copier paper in isopropyl alcohol, slide it under the open contact, then close the contact and carefully draw the paper strip out.  Do that a couple of times for each contact.

You may also have a bad attenuator module.  You can swap them from channel to channel if necessary.

There might also be a problem with the channel preamp or the channel switch.

But I'd start with the easy one.  Jumper the output of the two attenuators together, put the signal into ch2 and display ch1 and see if that cleans it up.  The wire you use for that will probably introduce some noise, so make some allowances for that.

[mskobier]

JacquessBBB,
   I don't know anything about that series of scope, but I have got to find out where you go dumpster diving! Really nice find, and it does not appear in too bad of shape. Maybe someday I'll be that lucky. Oh wait, I was... found a 2445B in the scrap metal yard earlier this year. Got it all tinkered up and working fine. Would love to have one of that series to play with. Maybe I'll find something fun in a couple of weeks at the annual SEAPAC ham radio swap meet in Seaside, OR. A bunch of friends , my wife and I go over to the coast to attend that one every year. Lots of treasures to be found. A lot of junk too.....
Anyway, please keep us posted on what you find wrong with it, and how you resolve the issues.

Mitch

[Grapsus]

Hi JacquesBBB,

Congratulations for your finding! I'm definitely jealous of your dumpster now...

I was confused about the ripple too when I checked my 465. I think this measurement needs to be performed with a reasonable bandwidth limit, something like 100 kHz, which DDMs have when measuring AC. Without a bandwidth limit, you'll be able to find arbitrarily large spikes due to HF pickup by the probe and the circuit itself. Maybe that's the problem with your measurements, since all values are consistently out by an order of magnitude.

[Smith]

Your attenuator switch may be dirty.

Inside the attenuator are little gold forks that get moved by a cam on the range switch.  Soak a small piece of copier paper in isopropyl alcohol, slide it under the open contact, then close the contact and carefully draw the paper strip out.  Do that a couple of times for each contact.

I second that.  Looks like the same problem I saw with a 465 a while back. The switches tend to become dirty and show eradicate behavior.

[JacquesBBB]

Thanks to all for your advices.
I have done a small video for the problem.
You will see that CH2 is fine. The problem is in CH1.

[Grapsus]

Looks like all the ranges that don't work in your video have in common the x100 attenuator. Either the range switch contacts are dirty or the attenuator is dead somehow.

[JacquesBBB]

I have started to dismount the attenuators,  but with  precaution.
I looked to the 100x attenuator, and it seems that one of the contact of CH1 is always open
while it is always closed in CH2 (the good one).



Is this part of the problem ?

If it is, what should I do ?  I try gently to push it down, but it seems to be stuck.

[Grapsus]

The first two switches that you're showing in your pictures are for AC/GND/DC selection, they're not actuated by the range switch so don"t force them.

Like others said, you should try to gently clean all the leaf contacts by sliding a piece of paper soaked in IPA under them:

https://www.eevblog.com/forum/repair/tektronix-465/?action=dlattach;attach=152364

[JacquesBBB]

The first two switches that you're showing in your pictures are for AC/GND/DC selection, they're not actuated by the range switch so don"t force them.

Like others said, you should try to gently clean all the leaf contacts by sliding a piece of paper soaked in IPA under them:

Thanks, I did the cleaning with the piece of paper soaked in alcohol,  and let  the last contacts in peace, after moderate attempts to close  the open contact.

Now everything seems to work, but I have also the impression that  the fact that the oscillo has been running for a while improved its condition.  I will eventually now make a calibration, but I am already quite happy with
the present status of  the device.

I have also checked again the ripple by several ways.
 
Using another oscilloscope (TDS460) and better probes (350 Mhz HP), with basically the same results as with the Rigol.

Using the Keithley 197 to measure the rms voltage in AC mode

(V)          nom. ripple    measured ripple   rms ripple (AC)
               p-p               p-p (DS1054z)     keithley 197

+55          4mV            27 mV                  0.27 mV
+15          2mV            20 mV                  0.15 mV
 +5           2mV            20mV                   0.23 mV
-8            2mV             23mV                   0.39 mV
+110      20mV             50mV                   4.8  mV

Also using the DS1054z with the bandwidth limited to 20 Mhz (I did not find another option). It gives results better than  with full bandwidth, but  still of the same order of magnitude.

So it  seems that I can leave it like that, thinking that the truth my be between the rms voltage and the ripple  at full bandwidth.

[tautech]

What's important is the Keithlys frequency response, those values are very good IMHO.
It's a linear PS isn't it? So ripple should be at ~100-120 Hz, well within the capability of most DMM's, so those measurements should be accurate.
Even if a SMPS the frequencies you listed are wrong, you'd expect ripple ~40 KHz.

[Paul Moir]

EDIT:  looks like you've got it. 

[JacquesBBB]

Today I verified and adjusted the calibration (external tuning).
Everything is fine and needed just small adjustment.

I verified also that the digital storage functions are working properly.
It is true that this  is not the strong point of the Tek 468.
It is given for 10 Mhz bandwidth in digital mode with a max sampling rate of 25 Mhz
and a storage memory of 512 points on 9 bits.

This is far from the performances of the DS1054z that I happened  to unbox just at the same time.
But  the tek 468 is really nice with its direct analog output, and very clean trace.

Although very limited, the digital functions  provide a bonus, as you can measure  time and volts
with cursors. I like the time cursor that is given by a bright dot moving along the trace.


The only thing to improve now is probably the level of noise. I will have to change the fan
to a quieter one.
Does anybody have a good suggestion for that ?

In any case, thanks for all for the decisive help that allowed me to put back this device into proper use in  a very short  time.

Jacques

[Grapsus]

Here is what I just measured on my 465 with a Fluke 87 III in AC mode:

rail	measured rms mV	converted peak-to-peak mV	reference peak-to-peak mV
110	17	68	20
55	1.2	4.8	4
15	0.9	3.6	2
5	0.9	3.6	2
-8	0.7	2.8	2

I converted the measured RMS values to peak-to-peak by multiplying by 4, assuming the ripple to be triangular. Your keithley 197 RMS measurments seem to be pretty close and if we apply the same 4 factor, your scope PSU seems to be within spec. Mine is not because it's a lot older.

[JacquesBBB]

Interesting.

I am glad that I do not have to change caps on this  one (at least not immediately), as although it is  said to be  serviceable,  it looks quite impressive  with respect to the other ones I have worked on.