Author Topic: Tektronix 2465B oscilloscope teardown  (Read 658581 times)

0 Members and 5 Guests are viewing this topic.

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #225 on: July 08, 2014, 09:16:43 pm »
So for those with a lost memory and a non-working 2465B, could the scope use "generic" cal data?
I.e. how poor would the scope behave if one used cal data from an other working 2465??

The horizontal, vertical, and other calibrations will be wrong but it should otherwise work.  I do not know if it would be any better than just clearing the calibration memory and operating the oscilloscope anyway.
 

Offline casinada

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #226 on: July 12, 2014, 05:46:59 am »
I lost the cal data from the Dallas Chip. I downloaded the backups available from the ko4bb repository as a starting point, programmed the data on to a fresh DS1225 and then calibrated my 2465BDM using different generators, signal sources, power supplies, multimeters until I was satisfied with the results.

 

Offline Bryan

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #227 on: July 12, 2014, 06:25:23 am »
Pretty much what I am faced with doing. Have myself a TG501 so that's half the battle. The rest I think I can do with my HP 50Mhz frequency generator, checking the outputs with a accurate DMM to test the amplitudes.
-=Bryan=-
 

Offline richhas

  • Supporter
  • ****
  • Posts: 11
Re: Tektronix 2465B oscilloscope teardown
« Reply #228 on: August 02, 2014, 08:18:36 pm »
Ok... Just received a 2464B and disrobed it for an initial inspection of A5 and the Main Board - having read this thread over several times.

I expected the worst wrt the A5 EC issues but think I'm pleasantly surprised - but I'm not an expert and thus my query here.

Both boards look super clean to me and I don't see any EC junk on A5. The thing that caught my eye have been the date codes. I thought that these units ceased production in '92-ish but I'm seeing date codes of '94, '95, and even a '97. Does anyone know how late these Beaverton units were produced? Or maybe these are replacement boards produced later than end-of-production. Could it be that Tek fixed the A5 EC issues late in the production?

In any case, it would great to get opinions from the experts here. Did I get lucky? Should I go ahead and replace the A5 caps? I'm going to heat sink U800 and socket/replace the damned Dallas ram - also update the LV/HV caps if needed.

This unit had 9330 hours spread over 1340 power cycles - approx on for 7 hours at a time. Makes some sense in a work environment - not too much cold start stress I hope.

The quality and experience from you guys has been super educational.

THANK YOU,

Rich
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #229 on: August 02, 2014, 09:40:49 pm »
My notes say that the 2465B and 2467B were available from 1989 through 1996.

I suspect the U800 failures were more related to stress caused by the poor mounting than heat.

I believe the later oscilloscopes did not suffer from leaking surface mount aluminum electrolytic capacitors.  If that is the case, then I would not risk of damage to the oscilloscope while changing them unnecessarily.
 

Offline richhas

  • Supporter
  • ****
  • Posts: 11
Re: Tektronix 2465B oscilloscope teardown
« Reply #230 on: August 02, 2014, 11:04:56 pm »
Thanks David... BTW the serial# is B065301.

I just went into the PS cage and many of the caps and other parts are marked '94 and some '95. The inspection sticker on the LV PS says mid '95.

Those transparent plastic covered caps on the PS pcb seem to not be cracked or such. Would your advise be the same - if is not broken, then...?

Best,

Rich
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 3163
Re: Tektronix 2465B oscilloscope teardown
« Reply #231 on: August 02, 2014, 11:25:24 pm »
Kind of a big SWAG but...

If you took all the Tek 24XX and 22XX scopes manufactured between 1990-1995 that were subsequently humanely treated:
- used (turned on) a few hours per day (20 years x 4 hours per day x 250 business days per year = 20,000 hours)
- maintained in a decent environment (temperature, humidity, etc.)
- operated with proper respect

What percent of these humanely treated Tek scopes would make it to 2015 without needing any parts (other than a battery) replaced?

What percent would make it to 2020 without needing any parts replaced?
 

Offline richhas

  • Supporter
  • ****
  • Posts: 11
Re: Tektronix 2465B oscilloscope teardown
« Reply #232 on: August 02, 2014, 11:50:56 pm »
No expert here but my guess is the cap plague with these EC lowered the average a great deal - amplified by the great tolerances tek tended to design in only served to hide problems until too much damage was done to pcbs... But just an old guy's guess.

Best,

Rich
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #233 on: August 03, 2014, 01:37:38 am »
I think it was only the early surface mount aluminum electrolytic capacitors which had this problem in the 2465 series and it was not capacitor plague.  I suspect it had to do with washing after soldering where the capacitors or the capacitor seals were sensitive to halogenated solvents.
 

Offline linux-works

  • Super Contributor
  • ***
  • Posts: 1993
  • Country: us
    • netstuff
Re: Tektronix 2465B oscilloscope teardown
« Reply #234 on: August 03, 2014, 03:27:36 am »
What percent of these humanely treated Tek scopes would make it to 2015 without needing any parts (other than a battery) replaced?

What percent would make it to 2020 without needing any parts replaced?

"In the year, twenty-five twenty-five
if tek is still alive
if agilent can survive
they may find..."



Offline richhas

  • Supporter
  • ****
  • Posts: 11
Re: Tektronix 2465B oscilloscope teardown
« Reply #235 on: August 03, 2014, 06:54:29 am »
Very good!
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #236 on: August 03, 2014, 07:36:48 am »
Rich, great score you got there !  :clap:

About the A5 board's electrolytic caps, have agree if I were you I will be facing the same hesitation too.

But, assuming you're very confident that you won't break the board while replacing them, I would say just replace them with best cap that is available now, just to have a good sleep knowingly they will not leak out any time soon, also those old SMD caps are made in early 90s, which already more than a decade old.

Still, its your call.


Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #237 on: August 03, 2014, 01:26:55 pm »
You could ask over on the TekScopes@yahoogroups.com email list or search the archive.  Someone there undoubtedly knows if the late 2465B series suffered from the surface mount capacitor problem.

I know of more than one instance now where someone changed all of the capacitors in one of these oscilloscopes and ended up breaking it.

Also, their is an error in the board layout diagram for the power supply which has two capacitors swapped so if you follow the service manual, it will short out the power supply.  It is a good idea to take notes and maybe photos as you go along.
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #238 on: August 03, 2014, 01:36:45 pm »
He is going to take out the whole big A5 board from the scope to desolder the big Dallas chip anyway, and solder the dip socket for new bbsram chip.

Imo, this task alone not a brief soldering job, and that will be the good opportunity to replace those caps at the board.

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #239 on: August 09, 2014, 06:10:20 pm »
I guess most of the Tek 24x5 owners are aware of the U800 chip vulnerability, just browsed Tek yahoo group pictures gallery and noticed this interesting pictures of A1 board captured using FLIR camera.

These pictures should speak for it self. Just remember to blow the whole A1 board with a strong fan when you're working without the enclosure, especially near the U800 chip.


Offline iDevice

  • Regular Contributor
  • *
  • Posts: 134
  • Country: be
Re: Tektronix 2465B oscilloscope teardown
« Reply #240 on: August 09, 2014, 06:53:03 pm »
That raises a question I wanted to post and forgot to.
A few weeks ago, I scored a mint 2445B with only 270 hours.
The only problem I could find was a failure of the beam finder switch that remained in the activated state.
It was fixed by cleaning.
But while the scope was open I left it running to check if everything was ok, obviously with a fan directed on U800.
And after about 15m, U800 was still cold.
So I stopped the fan a few minutes and U800 was still at about 30°C.
After about half an hour, I could still hold my finger on it and it felt just warm, I didn't think to take a real temp measure though.

So I'm wondering if there are different versions of U800, (based on IC dates, scope was build in '94) or if the legendary U800 meltdown only happen with certain settings ?
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #241 on: August 09, 2014, 07:04:31 pm »
So I'm wondering if there are different versions of U800, (based on IC dates, scope was build in '94) or if the legendary U800 meltdown only happen with certain settings ?
As far as I'm aware, currently only 2 versions, one made by Tek, another one from Maxim , posted both photos -> post #185 while ago.

Whats yours ?

Offline iDevice

  • Regular Contributor
  • *
  • Posts: 134
  • Country: be
Re: Tektronix 2465B oscilloscope teardown
« Reply #242 on: August 09, 2014, 07:26:36 pm »
I have the Tek labeled version.
Probably one of the last since they sold the factory to Maxim in '94, as it seems.

I'm waiting for a new DS1225, so I will open it again in a week or two.
This time I will take some shots with my TIC.
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #243 on: August 09, 2014, 07:44:19 pm »
This time I will take some shots with my TIC.

 :-+ , sounds great, looking forward to see further details at the temperature in A1 board, thanks !

Offline guido

  • Regular Contributor
  • *
  • Posts: 207
  • Country: nl
Re: Tektronix 2465B oscilloscope teardown
« Reply #244 on: August 24, 2014, 11:20:59 pm »

Had to calibrate the vertical after repairing the 2465. It got a new attenuator. As i don't have a PG506, i looked in the partsbin and found an AD CMOS switch. Put in on a pcb with a 555 and presto, a "chopper". Used a fluke calibrator to generate the DC voltages (20mV to 10V range) and managed to calibrate vertical and trigger. Not much of a circuit, but you need a decent calibrator to begin with  :D



The 20mV signal and source on a 2430A:



And since it was open (you need to calibrate the markers with pots), i also took a picture of my wannabee U800 cooler:


 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #245 on: August 25, 2014, 05:09:54 am »
Before I got a PG506, I managed to generate the calibration voltages using a stepped attenuator but I know of others that had good results using analog switches.
 
The following users thanked this post: Taviman

Offline Bryan

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #246 on: August 25, 2014, 08:20:02 am »
Does the vertical calibration have to be a 1khz square wave of varying amplitude like the PG506 or can it be a DC voltage. I know there is many cheap very accurate voltage references like a Ref5050 from TI .05% accuracy and 2.55 ppm/Celsius temp coefficient http://www.ti.com/product/REF5050/description that one could feed into a LTC1043 to divide/multiply to meet the required voltages. May need a couple different reference voltages to feed into the divider.  Suppose would need a buffer ie LTC1050
https://www.eevblog.com/forum/projects/should-the-ltc1043-voltage-divider-circuit-be-buffered/msg461531/#msg461531
-=Bryan=-
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #247 on: August 25, 2014, 01:28:12 pm »
I really should be a chopped square wave in order to reduce the effects of thermal drift and seven different voltages from 20 millivolts to 10 volts are needed.  Calibration of the automatic measurements absolutely requires a square wave.
 

Offline Bryan

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #248 on: August 25, 2014, 09:41:58 pm »
I do have a HP81161 pulse generator that I could use. Can output a 1khz square wave at 50% duty cycle and vary the amplitude to match the required calibration volatages for the 2465b. Only issue is I am not sure how accurate it is any longer. The performance checks for AC amplitude fall within the specs according to the measurements with my Uni-T multimeter. But can i really really on such a  cheap measuring instrument to confirm the HP outputs, especially on a a AC output.
-=Bryan=-
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #249 on: August 25, 2014, 10:41:15 pm »
There are two different easy ways to handle the calibration in that case.

An RMS reading voltmeter may be used in which case the peak to peak value is twice the RMS value.  A good meter will get you into the accuracy range necessary to calibrate an oscilloscope which is not a precision instrument for voltage measurements anyway.

Alternatively an average responding AC voltmeter may be used with a correction factor.  For square waves, it will read 11% high.  This is usually a better option because average responding meters are usually both faster and more accurate than RMS responding meters.

When I do this I use both as a sanity check.  All of my good meters handle 1 kHz square waves with no problem.

One thing to beware of though when doing the above is that the duty cycle needs to be 50% or at least known so a correction factor can be applied.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf