Author Topic: Tektronix 2465B oscilloscope teardown  (Read 658364 times)

0 Members and 3 Guests are viewing this topic.

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3316
  • Country: fr
Re: Tektronix 2465B oscilloscope teardown
« Reply #1450 on: March 01, 2019, 11:03:50 am »
Hello all: Working on another 2467B, SN B050xxx, ~ 8K hrs 1K cycles. Works fine, good CAL, perfect CRT and transient response.



1/ A5 NO SMD lytics, no corrosion, all Tant or TH bypass caps

2/ NVRAM original date 8852 = 31 years!
Passes all self test, No NVRAM errors

3/ A1 main orig TEK U800, no HS

4/ Fan replaced, works fine

a/Understand some  PROM readers misread or destroy data, also possible damage can occur to NVRAM in removal from PCB.

How to manually read out NV RAM contents before removal?

b/ Recommended DK and Mouse PN for NVRAM and low profile 28 pin machined pin socket?

c/ Precautions to desolder/remove  NVRAM and to read data on programmer (using XGecu TL866CS programmer) ?

MANY THANKS

Jon

« Last Edit: March 01, 2019, 11:06:49 am by jonpaul »
Jean-Paul  the Internet Dinosaur
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #1451 on: March 01, 2019, 12:17:20 pm »
Suggestion, just unscrew the two nuts that are holding the U800 chip, it is not going to fall down or separated from the board anyway, as the chip's pins all are soldered firmly on the pcb.

Thermal expansions and contractions actually stress out the whole U800 chip body with that 2 nuts attached, imo.

Offline AMR Labs

  • Frequent Contributor
  • **
  • Posts: 255
  • Country: an
Re: Tektronix 2465B oscilloscope teardown
« Reply #1452 on: March 01, 2019, 12:19:00 pm »
Hi Jon,

These are the parts I used from Mouser on my 2465B and they worked very well.

828-AG11D-ESL-LF IC Socket DIP .6CL 28P
MOUSER PN 571-2-1571552-9

DS1225AD-150+ NVRAM 64k Nonvolatile SRAM
MOUSER PN 700-DS1225AD-150

But I ended up using a FRAM instead of the SDRAM to do away with any future battery issues:

FM16W08-SGTR F-RAM F-RAM 64Kb 70ns 8K x 8 Parallel FRAM
MOUSER PN 877-FM16W08-SGTR

Fram is SMD type so it requires an adapter board:
IC & Component Sockets IC & Component Sockets 28P SOIC/DIP SOCKET
MOUSER PN 535-28-650000-10

I used the GQ-4X4 programmer which supports both the 1225 and the FM16W8.

Desoldering the 1225 was no big deal, just be extra careful as the A5 board is a multi-layer board. I used a regular 40W Weller pencil soldering iron dialed down to about 25W, and a soldapult. Make absolutely sure all pins are free by jiggling them a bit before trying to pull the chip. For me it was just straight forward, even pin 14 which is ground and usually requires a bit more heat came out just fine without any issues.

To back up the cal constants from the NVRAM before pulling it use EXER02 routine (check service manual section 10) and write the values down with paper and pencil, or take a video while you scroll trough all the individual values. Once you have the chip out read it and save it to file, then program the new chip.

Good luck with the procedures
Alex
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3316
  • Country: fr
Re: Tektronix 2465B oscilloscope teardown
« Reply #1453 on: March 01, 2019, 02:21:53 pm »
Hello again: RE the 2467B, I just checked H and V CAL, CRT and transient response, incredibly, its perfect.
For a machine built in KUDOS to the TEK engineers!

To BravO re A1/U800:

I wonder if the screws provide any heatsinking or better thermal contact to PCB?
Perhaps replacement of the lockwasher with a constant tension spring or Belleville washer?


To AMR Lab/Alex: re A5 NVRAM: Understand re the FRAM, will use the Dallas parts avail from DK.
For removal of the NVRAM, the internal ground and power planes make certain pins very difficult to remove.
For EXER 02, I used it but had only a few values. Which buttons to scroll up><down?


Irons for removal NVRAM: I found the best iron for such work:  Metcal SP-200 "smartheat"
http://www.testequipmentdepot.com/metcal/pdfs/sp200_manual.pdf

Just a few deg. tip temp regulation instead of the 50-100 deg overshoot and hunt of normal irons.




Enjoy!

Jon

PS: Full disclosure: in 1992,  I  designed the SP-200, resonant mode SMPS that drives 30W constant current sinewave to iron  the curie point tip (500 kHz),

Jean-Paul  the Internet Dinosaur
 

Offline Satbeginner

  • Regular Contributor
  • *
  • Posts: 151
  • Country: es
  • Dutch, early retired, living in Spain
Re: Tektronix 2465B oscilloscope teardown
« Reply #1454 on: March 01, 2019, 02:25:06 pm »
Same here, I could clean all pins but pin 14 using a vacuum soldersucker, I used a soldering iron for pin 14 when I removed the Dallas.

Using a small nose tip pliers to wiggle the pins first to make sure they are all loose is the trick.

I choose to replace the Dallas with a new Dallas I got from Mouser, using a TL866 MiniPro programmer.

Good luck,

Leo
You need a scope to repair a scope, and you need many multimeters to repair another multimeter!
*Tek 2467B, Tek 2465B, Tek 2465B, Tek 485, Tek 475A,  Keithley 175A, Keithley 2000, HP 3468B, HP 3457A, HP 34401A, PM 6671, PM 5716, Fluke 45, Fluke 75, Fluke 77, Fluke 79, AFX 9660BL, KPS 605D, etc. *
 

Offline med6753

  • Super Contributor
  • ***
  • Posts: 11313
  • Country: us
  • Tek nut
Re: Tektronix 2465B oscilloscope teardown
« Reply #1455 on: March 01, 2019, 03:17:19 pm »
Suggestion, just unscrew the two nuts that are holding the U800 chip, it is not going to fall down or separated from the board anyway, as the chip's pins all are soldered firmly on the pcb.

Thermal expansions and contractions actually stress out the whole U800 chip body with that 2 nuts attached, imo.

Please tell me you're not serious.  :o Just kidding, right?  :-// It's been in there 30 years and worked fine and now you want to monkey with it? Do so at your own risk.

Anecdotal evidence: This thread is 8 years old. Has ANYONE come in here and said that U800 took a dump? Not that I recall.
An old gray beard with an attitude.
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3316
  • Country: fr
Re: Tektronix 2465B oscilloscope teardown
« Reply #1456 on: March 01, 2019, 04:13:11 pm »
Hello Med6753, old but great thread.   Please see my previous posts today for full story.

This is preventive maintaince especially for the 1989 battery backup NVRAM.
Eventual battery failure was my motivation to open the scope up!

Agree to leave U800 asis.
If it ain't broke don't fix it!

With  2 other 2465/7B with the failed NVRAM and complete reCAL, I have learned the hard way.

Enjoy,


Jon

Jean-Paul  the Internet Dinosaur
 

Offline med6753

  • Super Contributor
  • ***
  • Posts: 11313
  • Country: us
  • Tek nut
Re: Tektronix 2465B oscilloscope teardown
« Reply #1457 on: March 01, 2019, 04:39:12 pm »
Hello Med6753, old but great thread.   Please see my previous posts today for full story.

This is preventive maintaince especially for the 1989 battery backup NVRAM.
Eventual battery failure was my motivation to open the scope up!

Agree to leave U800 asis.
If it ain't broke don't fix it!

With  2 other 2465/7B with the failed NVRAM and complete reCAL, I have learned the hard way.

Enjoy,


Jon

Understood. I was only referring to U800. I agree with the need to open it up and deal with the NVRAM issues.   
An old gray beard with an attitude.
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #1458 on: March 01, 2019, 04:41:02 pm »
To BravO re A1/U800:

I wonder if the screws provide any heatsinking or better thermal contact to PCB?
Perhaps replacement of the lockwasher with a constant tension spring or Belleville washer?

Jon,

As I said, its just my "suggestion". As I've concluded this thru lots of studying and also through close up photos of the U800, I do have few spare U800s (just in case you missed it, I did post them here too in this thread with close ups), and my conclusion is the two nuts basically are useless, and "PROBABLY" may do harm than good, again, this is just my own conclusion.

As I never summarize this up, so here we go. Again, this my own conclusion, and if you don't agree then don't touch them, as simple as that.

Ok, 1st, this photo below, is the standard untouched U800 chip with the locking nuts at the two embedded bolts, and watch closely there are serrated washers that are sandwiching the ic metal tab.

(click photo to enlarge)



Now, this is the close up of the PCB where I de-soldered my spare U800 chip off the board, and the two bottom serrated washers location, which are sort of lift the whole ic body not touching the big exposed pcb pad below it. So no effective direct contact for heat dissipation path on it.




Intuitively we will think that the generated heat dissipation path is from the chip through the ic metal tab, and then pass through the washers that are compressed by the nuts, and then to the two embedded metal bolts/screws.

This is how the U800 chip looks like at bottom side. If you watch closely, you can see there are dented "dots" in circle at the tab, that are made by the serrated washer tips, meaning, the metal to metal contact is really poor for heat transfer as its just "touched" by the washer's sharp tips.


Again, intuitively, we will think, why not we just fasten the two nuts that will squeeze the washers to the chip's metal tab to make it better thermal path. The answer is NO , as this below illustration, as you just CAN NOT tighten the nuts, as it will sort of peel off or push down the metal tab downward from the chip's body, once the bottom serrated washer are squeezed down, as the whole ic body is already secured firmly by 24 ic's pins that are soldered at the board.




Just simple question, say your U800 is toasted as this is the common problem, you replace it with a working one, then solder at it's 24 pins as usual, and then you put the two nuts. When you starting to fasten them, how much torque do you need ? Ever think about that ?

Ever imagine that once you starting to even slight put a force on the nut, it will starting to bend the ic metal tab downward or peel off from the chip black body as there is a gap there.

Also we all know, that this part and region experienced thermal cycles, what happened to the tightened nuts and washers as it expand or contract, basically its like micro bending the ic metal tab up and down on every cycle.

This is my conclusion, I could be wrong, but this what I believe, the two nuts basically serve no purpose, and if over tightened, will actually do more harm than good.

So, its up to you.  :-//
« Last Edit: March 01, 2019, 05:08:09 pm by BravoV »
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #1459 on: March 01, 2019, 04:43:41 pm »
Suggestion, just unscrew the two nuts that are holding the U800 chip, it is not going to fall down or separated from the board anyway, as the chip's pins all are soldered firmly on the pcb.

Thermal expansions and contractions actually stress out the whole U800 chip body with that 2 nuts attached, imo.

Please tell me you're not serious.  :o Just kidding, right?  :-// It's been in there 30 years and worked fine and now you want to monkey with it? Do so at your own risk.

Anecdotal evidence: This thread is 8 years old. Has ANYONE come in here and said that U800 took a dump? Not that I recall.

I'm serious med, and don't want to argue and I've posted my argument above.

Its my believe and my own reasoning, as I just "suggest" it to Jon, he is free to reject it if he doesn't buy it, as simple as that.

Offline med6753

  • Super Contributor
  • ***
  • Posts: 11313
  • Country: us
  • Tek nut
Re: Tektronix 2465B oscilloscope teardown
« Reply #1460 on: March 01, 2019, 05:01:14 pm »

This is my conclusion, I could be wrong, but this what I believe, the two nuts basically serve no purpose, and if over tightened, will actually do more harm than good.

So, its up to you.  :-//

Absolutely agree with that. As far as I can tell Tek never published the torque spec for those nuts. It would be all too easy to overtighten the nuts. And I while I have no direct evidence or experience otherwise let me give a counter theory. Isn't the most common failure mechanism of the U800 a delamination of the internals, especially on Maxim sourced chips? If that's the case I could argue that the nuts may help to PREVENT a delamination by keeping the IC together. Perhaps a dumb theory but who knows.  :-//     

Edit...I don't wish to argue it either. Peace.  :-+

« Last Edit: March 01, 2019, 05:03:29 pm by med6753 »
An old gray beard with an attitude.
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3316
  • Country: fr
Re: Tektronix 2465B oscilloscope teardown
« Reply #1461 on: March 01, 2019, 05:43:30 pm »
Hello Med and Bravo

ACK on all!

Any thoughts on the other questions I pose in my posts earlier today?

RE U800, In the 4 pcs 2465/7/B I have worked on, all have original TEK U800, and none overheat, just warm to the touch. 

Is there a worse case test condition for U800 heat, eg run at 5 ns/ div or use 10X mag?

Has anyone done temp test with thermocouple or IR probe?

Or calculated the worst case Pdiss?

Cheers,

Jon

« Last Edit: March 01, 2019, 05:45:14 pm by jonpaul »
Jean-Paul  the Internet Dinosaur
 

Offline AMR Labs

  • Frequent Contributor
  • **
  • Posts: 255
  • Country: an
Re: Tektronix 2465B oscilloscope teardown
« Reply #1462 on: March 01, 2019, 09:08:34 pm »
The U800 write up bellow is one of  the best ever explanation I came across, and it is from someone that has tons of experience repairing and calibrating 2465x/67x scopes on a very frequent schedule. He is THE authority on these scopes, or at least definitively one of "THE"...

Original message at:
https://groups.io/g/TekScopes/message/152372

==== Start Quote ====>

Everybody please listen!

U800 is not heat sensitive, and I have found that many of
the "tricks" folks have gotten from the internet on how
to "protect" their U800 chip actually harm the chip due
to over-tightening the nuts on the studs that support the
chip. I have also found that glued/taped on heat sinks
have raised the temperature of the chip by impeding the air
circulation around the chip.

I have also found glued/taped heat sinks floating freely
inside of scopes that have been sent to me for service!

Tektronix botched the mounting of U800. That, in my
experience, is why they fail.

U800 was designed to have its pins pass through a 2-3mm
thick aluminum heat sink mounted underneath the chip.
The belly of the chip is metal to engage this heat sink.

DIP style IC's have pins that are made with a wide part to
prevent the belly of the IC from bottoming onto the PCB.
This allows air circulation under the chip. U800 is no
exception.

The engineer that designed U800 was excessively cautious,
and overestimated the heat that would be generated by his
new IC. When it came back from fab, he found that it did
*NOT* need any heat sinking to meet the full +50C design
temperature rating of the brand new 2465 scopes. So, the
heat sink was left off of the board.

(Thought experiment: when was the last time you operated
your 2465 at 50C?)

The problem is the designer needed -5V to bias U800's
substrate, and he used the metal heat sink frame and tab
to make that connection.

He found that U800's stud could not clamp tightly to
the PCB (REMEMBER: shoulders on pins...) without putting
excessive pressure on U800's pins, so he put a pair of star
lock washers onto the mounting studs before mounting U800.

The washers served to make the electrical connection, and
to provide the needed spacing to protect the chip....

He hoped.

Everything is fine if nobody ever, ever, over-tightens the
nuts that hold the U800 to the board. If they do, the
extra pressure will collapse the star lock washers and
embed them into the PCB and the bottom of one end of the
epoxy body. When this happens, the lead frame of U800 will
crack at some of the pins

If you come along later, having listened to all the bloviation
about U800 fixes, and happen to tighten up the nuts a little
bit more: crack! You will have damaged the chip while trying
to protect the chip.

-Chuck Harris

<==== End Quote ====
« Last Edit: March 01, 2019, 09:16:22 pm by AMR Labs »
 
The following users thanked this post: nhzx0110wxd

Offline ezalys

  • Frequent Contributor
  • **
  • Posts: 322
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #1463 on: March 17, 2019, 01:44:26 am »
Hey folks, I'm really struggling with a 2467B with a "CT TEST 86 FAIL 02."

The options manual indicates this failure corresponds to "In Trigger After Delay mode with the delay time set longer than the delay, there was no sweep. Check B AUX TRG and HO at U6190 and output O1 at U6140." Is this a failure that others have seen? Could I have left a cable loose somewhere? U6190 appears to be some sort of gate array, and I'm not sure where I'd find a replacement. I'm also not sure how to get at the B AUX TRG/HO/O1 signals on the specified ICs, as the board needs to be installed face down to be powered by the power supply.

To me is really seems like I must've left a cable undone, but I've verified that I've connected, in the appropriate positions, ribbon cables C and E, and the pair of coaxes D, and on the opposite side of the instrument, a rather wide ribbon cable that passes through the side of the chassis. Seems like everything is connected OK. What's the next thing to check?
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2120
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #1464 on: March 18, 2019, 09:55:30 pm »
Hey folks, I'm really struggling with a 2467B with a "CT TEST 86 FAIL 02."
...
I don't have the 2467B options service manual, but the options from model to model in the 24xx family are very similar.

The 24x5A options service manual, which I do have, has a troubleshooting flowchart in the back that says (and I'll summarize the relevant thing here) if CT test 86 failed and any of the bits 0, 1, or 2 of the error code is a 1, it may indicate that the sweep speeds are uncalibrated.  If you have the actual 2467B options service manual, you could see if it says the same thing.

In reading the chart that you point out for "CT TEST 86 FAIL 02", I'm honestly having trouble deciphering exactly what behavior it's complaining about.  Can you determine what's not working with delay-by-events (besides the self-test)?

Is this a new problem or did it come like this?

Can you point to a copy of the 2467B options service manual on-line somewhere?
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Tektronix 2465B oscilloscope teardown
« Reply #1465 on: April 22, 2019, 02:30:24 am »
About a week ago I started on recapping the low voltage regulator and inverter boards in my 2467B after a few anomalies started showing up.

I would like to draw people's attention to an error on the board layout sheet fig 10-13 in the Service Manual 070-6863-01 that has C1115 and C1132 transposed on the inverter board (250uF/20V vs 10uF/160V). This took several hours to figure out, and a few burnt out parts too.

Necro posting time!
I just discovered the exact same fault. Measured C1115 ESR in circuit and it looked sussed, so sucked out only to find a 10uF 160V cap. Traced and the circuit a bit and figured it was swapped on the overlay diagram  |O
Sucky thing is, changing that cap (which was failed) made no difference to the 5V rail ripple (output cap C1280 seems fine), so  :-//
« Last Edit: April 22, 2019, 02:32:58 am by EEVblog »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Tektronix 2465B oscilloscope teardown
« Reply #1466 on: April 22, 2019, 04:25:05 am »
Oops

 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #1467 on: April 22, 2019, 04:28:15 am »
Geez... what happened Dave ?  :o

Offline kj7e

  • Frequent Contributor
  • **
  • Posts: 911
  • Country: us
  • Damon Stewart
Re: Tektronix 2465B oscilloscope teardown
« Reply #1468 on: April 22, 2019, 04:49:29 am »
At least the problem will be easy to find now.
 

Offline med6753

  • Super Contributor
  • ***
  • Posts: 11313
  • Country: us
  • Tek nut
Re: Tektronix 2465B oscilloscope teardown
« Reply #1469 on: April 22, 2019, 05:22:02 am »
Yep, known problem and Tek never fixed the manuals. I've done complete re-caps of the Inverter/Regulator boards on both my 2465's and I do one capacitor at a time (remove/replace) and I follow what's physically installed on the boards.

I haven't had the excitement of magic smoke.  :-DD   
An old gray beard with an attitude.
 

Offline 0culus

  • Super Contributor
  • ***
  • Posts: 3032
  • Country: us
  • Electronics, RF, and TEA Hobbyist
Re: Tektronix 2465B oscilloscope teardown
« Reply #1470 on: April 22, 2019, 05:41:25 am »
Ooof, that was a good bang!
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Tektronix 2465B oscilloscope teardown
« Reply #1471 on: April 22, 2019, 08:58:59 am »
Geez... what happened Dave ?  :o

You'll find out in the next video, about 30min in or something..
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Tektronix 2465B oscilloscope teardown
« Reply #1472 on: April 22, 2019, 08:56:54 pm »
I would like to draw people's attention to an error on the board layout sheet fig 10-13 in the Service Manual 070-6863-01 that has C1115 and C1132 transposed on the inverter board (250uF/20V vs 10uF/160V). This took several hours to figure out, and a few burnt out parts too.

I just discovered the exact same fault. Measured C1115 ESR in circuit and it looked sussed, so sucked out only to find a 10uF 160V cap. Traced and the circuit a bit and figured it was swapped on the overlay diagram  |O

That specific mistake has come up a couple times on the Tekscopes mailing lists.  Also, sometimes the board or layout will have polarity marked wrong.

I have tried to make it a habit to record part locations and orientation as I remove them to avoid those sorts of problems.
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Tektronix 2465B oscilloscope teardown
« Reply #1473 on: April 23, 2019, 12:27:05 am »
New Dave's video relating to 2465B, post here again as reference, originated from this thread -> EEVblog #1203 - REPAIR: Tektronix 2465B Oscilloscope


Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Fake Tektronix 2465B ?
« Reply #1474 on: April 24, 2019, 09:18:02 am »
Fake 2465B ? As Dave's 2465B's badge is different from mine.  :-//

Pointed out by Scopetechniques -> Fake 2465B in EEVblog #1203

Repost here as reference.

« Last Edit: April 24, 2019, 01:51:11 pm by BravoV »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf