Tektronix 2465B oscilloscope teardown

[SeanB]

A quick clean up of that anode connector, a generous application of polishing on the plastic and a spritz of anti corona spray and that will do for a long time. The caps will need cleaning of the board, looking for the dead bugs inside the case and soldering of new ones onto the board. As well you might want to read out the RTC chip and get a replacement one for when that one dies.

When I was working for the SAAF I got a 2223, and after about 6 months the digital board died. Unplugged the power connector and used it as a plain analogue scope for another 3 months till it's cal cycle time was due, and then plugged the board back in and sent it off. Came back with a new digital board.

[KD0CAC John]

Will have to use this thread to look for the next broken piece of gear for my bench
Being that disability is my regular source of income , I look for dead gear for my bench & ham radio shack .
And the next benefit to broken gear , is having to learn how to fix .
I've been in many trades and at least half of those , or more , were hobbies and then wanting to build / repair my own toys .

[iloveelectronics]

A few more pictures...

The fan (after some cleaning up). Curiously, like BravoV's unit, the date code (if that's what it is) indicates a 2004 date while everything else inside has 89/90 dates.


The probe power sockets (option 22):


Another look on the power supply board, with the black plastic shield on the left removed:


Some solder joints look like they have some corrosion going on but they are actually fine. It's just some sort of reflection in the photo. Unfortunately this unit has the HV caps with the transparent housing. This one in the picture still looks OK:


Not the other one though. It's already cracked open:



And here's a close-up view of the leaky/missing caps area after initial cleanup:


I've tested continuity and the one with only half a pad left seems to have lost connection to the trace. Shouldn't be difficult to fix given the proximity. The other pads are fine--ish.

I was trying to see if I could get the power supply board out but have had no luck. Can someone tell me how I should go about doing that? Does it come out together with the other board attached to it?

[linux-works]

the power supply board pair comes out as a pair.  loosen the rear screws since its physically tight and very hard to remove the way tek designed it.  I struggled a lot to get my board pair out.  its not meant to be taken out, so you -have- to loosen a few chassis screws to -create- clearance to remove the boards.

[iloveelectronics]

the power supply board pair comes out as a pair.  loosen the rear screws since its physically tight and very hard to remove the way tek designed it.  I struggled a lot to get my board pair out.  its not meant to be taken out, so you -have- to loosen a few chassis screws to -create- clearance to remove the boards.

Thanks a lot! Loosening up the rear panel helped. I've now taken the PS board out. Yet more pictures...

Better look on the crack opened HV cap:


I didn't see where this dangling connector go while I pulled out the board  It's coming from the TV option board. Does it connect to J305 in the second photo (the other PS board in the pair)?



Does anyone have the part numbers for those 2 HV caps handy, before I dig into the service manual? Thanks a lot for all the help given in this thread so far!

[mwilson]

Does anyone have the part numbers for those 2 HV caps handy, before I dig into the service manual? Thanks a lot for all the help given in this thread so far!

Yep, I'm actually putting together a DigiKey order for repairing my 2467B right now. They carry those exact same caps, Kemet part number PME271M568MR30 (CAP FILM 0.068UF 275VAC RADIAL). I'm debating whether to replace with new ones of the same type that have proven to be dodgy or going for something different that should be compatible like Panasonic ECQ-U2A683ML.

[BravoV]

Franky, thanks for sharing the photo shots, glad the corrosion caused by the leaked caps didn't do heavy damages there, and look like all traces still intact except for that half pad, btw great cleaning job you've done there. 

Btw, this confirmed that these particular hv transparent film cap is cursed, in this thread alone already two proven cases, hope this will help others in the future as reference and replace them before bad thing happened.

I guess you have many home works to do for this beast, btw just remember, when powering the whole scope without the enclosure, make sure you have a big fan blowing the "whole" giant A1 board with the scope up side down position, so those unobtainium hybrid ICs get cooled properly.

Looking at your fan's date code, year 2004 ?  Wow, if its the original fan came with the scope from the Japan factory, then your 2465B is definitely quite fresh imo.

Regarding your missing caps at A5 board, suggesting after you desoldered the old Dallas chip, install a good dip or sip socket for the new Dallas DS1225 chip, so it will be easy to backup & restore the calibration in the future.

Btw, spotted another missing cap at your A5 board at the right upper corner, besides the screw pointed by the red arrow. Also there are other A5 examples at this post -> Here for comparison.

[iloveelectronics]

Got a little bit of time today so decided to desolder the two 250V film caps on the power supply board. The one that still looks normal measures about 83nF, which is still within tolerance I guess. But the broken one measures only 170pF! I'm really surprised this scope was working despite the missing caps on the A5 board and this broken film cap!

BTW, thanks BrovoV, for spotting the other missing cap on the A5 board! I've ordered them now from Mouser (using the part numbers given by linux-works), and hopefully will receive them in a few days. The Dallas NVRAM is on backorder though so it looks like I will have to wait a couple of months Will certainly backup the data when I work on the A5 board though.

[BravoV]

Even at the normal one if you look closely at the clear plastic case, it is already has so many fine cracks at it's surface.

Btw, about the Dallas chip, actually you can start desoldering it now while you're working at A5 board, buy & solder a good quality DIP or twin SIP sockets at the pcb, and the most important is to back it up 1st using your TL866, and put it back while waiting for the new chip to arrive.

Remember, the lithium battery that is holding the scope's calibration in your Dallas BBSRAM chip is > 20 years old now (date code 90xx), imo its wise not to trust your luck on it for too long. 

[guido]

One more up and running

Fixed two zener diodes (reference supply protection and attenuator supply).
Replaced a broken 10nF cap in the HV section (and swapped a similar one), which 'shortened' one of the trimmer pots to gnd (40V to ground = broken trimmer).
Replaced the whole pots/switches pannel (cheaper than getting separate time/div and volt/div switches, which were bent).
Replaced a broken attenuator (did swap channels for test and verified that the relays were ok).
Replaced all elco's and X / Y Rifas in the PS section.

Got it quite cheap, so all in all a good outcome.

PS, no Dallas chip in these early models but an EEPROM (called EAROM). I've noted all the values, so in the future i might be able to replace it with a modern solution if it would fail.

I've read the temperature stories about the U800 chip. But i was also told it is not related to temperature, but because of outsourcing the manufacturing from tek to a third party. The original tek chips don't break. If that is true, i'm safe as it is a very early 2465 (Guernsey). It is also full of factory installed bodges (as in reworks).

A quick question for all the 2465 users here. One of the things i notice is that the brightness of the traces goes up if i switch on more channels. ........

I'm wondering if this is a bug or a feature. .............

Well, RTFM. A feature. It is just that mine is adjusting too much. Specially going from one to two channels. Minor thing.

[BravoV]

I've read the temperature stories about the U800 chip. But i was also told it is not related to temperature, but because of outsourcing the manufacturing from tek to a third party. The original tek chips don't break. If that is true, i'm safe as it is a very early 2465 (Guernsey). It is also full of factory installed bodges (as in reworks).

Well, my U800 chip was almost too hot to touch by fingers at the black ic body even there was a big fan blowing strong wind on it.

Yes, it was outsourced to Maxim IC after Tek decided not to produce it anymore, as far as I know, there are two versions of this U800 chip (TEK part no 155 0241 02) , photo of both attached below.

[w2aew]

I've read the temperature stories about the U800 chip. But i was also told it is not related to temperature, but because of outsourcing the manufacturing from tek to a third party. The original tek chips don't break. If that is true, i'm safe as it is a very early 2465 (Guernsey). It is also full of factory installed bodges (as in reworks).

Well, my U800 chip was almost too hot to touch by fingers at the black ic body even there was a big fan blowing strong wind on it.

Yes, it was outsourced to Maxim IC after Tek decided not to produce it anymore, as far as I know, there are two versions of this U800 chip (TEK part no 155 0241 02) , photo of both attached below.

Well, not really outsource.  Tektronix had its own internal IC fab - called the Integrated Circuits Operation (ICO).  Tek sold ICO to Maxim in 1994.  It is still co-located on the Tektronix campus in Beaverton.  So, the name changed, but the fab remained the same. 

At the same time, Tektronix sold half-interest in their hybrids design and manufacturing group to Maxim, which formed the group known as MaxTek.  Several years ago, Tektronix bought out Maxim's interest in MaxTek, returning this group to be wholly owned by Tektronix.

[e-pirate]

Hi folks!
A couple weeks ago I was lucky to get myself Tektronix 2467B. It's a end of 91 unit, produced in Japan by Tektronix-Sony (and seems to arrive somewhere from Japan because I found some papers with handwritten Japanese hieroglyphs inside pouch at the top of it). It's in a perfect visual condition and actually looks like a new device. TEST5 shows something under 2000 work-hours and about 900 power-on cycles, I'm not sure if I can trust this. But when I got into unit I was not able to find even a single tiny mark of dust on any of PCBs or parts, the only place that had just minor signs of dust is _original_ NIDEC fan, that is actually all most silent. Nothing soldered/repaired there. But dude, that sold it to me replaced A5 electrolytic caps and told me that the PCB is fine, and yes, it's virgin looking board without any signs of leakage. Besides, unit have active probe PSU option installed. I've had some time playing with this scope and seems that it work fine, all controls work as expected, all channels seem to be working fine to. So, meet another member in a 24xx club
But after reading this thread from the first page I came to conclusion that I still have some things to take care about.
Seems that I should start with A5 board (it is all SMD sure) and replace this pesky Dallas SRAM IC and then caps. Should I replace both electrolytic caps (I wish to replace them with solid state polymeric SMD caps) and this black one I don't know what type are they, seems that I've seen photos of them leaking to.
After I'm done with A5 board, I will go and check for HV and LV board and then recap A1 board.

[BravoV]

Well, not really outsource.  Tektronix had its own internal IC fab - called the Integrated Circuits Operation (ICO).  Tek sold ICO to Maxim in 1994.  It is still co-located on the Tektronix campus in Beaverton.  So, the name changed, but the fab remained the same. 

At the same time, Tektronix sold half-interest in their hybrids design and manufacturing group to Maxim, which formed the group known as MaxTek.  Several years ago, Tektronix bought out Maxim's interest in MaxTek, returning this group to be wholly owned by Tektronix.

Straight from the horse's mouth, thanks Alan for clarifying this. 

About the popular myth surrounding the Tek's hybrid ICs, is it true the chip design is gone caused by fire at MaxTek in the past ?

Alan, really wish if someday Tek will release the design of these hybrid IC's design, maybe begging contacting the PR dept about this possibilities ? Its like the good gesture in the past where Tek released all old scope's documentations to public legally. 

There was an attempt to design the U800 ic replacement with discrete components like this -> Tektronix 2465/2467 series U800 IC replacement design created by Thomas Lafay, but if Tek released the hybrid IC schematic design, that will be like breathing a new life for these great 246x analog scope series. 

Please consider it.

[BravoV]

Seems that I should start with A5 board (it is all SMD sure) and replace this pesky Dallas SRAM IC and then caps. Should I replace both electrolytic caps (I wish to replace them with solid state polymeric SMD caps) and this black one I don't know what type are they, seems that I've seen photos of them leaking to.
After I'm done with A5 board, I will go and check for HV and LV board and then recap A1 board.

The black rectangle box cap are tantalum, the ones that are affected are using common rounded smd electrolytic cap. Maybe a photo of your A5 will clarify that better.

Btw, congratulation on the score, really a great condition you got there. 

[e-pirate]

The black rectangle box cap are tantalum, the ones that are affected are using common rounded smd electrolytic cap. Maybe a photo of your A5 will clarify that better.

Well, unfortunately, I'm half a thousand kilometers away from my scope and I have no idea when I will return back. That what I'm doing now is preparing my road map for the feature, so I can do thins faster on the scope.
I always thought that tantalum caps are always enclosed in yellow plastic and black ones are just regular electrolytic of different type, but I wasn't right in that case. BTW, what does symbol "2" in a circle on the original black rectangle cap means? Kind of package size or value tolerance?

[e-pirate]

Actually, after googling a little bit and searching through SM a found that this a just regular general purposes tantalum 15 uF 25V 20% caps. And symbol "2" in circle is just a Vishay Sprague logo, nothing exciting here

[guido]

Had a go at U800; better safe than sorry.

I've had a look in my spare parts and found a heatsink that screws right onto one of the screws from U800. Used the one connected to the metal of it obviously. It is not as big as the examples in this thread, but i guess it is better than nothing.

It is from a 7A13 plugin; a beefy TO18 heatsink which has a screw mounting hole on the top. Fits like a glove. There is a bit less space than normal, as there are bodges around U800 which are not present in a more modern 2465 

Sorry, no picture. The battery of the camera is dead.

[BravoV]

Actually, after googling a little bit and searching through SM a found that this a just regular general purposes tantalum 15 uF 25V 20% caps. And symbol "2" in circle is just a Vishay Sprague logo, nothing exciting here

Yep, but they don't leak overtime and I think they should be pretty reliable.

Btw, for newly acquired 246x scope, suggesting to check these system voltage rails first at the J119 header when opening it for the 1st time.

Remember, get a big fan blowing the whole A1 board while its turned on without the enclosure to cool down those unobtainium hybrid ICs.

Of course ideally these rails should be measured with another scope instead of just ordinary DMM to see if the ripples are up to the specification.

The J119 header location and the voltages specification & tolerances required by the scope to work properly, quoted from Tektronix 2465B service manual.

[e-pirate]

Yep, but they don't leak overtime and I think they should be pretty reliable.
Btw, for newly acquired 246x scope, suggesting to check these system voltage rails first at the J119 header when opening it for the 1st time.
Remember, get a big fan blowing the whole A1 board while its turned on without the enclosure to cool down those unobtainium hybrid ICs.
Of course ideally these rails should be measured with another scope instead of just ordinary DMM to see if the ripples are up to the specification.
The J119 header location and the voltages specification & tolerances required by the scope to work properly, quoted from Tektronix 2465B service manual.

Well, after reading this particular tantalum caps datasheet I came to conclusion that there is no need to change them. They have no lifetime parameter specified at all, so, probably, as they intact, I will live them. But I'm going to replace aluminum "wet" electrolytic caps on A5 with solid state polymeric "dry" caps. The have longer lifetime, lower ESR and have nothing lo leak out. And I have to replace this pesky Dallas SRAM chip, but that's no so easy to get a fresh one here in Moscow and I have to get a good programmer before.
Concerning rails you told me about: I have another scope (just a cheepy chines Hantek DSO) and a DMM, so, I will take a look on this. Anyway, I'm going to completely recap the unit to make it stay alive for another couple decades, I hope. But I will need help and advices, just cause I had never my hands on anything complected like this scope before.

[e-pirate]

Everybody, who's going to replace DS1225 SRAM chip should take a look at Dr. Hugo Holden's document, concerning this procedure and possible chip replacement with FRAM and DRAM. Defiantly worth looking at http://www.worldphaco.com/uploads/TEKTRONIX_2465b_OSCILLOSCOPE_CALIBRATION___REPOWERING_THE_DS1225.pdf!

[BravoV]

Thanks for that document ! 

Yep, I bought few FRAM chips last year hoping I can throw away the DS-1225 (posted HERE), and the problem I have is my MiniPro TL866 universal programmer can not copy the SRAM content into this FM1608-120, since it has non standard CE signaling, so currently I'm stuck with this Dallas chip. 

[Bryan]

Everybody, who's going to replace DS1225 SRAM chip should take a look at Dr. Hugo Holden's document, concerning this procedure and possible chip replacement with FRAM and DRAM. Defiantly worth looking at http://www.worldphaco.com/uploads/TEKTRONIX_2465b_OSCILLOSCOPE_CALIBRATION___REPOWERING_THE_DS1225.pdf!

Great article. Thanks for sharing. Too bad the TL866 can not write to the FM-1608 though

[SeanB]

Turn off ID check and tell it it is a DS1225 and it probably will write to it.

[BravoV]

Turn off ID check and tell it it is a DS1225 and it probably will write to it.

Nope, it won't work, the problem is this FRAM needs the /CE (Chip Enable) line to be strobed at "EVERY" write or address change, while DS1225 is just like ordinary SRAM which only needs the /CE line to be pulled down "ONCE".

The reason for strobing at the CE on every write is because of the nature how FRAM works, this action will latch the new address and preparing the "pre-charge" while CE is HIGH which is needed for FRAM ferroelectric cell operation, CMIIW.