Tektronix 2465B oscilloscope teardown

[GerryBags]

I carried out the SMD cap replacement on the A5 of my 2467B (same as 2465B) as this one didn't appear to be in bad condition at all. As pointed out earlier by someone, appearances can be deceptive! The two uppermost (closest to the camera), a 10uF and a 33UF had both eaten their pads. For the 10uF (on the left in the pic) had only left half a via to solder to, hence the monstrosity you see extending past the via. Getting a pair of lock tweezers on that extension was the only way I could sink it to stop it moving when the cap went on top. It works anyway, the last shot is of the 'scope after I put it back together.

I tried to desolder the first cap, but there is not enough access to the negative pad, until you cut away the plastic seat, so for the other three I cut the plastic seat, snipped off the cap away from its pads, then desoldered the pads. By the last cap I'd gotten the hang of it, isn't that always the way? 

[GerryBags]

It seems that I have issues with U500, the A/B trigger hybrid. The readout got fainter as I put the 'scope through its paces, and I found that the A/B trigger button wasn't working properly.

Now I get no readout at all apart from a brief flash on start-up telling me it has failed "Test 05  40", which is the main board, a positive peak that is too positive, possibly a problem with the line trigger, but definitely something to do with U500. The readout issue may be related, as the trigger hold-off time or time spent waiting for a trigger event is used (2nd priority, I think) to display the readout, or completely separate. The main board is implicated in the readout troubleshooting chart at one point, but I can't get to that point without a scope to check for a 425mV sine wave. I'll let everyone know what I find, as all this is applicable to the 2465B.

[med6753]

In order to expand and preserve the knowledge base on the 24XX series in one place without attempting to cross post please refer to the Test Equipment Anonymous (TEA) thread pages 422 to 424. My saga of scoring a pristine 2465 for the pauper's sum of $60 USD.   

[med6753]

It seems that I have issues with U500, the A/B trigger hybrid. The readout got fainter as I put the 'scope through its paces, and I found that the A/B trigger button wasn't working properly.

Now I get no readout at all apart from a brief flash on start-up telling me it has failed "Test 05  40", which is the main board, a positive peak that is too positive, possibly a problem with the line trigger, but definitely something to do with U500. The readout issue may be related, as the trigger hold-off time or time spent waiting for a trigger event is used (2nd priority, I think) to display the readout, or completely separate. The main board is implicated in the readout troubleshooting chart at one point, but I can't get to that point without a scope to check for a 425mV sine wave. I'll let everyone know what I find, as all this is applicable to the 2465B.

If U500 took a dump you may be SOL finding one. But don't they plug onto the board? I think they do and I thought I read somewhere that they can develop contact resistance. Try reseating it and see what happens. It's worth a shot.

[GerryBags]

I've seen a couple for sale for around $40-60, but i shall make sure it is that first. I'll definitely try cleaning the contacts (Thanks!), and I also have a 2465B that I can swap the hybrids with to see if there is any change. I'm also going to check J119's voltages, to see what is too positive and whether the +/-5 V rail that supplies U500 and the 10 V reference are within tolerance. I'm not too hopeful that it will be as simple as cleaning contacts, the way it got got worse over a couple of hours. No readout or display at all by the end. I'll open her up again once I've had some sleep and see what I can find.

[tggzzz]

In order to expand and preserve the knowledge base on the 24XX series in one place without attempting to cross post please refer to the Test Equipment Anonymous (TEA) thread pages 422 to 424. My saga of scoring a pristine 2465 for the pauper's sum of $60 USD. 

Could you provide links to the specific start/end messages, please. That thread only has 213 pages in it - see screenshot.

[tautech]

In order to expand and preserve the knowledge base on the 24XX series in one place without attempting to cross post please refer to the Test Equipment Anonymous (TEA) thread pages 422 to 424. My saga of scoring a pristine 2465 for the pauper's sum of $60 USD. 

Could you provide links to the specific start/end messages, please. That thread only has 213 pages in it - see screenshot.

Entirely depends of your forum view settings for posts/page. 
For me TEA is 400+ pages.

[BravoV]

Just mention the post number instead of page number.

Still not fool proof though, if someone deleted their post before and screw up the sequence number.

[med6753]

Just mention the post number instead of page number.

Still not fool proof though, if someone deleted their post before and screw up the sequence number.

Oops, sorry...didn't think about some having different screen views. It starts at post 10542 and ends at post 10594.

[cheeseit]

It seems that I have issues with U500, the A/B trigger hybrid. The readout got fainter as I put the 'scope through its paces, and I found that the A/B trigger button wasn't working properly.

Did you run the scope for extended periods without a fan keeping everything cool while the case was off? If so that may explain any issues with the hybrids but hopefully it's just a bad contact.

[GerryBags]

Did you run the scope for extended periods without a fan keeping everything cool while the case was off? If so that may explain any issues with the hybrids but hopefully it's just a bad contact.

Hi Cheesit, no I didn't run it all with the cover off, until just now to check the voltages on J119, because I'd been warned about hot hybrids, U800 in particular with its poor thermal connection to the board. The voltages are all well within spec, with the 10V reference being spot on. Given the "positive level too positive" note on a 4X fail I was thinking it might be something I'd done in the PSU, but I'm also not getting more than a few mV ACV reading on the the test points (my only way of checking the ripple without a working scope).

Next step is to swap out the U500 for one that I don't even know is ok. Not ideal, as the other 'scope is show the same LED code on the front! I'm really hoping that the blown caps and corrosion on the A5 board around the DAC ref is the cause of that one, but for all I know that U500 could be shagged too. This is going to cost me £40-£50 whether it works or not.... at least if it works I'll know, won't I? 

Edit to add: No joy. Swapped out the U500 hybrids and get exactly the same behavior as before. The A/B Trig button won't clear the error state (add LED illuminated) immediately, it takes a couple of goes to do it which, coupled with the failure to switch to the B Trig menu, makes me think it is an issue with the control board, as it involves the readout, display and the A/B trigger. 

Another edit to add: I think I've narrowed it down to the area around the DAC on A5, the affected electro caps, and the ones near it are on the +/-15 V rails for the DAC and a load of op amps in the analog control section, and also one on the Vref for the DAC. It seems that most of the problems others have had with similar symptoms have all traced back there, and the fact I had to bodge up pads for one of the caps shows that the corrosion was worse than it looked on the surface (this was the "OK" board out of the two!  ) and I think that the slow onset was down to a corroded trace opening once it had warmed up. At least now I'm pretty sure where I am in the schematics (1, 2 & 12) and what's doing what so I can start probing and see what I can find.

[santosp]

Hello.
Does anyone have available a copy of the eproms for the 2465 oscilloscope?  I need specific the 160-162..-06 version, as to verify with mine originals.

Thank you.

[santosp]

Just for the history this is the firmware dump from my scope. 

[emece67]

.

[tggzzz]

Just bought this really nice 2465B, in excellent cosmetic condition and with only 750 working hours. It works superb. Being in such good cosmetic shape and working so flawlessly, I'm a bit reluctant to open it for a recap

If it is working, there are three principal issues: the RIFA mains filter caps, the three SMD electrolytic caps on the board containing the NVRAM, and the general electrolytic caps in the PSU.

Particularly on 240V, I'd always replace the RIFA caps. One detonated on my 2465, and the series resistor removed a 1cm diameter of the PCB's prepreg (and spread carbon elsewhere). This requires normal desoldering skill.

I'd replace the three SMD caps before they leak and mess up other components - as happened on my 2445B. This requires some skill with removing/replacing individual SMD components.

I wouldn't replace the electrolytic caps in the PSU unless you measure excessive ripple on any of the supplies. (OTOH, I have done just that on my 2465 Inconsistent? Me?)

and cal RAM refreshment, what do you think about postponing such maintenance? Serial is 134xxx.

I've recorded a video with the cal data thru EXER 02, is this enough to recover the cal data in an eventual backup battery drain?

My attitude is to record the data and replace the NVRAM iff it fails. But I haven't verified that is sufficient strategy.

[MarkL]

...
I've recorded a video with the cal data thru EXER 02, is this enough to recover the cal data in an eventual backup battery drain?

My attitude is to record the data and replace the NVRAM iff it fails. But I haven't verified that is sufficient strategy.

The EXER 02 dump is sufficient to restore the cal constants.  It was verified by someone on the Yahoo groups.io Tekscopes group a couple of months ago.  What EXER 02 doesn't do, however, is capture the various vertical, sweep, and other stored settings.  So you would lose any presets you've saved, but that's usually not of any consequence.


EDIT: I keep forgetting Tekscopes was moved from Yahoo to groups.io about 6 months ago: https://groups.io/g/TekScopes

[cheeseit]

I very much agree with what tggzzz suggests but would like to add that the PSU in one of my 2465B's had a few caps that were starting to leak electrolyte. It wasn't critical, nor visible before I pulled the caps, but there were a bit of corrosion. I don't think PSU caps leaking is a common problem though.

Neither have I performed the procedure but one or more in this thread verified that restoring cal data from a recording was working so you should be safe. On the other hand, if you do work on the A5 and have the board out, you might as well get it over with IMHO. Then you're safe for a dozen or more years and don't risk having it die some day when you need to use it.

[med6753]

About a month ago I picked up this 2465 (sitting on top my 2465 DMS). It is in excellent condition and it appears to have low usage/hours. (The 2465 doesn't have EXER5 so can't verify the hours). I verified the supply voltages to be in spec. But despite all that I've decided to re-cap the Inverter/Regulator boards anyway. The caps are on order from Mouser and should arrive tomorrow. Next month I have to take a week furlough from work that's when I'll perform the surgery.     

[emece67]

.

[med6753]

Thanks to all that answered my question.

So, well, I'll open it and perform the update, focusing on the RIFAs, PSU and the three electrolytic ones. Recently I updated the battery backed Dallas RAM in some hp54600 scopes (that was easy. All they were suffering from the vertical deflection cap syndrome), so I will also try to update it. Once opened there's no big reason to not to try a more ample overhaul, maybe including some heatsink for the U800. I'm wondering if Berillyum Copper may be a good choice for such custom made heatsink.

Thanks again and regards.

Buried in this thread you'll find the chart that lists all the caps you'll need for the Inverter/Regulator boards. But for your convenience I'll post it again. It originally came from the Tek Yahoo Group. I used it successfully 2 years ago to re-cap the 2465 DMS.

As I mentioned previously I will be re-capping the PSU on this recently purchased 2465. I plan on doing it the week of July 8-14 and I will do a full pictorial of the process. So if you can wait until that time it will give you a better idea on completing the job successfully. There are pitfalls to watch out for....like mistakes in the service manual...which can lead you to a dead instrument and magic smoke. I've been down that path before so I know how to avoid these issues.

U800. Be very cautious here. Your 2465B has a Maxium chip which is known to have issues. My take on it is that it's been in there 30 years, leave it alone. Don't operate the scope for long periods without the case installed and make sure the fan operates properly. But if want to install a heatsink make it one that just epoxies to the top of the chip. DO NOT under any circumstances mess with the screws or nuts that secure the chip to the board. Chances are you'll cause uneven torque and the chip will fail. And also be aware that those screws are not at ground potential.         

[tggzzz]

Once opened there's no big reason to not to try a more ample overhaul, maybe including some heatsink for the U800.

The U800 in my 2465 is cool to the touch. There's no way I will fool with it, given the reports I've seen about how easy it can be to introduce problems.

"If it works, leave it alone".

As for which capacitors to use, there's currently a long discussion about it on the TekScopes group - including a couple of lists of capacitors.

[emece67]

.

[vik]

Hi all,

Thanks for everyone who contributed to this thread, I have found it invaluable. I owned a 2465A for nearly 10 years but lately I've started to worry about the 30+ years old battery dying one day, taking my calibration data with it. After reading the part of this thread about backing up the memory using exerciser 02 or GPIB, I have built my own cheap GPIB adapter and backed up the data. Please allow me to share some findings.

First, a really cheap GPIB interface can be built from an Arduino Mega (or possibly an Uno even), a pack of jumper leads, a 26 pin ribbon cable cut in half and a rather plasticy but functional 26 pin Centronics to ribbon cable connector from Ebay / Hong Kong. I used this library: https://github.com/mathiashelsen/agipibi. I have forked it and added another Python program that lets you send commands directly:  https://github.com/viktorradnai/agipibi/blob/master/python/gpib_console.py

I have also accidentally discovered that the 'earom? f' command prints the entire contents of the memory at once:

vik@cygnus:~/priv/projects/arduino/gpib/agipibi/python$ ./gpib_console.py
2018-06-20 00:21:47,795 WARNING No reponse to ping, you should reset the board
2018-06-20 00:21:47,795 INFO Get instrument ID
2018-06-20 00:21:48,044 INFO ID: ID TEK/2465A,V81.1,SYS:FV06,BB:FV01,GPIB:FV02;

agipibi> key 0
agipibi> earom? f
EAR 0:11,1750,9930,1717,9906,9954,10102,10039,1822,9963,1718,647,8837,8838,644,8855,8828,648,8842,651,8868,97,31,31,8225,32,15820,7198,7459,15645,16295,16290,8111,16290,16310,8759,8759,565,570,7907,15667,7474,8192,16328,8099,8105,8111,8138,744,747,743,8933,2144,2134,10331,10314,10301,10328,10301,8298,2140,10334,10324,10310,2137,2108,110,842,823,837,1648,1627,1700,9834,696,8915,6238,11894,11795,1828,1819,566,583,302,8492,2042,10238,1435,9028,9025,830,9018,827,827,8883,8883,9935,1359,9018,8897,707,827,832,9020,832,8903,713,8885,8381,205,173,65312,65280,65298,254,631,511,32893,65282,65344,767,2303,63752,65344,32512,65280,16895,8703,255,255,65408,65280,47872,65280,255,255,65280,57120,255,254,255,255,65280,65282,48896,61328,222,2265,32991,32895,57104,65280,32963,39135,63232,32008,57088,64832,16831,229,254,255,65280,65280,65280,48896,255,255,65296,64256,9610,255,255,239,65280,65296,48916,47969,4319,33014,763,253,49026,65296,239,255,29440,65280,61184,65296,32891,239,255,191,65280,65280,65280,65280,255,255,65280,65280,255,33023,255,255,65288,65352,64802,65296,4351,16503,4792,16575,65280,65296,8959,255,65280,30464,65024,63360,9182,6399,511,255,65280,65280,65280,65280,255,255,65280,65280,255,127,8447,255,65280,65281,63232,32672,127,6399,127,16582,61188,65288,4351,37119,13104,32256,5376,40704,8005
agipibi> ^C
vik@cygnus:~/priv/projects/arduino/gpib/agipibi/python$

It also seems like the 2465A only has entries from 0-254. 255 is zero and from 256 onwards the values wrap around. Is that enough to back up all of my calibration data?

I don't feel like experimenting with restoring the settings, but I'm happy to write another script that does this over GPIB in one go if someone is willing to test.

I have looked around in my scope and it was surprisingly clean and dust free. I did not see many electrolytic caps and the couple I've found seemed OK. Is there anything else I would be advised to do to future-proof the scope? The guides I've seen about cap replacement, etc, were mostly for the 2465B.

Regards,
Vik

[emece67]

.

[MarkL]

...
It also seems like the 2465A only has entries from 0-254. 255 is zero and from 256 onwards the values wrap around. Is that enough to back up all of my calibration data?

I don't feel like experimenting with restoring the settings, but I'm happy to write another script that does this over GPIB in one go if someone is willing to test.
...

The 256 values should match what you see in EXER 02, in which case, yes, it has all of your calibration data.

Putting the calibration data back via GPIB works on a 2465(plain), but I haven't heard of anyone trying it yet on an A or B.  There's no reason to think it would operate differently, but until someone tries it I guess we don't know for sure.  I've been watching for an A or B junker on ebay for some time now to do this experiment.

Nice find on the "f" option to the EAROM command, thanks!