Author Topic: Tektronix 2465B oscilloscope teardown  (Read 455211 times)

0 Members and 2 Guests are viewing this topic.

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #450 on: April 03, 2015, 12:46:47 am »
malch,
    Yes, I am aware of that particular file. I downloaded it when I was not sure I was going to be able to save my calibration data from the DS1225 that was in my scope. Once I was able to recover my cal data, it became a moot point. The questions I was asking SoundTech-LG was to determine if he was able to salvage his cal data, and if not how he went about calibrating his scope.

I do not have access to the Tek cal fixtures. I have been studying all that I can find trying to determine what types (square wave, sine wave), amplitudes, and frequencies (1khz) I would need. To perform a vertical amp calibration on my scope. I have a Tektronix SG280 function generator that goes to 20mhz and up to 20v p-p that I can use to generate the square wave signal necessary to do some of the calibration, and a Boonton 102D that goes to 520mhz and up to 3v I can use for some of the other signals. I do not think my scope needs a full calibration, as the frequency measurements are dead on. The voltage measurements are where I am having some issues. When using the internal cal fixture, it is supposed to be exactly .4v p-p square wave. My scope measures .415v p-p. Not bad, but just a little out of spec. When I try to measure a sine wave signal, then things get a bit more interesting. Having a known RF signal of a know amplitude, the scope reads quite a bit higher than I think it should. About twice the expected value.  I still have a lot to learn about the operation of the scope. Might just be operator error. I'll figure it out eventually.

Mitch
 

Online MarkL

  • Supporter
  • ****
  • Posts: 1682
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #451 on: April 03, 2015, 02:18:16 am »
...
When I try to measure a sine wave signal, then things get a bit more interesting. Having a known RF signal of a know amplitude, the scope reads quite a bit higher than I think it should. About twice the expected value.  I still have a lot to learn about the operation of the scope. Might just be operator error. I'll figure it out eventually.

Mitch
RF and other high frequency sources are typically terminated in 50 ohms.  If it's not terminated (or loaded with a very high impedance like the scope 1M input), the amplitude will be double.

Are you using the 50ohm input coupling?
 

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #452 on: April 03, 2015, 03:54:37 am »
MarkL,
   Thanks for the reply. I tried it in both the 1meg and 50ohm inputs. At the moment, I do not remember what the differences were, it was way late and I should have been asleep. The next time I will put a 50ohm termination feed through as well as having the scope in 50 ohms to see if that makes things look better. Like I said, I still have a lot to learn about this particular scope. I do have another scope I have had for many years, but it is not nearly as much scope as the 2445B.

Mitch
 

Online MarkL

  • Supporter
  • ****
  • Posts: 1682
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #453 on: April 03, 2015, 04:22:32 am »
Hi Mitch,

You wouldn't need both the external 50 ohm and internal 50 ohm termination.  If everything is working and you did that, you would get an amplitude that's too low.

But your test is a good one since maybe your scope's internal 50 ohm termination isn't working.  You could also see how the other channel behaves.

 

Offline malch

  • Regular Contributor
  • *
  • Posts: 86
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #454 on: April 03, 2015, 04:16:02 pm »
I know my suggestions are sometimes off the wall, but they are search ideas.
Have you checked out w2aew's youtube    for ideas?
 and post #244 to #264 ?
 

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #455 on: April 03, 2015, 05:48:18 pm »
MarkL,
      I got a chance to play with the scope some more this morning. I fed the output of a Boonton 103D signal generator (.125mhz to 175mhz) set at 10mhz to a T connector which went to two different oscilloscopes. The 2445B and my Kenwood CS-5170. I bought the CS-5170 new in about 1996 or so. So I have a high degree of confidence in its accuracy. I set the 2445B to 50 ohms coupling, and used a 50 ohm feed through termination on the CS-5170 and compared both displays. Both scopes read exactly the same (within a couple of decimal points) in both frequency and amplitude. These readings also agreed with the expected output of the signal generator. So it looks like at least for these simple tests that the scope is working as expected. One test down and many more to go.

malch,
    I had seen that youtube clip, but had not watched all of it. Good general information on how to cal a scope.

So, using the info from post#252, I need a 1khz 50% duty cycle square wave signal at 20mv, 50mv, 0.1v, 0.2v, 0.5v, 1v, 10v. My CFG280 function generator has the ability to do all of that with the exception of the precision p-p voltages. For that, I can use my Amprobe AM-530 to measure the output voltages in RMS and multiply by 2 to get the actual p-p voltage. I believe my AM-530 will go that low. The spec sheet states the lowest range is 0-400mv. Not sure I will be able to set the lowest voltage with the level of precision necessary. I may be able to use a stepped attenuator I have (Jarrod AV-50), set the output voltage to a range that the AM-530 can read reliably, and use the stepped attenuator to drop the final voltage to a level that is appropriate. Of course, at these low levels, every connection can have a significant effect on the final input voltage. The challenges never end!

Mitch
 

Offline SoundTech-LG

  • Frequent Contributor
  • **
  • Posts: 527
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #456 on: April 03, 2015, 07:00:46 pm »
Unfortunately lost my CAL Data as I was placing the Dallas in the TL866. The external battery wire came unsoldered and touched another pin. Dallas is ok, just F'd the Data...  :palm:
I think I have all the gear together now to do the CAL. Just need to get up one morning and spend all day, and maybe the next twiddling it.
 

Offline malch

  • Regular Contributor
  • *
  • Posts: 86
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #457 on: April 03, 2015, 07:16:06 pm »
Be happy you're not doing a TDSXXX
 

Online MarkL

  • Supporter
  • ****
  • Posts: 1682
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #458 on: April 04, 2015, 01:40:29 pm »
MarkL,
      I got a chance to play with the scope some more this morning. I fed the output of a Boonton 103D signal generator (.125mhz to 175mhz) set at 10mhz to a T connector which went to two different oscilloscopes. The 2445B and my Kenwood CS-5170. I bought the CS-5170 new in about 1996 or so. So I have a high degree of confidence in its accuracy. I set the 2445B to 50 ohms coupling, and used a 50 ohm feed through termination on the CS-5170 and compared both displays. Both scopes read exactly the same (within a couple of decimal points) in both frequency and amplitude. These readings also agreed with the expected output of the signal generator. So it looks like at least for these simple tests that the scope is working as expected. One test down and many more to go.
If I understand your setup, you would actually have *two* 50 ohm terminators.  One in the Tek scope and one external on the other scope.  So that would be 25 ohms because they're in parallel.

I can't find a manual for it, but the 103D is undoubtedly a 50 ohm source and its output amplitude would be calibrated assuming a 50 ohm load.  So, it's not adding up that you're seeing exactly what you expect with 25 ohms loading.  And the two scopes are also in direct parallel with eachother, so they would always display the same amplitude whether they were terminated correctly or not (not counting effects from reflections).

I guess the 103D could also have its AGC pick-off at the RF output and it might compensate for a 25 ohm load, but I think that would be unusual.  Or maybe you've already calculated for 25 ohms and that's what you mean by "expected"?

Anyway, if still in question, there's a quick check for a bad terminator.  Just turn on the 50 ohm termination on the Tek scope and use a DMM to check for 50 ohms at the scope's input BNC.
 

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #459 on: April 04, 2015, 02:46:03 pm »
MarkL,
     You are correct in regards to the miss match. What I was trying to state is that the amplitudes and wave forms were the same on both scopes. Yes the amplitude was off, but I was trying to see if the 2445B responded the same as my CS-5170, which was the case. I'll round up a 50 ohm splitter and replace the  BNC T with it. Or just run the one 50 ohm termination and see what the scope shows. I'll have to pull down one of my Boonton 92C RF Millivolt Meters and measure the output so I have a good reference. I know the output meter on the 103D is not calibrated and is out by some amount. It was originally a 75 ohm unit. After inspecting the schematic, I discovered that the only difference between it and the 50 ohm version, is that it had a 50 ohm to 75 ohm unun in line with the output. I even have a new 50 ohm meter face that I need to install then do a cal on it.

Mitch

 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4966
  • Country: gb
Re: Tektronix 2465B oscilloscope teardown
« Reply #460 on: April 04, 2015, 09:30:20 pm »
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.

It was only while probing the underside of the caps during the fifth or sixth iteration of the fault finding flowchart (which also uses low voltage PSUs to diagnose but showed no fault) that I noticed that there was 30V on the supposed 5v unreg cap. This was after I deliberately disabled the over current protection that kept kicking in to have the lower voltage cap splutter itself unceremoniously when full AC was applied.

The symptoms are that on the low voltage tests no faults show, but on full AC the over current protection kicks in two or three times each second.

It made a two hour job into a three day event.

« Last Edit: April 04, 2015, 09:40:57 pm by Howardlong »
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4966
  • Country: gb
Re: Tektronix 2465B oscilloscope teardown
« Reply #461 on: April 04, 2015, 09:51:19 pm »
Looking back on it, I am wondering how I'd have figured it out without disabling the overcurrent protection and blowing the cap. Must've gone over that board checking parts visually and electrically dozens of times now.

 

Online MarkL

  • Supporter
  • ****
  • Posts: 1682
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #462 on: April 04, 2015, 10:56:58 pm »
MarkL,
     You are correct in regards to the miss match. What I was trying to state is that the amplitudes and wave forms were the same on both scopes. Yes the amplitude was off, but I was trying to see if the 2445B responded the same as my CS-5170, which was the case. I'll round up a 50 ohm splitter and replace the  BNC T with it. Or just run the one 50 ohm termination and see what the scope shows. I'll have to pull down one of my Boonton 92C RF Millivolt Meters and measure the output so I have a good reference. I know the output meter on the 103D is not calibrated and is out by some amount. It was originally a 75 ohm unit. After inspecting the schematic, I discovered that the only difference between it and the 50 ohm version, is that it had a 50 ohm to 75 ohm unun in line with the output. I even have a new 50 ohm meter face that I need to install then do a cal on it.

Mitch
Hi Mitch,

Sounds like the 103D needs a little work.

My only concern was from a couple of posts back when you said you had double the amplitude when you didn't expect it.  The first thing that pops into my mind when I hear that is an unterminated connection.

I would just double check your internal 50 ohm termination on your 2445 channels with a DMM to make sure they're ok.  You might not even need internal termination for your calibration procedure, but you'll want to know at some point if they're working or not.
 

Offline malch

  • Regular Contributor
  • *
  • Posts: 86
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #463 on: April 04, 2015, 11:30:24 pm »
Howardlong This is my board layout. Is it the same as yours?
C1116 is at 8B at the bottom of A3 board and C1132 is at 7.5C
« Last Edit: April 04, 2015, 11:40:23 pm by malch »
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4966
  • Country: gb
Re: Tektronix 2465B oscilloscope teardown
« Reply #464 on: April 05, 2015, 12:15:29 am »
Yes. That has the same error.

If you change the caps one at a time without referring to the layout diagram it isn't a problem of course. However if, like me, you removed all the caps first and then repopulated them using the layout diagram, you'll have the same problem :-(

C1114 and C1115 are the same value and voltage rating, but show as different sizes on the layout diagram if you use that as your guide.

If you look a picture such as this one from http://bradthx.blogspot.co.uk/2014/01/tektronix-2445b-capacitor-replacement.html before replacing the caps, you'll also see that C1114 and C1115 look different sizes if you use the layout diagram to identify them.

 

Offline malch

  • Regular Contributor
  • *
  • Posts: 86
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #465 on: April 05, 2015, 12:34:19 am »
In this repair the "before" pic has C1114 and C1115 different, but the "after" pic he replaced them with the same.
 

Offline malch

  • Regular Contributor
  • *
  • Posts: 86
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #466 on: April 05, 2015, 02:35:11 am »
OK I couldnt stand it anymore. I took my 2445B apart and this is what I found.

C1113 and C1116 are 180uf/40v
C1114 and C1132 are 250uf/20v
C1120 and C1130 are  10uf/100v
C1115 is 10uf/160v

C1132 and C1115 are flipped in the service manual.
 

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #467 on: April 05, 2015, 06:35:18 am »

[/quote]
Hi Mitch,

Sounds like the 103D needs a little work.

My only concern was from a couple of posts back when you said you had double the amplitude when you didn't expect it.  The first thing that pops into my mind when I hear that is an unterminated connection.

I would just double check your internal 50 ohm termination on your 2445 channels with a DMM to make sure they're ok.  You might not even need internal termination for your calibration procedure, but you'll want to know at some point if they're working or not.
[/quote]

MarkL,
     You read correctly. I did some more checking, and realized I was sending a 50 ohm RF sine wave (10mhz) signal into the 1meg ohm scope input. That had things all messed up. I did some more testing this morning with the function generator set to 1khz square wave, and sampling the output with my Amprobe AM530 DMM, (which is supposed to be a true RMS meter). This signal was fed into the scope set to 50 ohm coupling. I expected to see about twice on the scope display as what I was seeing on the DMM, which is roughly what I saw. Actually the scope reading was a bit more than twice the DMM reading. Both of my scopes read very close to each other, so I think the scope reading are good, but the DMM reading may not be as accurate as I would like. Now the frequency readout on the scope appears to be very accurate. Tomorrow, I'll verify the accuracy by feeding the output of a rubidium frequency standard I have. That frequency has been verified by a frequency counter calibrated against a GPS disciplined clock a friend has. I expect the scope to be dead on from what I have seen so far. As for the amplitude readings, I'm sure its something I am doing wrong. Just need to get my head out of where the sun doesn't shine and figure it out.

Yes, the 103D does need a bit of attention. It, the 102C, the 102D, as well as most of the other pieces of equipment I have around here. That's what happens when you play with old equipment. It may not be as fancy as the new stuff, but it does work, just needs a bit more TLC. Gives me something to do in my copious amounts of free time. Spent most of today re-plumbing the toilet in the guest bathroom to make the domestic supervisor happy. I'm getting way too old to be crawling around under the house for long hours. 

Mitch
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4966
  • Country: gb
Re: Tektronix 2465B oscilloscope teardown
« Reply #468 on: April 05, 2015, 07:25:09 am »
OK I couldnt stand it anymore. I took my 2445B apart and this is what I found.

C1113 and C1116 are 180uf/40v
C1114 and C1132 are 250uf/20v
C1120 and C1130 are  10uf/100v
C1115 is 10uf/160v

C1132 and C1115 are flipped in the service manual.

While I am delighted that I am not completely crazy, I really was questioning myself at times because surely this must be a known error? Perhaps it is, but I have not been able to unearth any references to it. I am sorry, but at the same time very appreciative, that you went to that trouble, you have to do a fair bit of disassembly to expose those caps.

When it finally came to putting mine back together after a few days being dismantled and a nasty dose of man flu later, it took me about four attempts to reassemble it and not have any spare screws/brackets/shields left over! After all, I had only anticipated a couple of hours for this repair.

I would imagine then that this affects 2445B, 2455B, 2465B and 2467B, all of which share that same inverter board, although I searched online and couldn't find board layout diagrams specific to all those models, I only have the service manual (including layouts etc) for the 2465B and 2467B.

Thank you once again for confirming my sanity, it is appreciated.

Edit: I guess we should be clear that they are only flipped in the board layout diagram, fig 10-13. The schematic and BOM are correct.
« Last Edit: April 05, 2015, 07:29:39 am by Howardlong »
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 19359
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Tektronix 2465B oscilloscope teardown
« Reply #469 on: April 05, 2015, 07:47:05 am »
OK I couldnt stand it anymore. I took my 2445B apart and this is what I found.

C1113 and C1116 are 180uf/40v
C1114 and C1132 are 250uf/20v
C1120 and C1130 are  10uf/100v
C1115 is 10uf/160v

C1132 and C1115 are flipped in the service manual.
I would imagine then that this affects 2445B, 2455B, 2465B and 2467B, all of which share that same inverter board, although I searched online and couldn't find board layout diagrams specific to all those models, I only have the service manual (including layouts etc) for the 2465B and 2467B.

Edit: I guess we should be clear that they are only flipped in the board layout diagram, fig 10-13. The schematic and BOM are correct.
Everybody take note, make a note and stick it to your scope. Ohh shit it rhymes.  :palm:
Avid Rabid Hobbyist
 

Offline malch

  • Regular Contributor
  • *
  • Posts: 86
  • Country: ca
Re: Tektronix 2465B oscilloscope teardown
« Reply #470 on: April 05, 2015, 01:49:07 pm »
Going through the schematic now. So the parts layout fig 10-13 should be remarked C1132 to C1115 and C1115 to C1132.


mskobier has a proper 2445B service manual. It would be handy if that manual proved to be incorrect.

I changed the fan solder point to a feed-through connector with plugs and sockets, so I can remove either the fan or LVPS without a soldering iron.
I cut out the fan punched-in area that interferes with the LVPS cards, so the cards come out in a minute.
« Last Edit: April 05, 2015, 02:41:42 pm by malch »
 

Online MarkL

  • Supporter
  • ****
  • Posts: 1682
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #471 on: April 05, 2015, 04:06:44 pm »
MarkL,
     You read correctly. I did some more checking, and realized I was sending a 50 ohm RF sine wave (10mhz) signal into the 1meg ohm scope input. That had things all messed up. I did some more testing this morning with the function generator set to 1khz square wave, and sampling the output with my Amprobe AM530 DMM, (which is supposed to be a true RMS meter). This signal was fed into the scope set to 50 ohm coupling. I expected to see about twice on the scope display as what I was seeing on the DMM, which is roughly what I saw. Actually the scope reading was a bit more than twice the DMM reading. Both of my scopes read very close to each other, so I think the scope reading are good, but the DMM reading may not be as accurate as I would like.
...
I think you're right.  The Amprobe AM530 spec sheet says its frequency range is only to 400Hz and it's unclear if that's a 400Hz sine wave 3dB point or the meter is good for measuring some number of harmonics past 400Hz if you're working with 400Hz AC line current.  Since you're using it to measure 1kHz, you can be suspicious of whatever reading it's giving you.  You could try 50Hz and see if they agree more closely.

The scope is going to have a better frequency response than the meter, so the fact the scope is higher doesn't surprise me. 

The 2445 has a basic amplitude accuracy of +/-2% for a 4 to 5 division signal display, and goes to +/-3% with the 50 ohm coupling on.


EDIT:  Oops "sine wave", not "since wave"....
« Last Edit: April 05, 2015, 05:22:12 pm by MarkL »
 

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #472 on: April 05, 2015, 04:08:48 pm »
All,
  My scope is currently fully assembled, so I could not look at the inverter board. However, I did consult my 2445B manual (also covers 2455B) and saw the same component size mismatch as shown in previous posts. I also have a copy of the service manuals for the 2445 and the 2465. Those have the diagrams with the physical size correct, but positioned slightly different. According to all three manuals I have, the parts list C1114 and C1115 as the same value. When looking at the 2445 board layout, the physical size of both C1114 and C1115 are the same, with C1132 being physically smaller. According to what I see in my manuals, C1132 is supposed to be 10uf@160v, C1115 is supposed to be 250uf@20v.

Mitch
 

Offline mskobier

  • Regular Contributor
  • *
  • Posts: 97
  • Country: us
  • Test Equipmentaholic!
Re: Tektronix 2465B oscilloscope teardown
« Reply #473 on: April 05, 2015, 04:15:05 pm »
MarkL,
      I knew it was something simple. For some reason  I was thinking the Am-530 went to 400khz, not 400hz. That would explain the discrepancies I have been seeing. I can try a lower frequency, or better yet, get a DMM that can handle higher frequencies. Any reason to get more test gear!!!

I need to go review the posts in the forum on DMM's.

Mitch
 

Online MarkL

  • Supporter
  • ****
  • Posts: 1682
  • Country: us
Re: Tektronix 2465B oscilloscope teardown
« Reply #474 on: April 05, 2015, 05:39:49 pm »
MarkL,
      I knew it was something simple. For some reason  I was thinking the Am-530 went to 400khz, not 400hz. That would explain the discrepancies I have been seeing. I can try a lower frequency, or better yet, get a DMM that can handle higher frequencies. Any reason to get more test gear!!!

I need to go review the posts in the forum on DMM's.

Mitch
Most DMMs don't go very far in RMS frequency.  Some of the best top out at a few hundred kHz.  And even then, the accuracy can be quite poor at the upper end.

Then you start getting into the realm of RF power meters, which can sink a pile of money if you like to buy test equipment.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf