Tektronix 2465B Square wave noise on active readout. #photo-notice


@Antonio90
 

Greetings,

I'm new to the group, and a beginner electronics hobbyist. Coming here looking for help repairing a Tek scope. I see most people have problems related to the power supply but, as far as I've been able to tell, it is working kind of fine.
It might have been repaired or refurbished already: It had a missing screw on the top metal shielding, and the caps seem new. No surface mount electrolytic either.

So, I recently bought a Tek 2465B for a very low price, it was sold as "good condition" and as working. I had the opportunity to test it for a brief time, but not knowing almost anything about the operation of this scope, I just checked the probe compensation square wave and fed a couple of waveforms from my AD2. It seemed to work fine, and apart from a slightly bent pot under the screen and a somewhat noisy fan, the apparent condition was good.

Then, testing it more thoroughly at home, I found a problem with the readout data and cursors, it is shown in the following video:

https://youtu.be/rFkmVqaTmHY

The problems are as follows:

1 - Readout intensity potentiometer in the "off" state: It works fine. Trace is stable, and the readings are more or less accurate.

2 - Pot turned clockwise, "scale factors on": The trace depicts a waveform that looks like a digital signal, an irregular square wave.

3 - Pot turned counter-clockwise "scale factors off":
- Measurement lines on: The same waveform as in point 2.
- Measurement lines off: A normal squarewave.

4 - Auto button pressed: Trace disappears on the upper side of the screen, readout and cursors activate. It can be moved down without problems, but the readout intensity has to be turned to "off" again to display the input waveform.

Some notes:
- This happens only to channels 1 and 2. Channels 3 and 4 work well with auto set up and everything else.
- Measured voltages are within specs according to my cheap DMMs (2 anengs and one WHA5000 something).
- Photos added in linked album.

Album: https://groups.io/g/TekScopes/album?id=264175

Also, don't know if it's relevant, but there is a jumper resistor and a small wire on U975 that goes under the ribbon connectors on the side of the scope.
I tried to look for U2810 and U2900, but the board in the service manual shows an earlier version that looked like an addon card, don't really know where it is on my scope.

If someone could help pointing in any direction to troubleshoot, I would be really grateful. Maybe this is a documented fault, but I haven't been able to find this reproduced in any of the forum threads or videos I looked at.

Thank you very much,
Antonio.


Jean-Paul
 

Hello: Bravo, what did it cost?

On turn on does it pass self test or indicate errors?

check the A5 control board on the side: Dallas 1227 NVRAM internal battery may be dead and loss of CAL constants.


SMD ELCO caps leak acid and corrode the PCB especially in the control area.

See many long threads on this forum and EEVBLOG, search 2465B A5, NVRMA, SMD caps.


Bon Chance,

Jon


Egge Siert
 

Hi Antonio,

Serialnumbers are the key to which Service Manuals are relevant. With Guernsey and Heerenveen Models I can sometimes help. Look for Production Dates on IC's and so on.

Greetings,

Egge Siert


@Antonio90
 

Thank you very much for your answers.

I *might* have found the problem. There is corrossion, gunk, and half-desintegrated components all around U650 (AMI 8843MBV 155-0244-01). I'm leaning towards external damage, IE water, some cleaning solution used around it, or something like that, because there are no electrolytics anywhere close, and the big filter ones would have leaked through the internal frame, which is pristine. Also, there is no corrosion or evaporation residue in the internal part of the PCB.

I will be making a list of the components that need replacing and looking out for equivalent replacements. In the meantime, any tips for cleaning? I read somewhere that a toothbrush and then isopropyl cleans it very well. Let's hope there's no pads or traces destroyed.

Also, what's the proper way of unsocketting ICs without destroying them? I'm guessing the AMI 8843MBV, could benefit from a little cleaning of the pins.

I'm posting some photos of the corrossion in the album, there are quite a few components damaged.

https://groups.io/g/TekScopes/album?id=264175

@Jean-Paul

It cost me 100€, which was a steal compared with the usual european prices for a working 2465B, not so much for a non properly working one I guess.
I want a bench scope for the lab, as it is not really kitted out yet, and this seemed like a good option. I have another non-working 466, will give a go repairing it when I finish with this one. It's a fascinating device really, never seen anything like it.

It self-tests OK at startup, and the rest of self-tests seem to be OK too, but I cannot see it in detail, as the readout screen is shifted up, I can only see the bottom line. Tried adjusting the pots for the position and gain, but the problem has to be elsewhere as I can only move it a little, also there aren't any SMD electrolytics, at least I haven't seen any, they are all through hole.

Will test the dallas ASAP, when I figure out what pins to measure for battery voltage.

@Egge Siert:

Are the production dates printed on the ICs? I will look up the IC markings anyway, to see If I can find production dates. It is a Guernsey-made model indeed, but I cannot find the scope's serial number anywhere.


@Antonio90
 

Sorry for posting again, don't know how to edit messages.

In the new photos added, there are a lot of 88's. So I'm guessing that's the year of manufacture of most of them. Maybe a 1990's or so scope?


Ozan
 

On Tue, May 18, 2021 at 07:36 AM, @Antonio90 wrote:


Sorry for posting again, don't know how to edit messages.

In the new photos added, there are a lot of 88's. So I'm guessing that's the
year of manufacture of most of them. Maybe a 1990's or so scope?
Usually two digit year, two digit week is marked. For example 8824 -> 24th week of 1988.
Ozan


@Antonio90
 

@Ozan
It's a 1988 scope then, or a little bit youger if the IC's are older than the scope.


Ozan
 

On Tue, May 18, 2021 at 07:06 AM, @Antonio90 wrote:

It is a Guernsey-made model
indeed, but I cannot find the scope's serial number anywhere.
Serial number location is between Ch2 and Ch3 BNC inputs. In your youtube video it is too grainy to tell but I think there is a serial number in that location on your scope too.
Ozan


@Antonio90
 

150140 it is then.
Thank you!


Egge Siert
 

Hi Antonio,

With many 1988 parts it is an early 2465B with 671-0722-00/05/07 as Main Board (A1) Partnumber. With this Partnumber you can look in the Service Manual (070-6863-01, SN B049999 & below) which Beaverton SN's are relevant for your Guernsey Model.

Greetings,

Egge Siert


Egge Siert
 

Hi again,

Just looking at the pictures in your Photo-Album. The Main Board has 671-0722-05 as PN and the A5 Controller Board is the new SMD-Version (1988 parts on it). Intriging mix. I have serviced several early 2465B's out of the same era and they all had the old style Controller Board with the separate A4 Read-Out Board.

Greetings,

Egge Siert


@Antonio90
 

I was just going to ask about that. Yeah, it is strange. And I don't think anyone swapped it, as all the smd ICs are from the 88s.


@Antonio90
 

Sorry if It's obvious, but is it normal to have around 42 ohms across the 5 & -5 volt rails and ground?
I was just checking continuity on the damaged area around U650, and the ceramic caps in the 5v power distribution, (in particular C675) appear as ~42R, with the resistance going up as I measure caps further away in the same rail. Also, pins 12 and 5 of J119 give out around 42R.


Ozan
 

On Thu, May 20, 2021 at 06:38 AM, @Antonio90 wrote:


Sorry if It's obvious, but is it normal to have around 42 ohms across the 5 &
-5 volt rails and ground?
I was just checking continuity on the damaged area around U650, and the
ceramic caps in the 5v power distribution, (in particular C675) appear as
~42R, with the resistance going up as I measure caps further away in the same
rail. Also, pins 12 and 5 of J119 give out around 42R.
I vs V transfer function of the supplies will be nonlinear because of many active devices on the supplies; however, measuring 42-ohm is not a concern by itself.

The resistance across the supplies should measure about the same on all points of the rail unless you are crossing one of the supply filtering inductors shown on sheet <11>.

Ozan


@Antonio90
 

Thank you very much Ozan. It worried me because the measured I vs V, taking into account the different voltage of the rails, was three times lower on the 5 and -5v rails. 42-46R stable, vs ~250-300R slowly increasing on the rest of the rails.
I also disconnected the power supply connectors on the backside of A1, and the measured Ohms were 50 on the 5V and 46 on the -5. A little bit higher, but still low. Maybe I'm missing a parallel resistor or voltage divider somewhere in the distribution rail.


Egge Siert
 

Hi to All,

Regarding my remark about the newer Controller Boards in early 2465B's. In my Files I read the following (out of a SMD-Board Description of Qservice):

" This board is backwards compatible with ALL SMD Versions having Part Numbers ending in -00, -01, -02 , -03 , -04, -05

It can also be used in older B Series Oscilloscopes to replace the older Discrete Component Controllers
and eliminate the need for a separate Readout Board as this board contains all readout circuitry,

Using it in any older serial number instruments (below B050000 ) is limited to standard instruments only.
Options fitted in low serial oscilloscopes will not work with new controllers. "

Greetings,

Egge Siert


@Antonio90
 

@Egge Siert

Thank you very much. I guess somebody replaced it. At least, the wima caps in the PSU and the tantalums in place of A5 electrolytics seem to point to this scope being serviced after it's production.

Sorry if I only post questions and not any progress or effort on my part, I'm trying to understand at least superficially the scope via the service manual, but I have very little time right now (new job, lots of extra hours to learn).
Anyway, I've been reading parts of the service manual, and provided that:
1. The scope works well as long as the readout is off, so the normal signal path is OK.
2. The "intruder" waveform varies normally in the X axis as the timebase and sweep vary, so it doesn't seem to effect the horizontal stuff.
3. The depicted amplitude is, visually, always the same. The volts/division knob doesn't have any effect, I always see the same 1/2 division amplitude in the CRT, so the noise is being injected like that directly, either to the Vertical Channel Switch or the Vertical Amplifier.
4. Channels 3 and 4 don't depict the alien waveform whether the readout is on or not, however, the readout values (timebase and volts/div) are still on top of screen instead of bottom.

The problem should be either between the readout and the vertical channel switch (Readout Vertical Out), or somewhere betwerr the VCS and the VA. Maybe there are more options, but that looked like a good place to start. So I have been probing around the READOUT VERT OUT line (figure 6) and, voilà, R480 (78,7KOhms,) between the 15V rail and the Diff Amp U475B failed open. I made a quick patch with a 80KOhm resistor to check, and powered it up just enough time to see the readout and cursors, and it seems to work well enough: depicts frequency, no strange waveforms, and it doesn't shift up out of screen when I click the "auto" button.

So, before I power it up again, I will buy some 0.1% resistors of the appropriate values to substitute R480 and, maybe, R490, and also list and replace all components affected by the liquid damage, or whatever it was.

I guess I will have to remove the A1 board. Will be reading the manual about that but, does anybody have any tip, video or post about the process? I suppose it is not a very good idea to do the component swap from the top.


@Antonio90
 

I've been reading the service manual to see the procedure for A1 removal. It doesn't look trivial at all. Also, the NVRAM has not died yet, but will die soon. I'm not confident enough on my skills to remove the A1 board without damaging anything, nor to stick stuff into the CRT to pull the connections out. I don't have a programmer for the Dallas nor decent desoldering equipment, and costs are going to add up.

So, I'm selling it. Will be listing it on ebay, unless someone here wants it.

Thank you everyone for your help.


satbeginner
 

What is your location please?
If you are in Spain maybe we can work together?

Leo


@Antonio90
 

I'm in Toledo, 1h from Madrid. I also travel sometimes to Tenerife and Valencia.