Topics

2465A CT Lessons and Leroy's Breadcrumbs


Brian Nordlund
 

I just resurrected a 2465A CT. I first want to express my appreciation for the threads here that were such a great help, especially Gregor's "noise on readout" adventure, I thought I would summarize my experience and list some important lessons. As has been noted elsewhere, the 2465A seems to be underrepresented in the boards, and the "CT" version is even less common.

Originally, it was not functional. However, there was some sort of noise on the CRT. I was happy to find that the 2465A uses the stand-alone memory battery, instead of the Dallas nightmares found in the later versions (and so many other models I have dealt with). The battery voltage measured 3.7V, so I left it alone, knowing it will eventually need to be replaced, but not for now.

Testing the power supply rails on J119, I found that the -5V and -8V were dead, so I dove into the power supply. I tested several bad electrolytics. Because of the work involved in extracting the power supply from the case, I decided to replace all electrolytics, the safety capacitors, and the large poly Sprague. After the reassembly, it powered up and appeared to be fully functional. However, the CRT seemed to be a bit out of focus. My rush to clean up the display led to some careless mistakes.

Early 2465A's (before serial number B014330), DO NOT have a CAL 08 (CRT Cal) menu item. Instead, the CRT is adjusted the old fashioned way, without any help from the control board. The procedure for those adjustments is at the very end of section 5 of the service manual.

If one (me) blindly hunts around in the menus on the 2465A CT, for the CRT (CAL08) procedure,they might see an entry called "CT CAL 81". It is important to understand that this is NOT a misprint of "CRT" just missing the "R"! By going into that item, the scope expects you to calibrate the timer settings for the high resolution "Counter Timer Trigger" option (so, I guess it is "CTT", but missing one "T").

If one finds themselves in a menu they don't want to be in, it is important to understand how to navigate with the front panel keys. It is even more important to not panic and shut off the power (yep, that was me). Worse yet, don't go back in to another calibration routine (CAL02, vertical cal) and do the same thing again.

After panicing, there was a "Fail 04 1x" error, and a "Fail CT CAL 81" error. These are indicators of a bad checksum in the battery-backed RAM configuration data, and also out-of-range configuration values. In many cases, these are interpreted as being an indication of a weak memory battery, but the battery still tested at 3.7V. My hope was, if I could successfully complete the two aborted calibration routines, I might get it back on its feet. My hope was that "Fail 04" could be fixed with a successful CAL02, and that the "Fail CT CAL 81" could be fixed by a successful "CT CAL 81", without having to perform a full calibration.

Lesson #1: The following Trigger buttons are how you navigate in the calibration routines:
1. "A/B-Menu" is like an "escape", used to exit the diagnostics / calibration routines
2. "Mode (Up)" and "Mode (Down)" step you through the menu items
3. "Coupling (Up)" is like an "enter" key, and causes the selected item to execute
4. "Coupling (Down)" is like a "backspace", used to back up and redo a step (if the step failed)

Lesson #2: The service manual is nice to have, but is missing some details that are important to know. I found the most help by searching for information here on TekScopes and on the EEVblog forum. Additional help was found on the wb0smx.net web page. The wb0smx page is specific to the 2465A, not the "CT", so it covers the main calibration routines, but not the "CT CAL 81" process. His observations and pictures are very helpful for verifying what the screen should look like before you proceed with "Coupling (Up)".

Lesson #3: If anyone has a working 2465 family scope, there is a simple way to "back up" the configuration data, found in the battery-backed RAM. After my experience, I would recommend that this be done on any working scope, before the cover is removed for any work, and to have on hand in case the memory battery does go dead, or if anything else goes wrong. There is not a clearcut way to restore that data, but having that data might lead to a way to restore operation if other methods fail.

I do not know what variables are associated with the RAM entries, but I think that there may be some interest in deciphering the data. Therefore, I have included a link to a table (below), which includes two sets of data:

https://groups.io/g/TekScopes/files?id=2620985&folder=0

1. The 2/21/21 column was read while both the "Fail 04" and "Fail CT" errors were still active.
2. The 2/25/21 column was read after successful calibration, so is without errors.

The entries in Red can be considered to be the "bad" entries, which were apparently fixed by the sucessful CAL02 and CT CAL 81 calibration procedures. There is a block of 26 entries, right at the beginning, that are all different. Then there are three smaller blocks later in the RAM.

I have read that others have not been able to clear even the Fail 04 1x error without completing ALL calibration routines in order. I can say that this was not true in my case. I only did the two calibration routines, and both of my errors were cleared. Therefore, a partial calibration can not only be saved, but can clear the errors. Since Fail 04 errors could possibly reflect a checksum error from any calibration data, the fact that I actually knew what data I had screwed up may have helped. I guess a point woudl be that "Fail 04" does not mean that CAL04 is invalid, it apparently means that any data could be corrupted. If one is dealing with a bad backup battery, causing much or all of the data to be bad, that may explain why a full calibration might be necessary much of the time.

Lesson #4: This scope seems to be far more forgiving in the calibration requirements than the later TDS models I have tried to "cowboy calibrate". Even with all the wrong calibration equipment, I was able to get it to accept the signals. Some steps did fail, by displaying "Level" instead of proceding to the next step. However, this is where pressing "Coupling (Down)" will back you up and let you "retry" the step. This can be done over and over, until it finally accepts it. I experienced this on the 10V step for the CAL 02 routine. To get the 10V level, I had to use an analog signal generator. Although I could verify and adjust for the 10V level, it may have not been stable enough to succeed. However, after three or four tries, it accepted it!

This scope, now that it is working, is shaping up to be a favorite of mine. It is new enough that it has great analog performance and very flexible triggering capability. However, it seems just old enough, that even I should be able to keep it running for years to come. My TDS digital models always scare me that they are going to suddenly lose calibration and be unrecoverable. I feel like I can use the 2465A as a "daily driver" and enojoy it.

Since this is the "CT" model, it has high-resolution time interval measurment, plus GPIB (woo-hoo!). It may not be state of the art, but it still impresses me. I know I am not the only person here that remembers how expensive this kind of equipment was in its day, and appreciate the value that a $20 purchase can bring today!

The "CT" also supports the P6407 "Word Recognizer Probe". Of course, I had to pick one up to try, and low and behold, it worked! For those who are not familiar with this feature, the Word Recognizer is a simplified logic analyzer trigger probe. It has 16 TTL inputs, an additional "Qualifier" bit, plus a Clock input. It allows digital events to be used as a sweep trigger, arm an analog trigger, etc.

I know I have dragged on too long, but I do have one more item that might be of interest. It has a Tektronix Property sticker, so at some point it was an internal asset. Once I got it up and running, I discovered that the Setup Memory is full, with 30 saved settings. What intrigues me is the Settings saved in location #1 is named "LEROY G". Position #5 is named "ROY". Other set names are more practical, such as "TIMEDLY" and "FUJITSU". This makes me very curious to know if there was a Tek engineer, named Leroy (Roy) G, who may have worked on Fujitsu gear. Does anyone here know Leroy?

Thanks for making it through my rambling,
Brian Nordlund


Thomas Voshell
 

Hey Brian,

I just went thru a similar excursion with my 2467B. Only mine was more fun: I had my 'scope on and the fan seized. I didn't know it and I went out to eat and when I returned, it was stone cold dead. I removed covers and found nothing on J119. PS was dead. I removed the ps card and eventually found a shorted shottky diode. I replaced it and the fan then the ps was working but no CRT activity. I had noticed the upper RHS of board A5 (processor) had corrosion on the mounting screw. Kinda like something caustic had got to it. I Started reading this forum and came across the discussion of failing capacitors on the A5 board. I also noticed that I had a Dallas Semi timekeeper RAM that was over 20 years old. I thought that this one is either getting ready to go bad or already is so I found a new one on DigiKey. I replaced it this time adding a socket to the A5. Still no CRT. Eventually I found a -1.25 reference signal which was developed by the DAC on A5. That signal was around +0.1V. many circuits were dependent on this -1.25 reference. After probing signals on the DAC I concluded the DAC is bad. I ordered in a couple more DACs and replaced the one on A5. No help. I found a flow chart that included setting CRTs grid bias to get an image. I was able to turn it and get the CRT to light up. Out of focus and bright but at least something was there. Still I had no -1.25. J504 is a test jumper that forces circuits downstream of the DAC to be -1.25. When I moved this jumper to the test position, I could re-adjust the CRT bias and see some characters! But things needed to see the DAC working and it wasn't. So, now I started probing analog voltages feeding the reference levels into the DAC. They were correct at the components but clearly wrong on the DAC pins. There were missing connections between surface mounted parts and the DAC! One of those leaky (electrolyte wise) caps was near the resistor network for the reference current into the DAC. I ohm'ed out the DAC pads to the resistors and NO Connections!. The traces looked ok but clearly several were not connected. The solder masking on the board looked ok but the copper underneath was eaten away by capacitor ooze. So new caps, & new 30 gauge blue wire. Finally I have an image and TEST 04 FAIL 02. Time to re-cal. It took me about a day each and about 3 iterations to get it right. Then I had a CT failure and it was a separate cal.

In the end: My scope had SM caps that were ready to die, explosively. My fan quit which probably turned to case into an oven this probably encouraged the caps to squirt out their juice and the shottky to short. The RAM was probably ready to go anyway but replacing it required a re-calibration. I am up and running now: This started just before XMAS.

Many thanks go to Chuck Harris who is also a contributor to this forum. Without his help I probably would have round file'ed this machine.


Brian Nordlund
 

Wow Thomas, that is certainly a rapid decline- going from a working unit one minute, to having multiple problems the next. Sounds like you were up to the challenge and stuck to it through the end- Good job!

Also, the link to my RAM file should be https://groups.io/g/TekScopes/files/Tek2465A_CT_RAM_022521.pdf (The one above doesn't work).

Thanks
Brian.


Brian Nordlund
 

I'll try the file one more time: https://groups.io/g/TekScopes/files/Tek2465A_CT_RAM_022521.pdf

Hopefully this works...


VK1GVC
 

Brian, that latest link works for me ... but do you have the data in a more user-friendly form for possible use by others, eg plain text, Excel or CSV?  The PDF is entirely legible but would need a lot of massaging to get it into a form digestible by EPROM programmers and PDF is not an ideal starting point - whereas data text or Excel is much easier to massage.  While I hope I never need it, could you possibly post again in one or more of those other formats?

Secondly, in your first post in this thread you said in Lesson 3 " ....there is a simple way to back up the configuration data ... " but you didn't actually say which method you used.  Is this the method where you step thru' the calibration data on the 'scope screen and take pictures or a video, or is there a better/simpler way?  I have a plain vanilla 2465, no GPIB.

Thanks in anticipation (from Down Under)

Graham

On 1/03/2021 3:28 pm, Brian Nordlund wrote:
I'll try the file one more time: https://groups.io/g/TekScopes/files/Tek2465A_CT_RAM_022521.pdf

Hopefully this works...




Thomas Voshell
 

The ram dump is certainly worthwhile. Getting it into excel is not bad: use a text editor that can do vertical selection and mark & copy columns.

Thanks,
--TomV


Brian Nordlund
 

Thomas & Graham-
Here is the data, in excel format. The corrupted data, from February 21st, is highlighted in Red for the values that differ from the data that is not corrupted (from February 25).

https://groups.io/g/TekScopes/files/Tek2465A_RAM_022521.xlsx

Graham-
I retrieved the data the hard way, by running EXER 02 and stepping through all the values while taking video with my cell phone. Later, I sat down with my cell phone paused, and scrubbed through the video, a couple seconds at a time, while I entered the data into the spreadsheet.

Thanks,
Brian.