Tektronix 2465B oscilloscope teardown

[JSC]

I like the 1502. In fact, I like it better than the "improved" 1503 with the LCD screen. The trace was sharper and quicker responding.

A coworker of mine flew RC controlled aircraft. He had a hi-tech Ni-Cad battery charger that could resurrect a bad TEK battery pak about 2/3s of the time by subjecting it repeated charge-discharge cycles. Took about a week each.

I never knew about the 1 ma drain. I always thought the main problem was the age of the batteries and the high temperature. The only places I ever used a 1502 were in Ft. Bliss, TX (gets sort of warmish), Saudi Arabia, and Bahrain (really warmish).

John

[tggzzz]

I like the 1502. In fact, I like it better than the "improved" 1503 with the LCD screen. The trace was sharper and quicker responding.

A coworker of mine flew RC controlled aircraft. He had a hi-tech Ni-Cad battery charger that could resurrect a bad TEK battery pak about 2/3s of the time by subjecting it repeated charge-discharge cycles. Took about a week each.

I never knew about the 1 ma drain. I always thought the main problem was the age of the batteries and the high temperature. The only places I ever used a 1502 were in Ft. Bliss, TX (gets sort of warmish), Saudi Arabia, and Bahrain (really warmish).

The manuals state "recharge every month", but don't explain why that is necessary

The chart recorder is permanently connected to the battery, and is responsible for a lot of the current drain. I can't remember what happens with the XY-output option or an empty slot, but the battery is not disconnected and there is still some drain.

[JSC]

Everyone always says, "RTFM.", but alas, hardly anybody does.

John

[AMR Labs]

I like the 1502. In fact, I like it better than the "improved" 1503 with the LCD screen. The trace was sharper and quicker responding.

A coworker of mine flew RC controlled aircraft. He had a hi-tech Ni-Cad battery charger that could resurrect a bad TEK battery pak about 2/3s of the time by subjecting it repeated charge-discharge cycles. Took about a week each.

I never knew about the 1 ma drain. I always thought the main problem was the age of the batteries and the high temperature. The only places I ever used a 1502 were in Ft. Bliss, TX (gets sort of warmish), Saudi Arabia, and Bahrain (really warmish).

The manuals state "recharge every month", but don't explain why that is necessary

The chart recorder is permanently connected to the battery, and is responsible for a lot of the current drain. I can't remember what happens with the XY-output option or an empty slot, but the battery is not disconnected and there is still some drain.

Guys, just a friendly suggestion for the benefit of everyone on this long and distinguished thread about Tek 2465 scopes, and try not to wander off too far to other unrelated things like the 1502 scope or RC controlled aircraft, etc. Suggest it would be best to open a separate new thread and move the discussion there of those other topics. Thank you.

[JSC]

Sorry. I have been a moderator. I should have known better. 

[jonpaul]

Please start a NEW thread eg "Tektronix 1502/1503 TDR".

j

[Scopetechniques]

[IIRC, that is one of a number of mods performed by well-known eBay vendor 2465b on all his "Refurbished/Guaranteed" 24xx scopes.]

I was searching online for images of a 2465B mainboard and saw this image, which I know to be my work.
My ebay ID is "scopetechniques" - same as EEVBLOG id.  The quoted posting attributed this work to ebay seller "2465B" and that is not true.

On the original question, I believe the black plastic piece stuck to the bottom of the scope was meant to protect the scope case when it was installed on a cart.

--Victor

[jonpaul]

risky to touch U800 HOR hybrid. Heatsink not needed and can short to case.

j

[AMR Labs]

Not worth the risk of messing with U800, besides no additional heatsink has ever been necessary.

[Scopetechniques]

No risk if you have the proper equipment, which I do.
Although I do agree with the limited value of the heatsink.
Unfortunately customers do research on the web and some want the heatsink.

In my research U800 runs at 112F (at 75F ambient), with the heatsink installed it runs at 102F.

Also, I have modified the attachment method for the heatsink. I install it below the U800 tab, so the high points of the whole assembly are the two studs.
So if there is a short to the case it will be on the studs not the heatsink.

--Victor

[AMR Labs]

The main risk, aside from a short (studs are not at ground potential) is physically damaging the chip legs where they come out of the package due to improper tightening of the two nuts. Very easy to do if not careful and failing to keep the proper hardware in the correct place. This topic has been explained many times here and on Tekscopes group, as it has been stated and explained why no heatsink is needed. Not to mention stick-on heatsinks falling off unexpectedly and possibly shorting other things inside the main board. Not worth trying to fix a problem that does really not exist.

[Scopetechniques]

The main risk, aside from a short (studs are not at ground potential) is physically damaging the chip legs where they come out of the package due to improper tightening of the two nuts. Very easy to do if not careful and failing to keep the proper hardware in the correct place. This topic has been explained many times here and on Tekscopes group, as it has been stated and explained why no heatsink is needed. Not to mention stick-on heatsinks falling off unexpectedly and possibly shorting other things inside the main board. Not worth trying to fix a problem that does really not exist.

I agree with what you say.  It does require care and some expertise to perform this work, along with a proper de-soldering workstation.
I have literally sold hundreds of 2465Bs and the U800 and A1 recap is only performed when the customer requests it.
I am not a fan of stick on heatsinks, only metal to metal and mechanically attached.

--Victor

[jonpaul]

See the many comments of Chuck Harris on Tekscopes2 and Tekscopes at groups.io, RE U800.

Note the earlier SN 2465B have original TEK ICs but later SN used to substute from Maxim, not  as good.

Jon

[Scopetechniques]

See the many comments of Chuck Harris on Tekscopes2 and Tekscopes at groups.io, RE U800.

Note the earlier SN 2465B have original TEK ICs but later SN used to substute from Maxim, not  as good.

Jon

Chuck and I communicate a lot.  He has sent me a lot of 2465B business my way since he retired.
Chuck and I do not agree on removing the A1 for complete recap and U800 socket install.
My opinion is that if a customer wants it I give to them. Worst case an A1 with a socketed U800 will allow a casual user to repair their beloved 2465B when a U800 failure occurs.

--Victor

[jorgino]

Hi all,

I'm also a "happy" owner of a 2465A CT. I was having issues with the display jumping all over the place:
https://youtu.be/g6iKEVwqY9M
So I took it apart to recap the boards, but then I decided to put it together so I could get the CAL constants before recaping the boards, however, now I get an error message and no traces and the LEDS above the display don't light up:
https://youtu.be/q4gS29u1FuY?si=lmVH3ULfq2Pwa76U
From your experience, did I re-connect something wrong, or did one of the boards go bad? I wasn't having any other issues before other than the display jumping. The manual says TEST 05 FAIL 24 is ADD Main Board - Positive Level is not Positive Enough. I didn't touch the High Voltage Power supply or the board that's underneath the O'scope, originally I only removed the 2 boards from inside and the one on the side with the NVRAM battery. Any help would be greatly appreciated!

Thanks,
George

[mnemotechnika]

Hi everybody, I'm new in this forum and a new analog oscilloscope user too. I wolud like to ask what is the problem in my 2467B if i see this on the screen. My probe is on the calibration signlal. THX

[AMR Labs]

Looks like you are using a generic probe. Switch input mode to DC coupling and see what happens and the waveform changes or clears up. Or there might be some interference or grounding problem that is distorting the waveform. Have you tried to compensate the probe until the square wave displays properly? (flat tops and bottoms). Are you sure the probe is working correctly? (tested in another scope). Also do you get the exact same waveform display if you use CH2 input instead?

Another possibility, since  you are at 50nS time base, and the calibration test frequency that the scope outputs at this setting might be higher than your generic probe can handle. Can't remember what frequency is used at this setting, but worth a try to go to a lower setting, like 1uS and see if the waveform clears up

BTW for a probe calibration setup you don't need to connect the grounding clip of the probe to one of the BNC shells, as the probe should already be properly grounded at the CH1 input that you are using. Just in case try disconnecting the ground clip from the BNC shell and see if the waveform changes. It should not.

Also keep in mind that if you are using a 1X/10X switchable probe you should set it for 1X to make sure amplitude readings are correct on screen. If its a 10X make sure to do the calculation in your head, as generic probes without the sensing pin will not properly adjust the amplitude scale on the scope when switched to 10X.

Ideally at some point you should get a proper Tektronix scope probe like the P6037 (or better yet a pair of probes) which will allow you to take fully advantage of this very fine scope you have.

Also since this is a '67B, be careful with the screen intensity setting, keep it on a not too bright (normal) level, as otherwise you would be straining and unnecessarily wearing out the super sensitive screen, which is the main feature of this scope over a regular '65B.

[tggzzz]

Hi everybody, I'm new in this forum and a new analog oscilloscope user too. I wolud like to ask what is the problem in my 2467B if i see this on the screen. My probe is on the calibration signlal. THX

No problem whatsoever.

On all other scopes the cal signal is a fixed frequency around 1kHz. For no solid reason on the 2445/2465/2467 ranges Tek decided to vary the cal frequency with the timebase - and did a poor job at faster timebase frequencies.

Realise the cal signal has only one useful function: to enable the user to adjust the low frequency compensation of *10 "high" impedance passive probes. It is not suitable to verify probes' high frequency behaviour; in some probes such adjustments are buried in the probe. (Exception: the strange 485 cal out has a clean 1ns risetime into 50ohm, and can be used to verify risetimes of 50ohm inputs)

So, RTFM (or just about any introduction to scopes!) to see how and why to use a ~1kHz cal signal. Ignore the misfeature at other timebases

[David Hess]

On all other scopes the cal signal is a fixed frequency around 1kHz. For no solid reason on the 2445/2465/2467 ranges Tek decided to vary the cal frequency with the timebase - and did a poor job at faster timebase frequencies.

The Tektronix 2440 series of DSOs also have the calibrator output frequency track the sweep.

There is nothing unusual here.  The calibrator output signal gets distorted at high frequencies.  It should only be used at low frequencies for probe calibration anyway.

[Kartika]

https://www.facebook.com/share/p/pvQjUnVevMGZETpQ/

[David Hess]

https://www.facebook.com/share/p/pvQjUnVevMGZETpQ/

And if you look under the SRAM or the board, there will be a Dallas or Maxim battery backup IC which implements switching between power sources and control of the chip enable.

[siggi]

For no solid reason on the 2445/2465/2467 ranges Tek decided to vary the cal frequency with the timebase - and did a poor job at faster timebase frequencies.

Well, actually , there is a very good reason for this behavior that is documented in the user manual, namely that this allows the user to do a quick and dirty validation of the accuracy of the sweeps. The calibrator signal is derived from a crystal oscillator that's orders of magnitude more accurate than the calibrated analog sweeps, and so having 5 cycles of the calibrator on-screen at most speeds allows easy eyeball-validation of the sweep speeds.

The manual also indicates that the calibrator is sufficient to also eyeball that the vertical deflection is good, though obviously that only applies where a 0.4V peak to peak signal occupies multiple - ideally 8 - divisions.

[tggzzz]

For no solid reason on the 2445/2465/2467 ranges Tek decided to vary the cal frequency with the timebase - and did a poor job at faster timebase frequencies.

Well, actually , there is a very good reason for this behavior that is documented in the user manual, namely that this allows the user to do a quick and dirty validation of the accuracy of the sweeps. The calibrator signal is derived from a crystal oscillator that's orders of magnitude more accurate than the calibrated analog sweeps, and so having 5 cycles of the calibrator on-screen at most speeds allows easy eyeball-validation of the sweep speeds.

The manual also indicates that the calibrator is sufficient to also eyeball that the vertical deflection is good, though obviously that only applies where a 0.4V peak to peak signal occupies multiple - ideally 8 - divisions.

Accepted.

Still don't think it is a good enough reason

[mnemotechnika]

Thank you for your answers. I have successfully calibrated my probes

[AMR Labs]

https://www.facebook.com/share/p/pvQjUnVevMGZETpQ/

Facebook in this day in age? sorry but no thank you.