Tektronix 2465B oscilloscope teardown

[calibrationfixture]

Hi chilternview,

Thus Calibration needed. In case of the Dallas NVRAM, you could fill a new chip with the contents of another 2465B Series NVRAM. Maybe somewhere in this Thread you could find a bin.file of it. Otherwise I will send you such a File. This will "delete" the string of ?'s, but it isn't a real calibration. By the way some Sellers ask serious money for such Preloaded NVRAM's.

Calibrationfixture

[chilternview]

Thanks Calibrationfixture, I see in some of the early posts in this thread there are NVRAM binaries. I'll give them a try when the new DS1225AD-150 I ordered arrives. In the meantime I'll desolder the old one and fit a socket.

[chilternview]

I desoldered the DS1225 and fitted a machined pin style 28 pin DIL socket. And just for fun took the binary file from post #33 in this thread and programmed the  old DS1225 and tried booting the scope with it... it worked, no error. Of course the cal data is unlikely to be correct but I can live with that until the other problem (triggering not working correctly) is resolved.

The triggering is odd. It works most of the time on slow sweep speeds (> 2mS) but on faster sweeps it seems increasingly either fail to trigger or the triggering drifts. It basically makes the scope unusable. Trying to wade through the service manual to check waveforms and understand whats going on.

Edit: I fixed the triggering issue. My 2465B has Option 05, TV line triggering capability. Something on the 05 option board is bad - unplugging J102 that links it to the main board allows the scope to work normally. TV triggering is not useful these days so not really worth trying to fix the option board.

[MathWizard]

So I've heard of these scopes, besides their high BW, and of course impressive build quality/etc....what's their place today ? Just checking ebay they aren't as much as I remember, but besides maybe the BW, are they much better at anything, compared to new scopes for the same +$600USD ?

The reasons for an average user, to get an old CRT scope, are running out. I still like mine, but I just never use it anymore.

[calibrationfixture]

Hi chilternview,

Many years ago I had to fix a 2445B Option 05 for an Audio Repair Firm. It was my first 24XX Series Scope Repair. I was again overwhelmed by Tektronix its engineering and thus I fell in love with these Scopes. It also had Triggering Problems after the fixing of the LVSMPS. In my case I forgot connecting the Option 05 Board Power Connector. That was all. In your case it is the other way. I suggest looking to all Option 05 related ribbon cables (the Header Boxes). Also between both Attenuators, there is a Jumper Connector (J103 on the A1 Main Board) with a Coax connected to it.

Calibrationfixture

[chilternview]

Calibrationfixture. you're absolutely spot on. When I came to put the case back on I noticed the wire from the TV board to J305 on the PSU inverter board (the +5V digital supply for the board). It was out of the way when I recapped the PSU and I had not noticed it to reconnect it. Having now done so, and reconnected J102, the scope is working fine. I will probably never need the TV option but its nice to know all is working now. I just tested the scope with a 500MHz sig gen and the trace was bright and clear and triggering rock solid, pretty good for a 30 year old scope.

[faraday]

Looks like last call

[faraday]

Battery 42nd week 88 and still going strong 3.5V 
Resistors 15ohm -12k
56k hours and beam strong

[tggzzz]

Battery 42nd week 88 and still going strong 3.5V 

For now, until it isn't

I suggest you make a video while you use the test routines to scroll through the calibration values, ready to re-insert them into the replacement RAM

[faraday]

I have it (video), battery is on delivery. I did backup with 3 2032 cells

[faraday]

new battery arrived

[Veketti]

Sorry stupid question, but is there a way to set 10x attenuation probe to show correct voltages? Didn't find in the manual. Or is it so that you just need to remember to multiply in your head by 10? Or is there any need to use the 10x probe and go with 1x all the time?

[AMR Labs]

There is a very good reason to use 10X probes, specially at higher frequencies or low signals as among other advantages this will greatly reduce the capacitive and resistive loading on the circuit to be measured, as it will get closer to an "ideal" probe. There are other advantages as well, try to look up a PDF by Tektronix called "ABCs of probes Primer" that has a lot of good information. There are also many other useful documents out there on this that explain all the probe details and proper usage. One should in general always try to use a 10X probe, unless a 1X would specifically be required.

If you use Tektronix 10X probes with the appropriate sensor ID pin (for example like the P6137 that comes with this scope) it will happen automatically, otherwise with generic probes that have a 1X/10X switch making a mental note is the only other way.

You could also trigger the change by grounding the ring around the BNC input connector through a specific resistance (I don't know the exact value required), but this would not really be easy or even practical to implement on a regular probe connector.

[faraday]

There is a very good reason to use 10X probes, specially at higher frequencies as among other advantages this will greatly reduce the capacitive and resistive loading on the circuit to be measured. There are other advantages as well, look up a PDF by Tektronix called "ABCs of probes Primer" that has a lot of good information. There are also many other useful documents out there that explain all the probe details and proper usage. One should in general always try to use a 10X probe, unless a 1X would specifically be required.

If you use Tektronix 10X probes with the appropriate sensor ID pin (for example like the P6137 that comes with this scope) it will happen automatically, otherwise with generic probes that have a 1X/10X switch making a mental note is the only other way.

You could also trigger the change by grounding the ring around the BNC input connector through a specific resistance (I don't know the exact value required), but this would not really be easy or even practical to implement on a regular probe connector.

11kohm

[awesomechapro]

This is a table of the attenuation settings and the resistance values that all Tektronix scopes follow.

From: https://w140.com/tekwiki/wiki/BNC_connector_with_readout_ring

[xardomain]

Hi Faraday,
where did you get the replacement battery? Mouser won't ship outside US...
TIA.

Giuseppe Marullo
IW2JWW - JN45RQ

[faraday]

Hi Faraday,
where did you get the replacement battery? Mouser won't ship outside US...
TIA.

Giuseppe Marullo
IW2JWW - JN45RQ

batimex.pl

[xardomain]

Many thanks.

So the exact replacement part is:

LTC-7P

I will explore the possibility to use an FRAM and avoid it, else I will just replace the battery.


BTW, is it any way to check that the values I dumped from EXT 2 are correct? I mean, does a simple check of memory area in the binary file could be attempted before removing/reprogramming the new SRAM?

Giuseppe Marullo
IW2JWW - JN45RQ

[faraday]

Many thanks.

So the exact replacement part is:

LTC-7P

I will explore the possibility to use an FRAM and avoid it, else I will just replace the battery.


BTW, is it any way to check that the values I dumped from EXT 2 are correct? I mean, does a simple check of memory area in the binary file could be attempted before removing/reprogramming the new SRAM?

Giuseppe Marullo
IW2JWW - JN45RQ

It depend on your A5 digital control board version. You may have ds1225.

[xardomain]

It depend on your A5 digital control board version. You may have ds1225.

I have a low serial 2465B with SRAM, not DALLAS. So the questions are:
1) Will it work with a FRAM?
2) Just in case will it work with a DALLAS?

TIA.

Giuseppe Marullo
IW2JWW - JN45RQ

[David Hess]

I have a low serial 2465B with SRAM, not DALLAS. So the questions are:
1) Will it work with a FRAM?
2) Just in case will it work with a DALLAS?

The SRAM models can be retrofitted with DALLAS without issues.  Some DALLAS models used the printed circuit board for SRAM with the battery backup parts removed.

I question upgrading a SRAM model to FRAM.  Just install a new backup battery.

[xardomain]

I have a low serial 2465B with SRAM, not DALLAS. So the questions are:
1) Will it work with a FRAM?
2) Just in case will it work with a DALLAS?

The SRAM models can be retrofitted with DALLAS without issues.  Some DALLAS models used the printed circuit board for SRAM with the battery backup parts removed.

I question upgrading a SRAM model to FRAM.  Just install a new backup battery.

Sure, but...what if something goes wrong? Still have to reload the values somehow and battery can fail anyway even after I replace it.
At this point, buying all the stuff to prepare for the "disaster" may be worth to change to FRAM and be done with the battery.

IF FRAM can be used on a low serial, this is the last piece of information I need.

[David Hess]

I have a low serial 2465B with SRAM, not DALLAS. So the questions are:
1) Will it work with a FRAM?
2) Just in case will it work with a DALLAS?

The SRAM models can be retrofitted with DALLAS without issues.  Some DALLAS models used the printed circuit board for SRAM with the battery backup parts removed.

I question upgrading a SRAM model to FRAM.  Just install a new backup battery.

Sure, but...what if something goes wrong? Still have to reload the values somehow and battery can fail anyway even after I replace it.
At this point, buying all the stuff to prepare for the "disaster" may be worth to change to FRAM and be done with the battery.

IF FRAM can be used on a low serial, this is the last piece of information I need.

As far as I know, the address and data bus access to memory is the same with all models, so if FRAM works in one, it will work in all.  The limit with FRAM is that unlike the JEDEC standard asynchronous SRAM interface, which applies to most asynchronous memory like EEPROMs and whatnot, FRAM requires a precharge cycle initiated by toggling chip select or equivalent, which these oscilloscopes generate anyway so nothing extra is required.

The only change you might want to make, which they did when they used the same printed circuit board with SRAM and DALLAS, is to disable the low batter detection so it does not report low battery all the time.

[xardomain]

As far as I know, the address and data bus access to memory is the same with all models, so if FRAM works in one, it will work in all.  The limit with FRAM is that unlike the JEDEC standard asynchronous SRAM interface, which applies to most asynchronous memory like EEPROMs and whatnot, FRAM requires a precharge cycle initiated by toggling chip select or equivalent, which these oscilloscopes generate anyway so nothing extra is required.

The only change you might want to make, which they did when they used the same printed circuit board with SRAM and DALLAS, is to disable the low batter detection so it does not report low battery all the time.

Perfect, so I can choose whatever option, based on risk/effort. Last question, is there a way to check that the checksum in the memory will be okay? I mean,I have all the 256 values but clearly I will know for sure only when I will put it in the A5 board, so it would be great if something could check the 512byres and confirm that at least at teoretichal level it should work. If it will really work you will never know until you put the FRAM inside the scope of course.
Is it a standard CRC something? At which location is the checksum in the image?
Any good soul that already created a utility that could validate a memory image?

TIA

Giuseppe Marullo
IW2JWW - JN45RQ

[AMR Labs]

I don't think there is any way to verify the checksum of the data outside of the scope. Once installed the scope will either tell you if it likes it or not, and definitively put an error on the screen.

But since you are able to zero out and rewrite/verify the chip as much as needed, so if checksum does not work, just burn another known good file onto the chip from another scope just to make sure the chip is actually working on the scope. There are 2465 calibration constant files floating around that you can download and burn into your chip. The scope will obviously not be properly calibrated, but at least it should pass self-test and boot normally.

If it where me, and since you are migrating from SDRAM, I would try first with the DS1225 NVRAM and make sure it gets accepted, and only then try with the FRAM once you know it will most likely work.