Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

One thing to point out though. The Nuvistors are considerably more robust than the FETs are  Oh and more fun

Given how the pre-amp stage is so nicely isolated from the rest of the scope... it seems to me one could pretty easily mod the circuit with more modern MOSFETs that are considerably tougher...?

mnem
 

[Zoli]

I had to pop the cover off my 54657A to iron out this hickey:
PIC
in the shell, so I took a couple pix of the inside for comparison and future reference.

Yup... DALLAS DS1244Y-120 & DS1245Y-120; I read the date as 2002, so maybe a few more years if I'm really lucky.

Ghaaaah. ~$90 just for the two NVRAM, then I need to get a working programmer. *sigh* I also need to do the same thing with the NVRAM in my 3478A... 

Nope. If I'm gonna spend that money, it'll be on the 54659A I actually want on my scope, not this module that I bought just cuz it was cheap and right there.

I'm just gonna button this one up and use it til it loses its mind.   

mnem
*tinker-tinker... putter-putter...*

I can send you over my old Willem  Then you can convince it to work from an RPI GPIO using Python 
As side inspiration, I've seen some Dallas replacement projects, but for TEK - sorry  Maybe is time to do it for HP? 

[bd139]

One thing to point out though. The Nuvistors are considerably more robust than the FETs are  Oh and more fun

Given how the stage is so nicely isolated from the rest of the scope... it seems to me one could pretty easily mod the circuit with more modern FETs that are considerably tougher...?

mnem
 

It was only really MOSFETs that had any robustness improvements really. Normal JFETs are about the same really.

[mnementh]

One thing to point out though. The Nuvistors are considerably more robust than the FETs are  Oh and more fun

Given how the stage is so nicely isolated from the rest of the scope... it seems to me one could pretty easily mod the circuit with more modern MOSFETs that are considerably tougher...?

mnem
 

It was only really MOSFETs that had any robustness improvements really. Normal JFETs are about the same really.

Well, yeah, I was figuring a depletion-mode MOSFET. Or is it unlikely one will find anything with appropriate input capacitance/leakage current compared to the JFETs in this application?

Forgive my ignorance; lots of time working with MOSFETs, very little working with JFET.

mnem
 

[m k]

So total we have 48KB of RAM. But where is the F/W stored, that implements all the extra features ? No EPROM on the board, strange...
I see a chip that might be a micro-controller, so I thought maybe it holds the F/W... it's SCN2661, found the datasheet... oh no, no MCU, just a standalone UART controller to drve the serial port...

So... that means the F/W is stored in RAM ?!!    How evil , twisted do you have to be to do that !
So when the batteries go, I don't just lose the user data/ waveforms stored in the RAM, I also lose the brains of the module, it bricks it !

Wow... I am truly amazed and shocked.... 

Vince,

I don't think there is any firmware in the module, the module is just the parallel and serial ports and the extra storage. The storage is needed for the extra functionality which is already built into the firmware of the scope.

I second that, MCU is not needed and is actually more difficult to implement.

What is U4?
My prediction is EEPROM.
If so then you can read it on the fly with PLCC clip, not nice really since you obviously need also a centronics stump.
And of course the clip is a real bargain for a one timer thing.

Well could be but not many pins on that chip, and very few traces going to it. So would have to  be serial memory that the scope would therfore need to spend time at boot up to "unpack" into RAM for fast parallel access. A bit awkward and convoluted... but still technically possible. So, just to make sure, I removed that sticker and had a look.... nope, not a memory. A humble 16V8 PAL chip for glue logic, hence why it's right by the expansion connector and so few lines are going to it. Just doing some address decoding.

So... there is no F/W in the module, so we don't give a crap about losing the contents of the NVRAM chips once they go bad, it's just user data/ stored waveforms, not code nor cal data or anything sensitive...

mnem's module definitely has some ROM but I understand the GPIB controller is self sufficient so the storage must be for something else.
Same U4 is there also, maybe a needed module ID is there, then those battery memories can be totally secondary.

[TERRA Operative]

Ghaaaah. ~$90 just for the two NVRAM, then I need to get a working programmer. *sigh* I also need to do the same thing with the NVRAM in my 3478A... 

All my 3478A's have a separate battery backed SRAM/etc. So replacing the battery is just replacing the battery, not the memory chip.

Were there different revisions using the Dallas chips?

[AVGresponding]

OK, I picked up the 5 meters. Here's the picture of the first...



I'll correct the "mad eye Moody" effect before flogging it for a fortune on fleabay.

I like it!

Not enough to buy it.

Needs a finishing touch I think, perhaps a big pink tongue on the needle, and then contrive to feed a varying voltage to it, such as a 0.5Hz sine with a DC offset in the centre of the scale.

[nixiefreqq]

hey mnem

not that it matters much.

but your dallas chips are more likely 1992 week 45.

that would be more consistent with U3 being 1989 and U15 being 1992 week 41.

[mnementh]

I had to pop the cover off my 54657A to iron out this hickey:
PIC
in the shell, so I took a couple pix of the inside for comparison and future reference.

Yup... DALLAS DS1244Y-120 & DS1245Y-120; I read the date as 2002, so maybe a few more years if I'm really lucky.

Ghaaaah. ~$90 just for the two NVRAM, then I need to get a working programmer. *sigh* I also need to do the same thing with the NVRAM in my 3478A... 

Nope. If I'm gonna spend that money, it'll be on the 54659A I actually want on my scope, not this module that I bought just cuz it was cheap and right there.

I'm just gonna button this one up and use it til it loses its mind.   

mnem
*tinker-tinker... putter-putter...*

I can send you over my old Willem  Then you can convince it to work from an RPI GPIO using Python 
As side inspiration, I've seen some Dallas replacement projects, but for TEK - sorry  Maybe is time to do it for HP? 

You mean like this and the similar mod for the 246x family?

    

Or you mean hacking the epoxy and doing one of these...? Given how expensive the chips are, might do that once any data is extracted; but not likely gonna wanna try that on a live NVRAM still on the board...

For extracting the data backups I've several times considered doing something like the AR488 GPIB:

https://www.eevblog.com/forum/projects/ar488-arduino-based-gpib-adapter/

But so far, I have not been able to convince myself to take the deep dive. 

mnem
 

[mnementh]

So total we have 48KB of RAM. But where is the F/W stored, that implements all the extra features ? No EPROM on the board, strange...
I see a chip that might be a micro-controller, so I thought maybe it holds the F/W... it's SCN2661, found the datasheet... oh no, no MCU, just a standalone UART controller to drve the serial port...

So... that means the F/W is stored in RAM ?!!    How evil , twisted do you have to be to do that !
So when the batteries go, I don't just lose the user data/ waveforms stored in the RAM, I also lose the brains of the module, it bricks it !

Wow... I am truly amazed and shocked.... 

Vince,

I don't think there is any firmware in the module, the module is just the parallel and serial ports and the extra storage. The storage is needed for the extra functionality which is already built into the firmware of the scope.

I second that, MCU is not needed and is actually more difficult to implement.

What is U4?
My prediction is EEPROM.
If so then you can read it on the fly with PLCC clip, not nice really since you obviously need also a centronics stump.
And of course the clip is a real bargain for a one timer thing.

Well could be but not many pins on that chip, and very few traces going to it. So would have to  be serial memory that the scope would therfore need to spend time at boot up to "unpack" into RAM for fast parallel access. A bit awkward and convoluted... but still technically possible. So, just to make sure, I removed that sticker and had a look.... nope, not a memory. A humble 16V8 PAL chip for glue logic, hence why it's right by the expansion connector and so few lines are going to it. Just doing some address decoding.

So... there is no F/W in the module, so we don't give a crap about losing the contents of the NVRAM chips once they go bad, it's just user data/ stored waveforms, not code nor cal data or anything sensitive...

mnem's module definitely has some ROM but I understand the GPIB controller is self sufficient so the storage must be for something else.
Same U4 is there also, maybe a needed module ID is there, then those battery memories can be totally secondary.

Yes, it is my understanding that these modules all contain software for various math functions (and the extra space in there is where the games are stored) as well as additional storage for waveforms. The difference between them is the additional I/O, whether RS-232 & Parallel port, GPIB, etc... and whatever FW/SW is needed to make that interface work.

mnem
 

[Zoli]

I had to pop the cover off my 54657A to iron out this hickey:
PIC
in the shell, so I took a couple pix of the inside for comparison and future reference.

Yup... DALLAS DS1244Y-120 & DS1245Y-120; I read the date as 2002, so maybe a few more years if I'm really lucky.

Ghaaaah. ~$90 just for the two NVRAM, then I need to get a working programmer. *sigh* I also need to do the same thing with the NVRAM in my 3478A... 

Nope. If I'm gonna spend that money, it'll be on the 54659A I actually want on my scope, not this module that I bought just cuz it was cheap and right there.

I'm just gonna button this one up and use it til it loses its mind.   

mnem
*tinker-tinker... putter-putter...*

I can send you over my old Willem  Then you can convince it to work from an RPI GPIO using Python 
As side inspiration, I've seen some Dallas replacement projects, but for TEK - sorry  Maybe is time to do it for HP? 

You mean like this and the similar mod for the 246x family?

    

Or you mean hacking the epoxy and doing one of these...? Given how expensive the chips are, might do that once any data is extracted; but not likely gonna wanna try that on a live NVRAM still on the board...

For extracting the data backups I've several times considered doing something like the AR488 GPIB:

https://www.eevblog.com/forum/projects/ar488-arduino-based-gpib-adapter/

But so far, I have not been able to convince myself to take the deep dive. 

mnem
 

Nope, I meant something like:
https://www.eevblog.com/forum/repair/reverse-engineer-dallas-ds1742w/
Challenge accepted?

[TERRA Operative]

Hey, that's my hand!

I had to pop the cover off my 54657A to iron out this hickey:
PIC
in the shell, so I took a couple pix of the inside for comparison and future reference.

Yup... DALLAS DS1244Y-120 & DS1245Y-120; I read the date as 2002, so maybe a few more years if I'm really lucky.

Ghaaaah. ~$90 just for the two NVRAM, then I need to get a working programmer. *sigh* I also need to do the same thing with the NVRAM in my 3478A... 

Nope. If I'm gonna spend that money, it'll be on the 54659A I actually want on my scope, not this module that I bought just cuz it was cheap and right there.

I'm just gonna button this one up and use it til it loses its mind.   

mnem
*tinker-tinker... putter-putter...*

I can send you over my old Willem  Then you can convince it to work from an RPI GPIO using Python 
As side inspiration, I've seen some Dallas replacement projects, but for TEK - sorry  Maybe is time to do it for HP? 

You mean like this and the similar mod for the 246x family?

    

Or you mean hacking the epoxy and doing one of these...? Given how expensive the chips are, might do that once any data is extracted; but not likely gonna wanna try that on a live NVRAM still on the board...

For extracting the data backups I've several times considered doing something like the AR488 GPIB:

https://www.eevblog.com/forum/projects/ar488-arduino-based-gpib-adapter/

But so far, I have not been able to convince myself to take the deep dive. 

mnem
 

Hey, that's my hand! 

I've been thinking of taking to a Dallas NVRAM with a bastard file, take the top off and expose then remove the old battery and glue a new battery holder on top.

The good thing about the TDS500, 600, 700 series scopes is the cal data is held in some I2C PROM chips on tje acquisition board so you don't need to recal after playing with the NVRAMs.
It's trivial to read out the contents of the NVRAMs via GPIB, have at the chips to make a Frankenstein's monster, then stick them back in and load the data back on via GPIB again.

Might be more dicey for equipment that has cal data and other critical data unless you rip and dump it, but once you have a copy of the data backed up, file away!

[AVGresponding]

Someone on here was after one of these iirc?  https://www.ebay.co.uk/itm/265737962421





This is confusing...  https://www.ebay.co.uk/itm/115426720808





Green LEDs FTW! Hopefully brighter than the pic suggests:  https://www.ebay.co.uk/itm/225030536254





This poor old thing seems past saving to me, but maybe someone can love it:  https://www.ebay.co.uk/itm/185465698603





Data logging, old skool:  https://www.ebay.co.uk/itm/295040057246





BNIB Sig:  https://www.ebay.co.uk/itm/334473340438





Used but clean 'Lent:  https://www.ebay.co.uk/itm/155034606838





Just to trigger the h_FE-on-a-DMM haters:  https://www.ebay.co.uk/itm/295035026225





I'm curious as to how they plan to ship these:  https://www.ebay.co.uk/itm/314034213063  https://www.ebay.co.uk/itm/314034213080

[nixiefreqq]

mine are from 1992 also

[mnementh]

hey mnem, not that it matters much. but your dallas chips are more likely 1992 week 45. that would be more consistent with U3 being 1989 and U15 being 1992 week 41.

Yeah, except I believe this whole family of scopes wasn't released until mid-late 90s... tho I suppose the module could've come from one of the older families of HP scopes...?

mnem
 

[mnementh]

Ghaaaah. ~$90 just for the two NVRAM, then I need to get a working programmer. *sigh* I also need to do the same thing with the NVRAM in my 3478A... 

All my 3478A's have a separate battery backed SRAM/etc. So replacing the battery is just replacing the battery, not the memory chip.

Were there different revisions using the Dallas chips?

Oh, it looks like I was thinking of this thread about the 3458A I read a while back: 

   https://www.eevblog.com/forum/metrology/3458a-how-do-i-read-the-dallas-non-volatile-sram-date-codes/   

From that pic, it appears nixie is correct about the date code, and I was looking at the wrong number. 

mnem
*toddles off to make some lunch*

[bd139]

Data logging, old skool:  https://www.ebay.co.uk/itm/295040057246

Not data logging. Fairly unknown and now useless device. It’s actually an off air frequency standard. Uses 200kHz and has a OCXO inside it. The the chart recorder actually records the drift.

So one for the time nuts. But only UK time nuts.

Useless because someone moved the reference to 198KHz.

If anyone else doesn’t want it I may have a go at it.

[mnementh]

...As side inspiration, I've seen some Dallas replacement projects, but for TEK - sorry  Maybe is time to do it for HP? 

https://youtu.be/ve-MnHakY6Y   

You mean like this and the similar mod for the 246x family?

    

Or you mean hacking the epoxy and doing one of these...? Given how expensive the chips are, might do that once any data is extracted; but not likely gonna wanna try that on a live NVRAM still on the board...

For extracting the data backups I've several times considered doing something like the AR488 GPIB:

https://www.eevblog.com/forum/projects/ar488-arduino-based-gpib-adapter/

But so far, I have not been able to convince myself to take the deep dive. 

mnem
 

Nope, I meant something like:
https://www.eevblog.com/forum/repair/reverse-engineer-dallas-ds1742w/ Challenge accepted?

If I had a working GPIB adapter and software to extract the data, and maybe a little better Kung-Fu to know  I were actually doing, yeah. That's all part of the "deep-dive" I was talking about. Of course I'm not at all afraid to do a little grinder abuse. 

mnem
*lunch-ify*

[bd139]

hey mnem, not that it matters much. but your dallas chips are more likely 1992 week 45. that would be more consistent with U3 being 1989 and U15 being 1992 week 41.

Yeah, except I believe this whole family of scopes wasn't released until mid-late 90s... tho I suppose the module could've come from one of the older families of HP scopes...?

mnem
 

1992. Little bit earlier. First got my hands on one in 1993

[factory]

OK, but what makes the Type 422 so special?

No...you can't have mine. 

No ICs. A late 453 will suffice too 

No fan either in the 422.  Knowing how much bd139 hates them.

David

[factory]

Data logging, old skool:  https://www.ebay.co.uk/itm/295040057246

Not data logging. Fairly unknown and now useless device. It’s actually an off air frequency standard. Uses 200kHz and has a OCXO inside it. The the chart recorder actually records the drift.

So one for the time nuts. But only UK time nuts.

Useless because someone moved the reference to 198KHz.

If anyone else doesn’t want it I may have a go at it.

I put an offer on that last night, will probably have to bid for it as no response so far.
We might not even have the off-air reference much longer, as the BBC don't seem very interested in the future of LW/MW broadcasting.
Edit: note these do not contain a OCXO.

David

[bd139]

Data logging, old skool:  https://www.ebay.co.uk/itm/295040057246

Not data logging. Fairly unknown and now useless device. It’s actually an off air frequency standard. Uses 200kHz and has a OCXO inside it. The the chart recorder actually records the drift.

So one for the time nuts. But only UK time nuts.

Useless because someone moved the reference to 198KHz.

If anyone else doesn’t want it I may have a go at it.

I put an offer on that last night, will probably have to bid for it as no response so far.

David

Ok will leave it in your capable hands then 

We expect tear downs 

[factory]

I should correct you, as there is no OCXO in those.

David

[tggzzz]

Heads up for Brits living in or close to Highbridge, Somerset, you might want to jump on these quickly, I would personally would but way too far away for me.

https://www.gumtree.com/p/for-sale/2-sonartron-digital-valve-voltmeters-and-3-avometers/1434901970

I have already been in contact

Did you get them or what?

The seller is busy tomorrow (oops today).

So does that mean that you haven't sealed the deal yet then  Well worth taking a risk on IMO, quick before someone else beats you to it.

Just arranged to pick them up tomorrow. On Friday I may be able to a post a picture that is as offensive as that 80yo female.

OK, I picked up the 5 meters. Here's the picture of the first...



I'll correct the "mad eye Moody" effect before flogging it for a fortune on fleabay.

It did make me think of that absolutely dreadful show. It was, apparently, extremely popular in the 60s, but I never understood the attraction.

OTOH yootoob vids of that show are a good way of pointing out to, um, oversensitive youngsters just how much progress has been made.

[bd139]

I should correct you, as there is no OCXO in those.

David

Ah