Test Equipment Anonymous (TEA) group therapy thread

[Specmaster]

Come on Noodleboy stick your Bryman onto 50,000 count and gain an extra digit?

[bd139]

Nah it’s slightly above the count range on that.

To all UK TEA members, just got a £1 max selling fees thing appear. That means all the good shit will appear today. Scripts at the ready! 

[Specmaster]

Yep, got my offer as well, so items will be listed later. 

[tggzzz]

Dammit I need 6.5 digits now.

I built a zener reference from scratch. It sits bang on 6.3305 volts across a supply range of 22-35 volts. But what about the digits I can't see! 

Now breathe on it, or log its voltage over the course of a month. You shouldn't be satisfied until you can:

• use the zener as a thermometer
• verify whether or not it suffers from popcorn noise
• determine that it has "stopped aging", for your definition of "stopped"
• measure its RMS noise

Yes, I have used a 1.0186V saturated Weston Standard Cell as a thermometer; 40uV/C is more than sufficient.

I'm currently being pissed off with the reference[1] in one of my 7.25 digit 7075s. It has a 10uV/C tempco, and takes forever to stabilise. I'm thinking of replacing it with an LM399, but that brings its own interesting challenges, e.g. getting ~7.014V +ve and -ve voltages and 1mA through the zener with minimal tempco.

[1] a 9.3V zener wrapped in a coil with a thermistor and a tiny bit of insulation. Then there are lots of discrete precision resisitors not far from a heat source to generate the +ve and -ve reference voltages. But on the plus side, there's a panaplex display no microprocessor - so I don't have to worry about it forgetting its cal constants.

[bd139]

Indeed. That will be the objective eventually.  Covered in glue at the moment in a polystyrene box. For ref the next step is move it to CA3096 silicon when I can find the tube of them.m I had lying around.

Basically I’m tying to make ICs without any fab equipment. I bought a linear integrated circuit design book a while ago and want to try some ideas out.

I noticed that LT had building block bits of silicon pre bonded and packaged that covered their process in dev. I’m trying to do the same with someone else’s silicon

Next is band gap reference.

Edit: there’s an unknown untested HP 3455A on eBay. I was tempted by this as it’s a Bob Pease favourite. But it’s large, noisy and unknown.

[tggzzz]

Edit: there’s an unknown untested HP 3455A on eBay.

"Untested" is the entrance to the voltnut rathole. Welcome.

[Specmaster]

Indeed. That will be the objective eventually.  Covered in glue at the moment in a polystyrene box. For ref the next step is move it to CA3096 silicon when I can find the tube of them.m I had lying around.

Basically I’m tying to make ICs without any fab equipment. I bought a linear integrated circuit design book a while ago and want to try some ideas out.

I noticed that LT had building block bits of silicon pre bonded and packaged that covered their process in dev. I’m trying to do the same with someone else’s silicon

Next is band gap reference.

Edit: there’s an unknown untested HP 3455A on eBay. I was tempted by this as it’s a Bob Pease favourite. But it’s large, noisy and unknown.

It also looks like its not been well treated or respected in its time.

[bd139]

Yes indeed. I might put a low offer in and see where I get

[Specmaster]

Yes indeed. I might put a low offer in and see where I get

I might stick a TEAC unit on there later, currently there's two H240s on loads more than I paid for my non working ones, and they have shitloads of watchers on them, could be some serious money in these units 

[Wolfgang]

Dammit I need 6.5 digits now.

I built a zener reference from scratch. It sits bang on 6.3305 volts across a supply range of 22-35 volts. But what about the digits I can't see! 

Now breathe on it, or log its voltage over the course of a month. You shouldn't be satisfied until you can:

• use the zener as a thermometer
• verify whether or not it suffers from popcorn noise
• determine that it has "stopped aging", for your definition of "stopped"
• measure its RMS noise

Yes, I have used a 1.0186V saturated Weston Standard Cell as a thermometer; 40uV/C is more than sufficient.

I'm currently being pissed off with the reference[1] in one of my 7.25 digit 7075s. It has a 10uV/C tempco, and takes forever to stabilise. I'm thinking of replacing it with an LM399, but that brings its own interesting challenges, e.g. getting ~7.014V +ve and -ve voltages and 1mA through the zener with minimal tempco.

[1] a 9.3V zener wrapped in a coil with a thermistor and a tiny bit of insulation. Then there are lots of discrete precision resisitors not far from a heat source to generate the +ve and -ve reference voltages. But on the plus side, there's a panaplex display no microprocessor - so I don't have to worry about it forgetting its cal constants.

There is no life without a 34470A. At least not for a TEA addict. 

[tggzzz]

Yes indeed. I might put a low offer in and see where I get

That's obvious and inevitable[1]: down the rathole!

You'll also need reference voltage(s) and reference resistors. You might add a 7decade 100kohm KVD to test other ranges.

You'll swear you don't need them, but when you see variations in your circuit's voltages, you won't know whether it is your circuit or the DVM's reference.

And to calibrate your voltage reference, you'll probably want to take it to the Hannover Maker Faire, to use the German-equivalent-of-NPLs HP3458

I'm still kicking myself for not getting a very cheap Fluke calibrator when I got my 7081. I saw how heavy it was and didn't realise how expensive they are.

[1] unless you shell out for a DVM with a valid calibration certificate.

[bd139]

Oh I’m well on the way there

I missed a 34401A a while ago (look at eBay sold listings) for £45 from a seller round the corner from me. Throwing errors galore but they were surmountable. Grr.

[tggzzz]

There is no life without a 34470A. At least not for a TEA addict. 

Yes, but that is ~20* what I paid for my 7081, voltage reference and reference resistors (OK, the KVD was another £200)

It would put me on the street, rather than keeping me off the street!

[bd139]

34470A isn’t a bad price for what it is. If I consider I was paying that per laptop a while ago and it’s 1/3 of a half decent motor then it doesn’t seem so bad.

Or am I just getting voltnutty there?

If you look at the design info for that class of meter you’ll see just the front end FET pair is £50 a pop so you can see where the money is.

[tggzzz]

34470A isn’t a bad price for what it is. If I consider I was paying that per laptop a while ago and it’s 1/3 of a half decent motor then it doesn’t seem so bad.

Or am I just getting voltnutty there?

If you look at the design info for that class of meter you’ll see just the front end FET pair is £50 a pop so you can see where the money is.

Ask SWMBO!

(And I'm not foolish enough to go searching for DVMs "wot i hav missed")

[bd139]

I don’t need to ask her I know what the answer is 

I look for sold items to look for sellers who have no idea what their stuff is worth so I camp them with scripts. Has been very profitable.

[Specmaster]

I don’t need to ask her I know what the answer is 

I look for sold items to look for sellers who have no idea what their stuff is worth so I camp them with scripts. Has been very profitable.

There's always a method in the madness 

[Kosmic]

I don’t need to ask her I know what the answer is 

I look for sold items to look for sellers who have no idea what their stuff is worth so I camp them with scripts. Has been very profitable.

Haha good strategy. On my side I'm concentrating on seller that accept ridiculous offer 

[med6753]

Here's a question for your smart guys...

Background: I have one of those cheap Chinese inverter boards that take 9-12VDC and steps it up over a wide adjustment range up to almost 300VDC. I currently have it set to 190VDC as that is the most common set point used to calibrate the 200V range on most DMM's. And in fact on the 2465 DMM it is an absolute calibration constant requirement. The problem with that board is that it's unregulated and bounces all over. One minute it's 188V and another 192V. So I've been looking for some way to regulate it and I found this which looks promising....

https://www.mouser.com/datasheet/2/268/20005399A-909415.pdf

The trouble is, how do I calculate the values of R1 and R2? It doesn't seem to give enough information other than one example for 5VDC. But I want 190.0V. Maybe I'm missing something. I sure would like to know what resistance values to start with and then tweak it to exactly 190.0V.   

[ruffy91]

Got a working Keithley 2001.
Owner wanted 200$ but I bargained him down to 100$.
He was a teacher at the university I graduated which gave me a little Bonus.
Calibration due date is a few years ago. What is the cheapest option to get a reference voltage or resistor good for 7.5/8.5 digits?

Next month I will move to our new house with 25 square meters workshop goodnes It's empty for now except for a 4mx0.6m table on one side.
My poor wallet..

[med6753]

Got a working Keithley 2001.
Owner wanted 200$ but I bargained him down to 100$.
He was a teacher at the university I graduated which gave me a little Bonus.
Calibration due date is a few years ago. What is the cheapest option to get a reference voltage or resistor good for 7.5/8.5 digits?

Next month I will move to our new house with 25 square meters workshop goodnes It's empty for now except for a 4mx0.6m table on one side.
My poor wallet..

Well, you're gonna slip down that rabbit hole real quick.  Welcome aboard! 

[Specmaster]

Here's a question for your smart guys...

Background: I have one of those cheap Chinese inverter boards that take 9-12VDC and steps it up over a wide adjustment range up to almost 300VDC. I currently have it set to 190VDC as that is the most common set point used to calibrate the 200V range on most DMM's. And in fact on the 2465 DMM it is an absolute calibration constant requirement. The problem with that board is that it's unregulated and bounces all over. One minute it's 188V and another 192V. So I've been looking for some way to regulate it and I found this which looks promising....

https://www.mouser.com/datasheet/2/268/20005399A-909415.pdf

The trouble is, how do I calculate the values of R1 and R2? It doesn't seem to give enough information other than one example for 5VDC. But I want 190.0V. Maybe I'm missing something. I sure would like to know what resistance values to start with and then tweak it to exactly 190.0V.   

Maybe I'm missing something here, if your sure that the current output voltage is correct, then I'd personally be tempted to live with it as it as because the tolerance on the HV regulator you found is +/-5% so that your 190 could be anywhere between 180.5V to 199.5V?

[med6753]

Here's a question for your smart guys...

Background: I have one of those cheap Chinese inverter boards that take 9-12VDC and steps it up over a wide adjustment range up to almost 300VDC. I currently have it set to 190VDC as that is the most common set point used to calibrate the 200V range on most DMM's. And in fact on the 2465 DMM it is an absolute calibration constant requirement. The problem with that board is that it's unregulated and bounces all over. One minute it's 188V and another 192V. So I've been looking for some way to regulate it and I found this which looks promising....

https://www.mouser.com/datasheet/2/268/20005399A-909415.pdf

The trouble is, how do I calculate the values of R1 and R2? It doesn't seem to give enough information other than one example for 5VDC. But I want 190.0V. Maybe I'm missing something. I sure would like to know what resistance values to start with and then tweak it to exactly 190.0V.   

Maybe I'm missing something here, if your sure that the current output voltage is correct, then I'd personally be tempted to live with it as it as because the tolerance on the HV regulator you found is +/-5% so that your 190 could be anywhere between 180.5V to 199.5V?

Yes, I noticed that. But I'd be willing to betcha that it would be STABLE...or at least much more stable than what I have now. And since it doesn't need to pass any appreciable current it may be possible to tweak it dead nuts or at least very close. I want to at least give a try if I can figure out the resistance values.

[Wolfgang]

Here's a question for your smart guys...

Background: I have one of those cheap Chinese inverter boards that take 9-12VDC and steps it up over a wide adjustment range up to almost 300VDC. I currently have it set to 190VDC as that is the most common set point used to calibrate the 200V range on most DMM's. And in fact on the 2465 DMM it is an absolute calibration constant requirement. The problem with that board is that it's unregulated and bounces all over. One minute it's 188V and another 192V. So I've been looking for some way to regulate it and I found this which looks promising....

https://www.mouser.com/datasheet/2/268/20005399A-909415.pdf

The trouble is, how do I calculate the values of R1 and R2? It doesn't seem to give enough information other than one example for 5VDC. But I want 190.0V. Maybe I'm missing something. I sure would like to know what resistance values to start with and then tweak it to exactly 190.0V.   

Trying to batter a Chinese upconverter into a stable source of measurement grade high voltage source ? I would say that resistance is futile

What about a proper postregulator ?

[bd139]

Here's a question for your smart guys...

Background: I have one of those cheap Chinese inverter boards that take 9-12VDC and steps it up over a wide adjustment range up to almost 300VDC. I currently have it set to 190VDC as that is the most common set point used to calibrate the 200V range on most DMM's. And in fact on the 2465 DMM it is an absolute calibration constant requirement. The problem with that board is that it's unregulated and bounces all over. One minute it's 188V and another 192V. So I've been looking for some way to regulate it and I found this which looks promising....

https://www.mouser.com/datasheet/2/268/20005399A-909415.pdf

The trouble is, how do I calculate the values of R1 and R2? It doesn't seem to give enough information other than one example for 5VDC. But I want 190.0V. Maybe I'm missing something. I sure would like to know what resistance values to start with and then tweak it to exactly 190.0V.   

Empirical finger in the air approach with casio's excellent solver so you don't even have to do any algebabooble.

Key is hiding in the datasheet fig 3-2.

Vout = 1.2 * (1 + R2 / R1)  + iAdj * R2

iAdj = 10uA
pick R1 around 2.2K (finger in the air estimate - could be totally wrong!), solve for R2 = 364.4k

Bingo. 360k in stock.

360k + 1% = 363.3, a good 1k less than the max.
360k - 1% = 356.4k. Span is about 10k or so potentially here.

Dig in junk box, find 10k multi turn trimmer.

Thus:

R1 = 2.2K
R2 = 360k in series with 10k multi turn pot.

desired output voltage is 200v, min load is 300uA so this is not good enough. Whack a 100k resistor on the output to stop it oscillating.

Edit: I solved for 200v there because I didn't read the question properly