Test Equipment Anonymous (TEA) group therapy thread

[Atomillo]

Well, here are the results.

The unit was turned on at around 9:30.
Measurements were taken starting at 10:30.

10:30 -> 0.010333mV
11:00 -> 0.013620mV
11:15 -> 0.014584mV
11:30 -> 0.015267mV
11:45 -> 0.015850mV
12:00 -> 0.016350mV

This was just using a very short piece of copper wire at the input. Connecting the input cable and then shorting didn't really change much. Placing over the short a piece of aluminum paper to avoid any possible interference also didn't change much.

The offset increased with time. Visually, the there was an increase of around 1 nanovolt every 2 or 3 seconds. Sometimes it would decrease by 4 or 5 nanovolts but then it would continue to increase.

I don't really know what to do. If it were noise, then it would not increase over time. Thus, I suspect a thermal effect. But the fact that it didn't change with the input connector and that it didn't stabilize after 2 hours makes me suspect the input connector.

 Might be it has some copper oxide. So, I opened it up to perform a quick check. Unfortunately, the connection is protected by shrink tube, so no way to tell right away. And I don't feel very comfortable going deeper.

Any more advice would be very valuable. Seeing existing material (mostly from TiN) shows an offset of around 100nV is normal. However, what I have here is two orders of magnitude bigger!

PS: the unit has a sticker saying it was calibrated in 2020. The seller also advertised this.

[mnementh]

I would still suggest that you cannot be sure one way or another until you buy/make/borrow the correct shorting plug. It really is that important. This would also give you some ammunition in starting a dispute if it actually is reading wrong; you can't expect the seller to take you seriously on a piece of gear like this when you try to fake your way through the zeroing procedure with a paper clip. 

mnem
 

[TERRA Operative]

Anyone here in USA able to do me a freight-forwarding favour?
I have a small-ish list of stuff I want to buy but shipping costs are stupid for most of it.. If someone can receive some stuff and send it on to me in Japan, I'll love you long time (and send you some extra $$ or the trouble).

[Vince]

Wow looks likze I finally found my bench DMM !

https://www.leboncoin.fr/equipements_industriels/2104828351.htm

A 34401A DMM for half price ?!  250 Euros ! Guy pretends it works... looks like new, it's a modern Agilent branded one, VFD looks even an bright... what's the catch ? These HP DMM/counter/DDS usually sell for twice that, 500 Euros..; or have prices plummeted recently and nobody told me ?

Maybe it's fucked .. but then same goes for all the other ones sold at 500 so...

i contacted the seller but after a small chat it looked like a scam to me ... be careful

More and more scam on this website regarding Lab Equipment.

Ah thanks, you got luckier than me then... guy still has not even read my message 

I guess I better stop dreaming and keep my money that I don't even have then...
DC loads and bench DMM are not in short supply, I can always buy one any time later.
It's not like it was a rare vintage piece of gear that pops up once every 20 years...

[Saskia]

received commercial invoice and tracking information for the N6700. He's sending per UPS and UPS already announced delivery for Thursday.

which brings me to the question: is 600 yankee bucks a good price for a Keysight N6700 ?

A quick Gurrgle shows the 6700C going for ~$3K; I see a refurb one for $2400. I think you got a decent deal, yeah; as long as it isn't going to sit unused. If you don't need it, obviously a good arbitrage acquisition.

mnem
 

*need* ... no
But I did not want to let the modules I have go unused. And without the mainframe they are quite useless.

[Kosmic]

Hi:

I'm here once again to ask probably stupid questions!

The Keithley 181 nanovoltmeter arrived today. After opening it to change the mains voltage to 230V as well as to inspect everything is ok, I power it up. I measure some normal voltages and everything is fine. Here comes the weird thing. I try to zero the meter by connecting the input cable and shorting it. When I made this measurement, the unit had been powered on for 40 minutes with the cable connected in order to let everything stabilize. And yet, look at that result! VERY far from zero. Is that normal?

I also tried a small piece of copper shorting the inputs. Letting it stabilize for 10 minutes gave me a similar result.

I'm unfamiliar with low level voltage measurements. I've watched Marco Reps video and he's able to get very close to zero. I've taken the same precautions and yet it doesn't seem to matter.

I would greatly appreaciate any advice

Might come from your test lead. Doesn't look original. The connector should be silver and cable black and relatively thin. Keithley was using a low noise triax cable.

Like so


When I had a Keithley 181, I was able to get +-60nV with my home made test lead.
https://www.eevblog.com/forum/metrology/possible-low-thermal-emf-connectors/msg1936489/#msg1936489

It might very well be... It doesn't look at all like what you show. I thought it was original because the alligator clips seem "coppery".
However the input connector just looks white, no sign of copper. Might be the alligator clips are original but the connector itself is not.

I've just turned the unit on with a short using a very short piece of copper wire. Reading the manual, it recommends a warm up of at least 1 hour for stated accuracy and, more interestingly, "To guarantee low drift, allow at least four hours".

Will post results in a couple of hours!

Yeah that test lead is clearly not original. Cables are way too big, look unshielded. I would try salvage the connector (look like a Amphenol MS3106A16-11P) and find some better cable.

[Zucca]

*need* ... no
But I did not want to let the modules I have go unused. And without the mainframe they are quite useless.

bd139 would buy it and flip it on ebay.... and drink beers with the profit.

[Neper]

I had a workmate with a Triumph Stag.............

One of the most beautiful useless cars ever built. I remember a dark red convertible with white leather at the British barracks in Hameln. A dream...

[Kosmic]

Well, here are the results.

The unit was turned on at around 9:30.
Measurements were taken starting at 10:30.

10:30 -> 0.010333mV
11:00 -> 0.013620mV
11:15 -> 0.014584mV
11:30 -> 0.015267mV
11:45 -> 0.015850mV
12:00 -> 0.016350mV

This was just using a very short piece of copper wire at the input. Connecting the input cable and then shorting didn't really change much. Placing over the short a piece of aluminum paper to avoid any possible interference also didn't change much.

The offset increased with time. Visually, the there was an increase of around 1 nanovolt every 2 or 3 seconds. Sometimes it would decrease by 4 or 5 nanovolts but then it would continue to increase.

I don't really know what to do. If it were noise, then it would not increase over time. Thus, I suspect a thermal effect. But the fact that it didn't change with the input connector and that it didn't stabilize after 2 hours makes me suspect the input connector.

 Might be it has some copper oxide. So, I opened it up to perform a quick check. Unfortunately, the connection is protected by shrink tube, so no way to tell right away. And I don't feel very comfortable going deeper.

Any more advice would be very valuable. Seeing existing material (mostly from TiN) shows an offset of around 100nV is normal. However, what I have here is two orders of magnitude bigger!

PS: the unit has a sticker saying it was calibrated in 2020. The seller also advertised this.

- Make sure your short is clean with no copper oxide.
- Put the short in place and protect from air draft.
- Let it cook (powered on) for at least 1 day.
- Then Zero the meter and note deviation over multiple hours.
- Keep ambient temperature constant.

Remember that each time you touch your copper short, you will induce thermal EMF over multiple minutes.

[Vince]

VARTA battery : woke up, I just disconnected it. Been disconnected for 27 Hours now, quite a bit... voltage  barely dropped. Only 50mV drop in the past 10 hours... it's indeed stabilized at about 9.6V.

So now I can discharge it and keep a close eye on it to make sure I don't let it go completely flat, and will take note of time stamps and voltage.

Stay tuned...

[BU508A]

Well, here are the results.

The unit was turned on at around 9:30.
Measurements were taken starting at 10:30.

10:30 -> 0.010333mV
11:00 -> 0.013620mV
11:15 -> 0.014584mV
11:30 -> 0.015267mV
11:45 -> 0.015850mV
12:00 -> 0.016350mV

This was just using a very short piece of copper wire at the input. Connecting the input cable and then shorting didn't really change much. Placing over the short a piece of aluminum paper to avoid any possible interference also didn't change much.

The offset increased with time. Visually, the there was an increase of around 1 nanovolt every 2 or 3 seconds. Sometimes it would decrease by 4 or 5 nanovolts but then it would continue to increase.

I don't really know what to do. If it were noise, then it would not increase over time. Thus, I suspect a thermal effect. But the fact that it didn't change with the input connector and that it didn't stabilize after 2 hours makes me suspect the input connector.

 Might be it has some copper oxide. So, I opened it up to perform a quick check. Unfortunately, the connection is protected by shrink tube, so no way to tell right away. And I don't feel very comfortable going deeper.

Any more advice would be very valuable. Seeing existing material (mostly from TiN) shows an offset of around 100nV is normal. However, what I have here is two orders of magnitude bigger!

PS: the unit has a sticker saying it was calibrated in 2020. The seller also advertised this.

- Make sure your short is clean with no copper oxide.
- Put the short in place and protect from air draft.
- Let it cook (powered on) for at least 1 day.
- Then Zero the meter and note deviation over multiple hours.
- Keep ambient temperature constant.

Remember that each time you touch your copper short, you will induce thermal EMF over multiple minutes.

I'd iike to add the following things:
- wear gloves
- clean ALL contact points! this includes the copper crocos, the copper piece, the connectors ...
- NO RF stuff around (WLAN, mobile phone, power switiching devices ...)
- thermal shielding on all sensitive points

Do the measuring several time. Let all the stuff settle.
Stop breathing, if possible. Ok, don't take this too seriously but you'll get the idea

[Atomillo]

Well, here are the results.

The unit was turned on at around 9:30.
Measurements were taken starting at 10:30.

10:30 -> 0.010333mV
11:00 -> 0.013620mV
11:15 -> 0.014584mV
11:30 -> 0.015267mV
11:45 -> 0.015850mV
12:00 -> 0.016350mV

This was just using a very short piece of copper wire at the input. Connecting the input cable and then shorting didn't really change much. Placing over the short a piece of aluminum paper to avoid any possible interference also didn't change much.

The offset increased with time. Visually, the there was an increase of around 1 nanovolt every 2 or 3 seconds. Sometimes it would decrease by 4 or 5 nanovolts but then it would continue to increase.

I don't really know what to do. If it were noise, then it would not increase over time. Thus, I suspect a thermal effect. But the fact that it didn't change with the input connector and that it didn't stabilize after 2 hours makes me suspect the input connector.

 Might be it has some copper oxide. So, I opened it up to perform a quick check. Unfortunately, the connection is protected by shrink tube, so no way to tell right away. And I don't feel very comfortable going deeper.

Any more advice would be very valuable. Seeing existing material (mostly from TiN) shows an offset of around 100nV is normal. However, what I have here is two orders of magnitude bigger!

PS: the unit has a sticker saying it was calibrated in 2020. The seller also advertised this.

- Make sure your short is clean with no copper oxide.
- Put the short in place and protect from air draft.
- Let it cook (powered on) for at least 1 day.
- Then Zero the meter and note deviation over multiple hours.
- Keep ambient temperature constant.

Remember that each time you touch your copper short, you will induce thermal EMF over multiple minutes.

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

[Atomillo]

Well, here are the results.

The unit was turned on at around 9:30.
Measurements were taken starting at 10:30.

10:30 -> 0.010333mV
11:00 -> 0.013620mV
11:15 -> 0.014584mV
11:30 -> 0.015267mV
11:45 -> 0.015850mV
12:00 -> 0.016350mV

This was just using a very short piece of copper wire at the input. Connecting the input cable and then shorting didn't really change much. Placing over the short a piece of aluminum paper to avoid any possible interference also didn't change much.

The offset increased with time. Visually, the there was an increase of around 1 nanovolt every 2 or 3 seconds. Sometimes it would decrease by 4 or 5 nanovolts but then it would continue to increase.

I don't really know what to do. If it were noise, then it would not increase over time. Thus, I suspect a thermal effect. But the fact that it didn't change with the input connector and that it didn't stabilize after 2 hours makes me suspect the input connector.

 Might be it has some copper oxide. So, I opened it up to perform a quick check. Unfortunately, the connection is protected by shrink tube, so no way to tell right away. And I don't feel very comfortable going deeper.

Any more advice would be very valuable. Seeing existing material (mostly from TiN) shows an offset of around 100nV is normal. However, what I have here is two orders of magnitude bigger!

PS: the unit has a sticker saying it was calibrated in 2020. The seller also advertised this.

- Make sure your short is clean with no copper oxide.
- Put the short in place and protect from air draft.
- Let it cook (powered on) for at least 1 day.
- Then Zero the meter and note deviation over multiple hours.
- Keep ambient temperature constant.

Remember that each time you touch your copper short, you will induce thermal EMF over multiple minutes.

I'd iike to add the following things:
- wear gloves
- clean ALL contact points! this includes the copper crocos, the copper piece, the connectors ...
- NO RF stuff around (WLAN, mobile phone, power switiching devices ...)
- thermal shielding on all sensitive points

Do the measuring several time. Let all the stuff settle.
Stop breathing, if possible. Ok, don't take this too seriously but you'll get the idea

I actually didn't think about the gloves. I have limited experience with low currents where such precautions are fundamental and indispensable. However I was unaware it was not necessary for low voltages. If I understand right, is it because the organic residue can also act as an union of two materials and generate thermal EMFs?

[Vince]

- I just got a cool HP 3200B 500MHz oscillator that's got a defective attenuator.
- A TDS 544A scope that can't trigger (or other problem ?) above 400MHz... reducing the usefulness of the scope somewhat...
- A TDS 694C that needs a keypad fixing...
- A 2467B that escaped the magic smoke.

I can tell you a little hack to increase the bandwidth of the TDS scopes, even upgrade them to higher models...
Maybe leaving them at the current model numbers but removing the bandwidth limiting capacitors on the acquisition boards would help a bit...

Did you recently run SPC too?

Also, one favour to ask, can you tell me what firmware you have on your units? I might task you to extract it for my firmware collection if they are versions I don't have..... It's easy with a GPIB interface and a command line app.

I don't think B/W is the problem here, plenty enough signal amplitude.. looks more like a tired trigger circuit and I have no idea how to fix / diagnose that... no schematics around for the 544A, well unless it leaked since 5 years ago when I worked on it. Think I found schematics from a 520 or something, something like that... better than nothing... but still only partial schematic, so don't don't know it they show the trigger circuitry...

Anyway regardless, would be cool upping the B/W if it's just a cap to remove on the front end.. even if in practice it's not going to make much of a difference since the sampling rate and sweep sweep won't increase accordingly to make use of the extra B/W....
But for repetitive signals I guess it could still be somewhat useful.

Bring it on, give me instructions ! 

SPC ? No I didn't do that... I ran it 10 times in a row 5 years ago after fixing the scope, to make sure it was working fine... but have not run it ever since... OK will try that, though from memory it has nothing to do with triggering, SPC only compensates for small temperature induced DC offsets in the signal path ?  Just from memory...
But it's free to try so will do !


FW versions ?
OK just powered them up to check... here goes :

TDS 544A:   v3.8.3e with options 13 1F M
TDS 694C:   v6.4e   with options 13 1F 2F


Would be cool to extract the FW through GPIB, never done it ! Gimme the S/W and instructions even if you are not interested in my crusty old FW ! 

[ch_scr]

(...)

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

I have seen one of these China "Keysight" USB GPIB... They are very impressive, come in a convincing cardboard box, all the paperwork is inside like expected, the device itself looks spotless.
If you unboxed new Agilent/Keysight stuff, it looks 100% the same.
There have been some troubles with too small traces breaking inside, but that was a few years ago - I would guess they have it worked out by now?

Edit: If you just want data from one device and are not dependent on Labview / other premade software, just go for Arduino GPIB - it'll show up as a usb serial and translate the messages.
I built this version twice and used it with 3 devices successfully.
"Built" is an overstatement, just wire an arduino onto the connector...

[Atomillo]

(...)

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

I have seen one of these China "Keysight" USB GPIB... They are very impressive, come in a convincing cardboard box, all the paperwork is inside like expected, the device itself looks spotless.
If you unboxed new Agilent/Keysight stuff, it looks 100% the same.
There have been some troubles with too small traces breaking inside, but that was a few years ago - I would guess they have it worked out by now?

Did they actually work?
I've seen this:
https://www.ebay.es/itm/Applicable-for-82357b-GPIb-to-USB-Card-82357b-IEEE-488-test-lead-/372691758243?mkcid=16&mkevt=1&_trksid=p2349624.m46890.l6249&mkrid=707-127634-2357-0

Where they pixelate the Keysight logo. It also helps that it has guarantee.

[Saskia]

*need* ... no
But I did not want to let the modules I have go unused. And without the mainframe they are quite useless.

bd139 would buy it and flip it on ebay.... and drink beers with the profit.

"Beastie Boys"

Flip it !!
you know you got stuff that you don't really need !
you got that ebay account so you better take heed !
...

[ch_scr]

(...)

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

I have seen one of these China "Keysight" USB GPIB... They are very impressive, come in a convincing cardboard box, all the paperwork is inside like expected, the device itself looks spotless.
If you unboxed new Agilent/Keysight stuff, it looks 100% the same.
There have been some troubles with too small traces breaking inside, but that was a few years ago - I would guess they have it worked out by now?

Did they actually work?
I've seen this:
https://www.ebay.es/itm/Applicable-for-82357b-GPIb-to-USB-Card-82357b-IEEE-488-test-lead-/372691758243?mkcid=16&mkevt=1&_trksid=p2349624.m46890.l6249&mkrid=707-127634-2357-0

Where they pixelate the Keysight logo. It also helps that it has guarantee.

Sadly, we did not test it thoroughly. At first glance it sure did seem to work, was recognized by all the software. It looked more like this.
IMHO this is only a "good investment" IF you are dependent on Labview / other premade software that expects this kind of device.
See the edit on my comment above if you are not scared to write e.g. a python script to get your data - less waiting & much cheaper.

[Kosmic]

(...)

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

I have seen one of these China "Keysight" USB GPIB... They are very impressive, come in a convincing cardboard box, all the paperwork is inside like expected, the device itself looks spotless.
If you unboxed new Agilent/Keysight stuff, it looks 100% the same.
There have been some troubles with too small traces breaking inside, but that was a few years ago - I would guess they have it worked out by now?

Edit: If you just want data from one device and are not dependent on Labview / other premade software, just go for Arduino GPIB - it'll show up as a usb serial and translate the messages.
I built this version twice and used it with 3 devices successfully.
"Built" is an overstatement, just wire an arduino onto the connector...

I did a comparison between an original 82357B and clone a while ago. They are different.
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3232212/#msg3232212

Also, the clone was getting a lot hotter. The cause of the suspectly high failure rate ?
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3232236/#msg3232236

[m k]

Wow looks likze I finally found my bench DMM !

https://www.leboncoin.fr/equipements_industriels/2104828351.htm

A 34401A DMM for half price ?!  250 Euros ! Guy pretends it works... looks like new, it's a modern Agilent branded one, VFD looks even an bright... what's the catch ? These HP DMM/counter/DDS usually sell for twice that, 500 Euros..; or have prices plummeted recently and nobody told me ?

Maybe it's fucked .. but then same goes for all the other ones sold at 500 so...

i contacted the seller but after a small chat it looked like a scam to me ... be careful

More and more scam on this website regarding Lab Equipment.

Ah thanks, you got luckier than me then... guy still has not even read my message 

I guess I better stop dreaming and keep my money that I don't even have then...
DC loads and bench DMM are not in short supply, I can always buy one any time later.
It's not like it was a rare vintage piece of gear that pops up once every 20 years...

You can ask Ray what he wants for Fluke 8505/6A with GPIB.

[Kosmic]

(...)

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

I have seen one of these China "Keysight" USB GPIB... They are very impressive, come in a convincing cardboard box, all the paperwork is inside like expected, the device itself looks spotless.
If you unboxed new Agilent/Keysight stuff, it looks 100% the same.
There have been some troubles with too small traces breaking inside, but that was a few years ago - I would guess they have it worked out by now?

Did they actually work?
I've seen this:
https://www.ebay.es/itm/Applicable-for-82357b-GPIb-to-USB-Card-82357b-IEEE-488-test-lead-/372691758243?mkcid=16&mkevt=1&_trksid=p2349624.m46890.l6249&mkrid=707-127634-2357-0

Where they pixelate the Keysight logo. It also helps that it has guarantee.

Sadly, we did not test it thoroughly. At first glance it sure did seem to work, was recognized by all the software. It looked more like this.
IMHO this is only a "good investment" IF you are dependent on Labview / other premade software that expects this kind of device.
See the edit on my comment above if you are not scared to write e.g. a python script to get your data - less waiting & much cheaper.

This GPIB project look interesting
https://github.com/xyphro/UsbGpib

You can't connect a long chain of devices, but apparently work with NI VISA, Labview, Matlab or PyVisa.

[TERRA Operative]

I don't think B/W is the problem here, plenty enough signal amplitude.. looks more like a tired trigger circuit and I have no idea how to fix / diagnose that... no schematics around for the 544A, well unless it leaked since 5 years ago when I worked on it. Think I found schematics from a 520 or something, something like that... better than nothing... but still only partial schematic, so don't don't know it they show the trigger circuitry...

Anyway regardless, would be cool upping the B/W if it's just a cap to remove on the front end.. even if in practice it's not going to make much of a difference since the sampling rate and sweep sweep won't increase accordingly to make use of the extra B/W....
But for repetitive signals I guess it could still be somewhat useful.

Bring it on, give me instructions ! 

SPC ? No I didn't do that... I ran it 10 times in a row 5 years ago after fixing the scope, to make sure it was working fine... but have not run it ever since... OK will try that, though from memory it has nothing to do with triggering, SPC only compensates for small temperature induced DC offsets in the signal path ?  Just from memory...
But it's free to try so will do !


FW versions ?
OK just powered them up to check... here goes :

TDS 544A:   v3.8.3e with options 13 1F M
TDS 694C:   v6.4e   with options 13 1F 2F


Would be cool to extract the FW through GPIB, never done it ! Gimme the S/W and instructions even if you are not interested in my crusty old FW ! 

Here's a link to TDS554A schematics, in the sidebar on the right:
https://w140.com/tekwiki/wiki/TDS544A

Here's a link to some info on the TDS694C, it mentions a common issue with the triggering, maybe yours needs some love?
https://w140.com/tekwiki/wiki/TDS694

For the downloading of the firmware (and uploading of newer firmware), check this link:
https://www.eevblog.com/forum/repair/unified-tektool-released-!-(firmware-flash-tools-for-old-tds-series)/msg3758918/#msg3758918

I'll take both those firmwares if you can extract them.
The newest TDS6xx firmware I have is 4.2e, and I don't have TDS5xx 3.8.3e yet either. I've attached 3.8.4e firmware if you want to do a little upgrade too. It will take about 45 mins to upload, downloading is much faster (make sure to pull down the NVRAM as well as the firmware as in the instructions, that way it's basically impossible to brick your scope).


Also, follow this one to enable extra options:
Try the 2M option on the TDS694C, if it doesn't work, set it back to 0 and try 1M, and set 2F on the TDS544A.
http://www.ko4bb.com/getsimple/index.php?id=enable-tds754d-options

[mnementh]

Since counters are a hot topic at the moment, I came across this one in the Toronto area:

https://www.kijiji.ca/v-buy-sell-other/city-of-toronto/hp-5248l-electronic-counter-for-parts-only/1559781767

"We specialize in electronics" and "Could not test it because we do not have a power cable" 

Some of their postings say they will accept any reasonable offer; some ads do not.
They probably "specialize" in finding a cheap source or are involved in disposal/recycling...

Did the Dwagon ever check out this store during his sojourn to the GWN?

They're EcoTech Electronics; a shoestring eCycling shop much like the one I did business with in San Antonio.



They do mostly reclamation of fleet computer gear that's too old/fuxxored up to sell on fleaBay as "off-lease". 25CPS came and visited me in the fall of 2020; we went there and I bought some laptop packs for cells and he got, IIRC, some hollow-state goodness similar to the one in this listing. They're legit, and willing to make a deal on most of what they have (cash is your best negotiation tool; nudge-nudge, wink-wink  ), but I wouldn't expect them to know anything about any TE that might come through their hands; certainly not how to test it in any manner we'd be comfortable with as being better than "where-is/as-is".

Shia's the resident nerd, but his specialty is recycled server RAM/CPUs, not TE in any flavor. I believe him when he says he doesn't have the cord for that; its a PH-163, IIRC. He'll have C5 & C13 cords by the MFT; any PH-163 he might have would be completely by accident.

 

They're also where I got the Dell server PSUs I made these from for my deal with you; BTW, have they found a good use yet, or still in the "pending" paddock?

mnem
 

[TERRA Operative]

(...)

I will do! I thought of covering it with a small pot filled with cotton to act as thermal insulator.
The thermal EMFs after touching the short are very visible. In my case it took around 5 or 6 minutes to "change direction".
Now I'm starting to really wish having a GPIB USB adaptor... But new are extremely expensive, and the options from eBay coming from China from around 150 euros as "new" seem suspicious.

I have seen one of these China "Keysight" USB GPIB... They are very impressive, come in a convincing cardboard box, all the paperwork is inside like expected, the device itself looks spotless.
If you unboxed new Agilent/Keysight stuff, it looks 100% the same.
There have been some troubles with too small traces breaking inside, but that was a few years ago - I would guess they have it worked out by now?

Did they actually work?
I've seen this:
https://www.ebay.es/itm/Applicable-for-82357b-GPIb-to-USB-Card-82357b-IEEE-488-test-lead-/372691758243?mkcid=16&mkevt=1&_trksid=p2349624.m46890.l6249&mkrid=707-127634-2357-0

Where they pixelate the Keysight logo. It also helps that it has guarantee.

Mine works perfectly as far as I can tell. No issues so far, including writing multiple scope firmwares without issue which takes around 45 minutes per scope.

[BU508A]

Well, here are the results.

The unit was turned on at around 9:30.
Measurements were taken starting at 10:30.

10:30 -> 0.010333mV
11:00 -> 0.013620mV
11:15 -> 0.014584mV
11:30 -> 0.015267mV
11:45 -> 0.015850mV
12:00 -> 0.016350mV

This was just using a very short piece of copper wire at the input. Connecting the input cable and then shorting didn't really change much. Placing over the short a piece of aluminum paper to avoid any possible interference also didn't change much.

The offset increased with time. Visually, the there was an increase of around 1 nanovolt every 2 or 3 seconds. Sometimes it would decrease by 4 or 5 nanovolts but then it would continue to increase.

I don't really know what to do. If it were noise, then it would not increase over time. Thus, I suspect a thermal effect. But the fact that it didn't change with the input connector and that it didn't stabilize after 2 hours makes me suspect the input connector.

 Might be it has some copper oxide. So, I opened it up to perform a quick check. Unfortunately, the connection is protected by shrink tube, so no way to tell right away. And I don't feel very comfortable going deeper.

Any more advice would be very valuable. Seeing existing material (mostly from TiN) shows an offset of around 100nV is normal. However, what I have here is two orders of magnitude bigger!

PS: the unit has a sticker saying it was calibrated in 2020. The seller also advertised this.

- Make sure your short is clean with no copper oxide.
- Put the short in place and protect from air draft.
- Let it cook (powered on) for at least 1 day.
- Then Zero the meter and note deviation over multiple hours.
- Keep ambient temperature constant.

Remember that each time you touch your copper short, you will induce thermal EMF over multiple minutes.

I'd iike to add the following things:
- wear gloves
- clean ALL contact points! this includes the copper crocos, the copper piece, the connectors ...
- NO RF stuff around (WLAN, mobile phone, power switiching devices ...)
- thermal shielding on all sensitive points

Do the measuring several time. Let all the stuff settle.
Stop breathing, if possible. Ok, don't take this too seriously but you'll get the idea

I actually didn't think about the gloves. I have limited experience with low currents where such precautions are fundamental and indispensable. However I was unaware it was not necessary for low voltages. If I understand right, is it because the organic residue can also act as an union of two materials and generate thermal EMFs?

It is as always, when dealing with delicate high precision stuff. Little things become more important, the nearer you are at the edges of what's physically possible.

Fingerprints contain not only grease, but also various acids, salts, etc. These can generate electrical voltages through chemical reactions, for example, but can also change the metal surface through chemical reactions. This can lead to substances that intensify the Seebeck effect. If measurements are to be repeatable and comparable, then one must ensure that as few parameters as possible change between measurements. This in turn leads to the fact that one should follow a protocol as far as possible to ensure this. This also includes, among other things, keeping everything as clean as possible. As already said above: every little thing becomes important when you get into the µV or nA range.