Test Equipment Anonymous (TEA) group therapy thread

[tggzzz]

Here's the improved version of a 190VDC reference completed and mounted in a case. I changed the 4 turn trim with a 25 turn trim pot to give a finer adjustment. The reference is not perfect but it is significantly better than my old version. The old one would drift around as much as 2 to 4 volts which made it almost useless as a decent reference. With the improved regulation it drifts about +/- 70mV or about 140mV total.

140mV in 190V is ~700ppm.

Whether that is acceptable depends on what you are trying to do. It wouldn't satisfy the voltnuts, but arguably nothing can satisfy them

[bd139]

Some other universe with better constraints might 

[med6753]

Here's the improved version of a 190VDC reference completed and mounted in a case. I changed the 4 turn trim with a 25 turn trim pot to give a finer adjustment. The reference is not perfect but it is significantly better than my old version. The old one would drift around as much as 2 to 4 volts which made it almost useless as a decent reference. With the improved regulation it drifts about +/- 70mV or about 140mV total.

140mV in 190V is ~700ppm.

Whether that is acceptable depends on what you are trying to do. It wouldn't satisfy the voltnuts, but arguably nothing can satisfy them

For DMM's with trim pot calibration on the 200VDC range it's perfectly acceptable as long as I monitor it with a known good DMM. But it is NOT acceptable to calibrate DMM's requiring a "Calibration constant" such as the Tek 2465 DMS with the DMM. That DMM requires constants of exactly 190.00mV, 1.900V, 19.000V, 190.00V, and 500.0V. The first 3 I have. The 190V was "pray I catch it" but checking afterwards revealed it was a little off. The 500V? Forget it. With this update I CAN get it to stabilize at 190.003V or so for a few minutes so I now have a better chance of setting an accurate constant. And I am going to change the carbon comp resistors per bd139's suggestion and hopefully it will become even more stable.

And believe it or not I have some thoughts on a 500VDC constant. Now that is venturing into real dangerous territory.     

[med6753]

... so I guess he's picked up a little of our Kiwi twang.

That twang is quite persistent.  I was at a garage sale a few years ago and heard a lady there say something.  I turned to address her stating "You're a kiwi."  She looked at me somewhat startled and told me she left New Zealand 17 years ago.  She obviously thought that would be long enough to lose that accent.

I don't think it ever goes away.

Just like a NYC (and upper New Jersey) accent. It NEVER goes away and I can pick it out a mile away.   

Yup, upper NJ for me.  I've been in Florida for 29 years.  The accent is less but still there.  Of course, I can easily exaggerate it if necessary.  Fuggeddaboudit.

My paternal grandmother was originally from Jersey City, NJ and she used to say "uril" for "oil" and "turlet" for "toilet". My younger brother would tease her by running through the house yelling "turlet, turlet!" and it would piss her off to no end. 

[med6753]

Try swapping out the resistors for ~1W metal film 1% resistors. Carbon ones drift terribly when they get warm. 250-500 ppm/oC at average. You can get that down to 50ppm/oC and use larger resistors that dont get warm to dodge it.

If you want to play with this idea, solder a 100 ohm cheap CF resistor on the end of a couple of bits of wire and plug it into a handheld DMM and wave it over a flame. Then do the same with an MF one.

Hmmm...maybe go one better with 0.1% resistors. The next Mouser order will include some. 

Don’t really need 0.1% as you can cal the thing out with your trimmer. Main thing is low ppm drift. If you can do better than 50ppm put your money into that.

Also if it doesn’t get any better then the IC is probably causing it. Glue it to a big chunk of metal

After running for a few hours I'm definitely seeing thermal drift. If I put the top on the case the voltage sags to about 189.082V. Remove the cover and it jumps back up to 190.035V and higher.

[bd139]

If you’ve got some freezer spray squirt the resistors and the IC carefully and independently and work out what happens. That should point to the biggest offender.

[GreyWoolfe]

... so I guess he's picked up a little of our Kiwi twang.

That twang is quite persistent.  I was at a garage sale a few years ago and heard a lady there say something.  I turned to address her stating "You're a kiwi."  She looked at me somewhat startled and told me she left New Zealand 17 years ago.  She obviously thought that would be long enough to lose that accent.

I don't think it ever goes away.

Just like a NYC (and upper New Jersey) accent. It NEVER goes away and I can pick it out a mile away.   

Yup, upper NJ for me.  I've been in Florida for 29 years.  The accent is less but still there.  Of course, I can easily exaggerate it if necessary.  Fuggeddaboudit.

My paternal grandmother was originally from Jersey City, NJ and she used to say "uril" for "oil" and "turlet" for "toilet". My younger brother would tease her by running through the house yelling "turlet, turlet!" and it would piss her off to no end. 

Small world.  I'm from Bayonne and went to what was then St Peter's College in Jersey City back in the 70's  I have never spoke like that, as far as I know and I have never said Joisey either.  My accent has more of a Brooklyn tinge to it as my father was born there.

[mnementh]

It's in the water. 

mnem
*X-Mas-ing like a foo*

[mnementh]

If you’ve got some freezer spray squirt the resistors and the IC carefully and independently and work out what happens. That should point to the biggest offender.

Squirt me with freezer spray and I'll point to the biggest offender too... 

mnem
Okay... it's me.

[med6753]

If you’ve got some freezer spray squirt the resistors and the IC carefully and independently and work out what happens. That should point to the biggest offender.

I do have freeze spray but the spray pattern is too damn big to really isolate individual components. I'll start replacing the resistors and take it from there. 

[mnementh]

Find a spray button from a WD-40 or similar lube that fits or that you can carve with a razor blade to fit in the valve hole on your can. Then you can use the tube to pinpoint spray individual components. I do this with the little canned airs from the dollar store.

mnem
X-mas tree is up! w00t! w00t! w00t!

[bd139]

You can also spray it on a cotton swab and touch the component with it as well. Also with the added bonus of a plastic insulator between you and the other end of it where all the volts are hiding.

No Christmas tree up yet. I just managed to undo about 2 weeks' worth of damage from my slightly over-creative 6 year old. Hopefully we'll get the tree up before she trashes the place again. Tree tomorrow, maybe.

[tautech]

You can also spray it on a cotton swab and touch the component with it as well. Also with the added bonus of a plastic insulator between you and the other end of it where all the volts are hiding.

No Christmas tree up yet. I just managed to undo about 2 weeks' worth of damage from my slightly over-creative 6 year old. Hopefully we'll get the tree up before she trashes the place again. Tree tomorrow, maybe.

Crayons ?

[bd139]

Actually no. Scissors, tape, paper, pencils, pens, paint, glitter, card, pom poms, stick on eyes, chalk, toilet rolls, slime, dry markers, books, shells, rocks. That's just what I can still see. Non stop craft machine.

[Mr. Scram]

I just got a bunch of GPIB cables I'd like to test. Considering the recommended maximum cable length between devices is 4 metres, I think I can just daisy chain them together and stick a device on the end as long as I don't go over those 4 metres total to test the lot. Right?

[Neomys Sapiens]

When I get back to the 6114A again I will have to do something.   

You need to get that power supply out of your way! Can I help?

[beanflying]

I just got a bunch of GPIB cables I'd like to test. Considering the recommended maximum cable length between devices is 4 metres, I think I can just daisy chain them together and stick a device on the end as long as I don't go over those 4 metres total to test the lot. Right?

I thought it was maximum of 4m between devices and 20m (I think) total length. So providing there is an active device every 4m should be fine daisy chaining leads.

[tggzzz]

I do have freeze spray but the spray pattern is too damn big to really isolate individual components. I'll start replacing the resistors and take it from there.

Old doctors technique for freezing warts...

Take some acetone, discharge a CO2 capsule through it to create acetone slush. Saturate cotton wool bud in the slush and apply to the wart.

The CO2 can be replaced by freezer spray. Acetone could be used, as could anything that turns to (preferably non-conducting) slush. Wart can be replaced by a resistor

[Mr. Scram]

I thought it was maximum of 4m between devices and 20m (I think) total length. So providing there is an active device every 4m should be fine daisy chaining leads.

That's what I thought too. It saves me going through every cable separately, providing everything works.

[tggzzz]

... glitter ...

That's worse than sand. It gets everywhere, e.g. into Tek leaf switches

When I get a Christmas card from one relative, I always open the envelope over the bin, so that the loose glitter/stars/etc doesn't hang around and annoy me until the new year.

[URI]

When I get back to the 6114A again I will have to do something.   

You need to get that power supply out of your way! Can I help?

Thanks, but -no. 

I'm just stuck with it cause my job ate all my free time the last months.   
Hope to have some free time for the 6114A over Xmas and new year.

I have a pair of them and more plans for the two of those...   

[mnementh]

Here's the improved version of a 190VDC reference completed and mounted in a case. I changed the 4 turn trim with a 25 turn trim pot to give a finer adjustment. The reference is not perfect but it is significantly better than my old version. The old one would drift around as much as 2 to 4 volts which made it almost useless as a decent reference. With the improved regulation it drifts about +/- 70mV or about 140mV total.

140mV in 190V is ~700ppm.

Whether that is acceptable depends on what you are trying to do. It wouldn't satisfy the voltnuts, but arguably nothing can satisfy them

Some other universe with better constraints might 

   We have perfectly serviceable constraints for them right here in this universe... 

mnem
*wibble*

[bitseeker]

..for bitseeker:

There's room on the bench reardless how crowded it is, e.g. on my bench there's now an inverted pyramid stack of TE...


...because I wanted to test the counter but wasn't eager for rearranging my bench/tidying it up.      
When I get back to the 6114A again I will have to do something.   

Well, I made some progress on rearranging stuff (still some junk in front of everything). At least I was able to get the listed-as-guaranteed-working 3456A into a new stack arrangement on my bench and plugged in. Unfortunately, it fails self-test #4 (A/D converter, inguard logic, input amplifier, or input switching). I haven't gone inside yet to narrow down the issue. Attempting to measure a small DC voltage makes the meter go into overload.



Another one for the queue!

It's very nice, cosmetically, though. And the buttons, although rubber rather than the old clicky kind, feel fine. Similar to the 3478A, but quite a bit larger and less wiggle when pressing them.

I still have to unbox the other 3456A, which was listed for parts. Maybe it'll be working, instead.

[bitseeker]

Here's the improved version of a 190VDC reference completed and mounted in a case. I changed the 4 turn trim with a 25 turn trim pot to give a finer adjustment. The reference is not perfect but it is significantly better than my old version. The old one would drift around as much as 2 to 4 volts which made it almost useless as a decent reference. With the improved regulation it drifts about +/- 70mV or about 140mV total. One heck of an improvement and I can live with that amount of drift. But I'm sure it can be improved. Perhaps an op-amp with a precision reference as a feedback loop to null out any drift. But I'm not smart enough to figure that out so for now I'll leave it as is. 

That's quite good, indeed. Nice job, med.

[bd139]

... glitter ...

That's worse than sand. It gets everywhere, e.g. into Tek leaf switches

When I get a Christmas card from one relative, I always open the envelope over the bin, so that the loose glitter/stars/etc doesn't hang around and annoy me until the new year.

It is. Terrible stuff. Also major pollutant as a lot of it ends up in the water supply processing and sea.  Being to grumpy old grinch bastard I am, I usually throw cards in the paper recycling straight away after surveying them for interesting contents over the bin. That’s the ones I get my hands on first anyway.

I think it was a couple of years ago an acquaintance sent me a “vegan” card (wtf) which had organic biodegradable glitter in it and proudly said so on the evelope. Saved me opening it