Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

...Must have had a major failure at some point, as the pass transistors have all been replaced with these.   

David

Seems pretty likely that Sprague doody-bomb died the way they often do... dead-short and kaboom. I'd shotgun that section m'self, including the one that was "replaced in the 70s".

mnem
Trust no one capacitor.

The two remaining Spraque caps on the regulator board were the originals from 1963, they weren't shorted and both me & the seller* had tested the counter with them still in place.

      Later on, the Peak ESR said they were past it, put some nice crappy NOS blue Philips ones in their place.

*there is a video on youtube of my counter from the seller.

David

You do like to live dangerously, don't you David...? 

mnem
*currently trying to think like a Scooba floor-mopping robot*

[mansaxel]

This is but the first step of a long journey, but  and keep  IMHO the direction is good 

Good advice from someone who obviously not is jumping on every Framework Trend. Thanks. I will run it, if I run it, on either Devuan or FreeBSD, in an ESXi vm. No Docker. That is not how we do things here!

[GreyWoolfe]

*well-known to tie a bobber and nothing else on the end of his line*

Didn't stop my son when he was 4 or 5 from catching a sunny with a stick and a piece of string.  Even pulled it up out of the water.

[ch_scr]

[...]
Yes stay away for any variant of evil raspi, they are design for having fun not to be stable.
I would like to smash HA on an old big and energy hungry server, which will also run FreeNAS.

oh man where I can find the time?

My impression is, that it's not the pi's (in this case an orangepi zero (node red) and an orangepi one (HA)) that are at fault but the software. An colleague has an orangepie zero in use as router/proxy in his home network for a similar timespan and he has not had stability issues.

[mnementh]

If it doesn't run on a QT-Py I'm not interested.

mnem
SAMD21: 32bits, 2MB and Python on a postage stamp... what's not to love? Even a total fuckwidget like me can make it do something.

[mnementh]

*well-known to tie a bobber and nothing else on the end of his line*

Didn't stop my son when he was 4 or 5 from catching a sunny with a stick and a piece of string.  Even pulled it up out of the water.

Yes, and those are the stories you tell the grand-kids. 

mnem
 

[factory]

...Must have had a major failure at some point, as the pass transistors have all been replaced with these.   

David

Seems pretty likely that Sprague doody-bomb died the way they often do... dead-short and kaboom. I'd shotgun that section m'self, including the one that was "replaced in the 70s".

mnem
Trust no one capacitor.

The two remaining Spraque caps on the regulator board were the originals from 1963, they weren't shorted and both me & the seller* had tested the counter with them still in place.

      Later on, the Peak ESR said they were past it, put some nice crappy NOS blue Philips ones in their place.

*there is a video on youtube of my counter from the seller.

David

You do like to live dangerously, don't you David...? 

mnem
*currently trying to think like a Scooba floor-mopping robot*

Well I figured it didn't blow-up when the seller powered it up, surprisingly those capacitors being pretty much dead didn't stop it from functioning when it arrived, did some quick checks to evaluate the condition at the time.
Oh this was in the middle of one of the lock-downs and I didn't won't to wait for back orders, plus I would like to use some of these parts I acquired from work, pity they weren't at least newer BC caps. 

David

[ch_scr]

If it doesn't run on a QT-Py I'm not interested.

mnem
SAMD21: 32bits, 2MB and Python on a postage stamp... what's not to love? Even a total fuckwidget like me can make it do something.

The joke is: after the initial struggle, it is a lot easier to do additional stuff. Example: add a sensor to the bathroom window to make the kitchen-window-lamp blink if you left it open for too long. That is very hard to do as an insulated problem solving thing (think microcontrollers, wires, relays...) - and all of your solution is non reusable in that case. If you have home automation, there exists a cheap (zigbee) sensor that is easy to integrate. You haggle with HA for an hour and it will make the ikea lamp blink now. But also, the lamp can be already turned on in the morning when you come to make breakfast and might turn off automagically in the evening at 10 PM when you have gone to bed. The timing can be additionally influenced by e.g.  solar sunset/sunrise. Stuff you don't do because it would need the really expensive timer clock and calculating in your DIY firmware would be nightmare. There are cheap ready-made wireless relays to mount behind your light switches.
Also we needed digital IO from HA to control 230V relays in the switchboard. For raspberry pi's there exist IO-pin drivers but not for orangepi's. But arduino to the rescue, there is an existing project called "Firmata" that standardizes "Digital IO over serial" that is as easy as uploading an sketch to an Arduino. And it already integrates seamlessly into HA as well. So your PC home server can have digital IO, with the added bonus that an mishap that fries the IO only killed a 5$ part...

[dl6lr]

Must have had a major failure at some point, as the pass transistors have all been replaced with these.

The rectifier and regulators boards are identical and they carry the numbers of the 5243.
Both counters share the same power supply with the three 2N301 pass transistors (I just found that I have 2 of them). The service manuals warns that these will go south along with the rectifiers on a short circuit (during measurements). The regulation is a bit strange with the +20V floating on the regulated +13V and the SM states to set the +20V with the control for the +7V regulator first. The SM states the reference zener to be 7V, but on my board it measures as 3.9V and it is different from the two other 7V zeners.

Don't you just love it when the documentation and what sits on your bench don't match? I see it ALL the time with Tek stuff. Tek was always changing/improving stuff and after all these years the correct docs to reflect those changes may not have been captured in someone's stash.

I am not to complain. The free BAMA manual is for a newer instrument with prefix 628. The manual says you should contact your -hp- sales representive for the service manual for your instrument prefix    I will order the Artek manual for my instrument.

[dl6lr]

The right way is to kill that "163" crap and install a "mickey mouse", ie. IEC 60320 C6. Unless you are building a museum. Me, I'm building a working shop with boat anchors, and I'll make them reasonably safe.

Have to order those plugs...

[dl6lr]

The two remaining Spraque caps on the regulator board were the originals from 1963, they weren't shorted and both me & the seller* had tested the counter with them still in place. Later on, the Peak ESR said they were past it, put some nice crappy NOS blue Philips ones in their place.



*there is a video on youtube of my counter from the seller.

David

Those are only taking any ripple from the feedback amplifiers reference. So if the voltage already is low ripple...
i am just collecting the caps from my inventory. Have most but not all, one of the cans have to be refilled with an axial instead of a radial type and for the boards it is the other way round for one of them. I just put down a note to restock the missing caps.

So for the weekend I hope to refill all the cans and replace the electrolytics on the boards.

[Zucca]

My impression is, that it's not the pi's (in this case an orangepi zero (node red) and an orangepi one (HA)) that are at fault but the software. An colleague has an orangepie zero in use as router/proxy in his home network for a similar timespan and he has not had stability issues.

could be, orange pi or not, everytime I touch a pi my face turn soon or later purple.
So I leave those teenager toys to somebody else I go like mansa straight to ESXi vm.

More than happy to pay more in electricity if this means mental sanity for me.

[mansaxel]

The right way is to kill that "163" crap and install a "mickey mouse", ie. IEC 60320 C6. Unless you are building a museum. Me, I'm building a working shop with boat anchors, and I'll make them reasonably safe.

Have to order those plugs...

Of course. There's a Schaffner part that's inexpensive and fits pretty well with minimal fettling in the "163" cutout.

[mansaxel]

My impression is, that it's not the pi's (in this case an orangepi zero (node red) and an orangepi one (HA)) that are at fault but the software. An colleague has an orangepie zero in use as router/proxy in his home network for a similar timespan and he has not had stability issues.

could be, orange pi or not, everytime I touch a pi my face turn soon or later purple.
So I leave those teenager toys to somebody else I go like mansa straight to ESXi vm.

More than happy to pay more in electricity if this means mental sanity for me.

I have a Pi for some 1-wire i/o via a i2c add-on card. It hangs partially from time to time, inasmuch as the OWFS is unreachable, and a reboot does not cure it. Only fix is to power cycle. Consequently, I would not run a server on it. An I/O unit, with some kind of watchdog function, that'll work.

My VmWare server is pretty slim in power consumption, much more so than the file server, which is a hog. I'm currently trying to remove one router from the power bill by reimplementing the routing I can't do in my core switch in a OpnSense firewall in a vm.  Power prices are insane right now, and will be for some time, I guess.

[25 CPS]

I need to go pick up the car from an oil change.  In the meantime, enjoy the pictures from today’s current clamp testing and I’ll edit the post once I’m back to cut in the commentary.  DC testing to follow on Sunday or Monday.

[mnementh]

If it doesn't run on a QT-Py I'm not interested.

mnem
SAMD21: 32bits, 2MB and Python on a postage stamp... what's not to love? Even a total fuckwidget like me can make it do something.

The joke is: after the initial struggle, it is a lot easier to do additional stuff. Example: add a sensor to the bathroom window to make the kitchen-window-lamp blink if you left it open for too long. That is very hard to do as an insulated problem solving thing (think microcontrollers, wires, relays...) - and all of your solution is non reusable in that case. If you have home automation, there exists a cheap (zigbee) sensor that is easy to integrate. You haggle with HA for an hour and it will make the ikea lamp blink now. But also, the lamp can be already turned on in the morning when you come to make breakfast and might turn off automagically in the evening at 10 PM when you have gone to bed. The timing can be additionally influenced by e.g.  solar sunset/sunrise. Stuff you don't do because it would need the really expensive timer clock and calculating in your DIY firmware would be nightmare. There are cheap ready-made wireless relays to mount behind your light switches.
Also we needed digital IO from HA to control 230V relays in the switchboard. For raspberry pi's there exist IO-pin drivers but not for orangepi's. But arduino to the rescue, there is an existing project called "Firmata" that standardizes "Digital IO over serial" that is as easy as uploading an sketch to an Arduino. And it already integrates seamlessly into HA as well. So your PC home server can have digital IO, with the added bonus that an mishap that fries the IO only killed a 5$ part...

Yes, but I don't want a smart home... especially a internet-connected home.

I've seen every episode of EUREKA!, and IMHO, it's invariably much less assache to just police doors and windows before I go to bed as I've done since I was twelve. That way I don't have to worry aboot my home deciding it's had enough of my dragon breath and directing the toaster to set fire to the drapes in the middle of the night. 

mnem
dumb home, smart family. 

[dl6lr]

The right way is to kill that "163" crap and install a "mickey mouse", ie. IEC 60320 C6. Unless you are building a museum. Me, I'm building a working shop with boat anchors, and I'll make them reasonably safe.

Have to order those plugs...

Of course. There's a Schaffner part that's inexpensive and fits pretty well with minimal fettling in the "163" cutout.

I have one from Schurter in my Digikey list. And Conrad has another one that could fit using a small adapter board.

[25 CPS]

So this was what was in the mystery Newark box that I was waiting for:



Fluke shrouded banana plug socket to male BNC adapters, extension leads and a set of Pomona test leads.  I thought about ordering regular Fluke test leads to replace the set missing from the 27 kit but decided a full Pomona set with the pouch might be a better idea.

With that in place, now I can get on with testing what I wanted to evaluate:





The Fluke current probe was a pawn shop find and the OTC was a Kijiji pickup.  First set of tests is into a multimeter on the mVAC range to see if they're working and giving plausible values.  They're into the X10 side of a line splitter with a 100W lightbulb as test load, bench lighting, and cold basement heat source.





Given the voltages and currents observed and taking into consideration these clamps are both being operated pretty much right at the bottom of their working ranges, the values which work out to a touch over 100W look reasonable for that lightbulb.  Now the next test was to see if it was possible to observe AC current waveforms with these clamps and that's where the Fluke adapters came in:



Fluke shows a nice clean trace on the Agilent.





The Fluke Scopemeter hits it out of the park compared to the Agilent though because you can define the probe as almost anything.  I've got it set up as a 10 mV/A current probe to take into account sitting in the X10 slot on the line splitter.  What I don't like is the trace thickness on the Scopemeter.  If it was a CRT scope, I'd be reaching for the focus control to tighten the trace up.

Up next is the OTC which was kind of disappointing:





The noise doesn't appear to affect measured values all that much but it sure is ugly when viewing the output graphically instead of numerically on a DMM.  It'll be interesting to see how it performs on direct current and at higher current values instead of loafing along at a bit less than an amp.  I did try toggling between the 200 and 2,000 amp ranges on the OTC to see if that made any difference but the output was unchanged.

As for the two portable scopes, I wish you could get it all in one place.  The Fluke's functionality combined with the Agilent's trace sharpness would be fantastic.

[mnementh]

I'm just waiting for bean to pop in and say MICSIG... 

mnem
 

[tautech]

I'm just waiting for bean to pop in and say MICSIG... 

mnem
 

These will be interesting when they get released.....and the isolated channel versions too. 

[med6753]

Speaking of measuring AC current. I have one of these. A Fluke LCA-10. Allows you to directly measure current of 120VAC devices up to 10A. I've only used it a few times. It's no longer available from Fluke and not sure why but I suspect it has run afoul with latest safety practices and inductive clamp is indeed safer. (It does have UL listing)

[Cubdriver]

Speaking of measuring AC current. I have one of these. A Fluke LCA-10. Allows you to directly measure current of 120VAC devices up to 10A. I've only used it a few times. It's no longer available from Fluke and not sure why but I suspect it has run afoul with latest safety practices and inductive clamp is indeed safer. (It does have UL listing)

It’s no longer available from Fluke?  Shocking!  (Pun intended, and given our litigious society and the prevalence of idiots these days I’m unsurprised by that.). Shame, it looks like a right handy little bit of kit.  Idiots and lawyers are IMO a large part of why we can’t have nice things.

-Pat

[med6753]

Speaking of measuring AC current. I have one of these. A Fluke LCA-10. Allows you to directly measure current of 120VAC devices up to 10A. I've only used it a few times. It's no longer available from Fluke and not sure why but I suspect it has run afoul with latest safety practices and inductive clamp is indeed safer. (It does have UL listing)

It’s no longer available from Fluke?  Shocking!  (Pun intended, and given our litigious society and the prevalence of idiots these days I’m unsurprised by that.). Shame, it looks like a right handy little bit of kit.  Idiots and lawyers are IMO a large part of why we can’t have nice things.

-Pat

Quick check of evil bay has 2 out there for real stupid money. 

[mansaxel]

The right way is to kill that "163" crap and install a "mickey mouse", ie. IEC 60320 C6. Unless you are building a museum. Me, I'm building a working shop with boat anchors, and I'll make them reasonably safe.

Have to order those plugs...

Of course. There's a Schaffner part that's inexpensive and fits pretty well with minimal fettling in the "163" cutout.

I have one from Schurter in my Digikey list. And Conrad has another one that could fit using a small adapter board.

Yes!    It's Schurter I mean. Sorry for thinking too much about exploding mains filters..

[Robert763]

I Built this for current measurement with a DMM or clamp meter.
It's just a C13/C14 lead cut in two and taken through a box.
The neutral is taken through the loop on top.
A variation is to replace the loop stain reliefs with 4mm safety sockets and a loop with two shrouded 4mm plugs. It can then be used with DMM or clamp meter.

Edit: my cased variac has C13/C14 connectors and I have leads with C14 on one end and a 13A socket of safeblock on the other.
Most of the power distribution around the workshop is by C13 outlet strips.