Test Equipment Anonymous (TEA) group therapy thread

[Cerebus]

I made progress on the little GPSDO NMEA data project. I managed to get some basic stuff  - time and # of satellites printed out the serial monitor (not on the tiny OLED screen yet) at least in a formatted manner. I'm using a neat library that makes it a lot easier - TinyGPS++ here -

http://arduiniana.org/libraries/tinygpsplus/

This is what my GPSDO sends out every second -

$GPVTG,,T,,M,0.128,N,0.237,K,D*2B
$GPGGA,111046.00,3243.81872,N,09728.40863,W,2,12,0.88,242.5,M,-24.8,M,,0000*60
$GPGSA,A,3,21,51,30,03,46,01,07,13,06,22,19,14,1.47,0.88,1.18*02
$GPGSV,5,1,18,01,31,048,28,03,07,093,20,06,26,190,28,07,15,155,28*75
$GPGSV,5,2,18,13,22,255,23,14,64,037,33,15,07,281,11,17,71,337,34*7D
$GPGSV,5,3,18,19,58,266,21,21,11,039,20,22,08,072,22,24,06,321,*74
$GPGSV,5,4,18,28,62,355,14,29,35,295,,30,48,170,31,46,39,229,29*7B
$GPGSV,5,5,18,48,42,224,33,51,50,198,31*7D
$GPGLL,3243.81872,N,09728.40863,W,111046.00,A,D*72
$GPRMC,111047.00,A,3243.81869,N,09728.40858,W,0.105,,160821,,,D*60

At first I was getting the same time stamp sent out from the TinyGPS "gps.time.isUpdated()" twice a second because it legitimately found the time listed two times in two different sentences (shown in bold above). I put in a test so that if the time had the same seconds to not parse it again. But I think now that I am more familiar with TinyGPS I can do it another way. Just started playing with it this morning. I need to see how much I can cram on the little screen. But if I need a bigger one I can obtain it at a reasonable cost.

I thought I'd try out the Arduino IDE Beta 2.0.0-beta11. When it is sent serial data from the running program its serial monitor starts getting real laggy, like maybe you might think you did something wrong (I thought I did a long time ago but that turned out to be an error on my part). Then I used another program to connect to the Arduino serial port and it ran perfectly fine for an hour. I guess that's why they call the Beta IDE "Experimental" 

Probabally bit banging a I/O pin (badly) rather than using the serial hardware on the chip. I once did firmware  for the engineering devlopment of a mechanical wiget that needed a sttere etc and simple Go/Stop control over serial (interface with legacy system). I'm no programmer can't get on with C and used PIC Basic Pro complied Basic. Took me a few hours to write. When the hardware was designed It went to the "proper" programers to write in C. Two weeks latr thay came back and asked for a PIC with more memory  My code used less than 1/4 of the available memory. I was of couse using the USART in the PIC for serial (simple 8 bit PIC with no pin mapping). Turns out they were using some bought in "universal" comms libary that did eveything by bit banging. I talked to them about the USART and they just looked blank.  The wiget went to production with the engineering code. It never needed any updates.

WTF. I wrote a little C++ class to drive those LED boards that I made to practice soldering tiny DFN chips. It abstracts the whole thing to set_led(number) and clear_led(number) calls and handles state storage and the whole bit banging thing behind the scenes. It's self contained apart from calls to GPIO routines, copes with either 1,2,3 or 4 boards daisy chained, has been tested at several clock rates and 'just works'.  It comes to a grand total of 124 lines including header file, with lots of comments and blank lines. If it compiles to more than 1k of actual code I'd be highly surprised. Just checked the linker files - 440 bytes dead.

When I wrote that I wasn't even bothering about efficiency in any sense. As long as it was simple, clear and worked I was going to be happy. Probably took 20 minutes and rounded up to a couple of hours, perhaps three, including building a physically running test rig and testing. Either I'm superdoubleplus good or, and I suspect this to be the actual case, you were dealing with people who would be better employed in Regents Park or Chessington eating peanuts.

[Cerebus]

Most suspect scope award goes to https://www.ebay.co.uk/itm/313671287713

Giblets... that is crusty...

Have a couple good ones here as well !

https://www.leboncoin.fr/bricolage/2037590332.htm

and

https://www.leboncoin.fr/collection/2035985058.htm


This one is clean but not so sure about value for money ! 

https://www.leboncoin.fr/collection/1845407573.htm

  ...?

Did everybody wake up and just randomly decide to have a "2-bagger" contest...?   

mnem
I never get the memo...

Is that a challenge to find a "Coyote" scope?

[Saskia]

made by ACME ?
or in Desert ( coyote ) coating ?
or a Wiley scope that has gone over the cliff ?

[tggzzz]

Ended up with a 5-decade resistance substitution box which I will probably flip (looks like it all works except the lowest resistor (10ohms). I can guess why that's now 70ohms, and hope the switch hasn't been buggered.

It wasn't buggered. The relevant resistor that was different to all the others, and there was a scorch mark on the case above it. Replaced.

All the resistors are 1% and well within spec, so that was a good buy at £5 - and much better than my previous 8-decade box that I promptly unloaded at a good profit.. Might not flip it; TBD.

Also rescued a Fluke 8300A DMM marked as "all nixies work". Without bothering to let it warm up, when my 34410A shows 10Vdc, this shows 10.0002V. The Vac settings look reasonable, the mVdc show 0, as do the ohms settings.

[img sidth=400]https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/?action=dlattach;attach=1269307[/img]

Presumably because it doesn't have the ohms/millivolt module (option 03) fitted.

Neither of mine have the ACV (01) module, one has the 03. Both are dead nuts against my recently cal'ed 8840A. Fluke for sure know how to build a meter!

Not having had time to look for a manual, I didn't realise those were options. Plus there's no indication of installed options on the case exterior. Opening it up shows that, as you expected, the mVdc/ohms option is not installed.

I'll get the manual, possibly replace the blue Philips electrolytics. I suspect the tweaking might be the zero offset and the scale for ranges other than 10V.

Since I got both just because they have nixies, I might have to decide whether to sell the 8300A or 8125A https://www.eevblog.com/forum/testgear/fluke-8125a-military-dmm-teardown/msg2540730/#msg2540730

[Specmaster]

I'm going back into the eBay pit again for yet another item, more details to come if I'm successful, not wishing to ruin my chances by one of you lot bidding against me 

Is it just me, or is eBay picking up again?

[Kosmic]

I made progress on the little GPSDO NMEA data project. I managed to get some basic stuff  - time and # of satellites printed out the serial monitor (not on the tiny OLED screen yet) at least in a formatted manner. I'm using a neat library that makes it a lot easier - TinyGPS++ here -

http://arduiniana.org/libraries/tinygpsplus/

This is what my GPSDO sends out every second -

$GPVTG,,T,,M,0.128,N,0.237,K,D*2B
$GPGGA,111046.00,3243.81872,N,09728.40863,W,2,12,0.88,242.5,M,-24.8,M,,0000*60
$GPGSA,A,3,21,51,30,03,46,01,07,13,06,22,19,14,1.47,0.88,1.18*02
$GPGSV,5,1,18,01,31,048,28,03,07,093,20,06,26,190,28,07,15,155,28*75
$GPGSV,5,2,18,13,22,255,23,14,64,037,33,15,07,281,11,17,71,337,34*7D
$GPGSV,5,3,18,19,58,266,21,21,11,039,20,22,08,072,22,24,06,321,*74
$GPGSV,5,4,18,28,62,355,14,29,35,295,,30,48,170,31,46,39,229,29*7B
$GPGSV,5,5,18,48,42,224,33,51,50,198,31*7D
$GPGLL,3243.81872,N,09728.40863,W,111046.00,A,D*72
$GPRMC,111047.00,A,3243.81869,N,09728.40858,W,0.105,,160821,,,D*60

At first I was getting the same time stamp sent out from the TinyGPS "gps.time.isUpdated()" twice a second because it legitimately found the time listed two times in two different sentences (shown in bold above). I put in a test so that if the time had the same seconds to not parse it again. But I think now that I am more familiar with TinyGPS I can do it another way. Just started playing with it this morning. I need to see how much I can cram on the little screen. But if I need a bigger one I can obtain it at a reasonable cost.

I thought I'd try out the Arduino IDE Beta 2.0.0-beta11. When it is sent serial data from the running program its serial monitor starts getting real laggy, like maybe you might think you did something wrong (I thought I did a long time ago but that turned out to be an error on my part). Then I used another program to connect to the Arduino serial port and it ran perfectly fine for an hour. I guess that's why they call the Beta IDE "Experimental" 

Probabally bit banging a I/O pin (badly) rather than using the serial hardware on the chip. I once did firmware  for the engineering devlopment of a mechanical wiget that needed a sttere etc and simple Go/Stop control over serial (interface with legacy system). I'm no programmer can't get on with C and used PIC Basic Pro complied Basic. Took me a few hours to write. When the hardware was designed It went to the "proper" programers to write in C. Two weeks latr thay came back and asked for a PIC with more memory  My code used less than 1/4 of the available memory. I was of couse using the USART in the PIC for serial (simple 8 bit PIC with no pin mapping). Turns out they were using some bought in "universal" comms libary that did eveything by bit banging. I talked to them about the USART and they just looked blank.  The wiget went to production with the engineering code. It never needed any updates.

WTF. I wrote a little C++ class to drive those LED boards that I made to practice soldering tiny DFN chips. It abstracts the whole thing to set_led(number) and clear_led(number) calls and handles state storage and the whole bit banging thing behind the scenes. It's self contained apart from calls to GPIO routines, copes with either 1,2,3 or 4 boards daisy chained, has been tested at several clock rates and 'just works'.  It comes to a grand total of 124 lines including header file, with lots of comments and blank lines. If it compiles to more than 1k of actual code I'd be highly surprised. Just checked the linker files - 440 bytes dead.

When I wrote that I wasn't even bothering about efficiency in any sense. As long as it was simple, clear and worked I was going to be happy. Probably took 20 minutes and rounded up to a couple of hours, perhaps three, including building a physically running test rig and testing. Either I'm superdoubleplus good or, and I suspect this to be the actual case, you were dealing with people who would be better employed in Regents Park or Chessington eating peanuts.

don't worry the size of your executable start growing quickly when you really start doing something with the MCU.

Bloated executable is normally cause by the usage of super generic libraries (different containers for example, array, vector, map etc ...). you end up linking super generic code that you don't necessarily use.

Specific code is always smaller and more efficient than generic one. It's a lot less reusable though. You need to balance both when developing software. I rarely used libraries when I was developing firmwares though.

[Saskia]

Just tested the newly aquired HPE 2910al switch.
It works, I am online while using it.
The noise it produces is quite excessive so this one will go to a room where the noise will not bother me.
Got hold of a 2 slot 10 gbe plugin card for cx4 cables for 24 Pesos, plus another HP Enterprise switch (a 5920 that produces around 43 or so db of noise for 65 Pesos.
This should cover my needs, except for possibly an office connect thingie. I should be more than fine with a 10 GBE backbone at home, though.

[mnementh]

...?

Did everybody wake up and just randomly decide to have a "2-bagger" contest...?   

mnem
I never get the memo...

Is that a challenge to find a "Coyote" scope?

made by ACME ?
or in Desert ( coyote ) coating ?
or a Wiley scope that has gone over the cliff ?

Ummmmmm... how to put this delicately because there's a lady present... 

"2-bagger" and "coyote-ugly" (pronounced kuy-yoat ugleee) are crude bro-speak meaning the same thing; a girl who is so ugly a) you put a bag over your head in case her bag falls off or 2) so ugly even a coyote wouldn't eat the carcass.

Such a contest would be a group of such ahem... men... who would makes bets on which would sleep with the best/worst example thereof.

I've never heard of women making such a spectacle... but sheer odds says it must have happened somewhere/somewhen...?

mnem
*toddles off to take 2 showers... for some reason, I just feel so dirty now*

[McBryce]

Most suspect scope award goes to https://www.ebay.co.uk/itm/313671287713

Oh, isn't it just darling...? I love its rustic charm... 

mnem
rust never sleeps...

It's even got its own external high voltage source hanging out the side. A true multi-functional device.

McBryce.

[0culus]

¹ Sadly, the lawyers from Keysight pulled an ugly stunt and called off Ben Riddle, because of NDAs, intellectual properties and other stupid reasons.
Keysight is working really hard to permanently damage their reputation.     

Not sure what's going on with Keysight. Apparently, they stopped selling critical parts for the 3458a to individuals. This is particularly a problem with the ADC board that tend to die in 50% of the 3458a.

Was always thinking of eventually getting a 3458a but recently changed my mind since I might well end up with an unrepairable brick.

Probably has more to do with the fact that they are making new 3458As again. 

https://www.keysight.com/us/en/product/3458A/digital-multimeter-8-5-digit.html

[Cerebus]

Bloated executable is normally cause by the usage of super generic libraries (different containers for example, array, vector, map etc ...). you end up linking super generic code that you don't necessarily use.

Not if you know where the linker's garbage collection switch is ("ld --gc-sections" or "ld -dead_strip") and the libraries are actually written to be modular, as opposed to "I've broken it into lots of tiny bits in modules. That's modular isn't it?"

[mnementh]

Most suspect scope award goes to https://www.ebay.co.uk/itm/313671287713

Giblets... that is crusty...

Have a couple good ones here as well !

https://www.leboncoin.fr/bricolage/2037590332.htm

and

https://www.leboncoin.fr/collection/2035985058.htm


This one is clean but not so sure about value for money ! 

https://www.leboncoin.fr/collection/1845407573.htm

  ...?

Did everybody wake up and just randomly decide to have a "2-bagger" contest...?   

mnem
I never get the memo...

Is that a challenge to find a "Coyote" scope?

Yeah, sure... why not.

I'd say simple rules:

Scope must have been at one time an actual mass-production scope made for resale. I've seen some horrifying home-brews; that's decidedly unfair; to the competition, and our eyeballs.

No "steampunk'd" or "Etsy-fied" examples, or pieces of scope hacked apart for salvage. Same reason as above.

All the bits that make a scope a scope have to be present, (except maybe socketed component parts like tubes/transistors); it has to be something that to a consensus in here would likely have at one time actually been a functioning scope and could reasonably be made to work as such again, even with bodges.

I would say that no visibly blown-up CRTs should be the conceptual go/no go dividing line... what aboot y'all?

mnem
 

[Saskia]

Most suspect scope award goes to https://www.ebay.co.uk/itm/313671287713

oh, like the typical antifa member ...
Giblets... that is crusty...

Have a couple good ones here as well !

https://www.leboncoin.fr/bricolage/2037590332.htm

and

https://www.leboncoin.fr/collection/2035985058.htm


This one is clean but not so sure about value for money ! 

https://www.leboncoin.fr/collection/1845407573.htm

  ...?

Did everybody wake up and just randomly decide to have a "2-bagger" contest...?   

mnem
I never get the memo...

Is that a challenge to find a "Coyote" scope?

made by ACME ?
or in Desert ( coyote ) coating ?
or a Wiley scope that has gone over the cliff ?

Ummmmmm... how to put this delicately because there's a lady present... 

"2-bagger" and "coyote-ugly" (pronounce kuy-yoat ugleee) are crude bro-speak meaning the same thing; a girl who is so ugly a) you put a bag over your head in case her bag falls off or 2) so ugly even a coyote wouldn't eat the carcass.

Such a contest would be a group of such ahem... men... who would makes bets on which would sleep with the best/worst example thereof.

I've never heard of women making such a spectacle... but sheer odds says it must have happened somewhere/somewhen...?

mnem
*toddles off to take 2 showers*

[AVGresponding]

Not having had time to look for a manual, I didn't realise those were options. Plus there's no indication of installed options on the case exterior. Opening it up shows that, as you expected, the mVdc/ohms option is not installed.

I'll get the manual, possibly replace the blue Philips electrolytics. I suspect the tweaking might be the zero offset and the scale for ranges other than 10V.

Since I got both just because they have nixies, I might have to decide whether to sell the 8300A or 8125A https://www.eevblog.com/forum/testgear/fluke-8125a-military-dmm-teardown/msg2540730/#msg2540730

Interesting, it should have a plate on the back listing the options. I can't think why anyone would remove it.

[capt bullshot]

Found this MMIC amplifier "board" that I had built some time ago for testing purposes and temporarily connected it to the pads of the defective UPD1677.
...So I'll just have to stash this MG3633 and its guts away for some weeks until the replacement parts arrive, put in a new amplifier, put everything together again and hopefully I'll have a working as normal sig gen again.

Good sleuthing/bodging Sherlock! So do you think random IC fail, or something PSU related or...? Do they run these hot/close to max specs like the AD811ANs in my B&K 4070A/BNC 630?

Given the age of this unit, it's most probably a random fail.

[bd139]

...?

Did everybody wake up and just randomly decide to have a "2-bagger" contest...?   

mnem
I never get the memo...

Is that a challenge to find a "Coyote" scope?

made by ACME ?
or in Desert ( coyote ) coating ?
or a Wiley scope that has gone over the cliff ?

Ummmmmm... how to put this delicately because there's a lady present... 

"2-bagger" and "coyote-ugly" (pronounce kuy-yoat ugleee) are crude bro-speak meaning the same thing; a girl who is so ugly a) you put a bag over your head in case her bag falls off or 2) so ugly even a coyote wouldn't eat the carcass.

Such a contest would be a group of such ahem... men... who would makes bets on which would sleep with the best/worst example thereof.

I've never heard of women making such a spectacle... but sheer odds says it must have happened somewhere/somewhen...?

mnem
*toddles off to take 2 showers... for some reason, I just feel so dirty now*

I’ve been on the end of it a couple of times 

Being honest the only true ugliness in people is their defective personality. Applies to scopes as well. Consider Heathkit 

[Kosmic]

Bloated executable is normally cause by the usage of super generic libraries (different containers for example, array, vector, map etc ...). you end up linking super generic code that you don't necessarily use.

Not if you know where the linker's garbage collection switch is ("ld --gc-sections" or "ld -dead_strip") and the libraries are actually written to be modular, as opposed to "I've broken it into lots of tiny bits in modules. That's modular isn't it?"

my point is mostly generic vs specific code. You are always going to have more lines and more layers in generic code.

specific is always going to be smaller. the cost is reusability.

Peoples can really create monsters by layering libraries on top of each others.

[mnementh]

Peoples can really create monsters by layering libraries on top of each other's.

mnem
 

[mnementh]

@Saskia - Why did you quote me and then leave no actual response...? 

*has vision of a dark-hooded figure very intently twiddling targeting vectors on the remote for my suborbital crowbar launcher*

mnem
Ummm.... I just remembered I have a thing... I need to do... very far away from here...   

[mnementh]

Not having had time to look for a manual, I didn't realise those were options. Plus there's no indication of installed options on the case exterior. Opening it up shows that, as you expected, the mVdc/ohms option is not installed.

I'll get the manual, possibly replace the blue Philips electrolytics. I suspect the tweaking might be the zero offset and the scale for ranges other than 10V.

Since I got both just because they have nixies, I might have to decide whether to sell the 8300A or 8125A https://www.eevblog.com/forum/testgear/fluke-8125a-military-dmm-teardown/msg2540730/#msg2540730

Interesting, it should have a plate on the back listing the options. I can't think why anyone would remove it.

To put on another unit they robbed the related parts out of this unit for...? 

mnem
 

[bd139]

Fluke 8050A appeared on eBay. Unstable readings due to beer accident   

[AVGresponding]

Not having had time to look for a manual, I didn't realise those were options. Plus there's no indication of installed options on the case exterior. Opening it up shows that, as you expected, the mVdc/ohms option is not installed.

I'll get the manual, possibly replace the blue Philips electrolytics. I suspect the tweaking might be the zero offset and the scale for ranges other than 10V.

Since I got both just because they have nixies, I might have to decide whether to sell the 8300A or 8125A https://www.eevblog.com/forum/testgear/fluke-8125a-military-dmm-teardown/msg2540730/#msg2540730

Interesting, it should have a plate on the back listing the options. I can't think why anyone would remove it.

To put on another unit they robbed the related parts out of this unit for...? 

mnem
 

Unlikely, it's a very thin foil, there's no way I can see of removing it without ruining it.

[Saskia]

@mnem
don't need to hide, I am fully aware that I am not a Miss World contestant. Don't intend to be. In fact, I find myself very  similar in character to Alice (the one from the Dilbert saga).
My fist of death is nothing to be sniffed at

But thanks for the quick explanation of that phrase. Had not seen that context, so ...

[mnementh]

Fluke 8050A appeared on eBay. Unstable readings due to beer accident   

LOL... that's what we get for standing around bitching aboot liars all over eBay... 

mnem
maybe a little too honest... 

[Cerebus]

Fluke 8050A appeared on eBay. Unstable readings due to beer accident   

Well, we've all been there....