Test Equipment Anonymous (TEA) group therapy thread

[bd139]

These are 100% works or doesn't work instruments.

Not entirely. The resistance ranges can be knackered with the rest working fine. I don't know why, in principle there's nothing obviously wrong with the designs, but HP always had problems with the current sources on resistance ranges. If you've an HP34xx and everything's working perfectly except resistance then it's always the current sources. It's a stock fault for HP meters.

Fair point but those are usually massive errors and these meters checked out close to spec before sending off for cal.

Edit: plus they had recently expired Keysight own calibration stickers all over them.

[Specmaster]

Yep. Divs.

It's not even analogue to cal it. You just:

1. turn the switch
2. stick it in the range you want to cal
3. short it.
4. press the cal button
5. apply 1/3rd scale (not even required on all ranges)
6. adjust the display to the calibration voltage if you need to.
7. press cal.
8. goto 2.

Eventually when you've done them all it'll start up with SELF TEST OK 

If it's reading low, it's basically a Y=MX+C transform where M and C are calibration variables. So someone put the wrong fucking M or C in.

When I did mine I did it with a PSU and connected the U1241 in series (current) or parallel (voltage) to it. Resistance I used some 1% resistors and frigged the display value to match the U1241's interpretation of the value. The linearity is amazingly good so was spot on when I cranked it up near FSD.

These are 100% works or doesn't work instruments.

That is my understanding of it as well, basically give it a known value and adjust the display by using the up and down arrow till the display reads the known value and press CAL which can mean holding the respective button in for a very long time to reach the desired reading, this is the reason why I did not decide to it myself using the method that you describe. I assumed that a proper calibration house would have short cuts to a way of dialing in the correct reading using the GPIB or something that we mere mortals do not have access to?

I must admit I have not RTM fully for that section but I assumed that the resistance would more or less fall into place once the DC voltage has been done as surely most meters are just reading the voltage drop across the load?

ACV does that have its own adjustment if so then that might be where I'd struggle if doing it myself as I don't have an accurate source voltage unless a variac is considered good enough.

I have the MS7221 coming early next week that will provide pretty reasonable DC voltage standards and also current unto 24mA so if all else fails I'll have a go myself 

[bd139]

It only requires one AC V I think. Was 3v. I used a function generator and connected U1241C in parallel.

[Specmaster]

These are 100% works or doesn't work instruments.

Not entirely. The resistance ranges can be knackered with the rest working fine. I don't know why, in principle there's nothing obviously wrong with the designs, but HP always had problems with the current sources on resistance ranges. If you've an HP34xx and everything's working perfectly except resistance then it's always the current sources. It's a stock fault for HP meters.

Fair point but those are usually massive errors and these meters checked out close to spec before sending off for cal.

Edit: plus they had recently expired Keysight own calibration stickers all over them.

No mine didn't have those stickers on it as I removed them as part of the cleaning operation but is calibrated by Keysight in Jan 2016 same as yours and they were able to tell me that it was calibrated about 18 months ago. A 10K 1% resistor was reading 10,0014.3 so yes its was just about spot on before the battery shorted out for a split second.

[Specmaster]

The fact that they were not instantly aware of the 3478A at the time of requesting a quote etc is now beginning to gnaw at me that they probably do not understand how to do it and either looking for manual adjustment pots or maybe the more modern meters are just a matter of plugging a source in and then telling it via a computer interface what the value of that source actually is and they weren't that far advanced when the 3478A's were made?

[Cerebus]

...and they weren't that far advanced when the 3478A's were made?

Yes they were. The remote calibration procedure is in the service manual.

The way most commercial cal labs process multimeters with a GPIB interface is just to hook them up to the calibrator with a set of leads, plug GPIB into a computer that drives both the calibrator and DMM, and press 'GO'. The computer identifies the DMM automatically over GPIB and then runs the whole process by 'magic'. You buy the software from your calibrator vendor. Fluke have MET/CAL, Wavetek/Datron used to have something similar and so on.

[Specmaster]

...and they weren't that far advanced when the 3478A's were made?

Yes they were. The remote calibration procedure is in the service manual.

The way most commercial cal labs process multimeters with a GPIB interface is just to hook them up to the calibrator with a set of leads, plug GPIB into a computer that drives both the calibrator and DMM, and press 'GO'. The computer identifies the DMM automatically over GPIB and then runs the whole process by 'magic'. You buy the software from your calibrator vendor. Fluke have MET/CAL, Wavetek/Datron used to have something similar and so on.

So which software should RS be using for the 3478 then, thats the question or could they be trying to wing it?

[med6753]

The fact that they were not instantly aware of the 3478A at the time of requesting a quote etc is now beginning to gnaw at me that they probably do not understand how to do it and either looking for manual adjustment pots or maybe the more modern meters are just a matter of plugging a source in and then telling it via a computer interface what the value of that source actually is and they weren't that far advanced when the 3478A's were made?

Wow, that cal lab sounds like a bunch of flunkies.  Sorry to hear you about your troubles. I would pursue getting some sort of refund but that might be a lost cause.

Update on the old beast. I keep making steady progress. Both channel 1 and 2 attenuator decks were FUBAR. Channel 1 responded perfectly to the Dexoit. The DMM is functional but needs a cal. Channel 2 still goes ape shit on .2V range and lower. Having one functional channel is definitely an advantage so I just have to compare readings. I'm sure I'll track it down. Hopefully that will be the last issue and then I can check the scope's overall calibration and button it up.

[Cerebus]

...and they weren't that far advanced when the 3478A's were made?

Yes they were. The remote calibration procedure is in the service manual.

The way most commercial cal labs process multimeters with a GPIB interface is just to hook them up to the calibrator with a set of leads, plug GPIB into a computer that drives both the calibrator and DMM, and press 'GO'. The computer identifies the DMM automatically over GPIB and then runs the whole process by 'magic'. You buy the software from your calibrator vendor. Fluke have MET/CAL, Wavetek/Datron used to have something similar and so on.

So which software should RS be using for the 3478 then, thats the question or could they be trying to wing it?

Generally things like MET/CAL have built in libraries of cal procedures for a selection of (generally more popular) instruments and a scripting language that you can use to add your own procedures for other instruments. You can see the list of built in procedures for MET/CAL at https://assets.fluke.com/metcalproc/html/procedure_list.html and, spoiler alert, the HP 3478A cal procedures are included for the Fluke 5520, 5700, and 5725 calibrators.

[med6753]

Forgot to add. In case you've not seen what the DM44 option looks like. And here's an oddity....it doesn't have the capability to measure AC volts. I wonder why Tek did that? 

[Specmaster]

Maybe they thought that the likelihood of using it to measure AC volts was a slim on in electronics that normally uses DC and thus left AC measurements upto the scope itself?

[GerryBags]

Not entirely. The resistance ranges can be knackered with the rest working fine. I don't know why, in principle there's nothing obviously wrong with the designs, but HP always had problems with the current sources on resistance ranges. If you've an HP34xx and everything's working perfectly except resistance then it's always the current sources. It's a stock fault for HP meters.

So.... if someone had just pulled the trigger on a 3455A that isn't working on the resistance ranges they wouldn't have just shot themselves in the foot, then? 

[Specmaster]

OT, I'm going to an evening air show tomorrow, is anyone interested in more aviation photos from it, I don't want wear anyone's patience by posting them here if they are not welcomed, it is after all an electronics forum 

[med6753]

OT, I'm going to an evening air show tomorrow, is anyone interested in more aviation photos from it, I don't want wear anyone's patience by posting them here if they are not welcomed, it is after all an electronics forum 

Yes, I'm interested. 

[Neomys Sapiens]

My guess would be that was produced by Pye Telecommunications in the 40s and later; evidently they enjoyed some significant popularity as mobile radio providers for police and military during and after the war. They even made some of the early RADAR receivers used in wartime according to the wiki.

Yes that's exactly what Pye used to do besides a few domestic radios and TVs. I lived in Haverhill for a time many years ago and they had a factory there solely making the Police radios. My ex worked there till they closed it down and I believe they relocated to Cambridge.

Rather W.G.Pye, as I remember the name. I even found a picture of the monster on the net!
 

[Cerebus]

Not entirely. The resistance ranges can be knackered with the rest working fine. I don't know why, in principle there's nothing obviously wrong with the designs, but HP always had problems with the current sources on resistance ranges. If you've an HP34xx and everything's working perfectly except resistance then it's always the current sources. It's a stock fault for HP meters.

So.... if someone had just pulled the trigger on a 3455A that isn't working on the resistance ranges they wouldn't have just shot themselves in the foot, then? 

I wouldn't panic, they usually easily fixed as long as it isn't something dead inside a hybrid, and those faults are rare and often hit more than one thing. As it is, I *think* that the 3455 has only standard components in it. The only gotcha on the older meters is that if you've got a dead transistor that's only listed under an HP part number you can have fun figuring out what to replace it with.

[med6753]

I wouldn't panic, they usually easily fixed as long as it isn't something dead inside a hybrid, and those faults are rare and often hit more than one thing. As it is, I *think* that the 3455 has only standard components in it. The only gotcha on the older meters is that if you've got a dead transistor that's only listed under an HP part number you can have fun figuring out what to replace it with.

Aren't there cross reference lists for HP transistors? I've been very successful in finding cross references for Tek transistor part numbers.

[bd139]

http://www.sphere.bc.ca/download/hp_xref-free.pdf

[GerryBags]

OT, I'm going to an evening air show tomorrow, is anyone interested in more aviation photos from it, I don't want wear anyone's patience by posting them here if they are not welcomed, it is after all an electronics forum 

Yes, I'm interested. 

Me, too. Even the old 'planes had some kind of electronics in them, so it counts in my book.

Thanks, Cerebus. I have the HP eqv. part list saved....somewhere. Downloaded from here.... somewhere. Aha! cheers, bd139, I'll save it in a few more places on my computer, just in case. I've downloaded the manual, but I've seen a paper copy for sale that I might grab. The scans are too low-res to see the numbers properly in the pdf and I find physical schematics much nice to work with.

I must remember NOT to send it to RS for a cal. It's hard to believe that nobody in that cal lab had the gumption to look the bloody thing up on online, even if they couldn't call up someone who'd know. You must be pretty miffed, Specmaster.

[Cerebus]

I wouldn't panic, they usually easily fixed as long as it isn't something dead inside a hybrid, and those faults are rare and often hit more than one thing. As it is, I *think* that the 3455 has only standard components in it. The only gotcha on the older meters is that if you've got a dead transistor that's only listed under an HP part number you can have fun figuring out what to replace it with.

Aren't there cross reference lists for HP transistors? I've been very successful in finding cross references for Tek transistor part numbers.

Yes, but they're not as comprehensive as I'd like. Case in point, the JFETs used on the 3458A aren't all decoded yet. There's still thousands of parts on the list though, lest I sound too damning. The usual file name for this is hp_xref-<varies>.pdf.

[Specmaster]

I must remember NOT to send it to RS for a cal. It's hard to believe that nobody in that cal lab had the gumption to look the bloody thing up on online, even if they couldn't call up someone who'd know. You must be pretty miffed, Specmaster.

That's one way of putting it. I really don't understand either when their web site states 5 working days, door to door that nobody could pick up the bloody phone or even send me an email about the problems they were having when they had both phone number and email address, that in my book ranks as sheer arrogance and not caring about the customer.

Edit It kind of makes you think that they somehow managed to completely screw it up and they are looking for an excuse to get out of a tight spot cheaply.

[mnementh]

Or maybe they weren't smart enough to know it came to them with the Cal data erased, and they thought they must've broken something when they couldn't get it close... and it snowballed from there.   


mnem
I can hear their techs reciting the "shit litany" from all the way over here...   

[bitseeker]

Forgot to add. In case you've not seen what the DM44 option looks like. And here's an oddity....it doesn't have the capability to measure AC volts. I wonder why Tek did that? 

Thanks, med. I hadn't seen inside one before.

Can't measure AC? That is odd. Do they figure that reading it off the screen is good enough?

[bitseeker]

OT, I'm going to an evening air show tomorrow, is anyone interested in more aviation photos from it, I don't want wear anyone's patience by posting them here if they are not welcomed, it is after all an electronics forum 

Yes, I'm interested. 

Me, too. Even the old 'planes had some kind of electronics in them, so it counts in my book.

We waver off topic from time to time. As long as it doesn't derail the thread (i.e., doesn't go on and on and on), I don't have an issue with it. Of course, since I like pics of planes and Spec takes great ones, I'm a bit biased.

[Mr. Scram]

I just found footage of Specmaster's DMM being processed.