Test Equipment Anonymous (TEA) group therapy thread

[AG7CK]

The second meter at around 24 min in the video has probably never been calibrated after its factory calibration by HP. Almost all new meters of similar software calibration class come with a calibration number count of some 30-50 (add or remove for type / brand). This probably does not indicate that the meter is very stable - it would rather indicate that the owner(s) over a couple of decades didn't feel any need for a calibrated meter.

The first meter in the video seems to have been calibrated on all ranges at least 7 times. If not so - a single or a few ranges could have been calibrated several tens of times. This could indicate precision use, or it could mean that the meter was/is drifting.

[Specmaster]

I found the video and here is that link as promised.

I've seen that one before. First of all, you can replace the battery without needing to re-cal it. I have the same meter in perfect working order which I got off Ebay. I replaced the battery and attached a voltage in parallel with the battery so it wouldn't lose it's cal memory for the short time I had the old battery out of the system.

Second - he keeps going on about the "SRQ" indication as meaning it needs "servicing" as in it needs to go to the shop. The "SRQ"  is a HPIB control function and pertains to the operations on that bus. Not sure where he got that from.

Oops,  seems I have got it all wrong, the SRQ he mentions is not a service request at all, it is an indicator of the stack used in the HPIB control interface which may have been accidentally set on the front panel push buttons it seems, or the previous owner had it thus connected with his computer and maybe failed to clear out all associated code before selling it on.

The calibration failure he shows when trying to tell it that the value of the resistor was 10M, is because he was using the incorrect command procedure for calibrating the resistance ranges, this I discovered by downloading the service manual and reading the section that deals with those areas. It is highly likely then, if he follows the correct procedures that the meter would once more be calibrated accurately enough for the average user and with the SQR flag cancelled, he will have bagged himself a real bargain there. If he is not convinced then he can chuck the meter in my direction, I'd love to have it, only problem being it's a little on the large size for my little room.

[bitseeker]

I remember that video from way back when. Since all the corrections have already been covered in recent posts, I'll just add that there is a lot of fine print that flashes by near the beginning of his videos. One part of it is a disclaimer that says, and I'm paraphrasing, that he's not an expert and is learning as he goes.

There's still good stuff that can be found in his videos, and I've watched many of them in the past, but do follow up with the documentation to verify the details before diving into things such as performance verification or calibration. In fact, that applies to any videos, not just his. Errors happen.

[URI]

Hi, all!

after being assured by bitseeker that I'm right at the TEA board here I post my second posting on the eevblog forum. 

I'm glad to be here amongst others with similar problems explaining others what you need all those gear for. 

I have to confess that I'm addictive to test equipment.

It startet all out that I wanted to repair a pioneer cassette deck CT-900S (change motor and most rubber pieces of the drive) and a kenwood DP-8020 cd-player.

That was three years ago.
I had a philips PM5134 function generator and a hameg 203 oscilloscope then.

I soon discovered that the PM5134 was not stable enough: it would change its frequency over hours and didn't reach a stable working point. I had to check that and I sniped a philips PM6680 frequency counter cheap on ebay to check the frequency.

I got carried away...
I realised the frequency counter needed a stable reference, too, since mine "only" had a tcxo. So I bought several oxco's from china and Russia for little money.
Also I had to replace my function generator because using my PM6680 I now knew how unstable it is.
I kept it because even if unstable its ok as an external sweep source for another (stable) function generator for example.
I came across a HP3325B with option 002 (high voltage output) for little money from the US (I got it for the price of shipping plus shipping..). It was offered as not working properly. After cracking it open I could not believe that inside some of the coax-patch-cables were not connecting the points they should have been.
I put the cables back in the right places and the HP3325 worked fine! 

Beeing already aware at that point that I needed a precise reference of 10 MHz I looked for a trimble gpsdo and finally got one cheap, also from china.
In the mean time I had acquired a PM6307 wow and flutter meter and even a reference tape with 3 kHz and 3.5 kHz test tones, but I haven't used those ever since.

I got carried away...
A second frequency counter (PM6685) found its way to me from ebay and I started to check oxco's and calibrated my two frequency counters.
I reached the possibilities of my old hameg 203 (20MHz) oscilloscope and went for a philips PM3065 (100MHz bandwidth) from bundeswehr surplus. I already had encountered the eevblog forum and several other forums and came across that I could further improve my home lab in the field of such basic things like power supplies, especially the HP6632B. I only had a simple, self-made power supply with 25V/3A cc/cv. But only 20V on the HP6632B? I could not stand an offer for two fully working HP 6643A (option J11:40V/5A) for 170$ each..
The HP6632B had a defective controller board that I had to repair but I lost the calibration data. My best DMM was a fluke 85 III. 

I got carried away...
An ERMIC G-1006.500 voltmeter was offered on ebay for 50Euro and I was the only bidder. 
The G-1006.500 is the best voltmeter ever designed and built by former Volkseigener Betrieb RFT in the german democratic republic.
It is driven by a clone of an intel controller, by the way. 
Because the GDR had no high-resolution ADC available, the G-1006.500 had many ranges to cover 200mV to 1000V and 200mOhm to 200MOhm with 4,5 digits resolution down to 10µV/10µOhm on the last digit.
Its impedance is >10G in the ranges up to 20V and above that it is >10M.
Summed up, it's an impressive peace of gear but the time had gone over it.
After the iron curtain fell in 1989 it became clear it was not competitive to western lab voltmeters at the time.
The test gear department of RFT, re-founded as ERMIC in the BRD, closed in the first half of the 1990s.
I liked the G-1006.500 even though I grew up in the BRD. 
Many peoply in the former BRD looked down on everything in the GDR. I did never share that because having relatives in the GDR.
The people in the GDR achieved impressive things under the circumstances they were living in/with.
The G-1006.500 was well designed and built in a way that it achieved the accuracy of western voltmeters but lacked the resolution as a consequence of the embargo in the technical fields.
Ok, I enjoyed to explore the G-1006.500 and also its limitations. In the end I sold it because of its size (19", 3U high, full depth) and weight of 15kg. Before I sold it, I bought a HP 3456A. Also "only" a voltmeter. 

But I had to measure current to calibrate the HP6632B, remember?

I got carried away...
A HP3457A found its way to me over the atlantic from canada. Quiet a nice pice of kit. It was ok for calibrating the voltage part of the HP6632B but I realised I needed a reference resistor of 1.1k for the current. Ok, it doesn't have to be precisely 1.1k if I can measure its value accurately enough. At that point I discovered how noisy the HP3457A is in comparison to the HP3456A. 

I already had specified for me that
- frequency-related test gear has to contain a 10Mhz reference input
- all test gear has to contain a GPIB port. (Only my PM6685 lacks a GPIB port but I'm working on it..)

I had hope to limit my TEA by that. I was wrong.

I got carried away...
Some day I came across an ebay-offer for several HP 3488A for 45Euro that I couldn't resist. I bought two of them.
Over several weeks also modules for those found its way to me. More than those two HP3488A can take..  Ok, I keep them as spare and for flexibility.
I acquired an USB to GPIB adapter and loads of GPIB cable and can now control my test gear and read results using python.
As I'm no professional programmer I have still much (almost everything) to learn in this field.

Eventually, I don't remember when, I won an HP4263A LCR meter.

My most recent acquisitions are another (my third) HP3456A voltmeter and a precision resistor decade 0.01-11111.1Ohm.
And 5x Agilent 8762B coaxial switches DC - 18 GHz.

Did you notice something?
Yep. I haven't repaired the cassette deck neither the cd-player yet.
I bought a used but working DP-8020 and put the defective one aside, still awaiting repair (since I have all parts needed lying around).

At an early stage I only had TEE but now it has become TEA, I'm afraid.
I think I got serious GAS in combination with ESD and DMM (wait: duplicate? multiple!  -> MMM).

Writing here gives me some relief. 

Oha, I forgot: The calibrators.
First I bought one from Ian Johnston. It's my portable voltage reference and I like it very much.
A few month ago I made a low offer on ebay that was accepted to my surprise. I now own a Knick JS 3010 5,5 digit DC voltage and current calibrator with 10nV - 140V and 10nA - 200mA maximum. Its the version with higher accuracy of 0.001%. 

I now think about a 3458A (leasing a calibrated one, it's to expensive for me to own one) for calibrating all my stuff to a valid standard for an acceptable amount of money..

I attached some photos showing most of my test gear on a shelf above my home-lab desk. The HP3325A has to sit under the table, because of its size and weight.
I'm running out of space a little bit but that's more a challenge than an obstacle.
It's not the end of my disease, I assume.

[tautech]

..........
Did you notice something?

Yes.

I got carried away...

Several times it seems. 

A big welcome, you'll feel quite at home here.

[Cubdriver]

URI, you'll fit in perfectly with this lot!!   

Welcome aboard!  And no, this is just the beginning of your disease.  Especially with us cheering you on.  (Not that we'd DO such a thing, mind you... )

-Pat

[Specmaster]

URI, a big welcome to you and I echo what the others have already said to you, seems to me that with us lot, the more buttons, LEDS and digits an item has, the more we feel compelled to obtain it and add it into our protective custody to prevent it from falling the hands of others 

As to getting carried away, well thats just par for the course until the money supply runs out or the physical space which ever comes first. 

[bd139]

Welcome URI to the land of the afflicted. Nice to see GDR equipment represented. I'm amazed at what the Eastern Bloc churned out with limited resources!

Physical space is the constraint here. Had to sell a couple of things (boo hiss) to get a printer in that I needed which I've spent all flipping day getting working

On a positive note, £150 up now so the moment a decent bench meter without an LCD display appears I'm on it like a seagull on chips.

[tggzzz]

Did you notice something?

Yes. You can't measure fast risetimes 

The people in the GDR achieved impressive things under the circumstances they were living in/with.

With skill and imagination, you can "do more with less"

[Specmaster]

On a positive note, £150 up now so the moment a decent bench meter without an LCD display appears I'm on it like a seagull on chips.

So what constitutes a decent bench meter and will it be VFD or LED display?

[Specmaster]

With skill and imagination, you can "do more with less"

Very true, but thats not the TEA way is it??

[tggzzz]

With skill and imagination, you can "do more with less"

Very true, but thats not the TEA way is it??

Actually, it is at the heart of it.

How else can you create the world's most accurate X, for any value of X you care to name? Just spending money buying something someone else built isn't that difficult a challenge.

[bd139]

On a positive note, £150 up now so the moment a decent bench meter without an LCD display appears I'm on it like a seagull on chips.

So what constitutes a decent bench meter and will it be VFD or LED display?

Anything that has a half decent AC bandwidth (20KHz), LED or VFD, 4.5 digits, bench, true RMS, auto range. Basically my 8050A with an LED or VFD display and auto ranging

Fluke 45 is spot on really.

Decided to build some test gear today. Started off with "hey lets try and measure some toroids with my chinese LCR meter" followed by frustration and failure. Now have dreamed up an LXI/SCPI rig and software I can run off my DG1022Z counter and an LC oscillator controlled via a relay from an Arduino which will measure the frequency as per http://www.qsl.net/wm5z/cq199301b.pdf

[Specmaster]

Decided to build some test gear today. Started off with "hey lets try and measure some toroids with my chinese LCR meter" followed by frustration and failure. Now have dreamed up an LXI/SCPI rig and software I can run off my DG1022Z counter and an LC oscillator controlled via a relay from an Arduino which will measure the frequency as per http://www.qsl.net/wm5z/cq199301b.pdf

How about one of these then, I have one and it works really well and is very accurate.

[bitseeker]

Welcome, URI. See, I knew you'd fit right in.

By the way, you know you're seriously afflicted as soon as you say, "Eventually, I don't remember when, I won a(n)..."

[xrunner]

Welcome, URI. See, I knew you'd fit right in.

Oh yea - good stacks of test gear, very nice height you got going there URI. And HP gear ... 

You defiantly deserve the TEA Seal of Approval. Welcome!

[URI]

>Quote from: URI on Yesterday at 22:47:21

Did you notice something?

Yes. You can't measure fast risetimes 

Got me! 

[BU508A]

Got me! 

That is nothing which cannot be fixed. 

[URI]

Thank you all for your warm welcome!  Quote>Quote from: bitseeker on Today at 04:09:31

By the way, you know you're seriously afflicted as soon as you say, "Eventually, I don't remember when, I won a(n)..."

Oh, that's not too serious, I think. I can exactly say, when:
In the past three years.
 It's just a matter of resolution..

As beeing carried away by remembering and presenting my life of suffering of the last three years I forgot more test gear sitting around here in hidden corners.. mostly portable "gadgets"

1x fluke 85 III
1x UNI-T UT139C because it fits into a carry case for an analog meter that I got for free without meter 
several dmm's at the 10,- Euro-range for having it on hand in the car, in the tool box, etc..
1x MK-328 transistor/diode/LCR/ESR tester from china performing very reasonable for 21Euro incl. shipping
1x Siemens Multizet analog meter 0.06A - 6A/6V - 600V (moving-iron movement)
1x Siemens Multizet analog meter 0.06A - 30A/6V - 600V (moving-iron movement)
1x HB/Elima Elavi 5n analog Multimeter 0.003A - 100A/0.1V - 1000V AC and DC and 1-10kOhm
1x Gossen Uphi special analog multimeter 12V - 600V/0.06A - 120A AC, frequency 40 -400Hz, cos phi and 1-100kOhm
1x water level 250mm long, resolution 0,04mm/m 

I like to be able to do basic physical measurements. On the other hand, analog meters show direct response unlike integrating high-digit-meters. So I like them for testing unknown devices or circuits.

I forgot to mention my Philips PM2534 and PM2535 Multimeters. One with an "ordinary" reference diode, one pimped at factory with an LM399. They are both 6,5 digit multimeters but not that accurate than the (HP3457A and) HP3456A. But they are versatile for me as they can measure different things, my HP volt/multimeters can't (out of the box without special extra gear).
For example they have a mode for measuring temperature with 0.1°C resolution using standard PT100 sensors. As there are used but accurate sensors out there for affordable prices this is an advantage over the HP3456A requiring a special NTC sensor of unknown characteristics.
And they both got GPIB ports! 

I also have a repair/modification queue as I already mentioned (cassette deck, cd-player).
Also waiting in there is my blue HP3456A because suffering from error -4 (very likely a problem in the power supply) and my three HP power supplies, that will be modified with binding post at the front.
Ah, of course my PM 2535 awaiting a LM399 or alike and my gpsdo and rubidium reference waiting to be housed. Sadly the rubidium reference has a problem in its logic part to fix first.
I don't really know if I have all gear needed for repair, really. 

[URI]

Got me! 

That is nothing which cannot be fixed. 

I know and that scares me a little.. nooop 

[BU508A]

Ah, of course my PM 2535 awaiting a LM399 or alike [...]

Do you have any further information which you can share here? Probably in a new thread?

Thanks.

Andreas

[URI]

Ah, of course my PM 2535 awaiting a LM399 or alike [...]

Do you have any further information which you can share here? Probably in a new thread?

I think I will, actually when I get there. 

A PM2535 or PM2534 you own fellow sufferer. 

[URI]

... Now have dreamed up an LXI/SCPI rig and software I can run off my DG1022Z counter and an LC oscillator controlled via a relay from an Arduino which will measure the frequency as per http://www.qsl.net/wm5z/cq199301b.pdf

Thanks for sharing that link! *saved* 

[bitseeker]

... LC oscillator controlled via a relay from an Arduino which will measure the frequency as per http://www.qsl.net/wm5z/cq199301b.pdf

Thanks! Saved for my reading list.

[VK5RC]

Welcome URI, That is the start of a nice collection - its something that is never finished! HiHi
In my volt-nuttery area I have  3455, 3456, 3457, 3458 as well as the 'standard' 34401 and 34461, I haven't included dot-LED, LED and Nixie volt measuring units (ARRRGHHH)
The 3456 is a real 'sleeper' unit - highly regarded,esp for the price - but can have issues - unobtanium parts.

I thought I should go down the Wheastone bridge method measuring 'rabbit hole'
Fluke seem to dominate in this area (most of my gear is HP - I fully support your goals of GPIB, 10MHz ref )
I have a Fluke 845AR Null Bridge (coming slowly from OS)
and found in Australia, a Fluke 720A Kelvin Varley Divider, A photo of a similar vintage unit below