Datron 1030 - a strange beast

[tggzzz]

So I picked up a "turns on and lights up" Datron 1030, because it was a bench voltmeter with Panaplex displays. Well, that's half right, but it isn't your normal voltmeter.

A few quirks:

• it doesn't have a negative sign; reverse the polarity and the display remains positive
• it cannot be zeroed, due to physics. ~40µV is the minimum possible
• on some settings it can take a while to settle to 0.1%: 1500s. Yes, 25 minutes!
• the reading is poorly damped, with overshoot and ringing. That's partly by design, but I could probably reduce it by fettling as per the manual
• 4.5 digits, but only 0.1%/0.3%/3% of reading plus 0.05%/0.1%/0.2% of full scale

So, what's the USP? It is an RMS meter with a response from 0.01Hz to 300kHz.

It settles to within 0.1% in 15cycles of the input frequency. For 0.01Hz, that implies 1500s.

The wide bandwidth implies noise: 10kHz and 30µV implies 0.3µV/Hz^1/₂, full bandwidth 80µV, and mains coupled from the transformer to the guard may be 120µV. Hence it can't be zeroed.

Manual available at https://www.opweb.de/en/manufacturer/da ... ments/1030 (someone asked for one at The Other Place)

Will I keep it? Dunno. Panaplex - yes; need - no; saleable - questionable.

[EC8010]

I wonder what they were used for? Someone must have identified a gap in the market and said, "I've an idea about how to solve that, lets develop a real instrument and sell it." I'm afraid I would leave it on the table if I saw one at a rally.

[tggzzz]

The manual says "it opens new [late 70s] fields of measurement capability, in control engineering, in vibration and noise studies, and in any applications involving complex or distorted waveforms". Nowadays a decent DAC plus software would be better.

Given the ringing due to an internal feedback loop, I'd want to carefully assess its use for control/vibration measurement.

I think I paid £1 for it. The Panaplex displays are probably worth more than that.

[tggzzz]

Photo of the 4.5 digits of goodness.

IMG_1129.JPG

[synx508]

The extra digit is an option, at least in the 1030A version. It's explained in this video, I wonder if the A was just the switch to LEDs and whether that extra digit is really worth having or mostly noise. Edit: Looks like the switch was *from* LEDs but I reckon the Advance DRM6 probably had LEDs or Nixies.

https://www.youtube.com/watch?v=rkJ9JH0flx8

I particularly liked the test-and-selected transistor array with Datron branded label. Selected arrays to do the RMS conversion also feature in RACAL's 9300 (and probably others). I prefer the look and display of the Panaplex instrument.

[tggzzz]

The extra digit is an option, at least in the 1030A version. It's explained in this video, I wonder if the A was just the switch to LEDs and whether that extra digit is really worth having or mostly noise.

These things can be subtle. I once worked for a company that sold a cable test set with 0.001dB resolution but only 0.1dB accuracy. It was used to test the changes in attenuation room-sized drums of sub-sea coax at various temperatures. Since each up-then-down measurement set took a week, stability was important. They had other tools for measuring the actual attenuation.

https://www.youtube.com/watch?v=rkJ9JH0flx8

Yup, that was the 16V tant on a 15V rail that I had to replace. I'm not going to bother with others!

The third vid in that series shows the ringing I mentioned in the other thread. Very confusing for someone with short-term memory problems

[AVGresponding]

I wonder what they were used for? Someone must have identified a gap in the market and said, "I've an idea about how to solve that, lets develop a real instrument and sell it." I'm afraid I would leave it on the table if I saw one at a rally.

Here's the company on the asset sticker: https://en.wikipedia.org/wiki/Roke_Manor_Research

[Zenith]

I wonder what they were used for? Someone must have identified a gap in the market and said, "I've an idea about how to solve that, lets develop a real instrument and sell it." I'm afraid I would leave it on the table if I saw one at a rally.

Here's the company on the asset sticker: https://en.wikipedia.org/wiki/Roke_Manor_Research

Three or four of us have treasures with that asset sticker on.

[EC8010]

Ah, the Crowdhill trove! I left that sticker on my Keithley 610C because it was neat and tidy, but I scraped all the damaged calibration stickers off. This website is staggeringly slow this morning.

When I worked in Lines Department, we had to take account of temperature variation in frequency response of music lines. Our lines were quite short (up and down the UK), but a cable across the Atlantic would be ten times the length. We had equalisers whose strappings would be changed quarterly based on ground temperatures at 500mm and 1m.

Our EP14/1 (Equipment, Portable, type 14, version 1) test set had resolution of 0.05dB, but I don't know what its accuracy was. Despite its description, the EP14/1 was 3U rackmount and quite heavy. Especially if fitted in steel carry case with lid and handle. The oscillator was a Wien bridge tuned by a humongous variable air-spaced capacitor having a 3/8" ceramic shaft feeding a MOSFET-input discrete component op-amp, stabilised by a precision rectifier and four-quadrant 1496 multiplier. Designs Department at their best. It went from 5Hz to 50kHz in four ranges. (The wide range was needed to cope with resonant equalisers whose resonance might have been out of the audio band.) The meter section was switchable between PPM mode (1-7 with 4dB between each division and a quasi-peak response with fast attack and slow decay) and an expanded scale of 0.1dB per division from +0.5dB to -0.5dB. The measurement was achieved by 600 ohm attenuators. It could measure distortion down to 0.05% by inserting 100Hz or 1kHz LC low-pass filters and measuring the residual. It also had a CCIR 468-2 noise shaping filter. It was the bee's knees in 1980, and I bought one from Redundant Plant in 1995. Turned out that mine had been tweaked (badly) and the ceramic shaft of its variable capacitor was broken. Fortunately, all the bits were in there and I managed to Araldite it all back together. To my amazement, when I reassembled the capacitor, it rotated smoothly. I could then see the problem that had caused the original interference, and added a couple of padding capacitors to bring the frequency calibration into specification.

[tggzzz]

So I picked up a "turns on and lights up" Datron 1030...

A few quirks:

• the reading is poorly damped, with overshoot and ringing. That's partly by design, but I could probably reduce it by fettling as per the manual

I fettled R113, first to no oscillation, then to just oscillating, then backing off to no oscillation.

Now it is a nice example of being critically damped, with a minimal overshoot for settling times of 1.5s, 15s, 150s. Haven't tried the 1500s setting; a storage scope would be useful for that

Better forum performance, BTW

[EC8010]

Yes, I've just noticed the (much) better forum performance, too. It was awful earlier this morning.

Interesting regarding your Datron 1030. Whether you'll actually have a real use for it is another matter, but if it only cost £1, that's scarcely an issue.

I have today learned of ratio measurement on a DMM. It turns out to be something found on 7 1/2 digit (and better) meters. To compare two resistors, you connect them in series, drive a current through them, then connect the sense terminals across the reference resistor and the input terminals across the unknown resistor. Invoke ratio in DCV and voila! Up comes the ratio of the resistances. I tried it on HPAK 34470A and it was easy to do. Mr Keithley remained true to from and was more difficult to set up, but a 2010 could do it. I saw 1.0030 from the 2010 and 1.00032 from the 34470A comparing two 1R resistors with 10mA flowing through them. That means only 10mV across each meter and I hadn't taken any real care about leads or thermocouple effects, I just wanted to try the experiment to see if it could be done.

[tggzzz]

I've used my Solartron 7081 (8.5 digit) in ratio mode to test my KVD (7 decade) and vice versa.

Impressive performance: max difference was 2.6e-6.

With somebody else's HP3456, the max difference was 3 digits (i.e. 3ppm)

[EC8010]

The Keyshite manual is rubbish about ratio mode and doesn't say what the 34470A is actually doing. Is it switching between inputs and computing their ratio? Or what? I had a rummage on the Keysight web site and it is now rubbish. Once upon a time, there were informative documents, now it seems to be aimed at 1st year electronics students. Oh, and selling stuff.

[tggzzz]

The Keyshite manual is rubbish about ratio mode and doesn't say what the 34470A is actually doing. Is it switching between inputs and computing their ratio? Or what? I had a rummage on the Keysight web site and it is now rubbish. Once upon a time, there were informative documents, now it seems to be aimed at 1st year electronics students. Oh, and selling stuff.

You could probably make a guess based on the time to make a measurement. If the time is doubled, it is probably multiplexing the two inputs. If unchanged, the external input replaces the internal reference.

The other tell-tale would be any difference between the reference and input specs, e.g. allowable voltages and offsets. IIRC, and it was a long time ago, the Solartron 7081 has significant (but not onerous) limitations.