Old or New multimeter?

[stickfly]

.... This struck me today when I was thinking of some problems I have with a vintage AVO meter : When you have a problem with a vintage multimeter & you have to take measurements, do you use a vintage multimeter or a new Digital meter?
i.e. I have a problem with an AVO Model 7 Mk2. Should I use one of my other AVO's for testing components etc, or should I be using a modern digital meter. Stupid, I know, but I just feel that it would be wrong to use a digital meter, but would the digital meter be more accurate ? Or does it really matter.
I think I'm just getting old I work in the photocopier industry & our apprentice said " What the hell is a multimeter "
Regards ..... Gary

[Bazz4CQJ]

Both the analogue and the digital meters each have their own merits. Digitals usually offer better resolution and accuracy, but when you need to make a measurement on something that is changing a lot or something with an intermittent connection, and analogue (moving coil) meter is often extremely useful. I use both types all the time.

B

[Boulevardier]

Agreed. I think both are essential in their own ways and for different purposes - I wouldn't be without either. The right horse for the right course!
Mike

Originally Posted by Bazz4CQJ

Both the analogue and the digital meters each have their own merits. Digitals usually offer better resolution and accuracy, but when you need to make a measurement on something that is changing a lot or something with an intermittent connection, and analogue (moving coil) meter is often extremely useful. I use both types all the time.

B

[Diabolical Artificer]

I'm use two vintage auto-ranging DMM's exclusively because they are very easy to read & use & have very good protection. I find the scales on Avo's etc hard to read, probably due to not using them much but also a test setup maybe sprawled all over a bench and a DMM is easy to read at a few feet.

I use a scope for everything else but do have an analogue meter but it's used rarely.

A half decent DMM with good protection, auto-ranging and no auto off is just so easy to use: no mucking about changing a range SW and no loud BEEP when your probing a HV circuit for instance. It is what your used to I suppose and also what your mostly working on also dictates choice.

Lastly although this isn't a recommendation thread Fluke 25's can be had for not much and do all the above. There's a manual available too.

Andy.

[Radio Wrangler]

If you're fixing vintage sets and using service sheet data, note that the information from the period was probably measured with a "20,000 Ohms per volt" meter such as an AVO 8.

So using such an instrument saves you having to keep pondering what different effect the loading of a modern electronic meter makes with its different loading.

They all have their strengths and weaknesses, but having the same combinations of such as the guy who wrote the service sheet can make life easier.

You also may wonder what accuracy you need readings to. I grew up in a lab where everyone went for an oscilloscope first. Enough accuracy for most purposes, but added clues if funny business was afoot. You went for a voltmeter as such only when you had a special need. They sat on benches, but rarely moved off their ohms ranges. The things I was working on were prototypes, things never made before and absolutely ANYTHING could be wrong. There wasn't the cosy thought of "Well, it used to work" or "Others work" so you have to work from scratch with no safety nets.

David

[Craig Sawyers]

It is a long time since I designed hard core gear, but back in the day I used to design stuff with everything fancy. PALs, ECL, bit slice processors, teletext generation and video insertion and processing.

So I know exactly where David is coming from.

Oddly enough everything I designed and built ended up looking like an audio product...

Craig

[lesmw0sec]

As I understand the original post, the question refers to checking the calibration of an existing meter, rather than 'what should I use in general'. The answer to that is: Anything which is more accurate than the device under test - preferably by around ten times. As to digital or analogue, it does not matter, provided that the foregoing can be achieved. Any chance of borrowing something with a certificate preferably traceable to NPL standards?

[Craig Sawyers]

Traceability is not straightforward. There is (a) repeatability, which is a statement about the quality of the multimeter and (b) accuracy relative to calibration standards, usually at 24h, 3 months and a year. At ppm of reading + ppm of full scale. All at a closely defined temperature (usually 20C +/- a couple of degrees)

Not surprisingly the cost associated with NPL traceability is significant (in the thousands of quid).

Last time I had a zener voltage standard measured cost £300. And that is just two terminals that generate 10V.

And transporting an NPL traceably calibrated instrument is not straightforward. Even a system DVM will take a day or two to settle down after transport. And warm up to within a few degrees of the test temperature. Even the 10V standard I sent for traceable calibration was returned exceptionally carefully.

FWIW If I had an NPL calibrated meter (I don't), and had recently coughed up a load of money for that, it would be staying put in the lab.

Craig

[Craig Sawyers]

To give some idea of the scope of traceability, I attach the specification of the Datron 1081.

Craig

[lesmw0sec]

Craig - costs have obviously risen since the last time I was near a standards lab! My suggestion was of course somewhat over the top, since an accuracy of around 5% is probably more than adequate for cheking out vintage radios.

[kellys_eye]

Update rate AND accuracy to 'multiple decimal paces' is, I find, often a distraction I could do without. In any valved circuit I'm looking for +/- a couple of volts (or even 10V) so waiting for a meter to settle - which I can't seem to ignore - is annoying.

Of course, switching to a higher range eliminates the decimal point issue and a digital readout is easier to read so it's swings and roundabouts for me.

'Probing around' for approximates means the analogue meter, looking for faults in semiconductor circuits means digital.

[emeritus]

Horses for courses. When my son's BT phone line stopped working for phone calls but broadband still worked, connecting my analogue meter to the incoming wires produced a kick followed by an expontial discharge to zero, suggesting a faulty DC supply at the exchange. Easy to see with a moving needle, whereas a digital readout would have produced a rapidly-changing number that would have been less easy to interpret.

[Roger Ramjet]

In the early days of working on burglar alarms, we used analogue meters that were brilliant for detecting wandering HR faults on circuits with reed switches or cotton lace closed circuit wiring. Tapping suspect bits of the circuit would cause the needle to instantly move, proving the fault location.

Later on we were given DMM's & these were brilliant for setting the charge voltage on SLA PSU's between 13.5 & 13.8 but of curse no good for rapidly wandering HR Faults.

Rog

[Lucien Nunes]

I always use the digital except when one of the following apply:

* Comparing results to service data where a specific meter was used, in circuits where meter loading makes a difference (often it doesn't)
* For effect / demonstration / vintage experience when working on vintage equipment where the AVO isn't a hindrance.
* Want the best trend-following visual response (bargraph on DMM is good but not as good as a real pointer)
* Want average-responding movement (e.g. unsmoothed battery-charging current)

In practice this means that the AVOs get used every week or two, compared to many times a day for the digitals. When working on test equipment such as other meters, as mentioned above one needs at least an order of magnitude better accuracy to calibrate, although not to faultfind. Doing anything to an AVO would automatically have me reaching for the best instrument available, either the bench digital or the Fluke 289.

[The Philpott]

I suppose the answer is to have an analogue meter and also 2 digital meters- one auto ranging and one manual ranging. I include the latter since (for example) when probing a moving coil meter using an ohms setting you need to be cautious with the current applied.

As already summarised the refresh rate and sampling rate of a DMM is a handicap in some situations, as (can be) the high impedance. Use of a low impedance voltmeter means you don't pick up phantom voltages from adjacent wires in a loom.

I understand some Flukes have a selectable low impedance function.

Dave

[Vintage Engr]

Agreed,
Regarding the low impedance function, this may seem odd to newcomers. It was added specifically for the needs of electricians, where a very high input impedance can result in non-existent "Phantom" voltage reading on a nominally un-powered cable. The so called phantom voltage is read due to capacitive and or inductive pickup along the cable route.

An fine example of this happened to a friend, who was staying in a flat, 'House-sitting'
The central heating failed, and although my friend was more than capable of tracing the problem, he obviously called out to a heating company whose name was on the boiler. An engineer arrived, spent much time prodding around with a Hi-z multimeter, saying that the fault appeared to be with the room-stat, but he couldn't understand why he was only reading 130 volts on the contacts... My friend put him right and then found the fault.

David.

[David G4EBT]

AVOs keep cropping up as the instrument of choice, which is fine for work on valved equipment with higher voltages, and AVOs are invariably what are quoted on data sheets. Sometimes forum members will say 'is it OK to use a digital multi-meter when an AVO is specified in service data?', and the general answer is 'Yes, but the readings might be a bit higher than stated', though without an in-depth explanation of why this is so.

It might be worth explaining why AVO 7s and 8s or similar low impedance meters when measuring low voltages across potential dividers will not produce accurate results. True, the meters will accurately measure the actual voltage that they see, but that voltage will be lower than we'd expect to see because the AVO will act like a resistor in parallel with the circuit under test. EG if we have a potential divider using two 100k resistors across a 12 Volt supply, we know from Ohms Law that the calculated voltage at the junction of the two resistors will be 6 Volts.

However, an AVO 8 (or any 20,000 OPV meter) on the 10V range will be like putting a 200K resistor in parallel with the 100K one in circuit, so will produce a reading of 4.8V - not 6V. However, an FET or Valve Voltmeter (such as the Heathkit V7A), with an impedance of 11 Meg Ohm will show 5.97V, close to what we'd expect to see.

Worse still if we apply an AVO 7 (or any 1,000 OPV meter) on the 10 Volt range across the junction.

It will be like putting a 10K resistor across the 100K one, so the reading will be only be 1 Volt – not 6 Volts. 1 Volt will the actual voltage due to the loading of the AVO in parallel with the resistor. Not so important when taking readings of higher Voltages, and there is the merit of using the same meter that was used when measurements were taken on valve radios by the manufacturers and shown on service sheets, so I do understand why many restorers like to use AVOs on old radios and TVs, but the lower the voltage being measured, the more significant this loading effect becomes.

The attached diagrams and notes showing the Ohms Law 'sums' are a bit more descriptive.

I'm not 'knocking' AVOs in any way, I'm simply saying that inexperienced users need to bear in mind these limitations, and how to interpret the Voltage readings they see.

As to analogue V digital meters, analogue meters come into their own when peaking up circuits and you want to see the needle rise and fall. A digital meter is pretty hopeless in that situation.

There was a rather nice High Impedance MOSFET Voltmeter project in December 1986 PW magazine to build a Voltmeter with 11 MegOhm input impedance, with eight ranges from 100mV to 500V. It included an AC Voltage probe to measure Voltages up to 10V RMS in the frequency range 50Hz to 200MHz, so it covered audio, video, and RF frequencies into the VHF range. It can be found at this link:

https://worldradiohistory.com/UK/Pra...PW-1986-12.pdf

Hope these wordy notes might be useful to someone.

[Vintage Engr]

Brilliantly explained David,

I was about to do something similar, but you beat me to it. Always good to read your most interesting posts,

David.

[ms660]

Originally Posted by David G4EBT

Worse still if we apply an AVO 7 (or any 1,000 OPV meter) on the 10 Volt range across the junction.

It will be like putting a 10K resistor across the 100K one

So far as the loading presented to the circuit by the AVO 7 is concerned it can be regarded as 500 Ohms per Volt not 1,000 Ohms per Volt for the range selected, this means that in this instance it will be 5k not 10k that's presented to the circuit when the 10 Volt range is selected as you will see in the schematic on the web page linked to below:

http://www.richardsradios.co.uk/avo7.html

Lawrence.

[G6Tanuki]

The problem with using any 'vintage' test-gear is that you really need to have modern test-gear on-hand to ensure it is working properly [and to fix it if it isn't].

I call this the "Tanuki's Paradox" [the Tanuki being one of Japan's trickster-creatures that was known for deceiving simplistically-minded humans into time-wasting and perplexing dead-ends].

OK, so use vintage test-gear if you must, but personally when I want to get the possibly-faulty equipment back into service fast, I'll use modern stuff first!

Don't chase your tail when 'odd readings' could be caused either by the actual problem or by inadequacies in your 50-year-old test-gear!