Old or New multimeter?

[majoconz]

Bit OT but the moral of the story fits.....Back in the 90's I worked for a small airline in the South Island. All the avionic guys were given a Weller bench soldering iron and a Fluke 77. Listening on the company chit-chat frequency, an aircraft was coming in with the port fuel pump inop. Minutes later an airframe engineer came into the shop and asked to borrow a Fluke and he went away happy. Half an hour or so he was back - "there's 28vDC on the connector so we changed the pump - it didn't go, so we put in another pump and that didn't go either. Now what?" I put my favourite bit of 28VDC test gear in my pocket and took a ride up to the tarmac. I disconnected the plug from the pump, connected my magic bit of test gear to the plug and called the boy in the cockpit to turn the pump on, he yelled back 'pump on'. I said to the engineer 'there's no 28v on the connector'. he replied "Well, there was when we tested it with the Fluke" to which I replied "well, there's not when I test it with my light bulb!" 5 minutes of work in the cockpit to change the switch and all was good. Sometimes the simplest test is the best!

[majoconz]

Stuarth said

With Analogue meters, the circuits are relatively simple, so you can see what sort of rectifier response it has, the equivalent circuit for Ohms ranges so you can see what voltage and current you are delivering to the device under test etc.

As well as an Avo Mk7 and a Simpson 260 I have a little Sanwa 390TR which is conveniently small for use on a crowded workspace. It does not have a capacitor in the AC switched ranges, only a diode and a different set of range resistors so on the AC ranges it reads "Peak voltage reading calibrated to show RMS value" which lead me completely astray more than once when I first got it! Trying to read the AC voltage on the anode of a valve gave some astonishing results! When I found a circuit diagram all was revealed - and I also found the the socket marked "output" had a capacitor to the +ve terminal which did the job nicely!

[Craig Sawyers]

If you want a truly high impedance meter, chose the precursor to the DVM, the differential voltmeter. Fluke were prominent in these, use 5 or 6 digit Kelvin Varley dividers to get that number of digits resolution - and are null reading.

The key thing is that when the meter is nulled, the input impedance is infinite. Well actually it is any internal leakage paths, which are tested at hundreds of G-ohms. From 0 to 1100V.

So if you want a non-loading DC voltmeter the 895A is the one to go for, and for an AC/DC voltmeter the 893A (infinite impedance on DC, and on AC 5Hz to 100kHz, with an AC input impedance of a somewhat low 1M//20pF)

Craig

[DrStrangelove]

Originally Posted by majoconz

Bit OT but the moral of the story fits.....Back in the 90's I worked for a small airline in the South Island. All the avionic guys were given a Weller bench soldering iron and a Fluke 77. Listening on the company chit-chat frequency, an aircraft was coming in with the port fuel pump inop. Minutes later an airframe engineer came into the shop and asked to borrow a Fluke and he went away happy. Half an hour or so he was back - "there's 28vDC on the connector so we changed the pump - it didn't go, so we put in another pump and that didn't go either. Now what?" I put my favourite bit of 28VDC test gear in my pocket and took a ride up to the tarmac. I disconnected the plug from the pump, connected my magic bit of test gear to the plug and called the boy in the cockpit to turn the pump on, he yelled back 'pump on'. I said to the engineer 'there's no 28v on the connector'. he replied "Well, there was when we tested it with the Fluke" to which I replied "well, there's not when I test it with my light bulb!" 5 minutes of work in the cockpit to change the switch and all was good. Sometimes the simplest test is the best!

^^^This. You can't beat a good old tungsten lightbulb.

Though I note with interest that Lidl(tm) have a fine stock of assorted DVM's at the moment.

[stuarth]

A lot of small analogue multimeters like the Sanwa mentioned above have a DC coupled half wave rectifier for the AC ranges. They repond to the average value of positive half cycles, but are calibrated for the RMS value of a sine-wave input assuming the the negative half cycles are present.

There is a possible safety problem to this in that if you probe something with a negative DC voltage while set to an AC range, the meter will read zero. There is however a very simple mod for this, which gives the meter a useful added feature.

The actual rectifier is a small signal germanium diode in series with the meter movement (germanium for low forward voltage drop) with a second small signal diode (usually silicon) in reverse across the diode/meter pair. The second diode is there to limit the reverse voltage across the meter diode, and does not otherwise affect the meter reading. The current is limited by the voltage range resistors, and neither diode sees more reverse voltage than the forward drop of the other diode, hence they are both small signal diodes.

If you replace the second diode with an LED, then for AC, you get a normal meter reading and the LED lights up. If what you're probing is DC, then for positive voltages you get a meter reading and no LED, and for negative voltages you get no meter deflection, but the LED lights up, and you know you need to switch to a DC range to make a measurement. So with one quick check you can see if there's voltage present, whether it's AC or DC, and if DC, what polarity it is.

Modern high sensitivity LEDs light visibly with only a few microamps of current; on my modified Robin OM70, with a reverse connected PP3 the LED (red for low voltage drop, and clear, not frosted type) is visible on the 50V AC range, and even on the 250V range if the lighting is not too bright. The mod is electrically easy, once you've found the germanium diode, the second diode you need to replace will be lurking nearby. The only real problem is where to put the LED so you can see it, but it doesn't look like a bodge.

Stuart

[The Philpott]

In reference to the original question posed, fault finding an old analogue meter (or just checking it over for precision) I will almost always be using a manual ranging DMM.
I use a similar philosophy when setting a clock right, rather than referring to another clock or watch I 'phone TIM. Errors can escalate otherwise.

The second I suspect a DMM of glitching I reach for another one, on the basis that it's rather unlikely they will both choose to glitch at the same time. If still in any doubt I hand the devices to someone else to do the test as I have been known to generate a lot of static and I don't think cheaper devices are always properly screened for this.

Dave

[Tomshep]

TIM? I use a radio clock. Good enough for my needs.
Regarding meters, a good Avo 8 costs from £15 these days and there is one on every street corner. A Fluke 25 usually costs between £25 and £40 so the only real issue is bench space. This is not a discipline where one needs to measure to fourteen decimal places and you get to know your test gear so buy one of each and use the other to confirm results when you get those you don't expect.

[DrStrangelove]

At least one of my dvms doesn't initialise properly every time it's switched on, so one needs to do a sanity check on a 1.5V AA to make sure it's in a sensible mood.

I won't mention the 5.5 digit solatron that came up with "FAULT" a fortnight after I bought it.

[6AL5W-Martin]

Originally Posted by Craig Sawyers

If you want a truly high impedance meter, chose the precursor to the DVM, the differential voltmeter. Fluke were prominent in these, use 5 or 6 digit Kelvin Varley dividers to get that number of digits resolution - and are null reading.

The key thing is that when the meter is nulled, the input impedance is infinite. Well actually it is any internal leakage paths, which are tested at hundreds of G-ohms. From 0 to 1100V.

So if you want a non-loading DC voltmeter the 895A is the one to go for, and for an AC/DC voltmeter the 893A (infinite impedance on DC, and on AC 5Hz to 100kHz, with an AC input impedance of a somewhat low 1M//20pF)

Craig

here you can see this John Fluke, a very trustful instrument !
The DC input impedance is a electrometer up to 10 volts, upper ranges are 10M.
This Oldie is not afraid to be compared with 5 1/2 digit DVMs.

[The Philpott]

Outstanding. We'd be lucky to find another one of those in working order!

Dave

[John_BS]

This is a "Standard" DC voltmeter made by Sangamo Weston, and used by the BBC Test Lab to calibrate AVO's.

It has a vernier scale which can be read to 0.2% or better, to complement it's 0.1% claimed accuracy.

John

[The Philpott]

What a clever idea on the scale. (Those tactile paignton knobs must have been hugely popular, they show up everywhere)

[ScottBouch]

Hi, too many messages to read, so just jumping in with an observation on the initial question...

Volunteering on old aircraft (EE Lightnings) with long cable runs from front to rear, we found that high impedance digital multimeters could give misleading readings.. Capacitive coupling between wires bunched together over long distances would result in a "phantom" DC voltage being measured (7 V to 9 V ish) at the end of a circuit, when it should have been 0 V.

Using an AVO, the load of the moving coil would put a stop to that and give sensible readings!

I've also heard Fairies of that era (official name for avionics trades) saying "never bring an oscilloscope near an aircraft" - I assume this may be for a similar reason....

Cheers, Scott.

PS: I too love that vernier scale! Genius!!

[wave solder]

I used to fit solar panels on Narrowboats and when it came down to fault finding I would use my AVO 8 because of the lower impedance which would reduce the erroneous results when there were poor connections in the wiring.

[G6Tanuki]

I'm happy to use a modern cheap-and-cheerful £10 DVM for mains and 'valve radio' type stuff where currents tend to be low and voltages high, but I also have a home-made volt/milliamp-meter using a 5mA ex-WWII-surplus meter, 5-Watt wirewound resistors and with a pair of back-to-back diodes across the meter for protection; this is my go-to for DC low-voltage-high-current stuff [vehicles, battery-powered gear, computer PSUs] because it takes a sensible current and is so unlikely to give a delusional reading due to induction or insulation-leakage.

I also have a 30KOhm-per-volt taut-band meter from the 70s by Taylor, which I use for those instances where I am not interested in accurate measurement of voltage/current but are adjusting for a voltage-peak/dip and where an analog indication makes it easier to get things set up 'on the nose' than looking at a flickering final digit on a DVM.

I've never liked stuff like the AVO8; they always seemed too big and clunky to me. You can't put one in your pocket!

[cmjones01]

Different multimeters for different purposes. Regarding input resistance, it's certainly something to be aware of when using a meter. My benchtop HP 34401 can be set up for infinite input resistance on the lower voltage ranges, which is sometimes handy when dealing with very high impedance circuits. On the other hand, my favourite handheld meter, a Fluke 12, usually has 10 megohm input resistance but has a party trick: on its ohms or diode check/continuity ranges, it'll automatically switch to reading voltage if it detects a voltage connected to it. In this mode (called "V-Chek"), its input resistance is somewhat lower (tens of kilohms, I think, akin to an analogue meter) so it's great for detecting "phantom" voltages.

A few years ago I used an Avo 8 to great effect when tracing undocumented mains wiring. It was in a house some relatives had bought which had reached "second fix" stage about 20 years ago but never been finished, so there were just wires hanging out of the walls and no indication of what was a socket, what was a light switch, or anything. I found that the Avo's AC voltage range would show 230V on a live wire, of course, but also show a little needle deflection on an unconnected wire core in the same cable due to capacitive coupling. It was unambiguous, unlike a digital meter would have been, and was great for figuring out which wires were intended to be switched how. I'm particularly proud of having figured out that one lighting circuit had a position for an intermediate switch as well as two two-way switches, and I don't think I could have done it without the analogue meter.

Avos are rather big and heavy, though, especially when working up a stepladder, and a pocket-sized analogue meter would probably have been just as good, but I didn't have one.

Chris

[G6Tanuki]

Another reason I like using a DMM over an old Avo or something is that most DMMs include an audio 'squawker' continuity-tester - great for those situations where you need both eyes on the probes and so can't also look at a display/dial to see whether you have continuity or not.

[One of the first electronic-things I built as a kid was just such a squawker: two OC71 as a multivibrator, a third one driving a telephone-earpiece as a sounder, all powered by one of those tubular 9V PP4 batteries that looked like a D-cell but which had a press-stud connectopr at each end. With experience you could get a feel for the resistance between the probes by the pitch of the squawk. Very useful - indeed so useful that some light-fingered ****** stole it!]

[The Philpott]

Agree. I find that a continuity squawk is almost a must-have especially for repetitive tests, and i also have a large digit DMM for ease of reading without reading specs.

Opinions are polarised on the Auto Power Off facility but it's a shame that very few meters with APO allow it's disablement. One notable exception is my Uni-T, which will beep at me every few minutes to remind me that i have defeated the APO...and that's fair enough.

Dave

[McMurdo]

My daily analogue driver is an Avo 7 from 1948, I use it almost exclusively on the ohms range for semiconductor junctions or very occasionally for looking at the output of inverter drives where it more reliably reads the PWM signals.

For everything else I use a middle-aged digital Fluke 27 which remains in top calibration and is very frugal on batteries.

I have other multimeters as I am somehow attracted to buying them and they sometimes get brought into play for more complicated work, load testing, monitoring, that sort of thing.

[ScottBouch]

I also have a nice Skywood (TMK) model 500 multimeter, with built-in buzzer for continuity testing! It's a little more convenient than the AVO.

But there is something a little bit special, almost ceremonial about getting the AVO out and setting it up on the bench compared to taking a quick Fluke DMM reading... Feels less transient / temporary, more meaningful somehow.