Analog multi meter - standard vs FET input

[nukie]

Hi all,
I need help in understanding how these FET input based analog meter works. I am thinking of getting one to read average current & voltage from a pulsed source, doesn't need to be very accurate because I will do some calculation.

I really like that FET based analog meter has very high input impedance but does the inclusion of FET circuitry turn it into a digital meter with an analog front end?

Obviously I have zero idea how these FET based meters work, all I can guess that it uses some sort of amplificaiton after the inputs. Should I be buying a standard high sensitivity meter or a FET input based analog meter? Please show me the way.

[Kiriakos-GR]

well I need to ask one question up front before I say anything .

How much money do you plan to spend for it ?

[Zero999]

I'd assume a FET input just means it has a buffer with a very high input impedance before the meter.

[Zero999]

The classic circuit is two FETs, paired and temperature coupled, to form a differential amplifier.  By todays standard worth a few cent.

I would've thought a modern analogue meter would simply use an FET input op-amp because the transistors will be better matched and it's cheaper.

[EEVblog]

I really like that FET based analog meter has very high input impedance but does the inclusion of FET circuitry turn it into a digital meter with an analog front end?

You could look at it that way, but you do still get that "analog" moving needle performance.
Having the FET input gives it a fixed high input impedance just like a digital meter, and it requires a battery to power it just like a digital meter.
The beauty of an analog meter on the other hand is that it doesn't require a battery (except for ohms), the input signal swings the meter directly.
Some can be quite sensitive actually. I used to have a 100K/volt analog meter.

I have a Tandy/Radio Shack FET meter in the lab.

Dave.

[Kiriakos-GR]

At this point I will add one question to the experts. 

nukie wrote that he need to measure a pulsed source.


At the users manual of my own Metrawatt MA4S,
its says ....... calibrated for sinusoidal 50Hz , still capable for square waves ( add 11% at your readings)
and triangle waves ( remove 3% from your readings) .

And the question is, are all the analog meters capable to measure all sort of waves ?   

[nukie]

hello all,
thank you for your replies very interesting insight on how they cut cost on cheap high impedance analog meters.

My concern was simply based on how DSO vs analog scopes as to these analog meters. Will the high input impedance circuitry alter the actual reading? Assuming we have an unobtanium ultra sensitive analog meter which matches those specs of FET based meters.

I know Sanwa makes very good analog meters having seen electricians use them when I was a kid. Sanwa curently has only one FET based analog meter.

[vk6zgo]

At this point I will add one question to the experts. 

nukie wrote that he need to measure a pulsed source.


At the users manual of my own Metrawatt MA4S,
its says ....... calibrated for sinusoidal 50Hz , still capable for square waves ( add 11% at your readings)
and triangle waves ( remove 3% from your readings) .

And the question is, are all the analog meters capable to measure all sort of waves ?   

Basically.FET & Vacuum Tube Voltmeters read Average values for AC,but have the dial calibrated for RMS of a sine wave,just like ordinary analog meters.
Some special units do have circuitry to read True RMS,but are very rare.
Others,rarer still,are calibrated in peak values,but this is just a special scale on what is really an average reading meter.

The advantage with these sort of meters is that,like a DMM,they have a very high input
impedance,& when connected across the normally much lower impedances of typical circuitry
have very little effect upon that circuit.
VK6ZGO

[Zero999]

Or maybe the Chinese companies just copy old designs.

I suspect that's more likely.

Having the FET input gives it a fixed high input impedance just like a digital meter, and it requires a battery to power it just like a digital meter.
The beauty of an analog meter on the other hand is that it doesn't require a battery (except for ohms), the input signal swings the meter directly.

That's true but surely you should get better battery life from a good quality analogue FET meter than a digital meter?

[Bored@Work]

Will the high input impedance circuitry alter the actual reading?

It depends.

For electronics you want a high input impedance (in the voltage ranges, that is). Thats why people once started using voltmeters with tubes and later with FETs. Still, the input impedance shouldn't be too high to avoid false alarms triggered by every electron happening to spent the meter a visit. And it should be affordable. Current high input impedance digital and FET meters are already confusing for some. There was a recent thread where someone did sent his new meter back as defective, because he didn't understand that it was just picking up random voltage when the leads were left open 

Physicist and other mad scientists sometimes require extremely high input impedance meters. These meters are called electrometers and are a class of its own. Modern ones are also build with FETs or ICs with FET input stages. But the key is in the construction, because e.g. the resistance of PCB material, which is typically so high that it is regarded as an isolator in normal electronics, ruins the input impedance. There you see stuff like teflon, air and black magic when they approach peta ohm input impedance.

Electricians on the other hand can often live without high input impedance, and sometimes they outright want a lowerish impedance voltmeter, to not get fooled by random picked up voltage.

[Zero999]

Yes an electrician's meter will have a low impedance input to eliminate ghost voltages.

[alm]

A higher input impedance will sometimes produce different readings because of lower circuit loading (depending on the circuit's output impedance). Usually this is a good thing. One exception (apart from the already mentioned ghost voltages) is that some old service manuals were written for eg. 10kOhm/V analog meters. Measurements with a 10Mohm input impedance may not agree with the values in the manual.

[Kiriakos-GR]

I did one quick check about impedance on my Metrawatt MA4S .

ranges : AC

75 mV ( 100 Ohm DC only)
750 mV  (1K DC only)

2.5V   340 Ohm  ( 2.4K on DC )
7.5     2.98K   ( 10K on DC )
25V    34K
75V    100K
250V  340K
750V  1 M Ohm. ( this range works up to 2.5KV for 30 seconds )

And I will agree with vk6zgo,
that this specific one it belongs in the group of the rare ones,
about the True RMS .  

And the most funny part is that I did not had any idea about it true specs when I found it on ebay,
I got attracted to it , mostly from the amount of the current ranges,
and the easy to read numbering on the meter.  
 

[Zero999]

A higher input impedance will sometimes produce different readings because of lower circuit loading (depending on the circuit's output impedance). Usually this is a good thing. One exception (apart from the already mentioned ghost voltages) is that some old service manuals were written for eg. 10kOhm/V analog meters. Measurements with a 10Mohm input impedance may not agree with the values in the manual.

Fortunately that's easy to get round with calculations, if you have the schematic or simply connecting a resistor in parallel with your volt meter.

[Bored@Work]

One exception (apart from the already mentioned ghost voltages) is that some old service manuals were written for eg. 10kOhm/V analog meters.

Such a spec. is of course not enough. You need to have the range given, too, to have a defined input impedance.

[alm]

Agreed, but I've seen something like that quite often on old schematics. For example from a Tektronix type 565 manual (page 115):

All circuit voltages were obtained with a 20,000ohm/V VOM. All readings are in volts.

No mention of range, neither in the notice nor in the schematic itself. Not sure if the ranges were standardized so people could infer them from the values? Otherwise it's quite useless indeed.

[tekfan]

Isn't there a list of required test equipment for calibration in the manual? All Tektronix manuals have a list of required test equipment. In the older manuals the Simpson 262 is used almost always, in the newer ones a digital 10Mohm input meter.

I would go for a FET input one also. You have no worries of loading the circuit. Ghost voltage may be a problem but you can elliminate that by just putting a 10meg resistor or a capacitor across the inputs. The other advantage is the sensitivity. A non buffered (passive voltmeter) is usable to about 1 volt full scale. Any lower range and it becomes an ammeter. FET meters can go down to millivolts full scale and still have more than 10Mohm input impedance.

[alm]

The required test equipment is usually just for the calibration / performance verification. I may be overlooking it, but the only list I can find is in the calibration section (page 46/5-1), which specifies 'A conventional DC voltmeter having a sensitivity of at least 20,000 ohms per volt and calibrated for an accuracy of at least 1% [...] and an accuracy of at least 5% at 3900 volts.' By the way, nothing specific about this manual, it wast just the first I found that wasn't a large download (no point in wasting bandwidth unnecessarily) and old enough to pre-date the DMM.

[EEVblog]

That's true but surely you should get better battery life from a good quality analogue FET meter than a digital meter?

Yes, you'd expect that.
It doesn't take much to power a front end FET amp and drive a meter needle.

Dave.

[Kiriakos-GR]

Ghost voltage may be a problem but you can elliminate that by just putting a 10meg resistor or a capacitor across the inputs.

Ghost voltage can be found only in AC 220V installations, and no its not a good idea to add anything on the probes.
The old school electricians just use one light bulb to test for Ghost or real voltage on the cables.

The most modern solution about Ghost voltages or stray voltages , found on the Agilent U1272A and called as Z low,
and adds an 2K Ohm resistance as load.
The resistance of a common light bulb 220V 150W it is 25 Ohm, but it can not read the voltage    

[tekfan]

The required test equipment is usually just for the calibration / performance verification. I may be overlooking it, but the only list I can find is in the calibration section (page 46/5-1), which specifies 'A conventional DC voltmeter having a sensitivity of at least 20,000 ohms per volt and calibrated for an accuracy of at least 1% [...] and an accuracy of at least 5% at 3900 volts.' By the way, nothing specific about this manual, it wast just the first I found that wasn't a large download (no point in wasting bandwidth unnecessarily) and old enough to pre-date the DMM.

I don't know why but the manual for the 565 doesn't seem to have such a list. I've looked in the other manuals no BAMA and they have a complete list of test equipment needed and with pictures included (atleast the older ones). The newer manuals just have a table that tells the required test equipment specs and usually Tek equivalents.

Well here's a comparison of the various analog (passive meters), all are 20kohm/V on the DC volt ranges. Scale size is also fairly important. It's like the resolution on digital meters.

Left to right:
Triplett 630 - the same one Dave's got
Simpson 260 - kind of a standard in America
Simpson 262 - huge meter movement