Megohmmeters - Analog or Digital?

First I wanted to recommend analog meter but your recommendation is absolutely correct.

Qqberci,

Analog meters are fine in the range of uA and mA. To derive such a high current through insulators, you also need very high voltage, which makes it hazardous and expensive.

200V power supply costs about US$250. 1000V to 3000V DC power supply will be about US$300. You also play high high voltage so need protected high voltage switches, high voltage cables, connectors, all adding to several hundreds of dollars more. uA meters are cheap.

If you go digital then $50 meter can measure nA current and you can then use simple 1V to 9V to excite and measure in 3-1/2 digit or even 4-1/2 digit. Putting slightly more you can read in 5-1/2 digit or even 6-1/2 digits.

There is no comparison between analog meters and digital meters. Digital is often low price, accurate and also permits lower limits to reach in current and voltage measurement.

Many working engineers are not aware of where to start and I think some kind of help can be extended by a blog, information link, and even by making available small easy to test circuits and devices for do it yourself experience. If some ready made solution exists then a reference to it will be a good idea. I think by now a lot of information exists on the net and some can be created by involving and becoming an active group without affecting others for those it may be very primitive.

I like to share information and many other do it here. However, it is not so planned in unique direction. Discussion itself is a good idea amd I find many here try to help. It is very good.

My preference would be for an analogue type meter. May be they are well-proven. And, in digital meters - there can be more problems due to interferences. Analogue meters are rugged.

We use a digital M-ohm Meter that registers up to 4,000 M-ohm. Easy to use and no problems to report. That said, accuracy is not a primary concern.

Cheers!

This is an electrical and when you mention IR then mostly it moves the subject to Telecom engineering's copper cable testing.

Recommendation:

1. Megger of Megger Corporation www.megger.com

2. Preference is Digital where Analogue can also do. Share more details of your work as specific model can be recommended.

3. Recommended Moels: [1] BM-80, [2] BM-15, [3] MJ-15, [4] MJ-159, [5] MIT-400,,,

Wishing you success --- Furqan Mohammad

Sapper,

You often have a very sensitive current meter and a constant voltage source to measure resistance.

Assuming that you have 1nA to 1uA range current meter with say 0.1pA accurate then 1V across 1G Ohms or 1000 M Ohms will give you that 1nA reading. 1M Ohms will be at 1uA for the same 1V applied across the resistance.

Lower voltage is better as higher voltage can cause breakdown in insulator. Some time you can apply as high as 1000V without having breakdown in the insulator, but that is what it is designed for. Some insulation material may be very thin and may breakdown even at 100V. Hence, use only say 10V reference voltage source and measure current in the range of 10nA to 10mA for resistance variation from 1G Ohms to 1KOhms.

I think digital measurement in six digits is much better or use a Logarithmic analog meter for this purpose to have all ranges on single scale.

9V battery can also be used as a voltage source as current through insulator will be limited to uA. For safety purpose add 1k ohms resistance in series to limit the current to 9mA even in short circuit condition.

All you need is a 9V battery, 1K resistance and a current meter for 10nA to 10mA range or 10nA to 10uA range for only high Meg measurement.

If you have a milli volt meter then you can also make a current source out of 9V.

Digital is preferable for accuracy , as the readings are usually stable.Also a meter like BM80/2 has analog combined and 50 to 1000V selectable so you can cover LV & HV gear.

The variable test voltage is handy if you want tweak a suspect circuit with a slightly higher voltage.

Garth,

50V-1000V is very high excitation voltage. It may be breakdown voltage for some materials or can cause self heating. Multimeters use only very small voltage and can measure 10M easily. For 1000M 9V is more than enough. High voltages are heath hazards and are used only when one needs to measure Terra Ohms level resistances.

Hi Shyam;

High voltage is not exclusively used for measuring terra ohms.

Remember that an IR test is a safety test as well as Hi-Pot, thus a breakdown in insulation shouldn't be tryed to avoid during test, because that is exactly what you're trying to discover before the customer gets zapped. The Mfr. specs are clear as to what voltage shuld be used when testing insulation of a device or appliance and they respond to industry standards and CE, UL or TUV requirements, you can't arbitrarily bring them down.

As per the heating issues, I've never seen a D.U.T. (in my case, servo, stepper and household electric motors) getting hot even when shorted and the test voltage as high as 2600 VAC.

Though, I've seen them warm up a little, but during resistance test of windings (next measuring will be slightly higher) but thats entirely another matter.

If you're specifying the test voltages on your products, I sugest you consult another professional for advise before your company is in trouble.

Regards

Yahlasit

Hi Guest,

You are mixing too many things and also putting adverse comment which is unprofessional. Where in the thread did you see CE, UL and TUV specs under discussion? None of my comments are for certifications, and they are for this thread and for a person who finds hard to find difference in analog and digital meter preferences. You are writing under name gust so I will not know which guest you are and why you have such comments and what is your point here in this thread.

Shyam

Wow ! just compare your explosive response to what I posted.

If the original poster is asking about test equipment for IR, then such specs. matter.

Read a little about IR and Hi-Pot tests and how important they are.

If you think I'm mixing things then you have little to no understanding at all, about safety certifications, they imply some destructive tests, wether you like it or not, you can't just do your will in such aspects. You can't be cheap and lousy abou it !

Now I'm writing in a not so friendy fashion, but just because you saw harsh comments where there weren't. Can you see the difference?

If you don't care about someone getting killed, then I don't care about your comments at all. Make whatever recomendations that you imagine to be correct.

And God help us from "professionals" like you.

Yahlasit

No, digital and analog will have the same level of accuracy. Digital will often (not always) have a higher degree of resolution. These are two different measurement qualities.

I agree with Shayam. Insulation testers can cause damage and can be expensive.

I perform maintenance on towed electrical marine cables and I use a small 20,000ohms/volt analog Triplett model 310 meter and it has always performed well in finding leakage. The reason for this is; unlike a digital meter that has a 1Mohm input impedance and requires very little current for its resistance tests, the 20Kohm/volt meter requires a larger current for it's hi-resistance readings. This is supplied by a separate 15-volt battery in the meter and provides sufficient current to find those elusive ground-faults and leakage. The Triplett 310 will display resistance up to 20Mohm.

Of course, i also carry a digital multimeter for voltage and resistance measurements, but the meter movement of an analog meter will display those elusive intermittent connections as the meter needle will show a "kick" whenever an intermittent connection is encountered - a digital usually will not. Something to think about.

It's horses for courses - the regs. say that certain mains powered products have to be tested at such-and-such a voltage and show an earth leakage of less than so-and-so. If what you're selling breaks down with those parameters, tough - you can't sell it.

Testing semiconductors and such like is another kettle of fish.

True but, Sapper didn't not give any specifics so, in answer to a general question one should be expected to receive only a general answer. Do you agree?

Regards

I agree that the OP wasn't specific (tho' it's arguable that his words "Insulation Resistance (IR), and earth continuity testing" suggest that product or system electrical safety testing are likely to be the issue).

The point I was making was that Shyam was talking about one aspect of insulation testing etc., while others (myself included) were leaning towards the safety testing side - and there's no point in an argument about which is correct, or even best, as there is no "correct" or "best" without more specific info from Sapper.

Cheers,
John

JohnDG,

You are correct, I am primarily intertested in Electrical Safety Testing for LV domestic or commercial environments.

Thanks & Regards,
Sapper

Dear JohnDG

I also have been asking for high isolation (6kV) AC-DC power supplies and lower leakage current types. I have been working safely on 10kV power supply design using these AC-DC to power them. However, now working with new 20kV and 30kV designs everything is going wrong when 100uA to 200uA leaking through thick PTFE cables charging the instrument body and looking for easy path to reach earth to dump all charge at high potential into Ground which it can not do unless corona breakdown takes place through whatever is near to it with AC mains Live-Neutral-Earth points. What if earth is open. It connects body to live or neutral which ever is nearby.

Even while I use isolation, I am also looking for safe discharge of unwanted charge from places where it should not be collecting and building up to unsafe limits.

Worst thing happened that it found easy path through voltage regulators mounted on instruments casing to get easy heat dissipation and were conducting through body became discharge path and burnt out each time I exceeded 16kV. Now I am removing those regulators and looking to design earth fault relay inside the instrument to trip high voltage if there is any earth fault. If I don't do that them instrument itself will burn out or may kill someone using it as it may discharge through control panel.

I need to put Faraday cage to all my electronics separately from high voltage electronics and make sure earth safety system is in place by automatic detection.

It is not a question of analog or digital alone but also of instrument and operators survival.

Cable and other insulation testing to be done at high voltage applied for short time and by monitoring pulse current for short period to avoid damage. Often us time is enough. 20kV cables to be tested for 40kV or even 60kV not only to withstand high DC voltages but to take rated AC voltages also.

Many manufacturers raise their hands for real 20kV or 30kV isolated DC-DC converters. I now understand that it requires 60kV design and test environment to know what will go wrong very easily.

Lots of problems may come if high voltage return is connected to common point of the instrument or its body even if by design HV return is at Ground potential. High voltage discharges are very funny and things go wrong in ns and us period.

I think you young engineers are discussing the right thing. I appreciate it else you will learn through bad experience or may not be there even to tell your story. One of my colleague died in fraction of seconds working with OFF state laser power supply which had enough charge on HV capacitors even in OFF state of instrument days after to finish the story of his life forever. This can happen to anyone. Make sure you all know things in advance and play safe.

Exactly the point I've found in general between analog and digital measurement equipment. Analog equipment does a better job at troubleshooting a device when used by a skilled operator. The analog display implies many more parameters than just a single numeric value. This permits a troubleshooter to gain a qualitative understanding of what is or is not happening during a failure. (The technician's perspective.) Digital instrumentation with higher resolution and ease of data capture for presentations and archives, works far better for quantifying how well a working system works. (The engineer's perspective.)

redfred,

Even though digital reading suffer from quantization error and INL and DNL errors, digital auto-ranging is more like multiple gain switched analog. Analog even though is near perfect continuous information, but we can not know what it is very accurately unless we acquire information and analyze it with multiple measurement. In some applications where we reach the limit of digital due to very fast measurement, you can say analog is the only way. I think, here we have slow measurement and there is much easy way to go digital.

Yes, digital provides more accurate readings and performs as an auto-gain device where it it automatically adjusts to utilize the full dynamic range of the A/D converter. But again, an analog meter movement will provide a better indication of intermittent problems weather it is a multimeter or a megger.

tropicalspeed,

Can eye on scale resolve better than what precision ADC can do and what signal averaging can do? When needle swings this way and that way head of the operator also goes the same way. Digital computation for measurement can do wonders. Think again.

You can easily place squrt(2) on scale but you can never resolve it in digital, but tell me which one gives a much better understanding of position on scale to you? Uncertain analog or many digits of digital?

Well, it all depends on what you propose to do with the meter. If you are looking for quantization for reports then a digital is your answer but, if you are looking for trouble areas then an analog is your best bet.

you said it right in your last post, "When needle swings this way and that way....", wouldn't you agree that your instrument is trying to tell you something? A digital, on the other hand, is not likely to.

You are making a lot of asumptions to claim that digital is always better. From a phylosophical point of view, how can any single thing be better than everything else all the time. I was trying to give the OP a paradigm on when an analog display is preferred and when a digital display is preferred. I was not making the choice between displays for the OP.

When one is troubleshooting a system, accuracy often is the very last consideration.

redfred

I agree, that analog contain everything of which digital makes use in part with quantization error and many other errors associated with ADC.

redfred,

It looks to me that analog meters made old galvanometer coil type meters will be more reliable in high voltage applications as they do not likely to fail due to breakdown voltages. Earth connection to instrument body is also very essential to remove charge build up on the instrument body. Digital circuits can't take HV corona discharges.

shyam,

I think you are beating this post to death, check this link for HV testing . The company would not be selling somthing which would fall apart with its own test voltages.

http://www.ledaelectronics.com.au/kyoritsu_3121_tester.htm

Dear Garth,

Right now on the tale for me are several 20kV HV Supplies and white testing above 15kV I found that corona leak through cable was increasing to about 50uA and at 20kV it exceeded 200uA and then due to floating power supply it tried to find path through electronics to mains power and killed whatever came in between. If there were electromechanical meters, I am sure they would have survived.

Electronics becomes very sensitive to high voltage charge and even small current is too much for these when it comes through high voltage static discharge where peak current can be very high or voltage developed due to charge collection over other element can easily kill electronics.

I think, many engineers who are here with electrical circuit experience know much better about it. Have you see high tension line cleaning using helicopters. Watch that to know what an open circuit can do even to the floating flying helicopter.

To conclude every electrical engineering in one line will be a serious mistake. I do not doubt sincerity of real working engineers looking for valuable information. I know what exactly they are talking about. I do not react to their young blood reactions, but have to see myself what it means in real engineering.

So far I was making floating HV Supplies, but now I am looking for some path for this static build up to go away. What if there is an earth fault? It will kill everything again if there is no other means to discharge the accumulated charge. I am not relying 100% on the earth circuit alone.

Normally electronics circuits / ICs divert charge either into common / ground or to Vcc the positive power source. These are having very limited capacity to handle charge and voltages even though lot has been tried out in current era.

Hi Sapper

Is good to see you don't want to become a Zapper; I've been a customer of Quadtech (formerly GenRad) for many years, and I'm sure they have a readily solution for you, e-mail your question to: marketing@quadtech.com they're honest and don't try to sell you the most expensive equipment they can, they will give you a linecard and also catalog prices, or you visit their homepage and see for yourself.

See this archive, I hope it helps.

Yahlasit

Sapper,

If your intent is to determine the ability of the insulation in a particular piece of electrical equipment to insulate or prevent the normal operating voltage from going to ground or between components and you want to record or log this data for the purpose of spotting a "trend" or deterioration of the insulation, moisture intrusion or contamination over time, I would suggest a digital version for ease of reading.

As for earth continuity, it would depend on what amount of current you would want to apply.

Most "Meggers" have ohmmeters that will show continuity but at very low current levels.

here's the question. What do you want to do with it once you got it? do you just want to test or test and record the results? If you want to just test then get analog, if you want to test and record for certification purposes get a digital with computer connections and software so you can download and do all kinds of wonderful things. The only thing you need to decide on is what voltage you need/want 500v, 1000v or 3000v or multi-voltage. It really isn't rocket science, have a look at Radio Spares (you got RS in Oz?)

Digital because it has high Z and is easier to read. I will even help you by the right brand more expensive though - FLUKE

I would say go digital. Reading numbers is so much easier than trying to see the location of a needle with my sorry ol' eyes. Quantizing error and such mentioned above (in a good instrument) are miniscule compared to parallax error when reading a needle.

Bill

Depends on what you are doing,if your in the field trouble shooting motors,feeders,ect

I wouldn't trade my hand crank megger for anything,never a dead batttery,and rugged as a rock.