So I've got 3 DMMs around, which complately disagree with the voltage output of a transformer I've got. They're creeping me out a bit.
The transformer's rate output is 12VAC. I measure it unloaded and:
Multimeter 1 says 10V
Multimeter 2 says 12.2V
Multimeter 3 says 14.6V
I know DMM 1 is old and its calibration is off a bit, but the other two are just a few years old and fairly good.
Scumbag DMMs.....
The transformer's rate output is 12VAC. I measure it unloaded and:
Multimeter 1 says 10V
Multimeter 2 says 12.2V
Multimeter 3 says 14.6V
I know DMM 1 is old and its calibration is off a bit, but the other two are just a few years old and fairly good.
Scumbag DMMs.....
My experience with DMMs are quite different. I have a $12-13 one (Biltema.no .. hehe), a Gold tool ($25), a UNI-T ($50) and a Fluke ($250). There is no difference in readings. At least not AC, DC and resistance.
Kinda funny... A DMM that costs the same as a pack of siggarettes is just as accurate as the Fluke.
However, after a few months the probes came apart on the cheap one. With cheap stiff leads, it goes with out saying it wont last.
Kinda funny... A DMM that costs the same as a pack of siggarettes is just as accurate as the Fluke.
However, after a few months the probes came apart on the cheap one. With cheap stiff leads, it goes with out saying it wont last.
hi Atilla,
Any meter unless calibrated recently can be off some meters read AC as a RMS value witch is a average of the voltages,some read peak that will throw you off too.
Because that can vary with your house main load. If you have another exact trans former compare it ,most of the time if you get a reading it;s good and if you don't it's open ,and one that gets extremely hot is a bad too.
Leads can be bought anywhere ,get a set made by fluke they last years.
I hope this has helped you some ,you can measure (with good leads) your house voltage and each meter should be close to all the same reading,also a battery 1.5 v for DC all should read the same , more cost, better parts= better meter.
Any meter unless calibrated recently can be off some meters read AC as a RMS value witch is a average of the voltages,some read peak that will throw you off too.
Because that can vary with your house main load. If you have another exact trans former compare it ,most of the time if you get a reading it;s good and if you don't it's open ,and one that gets extremely hot is a bad too.
Leads can be bought anywhere ,get a set made by fluke they last years.
I hope this has helped you some ,you can measure (with good leads) your house voltage and each meter should be close to all the same reading,also a battery 1.5 v for DC all should read the same , more cost, better parts= better meter.
Things are getting better as True RMS converters find their way into more and more meters. But at one time an AC measurement was best viewed with more skepticism than confidence. There were three different schemes: Peak, Average and RMS. I recall one incident where a panel meter on a test fixture wouldn't calibrate correctly when used to monitor the output of a variac. I was just a young buck and was saved by a wise old sage who looked at the output using differential scope probes. The waveform was fine until the variac passed the input winding into auto transformer mode; then it visible distorted from a more perfect sine wave, which is where the average responding meter circuit lost linear tracking. Sine, Triangle and square waves all read different depending on the type of rectifier circuit used. True RMS (the equivalent DC heating value ) converters are now about $4.50 a pop[. So even now you wont find them in a $2.99 cheapo DMM. The very best I know of is the Agilent 3458A, which has an unimaginable AC accuracy of 100 parts per million from 10Hz to 10Mhz. Of course they sell for $7,000 used. But it is easy to understand why many of us old coggers have old HP 3400C's on our benches, or even older HP 400's.
Doc
Doc
hmm...
I dont know how you did it... but you completely lost me and made perfect sense at the same time.
Talking about rectifier circuit .. IE. half wave and full wave ?
Wouldnt a large enough cap even out the difference?
EDIT: That is, a cap integrated in the DMM? For RMS measurements, a cap is needed, right? With out, the value is either peak or average?
I dont know how you did it... but you completely lost me and made perfect sense at the same time.
Talking about rectifier circuit .. IE. half wave and full wave ?
Wouldnt a large enough cap even out the difference?
EDIT: That is, a cap integrated in the DMM? For RMS measurements, a cap is needed, right? With out, the value is either peak or average?
Last edited by a moderator:
Every value cap has its own time constant that will effect the final result. AC measurement conversion is a real evolving science unto itself. The old simpson and triplett analog meters used just simple diode and cap rectifiers. The smaller the cap the more peak responding the result. Early true RMS conversion was performed using resultant heat measurement "Bolometer" systems, as many RF power measurement systems still do (Agilent/HP RF power sensors such as 8481 & 8484) The Agilent 3458A DMM uses a unique high speed phase lock to determin A/D sample rates then superimposes a high speed jitter clock that essentially moves the sample points around for a closer approximation of the final result. The HP 3400 used a Bolometer type system. Other various schemes have used double balanced demodulators before rectification. But I fear I've just muddied the waters instead of providing clarity...
Doc
Maybe this will help: Measurements of AC magnitude : BASIC AC THEORY
Doc
Maybe this will help: Measurements of AC magnitude : BASIC AC THEORY
Last edited:
Meters have impedance, but not low enough to upset the AC voltage leaving a power transformer.
A fair test would be to connect all three meters to the transformer at the same time - in parallel. That way you can watch the readings simultaneously. And that will eliminate any error from changing mains voltages. and differential loading. They still may give wide ranging readings, but at least certain errors would be eliminated.
We used to have some cheap meters that worked well enough for what we used them for, but the AC function basically just put a rectifier in front of the DC input, and put a correction factor on the display reading. Worked OK for just reading the mains voltage or something. Unfortunately it also could read DC voltage that way, of course giving the wrong result. SO one could be distracted and not set the meter correctly and get a funny voltage reading. Also couldn;t measure ripple on a DC line.
A fair test would be to connect all three meters to the transformer at the same time - in parallel. That way you can watch the readings simultaneously. And that will eliminate any error from changing mains voltages. and differential loading. They still may give wide ranging readings, but at least certain errors would be eliminated.
We used to have some cheap meters that worked well enough for what we used them for, but the AC function basically just put a rectifier in front of the DC input, and put a correction factor on the display reading. Worked OK for just reading the mains voltage or something. Unfortunately it also could read DC voltage that way, of course giving the wrong result. SO one could be distracted and not set the meter correctly and get a funny voltage reading. Also couldn;t measure ripple on a DC line.
(I have posted somethig very similar in another thread but I think it is relevant here as well)
If one or two of your multimeters are of the very-cheap type they can give you very unreliable values at low AC voltages as described later in this post
True RMS DMM will show the true RMS voltage for most waveforms within specified frequency range. Remember this is often only up to a few kHz or less for not-so-expensive TRMS meters..
Some (most?) "Semi-cheap" multimeters often used the rectified average AC voltage (measured after a input cap to remove the DC component) using a OP-based precision rectifier. The average voltage is then multipled a with 1.11 to calculate the RMS voltage. 1.11 is the ratio between the RMS voltage and the rectified average for a sinus wave. This means that these DMMs are good only for clean sine-waves at one frequency. They will show wrong values for all other waveforms, except pure sine waves. You can easily test this by applying a squarewave with 1V amplitude at a low frequency 50-100HZ to the DMM. For a square wave the RMS voltage equals the average voltage, and hence the multimeter will show ~1.11V if it is of the averaging type. The measured voltage could be one or a few hundreds of a volt lower due to that the normally low bandwidth of these DMMs cuts out some of the harmonics of the square wave from the measured voltage.
I have also investigated several "dirt cheap" DMMs ( up to10-12€) of different types. If you measure a pure 9V DC voltage in AC-mode with these, in one polarity they will show 0V as expected, in the other direction the ones I have tested show 20V(!). I haven't checked the actual PCB, but to me this means two things. First the input DC-block capacitance is omitted. Second, the multimeter uses only one diode to make a half wave rectification, measures the average and multiplies with ~2.22. I do not think they even take care of the voltage drop of the diode if looking at the numbers carefully. For the main use of these multimeters it does not matter very much. (i.e., measuring if 12V or 6VDC is properly connected in cars, boats, bikes.... and measuring if 230V is connected at home). The first time I discovered this was when measuring the output from a voltage converrter with the multimeter by accident put in AC-mode. to my surprise I got almost the double voltage I had expected. since then I have seen it in several different types of cheap multimeters
However, I would be very reluctant to trust the cheapest multimeters for measuring any low voltage AC-signals, especially non-pure-sines or signals with a DC level. You will get what you pay for.
If one or two of your multimeters are of the very-cheap type they can give you very unreliable values at low AC voltages as described later in this post
True RMS DMM will show the true RMS voltage for most waveforms within specified frequency range. Remember this is often only up to a few kHz or less for not-so-expensive TRMS meters..
Some (most?) "Semi-cheap" multimeters often used the rectified average AC voltage (measured after a input cap to remove the DC component) using a OP-based precision rectifier. The average voltage is then multipled a with 1.11 to calculate the RMS voltage. 1.11 is the ratio between the RMS voltage and the rectified average for a sinus wave. This means that these DMMs are good only for clean sine-waves at one frequency. They will show wrong values for all other waveforms, except pure sine waves. You can easily test this by applying a squarewave with 1V amplitude at a low frequency 50-100HZ to the DMM. For a square wave the RMS voltage equals the average voltage, and hence the multimeter will show ~1.11V if it is of the averaging type. The measured voltage could be one or a few hundreds of a volt lower due to that the normally low bandwidth of these DMMs cuts out some of the harmonics of the square wave from the measured voltage.
I have also investigated several "dirt cheap" DMMs ( up to10-12€) of different types. If you measure a pure 9V DC voltage in AC-mode with these, in one polarity they will show 0V as expected, in the other direction the ones I have tested show 20V(!). I haven't checked the actual PCB, but to me this means two things. First the input DC-block capacitance is omitted. Second, the multimeter uses only one diode to make a half wave rectification, measures the average and multiplies with ~2.22. I do not think they even take care of the voltage drop of the diode if looking at the numbers carefully. For the main use of these multimeters it does not matter very much. (i.e., measuring if 12V or 6VDC is properly connected in cars, boats, bikes.... and measuring if 230V is connected at home). The first time I discovered this was when measuring the output from a voltage converrter with the multimeter by accident put in AC-mode. to my surprise I got almost the double voltage I had expected. since then I have seen it in several different types of cheap multimeters
However, I would be very reluctant to trust the cheapest multimeters for measuring any low voltage AC-signals, especially non-pure-sines or signals with a DC level. You will get what you pay for.
I think you are confusing what you suspect with what is true.
1. No meter I've ever worked on (literally thousands from crap to highest state of art) has ever used a feed through DC-blocking cap. No input caps.
2. Virtually all multi function DMMs in use today use the same basic desigh of a DC measuring circuit and all other functions feeding that DC measuring circuit, predominantly passively on cheaper meters. None of the DC A/D circuits multiply anything. They just read DC. Ohms uses a constant current source across leads and measures DC. High end meters may contain math functions, but they all work on output end of DC measurement.
3. You are correct that most cheap DVM's use passive, simple rectification, and quite often only half wave for AC readings. They all are frequency relative. The AC response is stated in the specs.
4. You can't make generalities based on just a few samples. "You will get what you pay for." doesn't really apply without qualifiers. I have seven digit meters that resolve tenths of a microvolt that I only paid $50 for. I own a handful of cheap $2.99 DMMs (and keep one in every car). I also own multiple Flukes, HP's and other known brands. For under $10 you can make one of those $2.99 DMMs read as acurately on AC as a 3.5 digit Fluke or HP (assuming DC is accurate). Just add an AD536,636,736 or LTC1099 in place of AC rectifier. There are probably several other chips that would work as well.
For 50-60Hz <500 VAC -sine- measurements, virtually any handhelp DMM ought to read fairly accurately. I'd toss or repair any that varied by a volt at twelve.
Doc
1. No meter I've ever worked on (literally thousands from crap to highest state of art) has ever used a feed through DC-blocking cap. No input caps.
2. Virtually all multi function DMMs in use today use the same basic desigh of a DC measuring circuit and all other functions feeding that DC measuring circuit, predominantly passively on cheaper meters. None of the DC A/D circuits multiply anything. They just read DC. Ohms uses a constant current source across leads and measures DC. High end meters may contain math functions, but they all work on output end of DC measurement.
3. You are correct that most cheap DVM's use passive, simple rectification, and quite often only half wave for AC readings. They all are frequency relative. The AC response is stated in the specs.
4. You can't make generalities based on just a few samples. "You will get what you pay for." doesn't really apply without qualifiers. I have seven digit meters that resolve tenths of a microvolt that I only paid $50 for. I own a handful of cheap $2.99 DMMs (and keep one in every car). I also own multiple Flukes, HP's and other known brands. For under $10 you can make one of those $2.99 DMMs read as acurately on AC as a 3.5 digit Fluke or HP (assuming DC is accurate). Just add an AD536,636,736 or LTC1099 in place of AC rectifier. There are probably several other chips that would work as well.
For 50-60Hz <500 VAC -sine- measurements, virtually any handhelp DMM ought to read fairly accurately. I'd toss or repair any that varied by a volt at twelve.
Doc
Last edited:
If you measure a DC voltage in AC mode most meters I have used show 0V. Hence DC blocking somwhere down the line. A cap would be the cheap solution but ther might be others
Even today there are still many non-TRMS meters sold (check specs carefully) and if you are not measuring True-RMS internally in the meter you have to measure using another method. The cheapest way is to measure some rectified voltage. The ratio between RMS and the average for a sinewave is 1.11. If the peak is mesured you have to divide by 1.414. Since some non-RMS meters show 1.11 on the suggested 1V amplitude sqare wave I tend to belive in the former. It might be done internally in a AC measuring chip but some conversion must be made
I agree that DC on the cheapest meters are accurate enough and that adding a AC measuring circuit is a good idea to get good reading on these. Depending on your demands att low frequency sines AC accuracy might be good enough if you are sure you have no DC.
When referring to prices I thought of new instruments. I guess you can get second hand ones at much better prices.
Even today there are still many non-TRMS meters sold (check specs carefully) and if you are not measuring True-RMS internally in the meter you have to measure using another method. The cheapest way is to measure some rectified voltage. The ratio between RMS and the average for a sinewave is 1.11. If the peak is mesured you have to divide by 1.414. Since some non-RMS meters show 1.11 on the suggested 1V amplitude sqare wave I tend to belive in the former. It might be done internally in a AC measuring chip but some conversion must be made
I agree that DC on the cheapest meters are accurate enough and that adding a AC measuring circuit is a good idea to get good reading on these. Depending on your demands att low frequency sines AC accuracy might be good enough if you are sure you have no DC.
When referring to prices I thought of new instruments. I guess you can get second hand ones at much better prices.
A meter will read DC on the AC funtion if it is configured to read AC+DC, most as you say are not.
Unless a meter specifically states it reads TRMS, it almost certainly doesn't
Setting meter on DC when measuring an AC signal will read net offset. Likewise Measuring AC on a DC signal will read ripple.
Here is a complete RMS adapter circuit that will cost about $10 to build:
RMS-TO-DC_Adapter_For_DVM
Unless a meter specifically states it reads TRMS, it almost certainly doesn't
Setting meter on DC when measuring an AC signal will read net offset. Likewise Measuring AC on a DC signal will read ripple.
Here is a complete RMS adapter circuit that will cost about $10 to build:
RMS-TO-DC_Adapter_For_DVM
I agree.
Still I am not sure the thread starter has got a good answer to his question about the large differences between different meter. The output of the tranformer should be pretty close to a clean sine wave if the transformer is properly built, with no or very small DC voltage. The output impedance is quite low so loading from the meter is neglible, any meter should show fairly correct values no matter what type of AC measurement they are using. Was any of the meters of the TRMS type? I guess at least one of them is not working properly
Still I am not sure the thread starter has got a good answer to his question about the large differences between different meter. The output of the tranformer should be pretty close to a clean sine wave if the transformer is properly built, with no or very small DC voltage. The output impedance is quite low so loading from the meter is neglible, any meter should show fairly correct values no matter what type of AC measurement they are using. Was any of the meters of the TRMS type? I guess at least one of them is not working properly
My "guess" would be the one that reads lowest only uses half wave rectification. The one that reads highest uses full wave rectification but is not scaled to read RMS. Most non true RMS AC meters are average reading with AC output scaled -by components- to read the same as RMS. This is fine but unpredictable on non pure sine signals.
The only way to resolve it is to test all three on a known good pure sine calibrator or directly compare them to a known good true RMS meter.
Doc
The only way to resolve it is to test all three on a known good pure sine calibrator or directly compare them to a known good true RMS meter.
Doc
Well I don't know if it is accomplished with a blocking cap or not, but my B&K 5360 handheld DMM (True RMS) has an "AC" mode, and an "AC+DC" mode. Feed 10 VDC to the AC mode, and it reads... 0 V. Feed 10 VDC to the "AC+DC" mode, and it reads 10 V. Likewise, superimpose a 1 VRMS AC onto that 10 VDC, and the AC mode will read, surprise, 1 V. DC blocked.
I also have another True RMS bench meter (Keithley 194A) which uses high speed sampling and mathematical calculation in order to come up with the RMS value. It has selectable AC or DC input coupling, accomplished with a DC blocking input cap, listed only as 0.1 uF in the service manual. Another of my True RMS bench meters, Keithley 199, always uses AC coupling in AC mode, also accomplished with a input cap (0.1 uF 630 V polyester according to the service manual). My Thurlby 1905a (with true RMS option) uses a DC blocker in its AC rectifier section, so technically that one is not an input cap. The two Keithley meters definitely use input caps.
Does this help, post #121 which kind of got on to something like this.
http://www.diyaudio.com/forums/full...-power-do-they-really-need-7.html#post2807772
http://www.diyaudio.com/forums/full...-power-do-they-really-need-7.html#post2808771
http://www.diyaudio.com/forums/full...-power-do-they-really-need-7.html#post2809209
http://www.diyaudio.com/forums/full...-power-do-they-really-need-7.html#post2807772
http://www.diyaudio.com/forums/full...-power-do-they-really-need-7.html#post2808771
http://www.diyaudio.com/forums/full...-power-do-they-really-need-7.html#post2809209
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.