EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: W6EL on April 25, 2023, 05:36:08 pm
-
I was surprised the other day when I tried measuring the RMS value of a rather standard TTL square wave on several "True RMS" meters.
The waveform was 0 to ~4.1 volts, 50% duty cycle. I ran the test at several frequencies (100Hz, 2 KHz, 200KHz) just to make sure. Every meter I had basically failed to even get close to the correct RMS value, which should be ~2.90 volts (Vpp * sqrt(D)). The only gear I have which measured it correctly was my HP 54645D scope (not really surprising).
I figured a TTL waveform has to be one of the more common waveforms people would measure using a handheld meter. After all, TTL appears in many low-speed digital circuits. I wouldn't expect to measure more complex waves accurately or higher-speed waveforms (especially with a handheld unit), but come on!
Of particular disappointment is the more modern Owon meters, which surely could take a few readings and do an actual calculation? One would think? Even the scope mode on the HDS272S (a great handheld scope) lacks an RMS readout, providing only Vpp, Vax, Vmin, and Vamp (which reads the same value as Vpp).
The HP 400EL is excused of course, since it is calibrated to read RMS only for a sine wave (like many analog meters of its time). Somewhat ironically it provided a closer measurement than most of the other equipment though.
I get that RMS requires a calculation. But come on. How much trouble is it to take a few consecutive samples at 5 bits resolution and do the calculation? Surely modern meters can do such a thing?
Does anyone know of a good meter that would pass the TTL test? Maybe one of the EEVBlog models? Agilent? Fluke?
Here are my results at 2 KHz. I have measured using both DC and AC since this is a fully-positive signal:
- HP 4645D scope: 3.035V RMS, 4.094Vpp
- HP-400EL: 2.22V (3V scale used, read 74% of full scale)
- Tenma 72-410A True RMS: 2.224 (AC), 2.289 (DC)
- Owon B35T TrueRMS meter: 2.076 (AC), 2.278 (DC)
- RadioShack TrueRMS meter (can't find the model): 2.064 (AC), 2.278 (DC)
- Owon HDS272S: 1.761 (AC), 2.281 (DC)
I'm quite disappointed. I had thought the so-called "True RMS" would be a bit closer than this. How do the expensive meters stack up against my hobby-lot?
-
You have to be very carefull when selecting a DMM that can measure odd waveforms and other frequencies than 50Hz. Most handheld meters are 50Hz only.
But you don't have to buy extremely expensive meters. Vici VC8145 is one that can measure RMS in the audio frequency range because it has a dedicated RMS converter chip inside.
-
You have to be very carefull when selecting a DMM that can measure odd waveforms and other frequencies than 50Hz. Most handheld meters are 50Hz only.
But you don't have to buy extremely expensive meters. Vici VC8145 is one that can measure RMS in the audio frequency range because it has a dedicated RMS converter chip inside.
Have you actually tested this with something other than a sine wave?
I tried down at 60 Hz as well, same results. I believe at least one of these meters I mentioned has one of those "RMS chips" inside.
-
The TTL signal is a mixed AC+DC signal.
For the RMS reading there are 2 ways to handle this:
1) have AC coupling and thus only show the AC part. For the 50% square wave this would be the same AC as DC reading (50% of peak voltage if the low voltage is at zero).
2) a combined DC+AC RMS values and thus the AC part + DC part as the geometric sum. In this case 1.41 * the DC reading.
A few meters offer both version and many meter offer only the AC coupled case. Usually the manul will tell.
So the Tenma, Owon B35T and radioshack meters don't look that bad.
Many handheld DMMs don't work to very high frequencies. So 2 kHz square wave can already be a bit on the fast side and thus a lower than expeced reading.
-
The TTL signal is a mixed AC+DC signal.
For the RMS reading there are 2 ways to handle this:
1) have AC coupling and thus only show the AC part. For the 50% square wave this would be the same AC as DC reading (50% of peak voltage if the low voltage is at zero).
2) a combined DC+AC RMS values and thus the AC part + DC part as the geometric sum. In this case 1.41 * the DC reading.
A few meters offer both version and many meter offer only the AC coupled case. Usually the manul will tell.
So the Tenma, Owon B35T and radioshack meters don't look that bad.
Many handheld DMMs don't work to very high frequencies. So 2 kHz square wave can already be a bit on the fast side and thus a lower than expeced reading.
It's not the frequency. I tried even 60 Hz, same results. I'm actually surprised how well all these meters handle higher frequencies. The Tenma bench meter works well above 200 KHz, and the HP 400 really does work fine at 10 MHz! The others are pretty flat in the usual audio range at least, tapering off some as you get higher.
While I can understand RMS being different for AC and DC coupled measurements, the RMS calculation should be conceptually a differential measurement from Vmin to Vmax at duty cycle D.
What would an HP-3400 (a meter that converts the energy to thermal energy) read for a square wave about zero versus one with DC bias?
-
square wave has too many harmonics, so RMS measurement will depends on bandwidth. Since different DMM have different bandwidth, they show different results.
-
square wave has too many harmonics, so RMS measurement will depends on bandwidth. Since different DMM have different bandwidth, they show different results.
Indeed, but a meter that is shown to be flat with a sine wave up to and beyond 20 KHz should be able to handle many harmonics of a 60 Hz square wave with full fidelity, preserving the RMS level to at least three significant figures. Any meter claiming "True RMS" ought to at least be able to handle 60 Hz, right? The tenth harmonic of 60 Hz is only 600 Hz after all.
-
With analog RMS it is not just the harmonics, but also an effect of the slew rate. Usually there is some amplification / buffer with a limited slew rate. A 2nd point may be the precision rectifier, that can also have problems with a very high slew rate and this way loose some of the amplitude. The RMS part is by design nonlinear and thus does not handle the harmonics separate.
-
With analog RMS it is not just the harmonics, but also an effect of the slew rate. Usually there is some amplification / buffer with a limited slew rate. A 2nd point may be the precision rectifier, that can also have problems with a very high slew rate and this way loose some of the amplitude. The RMS part is by design nonlinear and thus does not handle the harmonics separate.
Alright but what would be the use in an RMS measurement that only works on pure sine waves? I mean, any meter that isn't RMS can be calibrated to read the RMS of a sine wave correctly. The entire reason to have RMS circuits is, indeed, to be able to integrate and calculate over some period of measurement. The slew rate for a 60 Hz square wave should not be an issue given how long the period of a 60 Hz signal is. Any op-amp from the 1970s could easily slew up and down for 60 Hz in under 1% of a period.
-
The EEVBlog BM235 would no doubt disappoint you. Dave points out that with a limited bandwidth, a few harmonics of 60Hz is not typical with low cost.
Quote from manual:
True RMS
RMS (Root-Mean-Square) is a term used to describe the effective or equivalent DC value of an AC signal. True RMS is the term which identifies a DMM that responds accurately to the effective RMS value regardless of the waveforms such as: square, sawtooth, triangle, pulse trains, spikes, as well as distorted waveforms with the presence of harmonics. Harmonics may cause :
1) Overheated transformers, generators and motors to burn out faster than normal
2) Circuit breakers to trip prematurely
3) Fuses to blow
4) Neutrals to overheat due to the triplen harmonics present on the neutral
5) Bus bars and electrical panels to vibrate
Dave's Note: Whilst this is a True RMS multimeter (that's good), like most lower end multimeters it does not use a separate True RMS converter chip. It relies upon the internal multimeter chipset capability. This gives the meter a low True RMS frequency response of only a few hundred Hz (check the specs), and is typical of other meters in this class. Don't be fooled thinking that “True RMS” automatically means “high frequency range measurement”.
-
You have to be very carefull when selecting a DMM that can measure odd waveforms and other frequencies than 50Hz. Most handheld meters are 50Hz only.
But you don't have to buy extremely expensive meters. Vici VC8145 is one that can measure RMS in the audio frequency range because it has a dedicated RMS converter chip inside.
Have you actually tested this with something other than a sine wave?
Yes. I can set my generator to RMS and the VC8145 tracks the level nicely. But it measures only the AC part of the signal.
-
I figured a TTL waveform has to be one of the more common waveforms people would measure using a handheld meter. After all, TTL appears in many low-speed digital circuits.
It's a common signal but nobody measures TRMS of it. :-DD
Stick to your oscilloscope for looking at TTL signals - shape is more important.
-
While I can understand RMS being different for AC and DC coupled measurements, the RMS calculation should be conceptually a differential measurement from Vmin to Vmax at duty cycle D.
What would an HP-3400 (a meter that converts the energy to thermal energy) read for a square wave about zero versus one with DC bias?
I'm not sure what you are trying to get at in the first statement. The RMS value of a 0 to 4.1V 50% duty cycle square wave would be 2.05V AC and 2.90V AC+DC, or DC coupled, RMS. In addition, the integrated DC value should also be 2.05VDC, meters not reading close to that on a 2kHz signal as you've described probably have terrrible normal mode rejection. If you have a meter that does AC and DC, but not AC+DC (or DC-coupled RMS if you prefer) then you can measure DC and AC separately, square the results and add them and the square root of that sum is your AC+DC result.
A thermal transfer measuring instrument would measure the latter if it had an DC-coupled input, but the HP 3400A has an AC-coupled input with a 10Hz lower cutoff (or spec anyway, the cutoff may be lower) so I'd presume it would read the former. I don't have one here at the moment to confirm that. I do have a different thermal transfer meter, but there's no room on the bench for it at the moment. I don't think it is necessary to go to that extreme for such a simple example, but it is probably the only meter I have that is going to read the TRMS of a 200kHz square wave with any accuracy.
Every TRMS meter I have will perform as I've stated and the only other limitation would be the bandwidth, obviously there will not be very many meters that will read a 200kHz square wave accurately.
How do the expensive meters stack up against my hobby-lot?
Frankly yours are looking like rubbish! :)
Seriously, they're just wrong. Or something is wrong in any case. If your signal is actually 0.00 to 4.10V and there isn't a loading issue, then only the HP 400EL is actually showing what it ought to with any reasonable accuracy. Every other instrument is off by enough to be considered a fail (IMO) except maybe the scope which would need further discussion to conclude anything.
-
The TTL signal is a mixed AC+DC signal.
For the RMS reading there are 2 ways to handle this:
1) have AC coupling and thus only show the AC part. For the 50% square wave this would be the same AC as DC reading (50% of peak voltage if the low voltage is at zero).
2) a combined DC+AC RMS values and thus the AC part + DC part as the geometric sum. In this case 1.41 * the DC reading.
A few meters offer both version and many meter offer only the AC coupled case. Usually the manul will tell.
So the Tenma, Owon B35T and radioshack meters don't look that bad.
Many handheld DMMs don't work to very high frequencies. So 2 kHz square wave can already be a bit on the fast side and thus a lower than expeced reading.
I read and re-read what you said. You were exactly correct. The DMM in AC mode is ac-coupled, and thus the result is given for the AC-coupled version of the signal.
Placing the scope's input in AC-coupled mode shows 2.045V RMS, which means that my list was sorted in reverse! The Owon B35T and radioshack meters were almost dead-on, and all others within about 10%.
This makes a lot of sense. So to sum it up:
1. There was not a problem with bandwidth or high frequency cutoff
2. There was no issue with slewrate
The bottom line is that the DMM places the input in AC coupled mode, and thus, any result must be considered from that point of view. There are still some discrepancies, but they are much less severe once you consider the point of view of the measurement.
-
Interesting topic... It got me a little puzzled.
Sorry for my ignorance, I fail to see where the theoretical 2.9V RMS or the 3.035V RMS measured on the DSO come from.
Is this about a "0" to 4.1V amplitude 50% duty cycle square wave?
How far off zero was the real low level?
-
What is the source of your signal? One possible explanation for the discrepancies that I see (discrepancies that seem fairly consistent, so perhaps I've maligned your meters unnecessarily) is that you have about a 135mVDC offset on the signal, so it is 0.135 to 4.235V.
-
Sorry for my ignorance, I fail to see where the theoretical 2.9V RMS or the 3.035V RMS measured on the DSO come from.
50% duty cycle, so half the time it is zero, the other half 4.10V. So (4.1)2 is 16.81 (the squares), half of that is 8.405 (the mean), the square root of that is 2.90V (the root).
How far off zero was the real low level?
I think that's the issue with the readings being off from what I'd expect, but I don't think it was the OPs primary question. As I wrote, it looks like a ~135mVDC offset would account for all of the errors except the AC readings of the Tenma and the OWON scope-thingy.
-
Interesting topic... It got me a little puzzled.
Sorry for my ignorance, I fail to see where the theoretical 2.9V RMS or the 3.035V RMS measured on the DSO come from.
Is this about a "0" to 4.1V amplitude 50% duty cycle square wave?
How far off zero was the real low level?
The source was single-ended and grounded at the scope's input, and zero was definitely zero. It's a test generator with a TTL output (among others).
And yes, it's 0 to about 4.1 volts.
The real RMS value is Vmax * sqrt(D) for a square wave, where D is the duty cycle, so basically 4.1 * sqrt(1/2) = 4.1*0.707 = 2.90 volts.
However, the DMM has a capacitor in series when it does AC readings. So the DMM is showing the RMS voltage for an AC coupled version of the input signal. This means it is measuring a square wave with a total amplitude peak to peak of 4.1 volts, centered at zero volts, with vmin = -4.1/2 = -2.05 and vmax = 4.1/2 = +2.05 volts. It's easy to visualize how the RMS is calculated from this data, and it's no wonder that the AC-coupled meters all reported values around 2 volts.
Placing the scope in ac-coupled mode put it at the same vantage point, and read about 2 volts as well.
The lesson here is to do tests just like this and learn more about your test equipment. Who would have thought a DMM from RadioShack purchased in 2002 would be that good, even at 200 KHz?
-
Welcome to the forum. Assuming that is your ham call sign, your photo of your lab looks nice. I had one of those 141Ts many years ago with the 1.3ish GHz plug in, and a couple others. I had the tracking generator for it as well.
Anyway, I setup my arb with a 50ohm load. Peak was measured with my HP34401A at 4.9155 or roughly 3.476 VRMS.
I measured seven different meters at 100Hz, 2kHz and 200kHz. Attached showing the measured values and their error relative to the 34401A.
The CEM is the lowest cost out of the group at $120 on sale. The Gossen Ultra was by far the most expensive, now rebranded as Prime after my review of it. The BM789 is an early pre-release and I have done a some rework to bring it up to the latest revision (using factory parts). The BM689s shown is the first one I purchased several years back. It was damaged during my testing and I did repair it. None of these meter have been realigned.
Once the UT181A with it's odd ball rechargeable battery gets a charge, I will measure it. Bad design but still one of my favorite products from UNIT.
***
The UT181A was allowed to charge to 40% and I retook the first measurement plus the others.
Because you seem to like old hardware, I tried an old Fluke 97 scope meter but like the Gossen, it could not read the value at 200kHz in DMM mode.
I saved an old Fluke 8506A Thermal RMS meter from the recycle bin that needed repairs. I aligned the DC stages using my HP34401A as a reference. The AC stages are still factory set as I don't have anything near this accurate and thought I would do more harm than good. For fun, I show the signal at 2MHz compared with the UT181A.
I'm not sure what accuracy you need but my personal pick of these meters is still the BM869s. I would take the Fluke 189 if they still offered them new. The 789 is a nice meter as well. Has a few things on the BM869s but I like the multi displays.
Also note the previous percentage was off 100X and has been corrected.
-
btw .. speaking of Tenma 72-410A and Owon B35T , such devices doesn't have dedicated RMS\DC converter like popular AD637.
so true RMS in some narrow brackets of frequency and AC waveform.
I'm guessin HDS272S , and RadioShack same story.
most basic dmm AC up 5kHz , if something add it would be 20K , maybe 50K, who claim 1% accuracy up to 100K usually dedicated RMC-DC converter.
like owon bt41+ has such dedicated chip
-
Welcome to the forum. Assuming that is your ham call sign, your photo of your lab looks nice. I had one of those 141Ts many years ago with the 1.3ish GHz plug in, and a couple others. I had the tracking generator for it as well.
Anyway, I setup my arb with a 50ohm load. Peak was measured with my HP34401A at 4.9155 or roughly 3.476 VRMS.
I measured seven different meters at 100Hz, 2kHz and 200kHz. Attached showing the measured values and their error relative to the 34401A.
The CEM is the lowest cost out of the group at $120 on sale. The Gossen Ultra was by far the most expensive, now rebranded as Prime after my review of it. The BM789 is an early pre-release and I have done a some rework to bring it up to the latest revision (using factory parts). The BM689s shown is the first one I purchased several years back. It was damaged during my testing and I did repair it. None of these meter have been realigned.
Once the UT181A with it's odd ball rechargeable battery gets a charge, I will measure it. Bad design but still one of my favorite products from UNIT.
***
The UT181A was allowed to charge to 40% and I retook the first measurement plus the others.
Because you seem to like old hardware, I tried an old Fluke 97 scope meter but like the Gossen, it could not read the value at 200kHz in DMM mode.
I saved an old Fluke 8506A Thermal RMS meter from the recycle bin that needed repairs. I aligned the DC stages using my HP34401A as a reference. The AC stages are still factory set as I don't have anything near this accurate and thought I would do more harm than good. For fun, I show the signal at 2MHz compared with the UT181A.
I'm not sure what accuracy you need but my personal pick of these meters is still the BM869s. I would take the Fluke 189 if they still offered them new. The 789 is a nice meter as well. Has a few things on the BM869s but I like the multi displays.
Also note the previous percentage was off 100X and has been corrected.
This is great, what fun!
Yes, there is something attractive about the older stuff, I don't even know what it is. At work we have scopes that cost more than my house, but I just don't find it as fun.
Was your square wave with or without DC bias? I am seeing that without DC bias my meters agree much better.
--E
-
I was surprised the other day when I tried measuring the RMS value of a rather standard TTL square wave on several "True RMS" meters.
The waveform was 0 to ~4.1 volts, 50% duty cycle. I ran the test at several frequencies (100Hz, 2 KHz, 200KHz) just to make sure. Every meter I had basically failed to even get close to the correct RMS value, which should be ~2.90 volts (Vpp * sqrt(D)). The only gear I have which measured it correctly was my HP 54645D scope (not really surprising).
I figured a TTL waveform has to be one of the more common waveforms people would measure using a handheld meter. After all, TTL appears in many low-speed digital circuits. I wouldn't expect to measure more complex waves accurately or higher-speed waveforms (especially with a handheld unit), but come on!
Of particular disappointment is the more modern Owon meters, which surely could take a few readings and do an actual calculation? One would think? Even the scope mode on the HDS272S (a great handheld scope) lacks an RMS readout, providing only Vpp, Vax, Vmin, and Vamp (which reads the same value as Vpp).
The HP 400EL is excused of course, since it is calibrated to read RMS only for a sine wave (like many analog meters of its time). Somewhat ironically it provided a closer measurement than most of the other equipment though.
I get that RMS requires a calculation. But come on. How much trouble is it to take a few consecutive samples at 5 bits resolution and do the calculation? Surely modern meters can do such a thing?
Does anyone know of a good meter that would pass the TTL test? Maybe one of the EEVBlog models? Agilent? Fluke?
Here are my results at 2 KHz. I have measured using both DC and AC since this is a fully-positive signal:
- HP 4645D scope: 3.035V RMS, 4.094Vpp
- HP-400EL: 2.22V (3V scale used, read 74% of full scale)
- Tenma 72-410A True RMS: 2.224 (AC), 2.289 (DC)
- Owon B35T TrueRMS meter: 2.076 (AC), 2.278 (DC)
- RadioShack TrueRMS meter (can't find the model): 2.064 (AC), 2.278 (DC)
- Owon HDS272S: 1.761 (AC), 2.281 (DC)
I'm quite disappointed. I had thought the so-called "True RMS" would be a bit closer than this. How do the expensive meters stack up against my hobby-lot?
Brymen BM859S (europa version)
Input
2kHz 4.10 / 0.00V square wave (no need tell 50% because if it is not 50% then it is not square wave).
Brymen display RMS 2.8932 And RMS is RMS and it naturally include also DC. 4.1V DC RMS is 4.1V. If some meter do not display 4.1V RMS for 4.1V DC then designer need doctor. Or some extra lesson for math.
-
And RMS is RMS and it naturally include also DC. 4.1V DC RMS is 4.1V. If some meter do not display 4.1V RMS for 4.1V DC then designer need doctor. Or some extra lesson for math.
Or maybe they had one lesson more than you and learnt that calculating the RMS of an AC coupled signal is still RMS. Would you say that if you enable a bandwidth limit on your scope, it's no longer measuring RMS? RMS just means the root of the mean of squared values. It can be for AC or AC+DC. Some meters are marked like that and can measure both. Other meters can only measure the AC part and you have to do the math yourself to add the DC part. And yet other meters can only measure AC+DC, and you need to do math to subtract DC to get the AC value. As long as meters are clearly marked as AC or AC+DC I don't see a problem.
-
...
Anyway, I setup my arb with a 50ohm load. Peak was measured with my HP34401A at 4.9155 or roughly 3.476 VRMS.
...
...
Was your square wave with or without DC bias? I am seeing that without DC bias my meters agree much better.
...
In the lower left of the spreadsheet is the reference. It was a squareish sort of waveform with a 50% dutycycle, 0 volt minimum and 4.9155 volt peak. The calculated RMS is last. The arb was loaded to 50 ohms and the peak level was measured with my old HP.
That old thermal RMS meter was set to normal mode when making these measurements. Nothing was allowed to warmup. The Arb isn't clean. Simple test wasn't meant to as a dive down the metrology hole.
-
Brymen BM859S (europa version)
Input 2kHz 4.10 / 0.00V square wave (no need tell 50% because if it is not 50% then it is not square wave).
Brymen display RMS 2.8932 And RMS is RMS and it naturally include also DC. 4.1V DC RMS is 4.1V. If some meter do not display 4.1V RMS for 4.1V DC then designer need doctor. Or some extra lesson for math.
You seem to get a little hung up on semantics and while you have one interpretation of the terms 'square wave' and AC vs DC RMS, others may think differently. You can argue that they are wrong or you can take care to clarify exactly what is meant. I prefer the latter approach, the former somehow seems small minded.
So, since your 859s is 'correctly' designed and since you state that any meter that doesn't display what I would call TRMS AC+DC is demented, which of these ranges did you select to get that reading?
(https://www.eevblog.com/forum/testgear/which-dmms-can-reasonably-measure-rms/?action=dlattach;attach=1769246;image)
-
New users of the vintage Fluke may have to read the manual. :-DD
-
It you like glitter, the UNI-T may be a good choice. Be VERY careful with the front end as it has been my experience that products from this company are not robust. One small ESD may end up trashing the meter.
Anymore, if I pull it out, the battery is dead (it only gets used for these demonstrations). Here it is, 10 hours on the charger and it's ready to use. Of course, you can't use the meter while it is charging. Oh, and a bit of advice about charging. What ever you do, DO NOT CONNECT THE CHARGER UNTIL YOU HAVE THE METER'S SELECTOR SET TO THE CHARGE MODE!!! The charger uses the low current safety fuse and if you set the meter to current, you will blow that fuse with the charger attached!! Ask me how I learned that right after I bought the meter.. :-DD
Still, it is very stable over temperature and accurate. It also has a few improvements over the Fluke 289 (its a copy if you didn't notice). Drives just like the 289. You can buy a BLE interface for remote data logging but sadly, there are no working applications for it. I started to hack the protocol and ran into a few others who had already been down that path who helped me out. There is now a document that covers it.
Cost is about $400 now. If the meter had a few improvements, it may be worth it but I would rather use the Brymen BM869s. It's just a nice basic meter and one of the more robust meters I have looked at.
-
Well, this is all very interesting.
I think we can decouple the bandwidth argument pretty easily. Of course a square wave has an infinite number of terms, so there is no way to measure it (or even create it) perfectly. But we get to 20 dB down within about 4 or 5 terms, so for low frequency square waves, your garden-variety DMM should be acceptable bandwidth-wise, at least down to 1% (power-bandwidth wise, not accuracy). Many meters have surprisingly wide bandwidth. I saw very little difference with my listed equipment across the spectrum of my testing. This is not to say it is not a factor, rather, it is not a major limiting factor in this particular test.
There was in fact a little DC offset on the low-side of the output of my generator. Just about 150mV above zero. Thank you to those that pointed this likely issue out.
Since this test has become more interesting than I anticipated, I'll run it again but with more attention to detail. I can now appreciate the AC vs DC coupling issue in how the DUT sees the incoming voltage. With that in mind, a purely AC measurement produces a different result than a DC measurement, but this is acceptable so long as the operator is aware of it.
It's also worth mentioning that any RMS measurement has a large dependence on the integration time of the reading. For high frequency signals, short integration times are probably just fine, but for lower frequency waveforms, the integration time needs to be sufficient so as to capture the full period of whatever nonsense is going on in the signal. For those using thermal conversion measurements, this is essentially the heat capacity and the thermal conductance of the module to the environment which sets the time constant.
-
The cheaper meters may not have a hard integration time for the AC mode. With the analog RMS converters there is more like a 1st or 2nd order response and thus possibly relatively slow settling. This is especially true for diconnecting from an AC source. So RMS converters (e.g. AD737 /736) can be awfully slow in this case.
There is a compromise between response time and the ability to measure low frequencies / accuracy at low frequencies (like < 30 Hz)
The digital RMS type (e.g. part of the DMM chip set) has usually much faster response, more similar to the DC mode. They may still be more accurate with low frequencies.
-
I don't know if it's right or wrong but with my Fluke 189 and a square wave 0-4.1V it reads a bit more than 2V (2.09V) on AC. It reads 2.9V on AC+DC mode.
The meter manual said"
" Your meter features
true rms readings, which are accurate for sinewaves and
other wave forms (with no dc offset) such as square
waves, triangle waves, and staircase waves. For ac with
dc offset, use ac+dc."
So a square wave with the minimum at 0 and max at 4.1 it's considered to have 2.05 VDC offset.
-
I don't know if it's right or wrong but with my Fluke 189 and a square wave 0-4.1V it reads a bit more than 2V (2.09V) on AC. It reads 2.9V on AC+DC mode.
That seems right to me from what I've learned so far.
-
Alright, well, I did some more careful testing.
I decided to leave the scope in AC coupled mode just to put it into the same point of view of my cheap multimeters. I allowed everything to warm up over an hour and then did some careful measurements.
In order to sort of establish how well things were accurate, relative to each other, I ran a 60 Hz sine wave first. That ought to be something any reasonable meter can measure, right? And indeed, all instruments were under 2% of error, which is pretty good for my bunch. I'm using the HP scope as my reference since it actually arrived calibrated (it's since lapsed, but not that long ago).
I then did a basic sweep to see where each meter fell off 3dB (power). Most of my meters provided plenty of bandwidth for audio-range signals, but the Owon HDS272S surprised me at only 4 KHz. After about 5 KHz it's basically deaf. The Tenma 72-410A went out all the way to 236 KHz before losing 3dB, however, it was not very flat along the way, showing little bumps of a 100mV here and there. The Owon B35T went out to 600 KHz before dropping 3dB, which surprised me. This little meter seems to hold its own pretty well! Even the RadioShack (22-174B) held out to a respectable 60 KHz.
Next came the square wave test. I did this at 60 Hz and at 1 KHz. Given the bandwidth of some of these meters, there was no point going out much further. The HP400EL didn't do so well on the square wave test, despite its massive bandwidth advantage, simply because it is not an RMS meter. It is calibrated to show the RMS level of a sine wave, but it lacks a circuit to actually compute or measure RMS. So despite it's massive bandwidth (well over 10 MHz), it is simply not accurate for a square wave.
As for the other meters, they did pretty well except for the Owon HDS272S, which probably did poor simply due to it's rather pathetic 4 KHz bandwidth. This is somewhat ironic given the scope portion of this meter can read out to 70 MHz... Add to this that the scope lacks an RMS measurement readout. Go figure. It's still a nice meter though, lovely bright display and very handy in tight spots to have a scope.
The standard deviation on the sine wave test was 14.6mV, the 60 Hz square wave was 106mV, and the 1 KHz square wave was 129mV. So you can see that indeed there is far more variation as you move up in frequency and some of these meters start to fall apart. I wouldn't read too far into this though since my meters are not exactly high-quality nor are they even similar to each other.
A little story on the radioshack meter. I got it when I worked at radioshack and it went on clearance for $29.97. With my employee discount it was probably half that price. I thought it was pretty great, it had HFE, capacitance, a dB scale, a bargraph, all the sorts of things a kid in the 90s would consider pretty premium stuff. Years later I was working as a tech for a guy and I noticed that the meter had poor performance over about 10 KHz. The guy I was working for had a basement full of old vacuum tube HP equipment, so I dragged up an enormous HP meter and compared the two, and indeed, the RS meter was off a few dB at the high end. Now comes the fun part. I had a warranty on the meter because, why not, it was dirt cheap as an employee, so I drove to the local radio shack store. I walked in and proceeded to plug in an HP oscillator and an HP meter, and planned to demonstrate how poor the performance was. The poor guy behind the counter had no idea what was going on and just offered to let me pick a similar model to replace it (since they didn't have it anymore). I picked the most expensive one they had. But! When I tested it, it was worse!! So I left with my original meter. That evening, I took the meter apart and discovered several pots inside it. After a few hours of playing around in the basement of HP equipment, I had it flat to 20 KHz (which was all I needed at the time). What surprises me about all this, is that "calibration" happened in around 2003. The meter's plastic melted in my car and several sets of batteries corroded up inside it. And yet, here it is, 20 years later, showing under 2% of error. It still has holes drilled in the back over the pots with little bits of tape over the holes.
Anyway, that's my story! Results are attached.
-
btw .. speaking of Tenma 72-410A and Owon B35T , such devices doesn't have dedicated RMS\DC converter like popular AD637.
so true RMS in some narrow brackets of frequency and AC waveform.
I'm guessin HDS272S , and RadioShack same story.
most basic dmm AC up 5kHz , if something add it would be 20K , maybe 50K, who claim 1% accuracy up to 100K usually dedicated RMC-DC converter.
like owon bt41+ has such dedicated chip
Hi GigaJoe,
I found the schematic for the Tenma 72-410A. It indeed does have an RMS converter, it uses the Maxim MX536A, a 2 MHz-wide RMS converter. This converter's value then goes into a 14-bit ADC (TC835) running at 1 MHz. I think, given that it was only off by 2-3%, it just needs a calibration.
Also the Owon B35T has an AD8439JCPZ RMS chip, see here: https://lygte-info.dk/review/DMMOwon%20B35T%20UK.html (https://lygte-info.dk/review/DMMOwon%20B35T%20UK.html)
--E
-
I have used various RMS voltmeters in combination with a square wave generator for oscilloscope calibration. From 100 to 1000 Hz, my Tektronix DMM916, Beckman RMS225, and HP3478A all read correctly, whether in AC averaging mode, AC RMS mode, or AC+DC RMS mode. The average AC reading should be 11.1% higher than the RMS measurement, and is a little more convenient because it settles faster so adjustment is easier.
-
Pretty interesting story about the Radio Shack meter, W6EL.
Radio Shack’s test equipment was average or below average, but there was the ocasional golden nugget.
Before affordable DMMs, the silver-standard were FET-VOMs, and I worked for several months at a McDonalds to be able to purchase one.
I loved the meter, its average responding AC function frequency response would extend all the way to 1 Mhz.
But then after a few years, the plastic case started to disintegrate. I ignore what the cause was, but the plastic became very brittle and little pieces would fall off. I mended it with tape, until the center post where the rotary switch actually rotated cracked.
-
In order to sort of establish how well things were accurate, relative to each other, I ran a 60 Hz sine wave first. That ought to be something any reasonable meter can measure, right? And indeed, all instruments were under 2% of error, which is pretty good for my bunch. I'm using the HP scope as my reference since it actually arrived calibrated (it's since lapsed, but not that long ago).
A couple of things that jumped out at me and I'm curious about them.
First, does the HP 54645D actually give you an explicit VRMS reading or are you calculating it somehow? If it gives you a reading, what is its specified accuracy? Scopes are typically not competitive with good DMMs in basic accuracy and I think the basic DC-gain tolerance is 1.5% or so for this model.
Second, you give readings for the HP 400EL to four significant digits which seems impossible on an analog meter--that last digit corresponds roughly with the thickness of the paint on the needle! Are you just squinting extra hard or do you have some other method?
I suspect your hand-tuned Radio Shack meter may be even more accurate than you think.
-
UniT 181A gives 2.8995 for 100Hz, 2.89 for 1KHz and 2.895 for 200KHz.
4.1Vpp square wave with 2.05V offset...
-
In order to sort of establish how well things were accurate, relative to each other, I ran a 60 Hz sine wave first. That ought to be something any reasonable meter can measure, right? And indeed, all instruments were under 2% of error, which is pretty good for my bunch. I'm using the HP scope as my reference since it actually arrived calibrated (it's since lapsed, but not that long ago).
A couple of things that jumped out at me and I'm curious about them.
First, does the HP 54645D actually give you an explicit VRMS reading or are you calculating it somehow? If it gives you a reading, what is its specified accuracy? Scopes are typically not competitive with good DMMs in basic accuracy and I think the basic DC-gain tolerance is 1.5% or so for this model.
Second, you give readings for the HP 400EL to four significant digits which seems impossible on an analog meter--that last digit corresponds roughly with the thickness of the paint on the needle! Are you just squinting extra hard or do you have some other method?
I suspect your hand-tuned Radio Shack meter may be even more accurate than you think.
The scope has an RMS "measurement" you can bring up (among other things). I turned on 8x averaging as well. Yes, scopes are not exactly known for their vertical accuracy, but it's the best I can do. That it agrees (sine wave) with the HP-400EL does say something.
The HP 400EL has more sig figs than you might expect, provided you are in the correct range. I used the 3 volt scale. Each of the most minor tic marks near the 2 volt mark is 0.020 volts, and I can discern values in between to a degree. Years of working in laboratories has refined this skill, although with most equipment being digital I don't get to use it as much as I wish. I'll grant you that maybe the 0.005 is generous though.
I looked inside the radioshack meter this evening. Nothing fancy really. A single quad-pack chip seems to handle most of it. I did not see a dedicated RMS chip, although I did not look under the LCD for fear of breaking it. I do note that the six batteries are tapped such that a negative rail is available inside. The PWB reads "Tandy 1992" :)
Also, the Owon B35T does indeed have an RMS chip (AD8439JCPZ, see here for the insides: https://lygte-info.dk/review/DMMOwon%20B35T%20UK.html (https://lygte-info.dk/review/DMMOwon%20B35T%20UK.html)).
-
UniT 181A gives 2.8995 for 100Hz, 2.89 for 1KHz and 2.895 for 200KHz.
4.1Vpp square wave with 2.05V offset...
I like the readout showing AC, DC, and the combined result. Fantastic. Tells a lot more of the story!
-
The scope has an RMS "measurement" you can bring up (among other things). I turned on 8x averaging as well. Yes, scopes are not exactly known for their vertical accuracy, but it's the best I can do.
Not having the vertical scale set as near as possible to ADC's full scale would add inaccuracy, I'm not sure what the optimal horizontal scaling would be for RMS measurement on a DSO although I'd suspect one full cycle and that it may vary from one scope to another.
-
Those multimeters that have AC+DC mode will do it.
-
And RMS is RMS and it naturally include also DC. 4.1V DC RMS is 4.1V. If some meter do not display 4.1V RMS for 4.1V DC then designer need doctor. Or some extra lesson for math.
Or maybe they had one lesson more than you and learnt that calculating the RMS of an AC coupled signal is still RMS. Would you say that if you enable a bandwidth limit on your scope, it's no longer measuring RMS? RMS just means the root of the mean of squared values. It can be for AC or AC+DC. Some meters are marked like that and can measure both. Other meters can only measure the AC part and you have to do the math yourself to add the DC part. And yet other meters can only measure AC+DC, and you need to do math to subtract DC to get the AC value. As long as meters are clearly marked as AC or AC+DC I don't see a problem.
AC coupling balances the signal either side of zero volts, so to all intents & purposes, an ac coupled 0-4.1v square wave becomes
a continuous voltage of 2.05 volts, as power doesn't care about polarity.
A "true RMS" meter should display it as 2.05 volts RMS.
A "non-true RMS" meter will display 0.7071 of peak, yielding 2.899 volts if DC coupled, or 1.496 volts if ac coupled------- if, of course, the correction factor does work for non-sinusoidal signals.
-
The scope has an RMS "measurement" you can bring up (among other things). I turned on 8x averaging as well. Yes, scopes are not exactly known for their vertical accuracy, but it's the best I can do.
Not having the vertical scale set as near as possible to ADC's full scale would add inaccuracy, I'm not sure what the optimal horizontal scaling would be for RMS measurement on a DSO although I'd suspect one full cycle and that it may vary from one scope to another.
Which data are used for the RMS calculation depends and cause some trouble. Some use all the data in the memory and some only the data on the screen. Multiple full periods may be less sensitive to a rest of a partial period.
Using averaging mode will remove some of the noise, but the noise in the signal is also part of the RMS value. So averaging (over multiple triggered sections) mode is more like a bad idea.
-
Many scopes (Tektronix and Lecroy scopes I have used, but also Siglent I believe) have a cycle RMS/cRMS option that just measures the RMS of a single cycle. This is less sensitive to partial cycles and horizontal scaling. For this you should adjust the horizontal scale for a bit more than one period of the signal.
-
The scope has an RMS "measurement" you can bring up (among other things). I turned on 8x averaging as well. Yes, scopes are not exactly known for their vertical accuracy, but it's the best I can do.
Not having the vertical scale set as near as possible to ADC's full scale would add inaccuracy, I'm not sure what the optimal horizontal scaling would be for RMS measurement on a DSO although I'd suspect one full cycle and that it may vary from one scope to another.
Which data are used for the RMS calculation depends and cause some trouble. Some use all the data in the memory and some only the data on the screen. Multiple full periods may be less sensitive to a rest of a partial period.
Using averaging mode will remove some of the noise, but the noise in the signal is also part of the RMS value. So averaging (over multiple triggered sections) mode is more like a bad idea.
Many scopes calculate RMS on whole buffer or screen AND have Cycle RMS and Cycle Stdev (AC RMS), that will calculate for either one single full period OR each period on screen.
Problem with accuracy on RMS calculation on full screen is what gets into calculation and it is more complicated with nonperiodic signals.. If you set timebase so it chops off piece of period RMS is going to be wrong.. If you have signal with large crest factor (large pause and then short spike) you have to capture exactly from same point at beginning and end of one period so you have exactly one period.
This is where Cycle RMS is useful, or you use gating and select which part of the curve you are measuring. Also, if scope can do it, use measurement cursors connected to measurement so scope will show you what exactly it is calculating on...
Or you simply go with deep memory, and set timebase slower to put hundreds of periods on screen. That will push down errors caused by chopping the signal off at beginning and end of the screen. It's good when you pretty much can't see much of the signal and it looks like a continuous block... For nonperiodic signal (like switcher going into discontinuous mode) that is only way to get any real estimate...
As for statistics, some scopes calculate exactly one measurement per screen (usually leftmost on the screen) and calculate stats between triggers (repetitive captures) and some are capable to capture long capture and if you do Cycle RMS, it will calculate it for every cycle it detects, 10s of thousands at the time from single capture....
Both of those will give you good estimate in slightly different way...
Averaging the measurements is good idea, because it does average of measurements of full signal not the measurement of average of signal...
16bit and 12bit scopes will give you excellent RMS results, because they have wider BW than any multimeter and they show you what are you measuring... So you get data and understanding of signal.... Even decent 8bit scopes will give you good accuracy compared to average meters in RMS measurements...
That being said, all my meters do good job in their BW. AC+DC TRMS is reason why I like dual or triple display meters...
-
Oscilloscopes have both DC RMS and AC RMS measurements for a reason.