Demian wins the prize! The first true rms meters used a bolometer. A technique that was good to very high frequencies, but suffered limited range and required a bit of protection to prevent over voltage burn out.
Modern true RMS meters often have a wider and safer input range but are frequency and crest factor limited. With a Fluke model 87 the value of a good noise source will change a bit with the measurement range selected!
Modern true RMS meters often have a wider and safer input range but are frequency and crest factor limited. With a Fluke model 87 the value of a good noise source will change a bit with the measurement range selected!
Scott beat me to it, but of course it is based on a sine wave. And measuring the AC line voltage is, for all intends and purposes, a sine wave. Same difference.
BTW just come back from the local market, scored some nice, almost pristine LPs: 10CC - windows in the jungle, and Crosby & Nash - wind on the water.
Great for a Sunday morning
BTW2 is it just me or does Dave Crosby look like Nelson Pass' twin??
Jan
Last edited:
Don't change the issue Ed! The issue you brought up was how much difference it makes on the true RMS meter re: a peak reading RMS calibrated one. You want to hazard a guess?
And of course, if that 5% is an issue for SMPS'es, you surely won't determine that by measuring the RMS value anyway....
jan
Last edited:
Ed, I envisage using separated windings and two bridges, and extending four power wires into the console.
I should have junked bits to do the CRC filtering.
I envisage that it should be idiot proof also...ie no opportunity for short circuit load.
Power requirement should be something like 15V or 16V, 1.5A..ie 24W per rail supply.
Why do you state 5%/1V ripple.
Dan.
Just what I said crest factor agnostic, if it's not a pure sine wave it's not the RMS.
LTC has an IC "bolometer", the IC true rms chips have pretty low errors up to crest factors of 5 or so and limited ranges, need to read the data sheets.
There are correction factors for other waveforms than sine. But, here's a timely factoid --
12-13-14
yesterday's date is the last time in our life time we will see a consecutive date occur.
An auspicious date. One we will all refer to and ask 'what were you doing on this date'
Right.
-RNM
12-13-14
yesterday's date is the last time in our life time we will see a consecutive date occur.
An auspicious date. One we will all refer to and ask 'what were you doing on this date'
Right.
-RNM
Last edited:
I am not changing the subject!
A peak reading RMS calibrated meter will read 3.5% or more lower than a true RMS meter on a typical power line.
RMS meters are accurate no matter what the waveform (within some peak/average limits). They show what DC voltage corresponds to create as much heat into a given resistive load as any type of AC waveform.
The AC power line is not a sine wave. Typical distortion is 5%. When the distortion reaches 10% power companies start getting lots of complaints. It is not just the distortion, but that not all customers get their nominal 120 or 230 volts. Those at the end of a line often get less. So adding the drop from line distortion means some items with power supplies will not see enough peak voltage to work correctly.
In my shop there is 3 phase nominal 120/208 VAC. The idea is that from any phase to neutral (common) there will be 120 volts. (sin(120)*240 =208) However as they also feed some houses from the same transformer my single phase voltage is actually 127 VAC! (sin(120)*254=220) The upper limit is now 132 VAC for US power lines. The lower limit is 110 VAC.
Why use known flawed equipment, I know you are ignoring me just sayin'? BTW we wrote the book at least one book, Jim Williams wrote a lot too. http://www.analog.com/static/import...283758302176206322RMStoDC_Cover-Section-I.pdf
Because the flawed equipment gives the right answer. The true RMS meter gets it wrong! The peak reading RMS indicating uses a rectifier just like in the power supply so it gives a more useful reading. True RMS is actually less useful for looking at power lines!
Right answer? It's not the RMS which has a specific definition, as I said there is ample literature for those who want to understand the issue. This is a made up argument with little useful purpose.
Do you really not get it? A distorted power line will deliver less peak voltage and cause regulated linear power supplies to not work properly. But if you check the voltage with a true RMS meter it will not show the problem.
True RMS is useful for measuring complex signals such as pink noise to determine loudspeaker sensitivity and power handling.
Now an oscilloscope will show the problem with the power line but is not very useful for reading accurate levels on signals such as pink noise. There are purpose built oscilloscopes that will also do a bit of FFT intended for power line use but they aren't as wideband as a full FFT analyzer. But few folks would hook up the full FFT analyzer straight to the AC mains.
No I stare at the floor drooling with nothing but a $15 Radio Shack VM constantly forgetting everything I and my colleagues have done for the last 40yr.
It used to be you wanted at least two volts across the regulator IC. Now if you have a 15 volt regulator that would be 17 volts. Allow for 1 volt of ripple 1.4 volts for the rectifiers and an AC line that is 10% low gives you a transformer secondary rating of 15 VAC. Now if there is even 5% line distortion that transformer will drop low enough that the regulator will run out of headroom. So the next size up would be 18 VAC secondary to work at minimum line voltage. The down side is that under high line conditions your regulator now has double the power to dissipate.
But to make it simple a 100 watt power amplifier will drop to 90 waste at 5% line distortion and 81 watts at 10% distortion from a full RMS voltage AC line.
So a very small 5% line voltage error can significantly affect the cost and performance of a linear power supply.
Now the simplest work around is to use a transformer that is designed to work from a 110 VAC line voltage. Due to hysteresis it will begin to saturate and limit the secondary voltage under high line conditions. The big but is that in doing so it increase the harmonic distortion of the incoming power.
Last edited:
Measuring line voltages accurately can be quite important in a large network or on a transmission line. And Ed's issues around scaling a linear power supply are important. You can't always have unlimited excess heat sink or power dissipation in the pass devices. understanding the wide range of actual power really pushes the supply's requirements. I use for the US 130 to 95 as the must always work range. That's up to 30% over the min voltage requirement, all going away as heat.
And (this is from experience) if you make an AC regulator do you stabilize against TRMS voltage or peak voltage? If you are supporting rectified loads the peak voltage will ensure the devices are operating at their design values. An RMS regulated source with 5% (and as much as 15%) distortion would potentially adjust the output voltage lower and possibly out of regulation. However if you are supporting motors and incandescent light bulbs RMS is appropriate. Incandescent bulbs are on their way out (possibly a mistake of large proportions as the environmental impact of the electronics becomes more apparent) and traditional motors are getting replaced with electronically controlled motors. The peak voltage would be a concern in a large industrial plant for example since it can translate into power factor charges and fried expensive stuff.
Actually the traditional rectified power supply may be retired soon since it doesn't meet the newer power factor requirements.
And (this is from experience) if you make an AC regulator do you stabilize against TRMS voltage or peak voltage? If you are supporting rectified loads the peak voltage will ensure the devices are operating at their design values. An RMS regulated source with 5% (and as much as 15%) distortion would potentially adjust the output voltage lower and possibly out of regulation. However if you are supporting motors and incandescent light bulbs RMS is appropriate. Incandescent bulbs are on their way out (possibly a mistake of large proportions as the environmental impact of the electronics becomes more apparent) and traditional motors are getting replaced with electronically controlled motors. The peak voltage would be a concern in a large industrial plant for example since it can translate into power factor charges and fried expensive stuff.
Actually the traditional rectified power supply may be retired soon since it doesn't meet the newer power factor requirements.
All good points Demian but has nothing to do with RMS or otherwise measurements. You will not measure the line voltage in your home and then size your powersupply to it. Unless you are 100% sure you want to use your equipment only in your room, at that day, at that hour.
Sounds sensible.
Jan
Industrial/large multi kW line xfmr can be expected to be better designed, often have multiple taps for localization but not the stuff most hobby diy home audio system builders buy
few hundred W and lower off the shelf line xfmr may not work well if you adopt a high line V requirement - most are initially designed to run deep enough in saturation that 10% higher V*t can push Im up >10x, especially with 50 Hz operation of "universal" xmfr
few hundred W and lower off the shelf line xfmr may not work well if you adopt a high line V requirement - most are initially designed to run deep enough in saturation that 10% higher V*t can push Im up >10x, especially with 50 Hz operation of "universal" xmfr
- Status
- Not open for further replies.