What sorts of levels of currents & voltages are we talking about here? What's the application? Switching converter, class D Amp? Measurement types - single ended voltage, ie wrt ground, or differential? If you're wanting RMS value of a varying DC current, then a current transformer is no good, a hall effect probe is more like it. The more info you give, the easier it is to answer.....
linear technologies LTC6968 == the late Jim Williams wrote an application note on it: http://cds.linear.com/docs/Application Note/an106f.pdf I compared these to the HP3403C I own and they are very good indeed. You can buy the dev board from LTC and it will save you many hours of burning a pcb yourself.
Analog devices also has a number of RMS converter chips useable to a few hundred kHz, and a slew of logarithmic converters useable to several hundred MHz.
Stateside you can occasionally get a Hewlett Packard 3403C or Fluke 8620 calorimeter meters at attractive prices.
Analog devices also has a number of RMS converter chips useable to a few hundred kHz, and a slew of logarithmic converters useable to several hundred MHz.
Stateside you can occasionally get a Hewlett Packard 3403C or Fluke 8620 calorimeter meters at attractive prices.
If you know it's sinusoidal, yes, rectify, measure, scale for RMS, bit higher than 200mV for schottky diodes I think. Feed the rectifier into a cap, say 1u, measure that, otherwise the 200kHz will still freak your meter.
Jim Wlliams is/was a fascinating writer, very addictive, beware!
Jim Wlliams is/was a fascinating writer, very addictive, beware!
Last edited:
Many thanks. I think for the time being and since this is for specific measurements only, and as Jim Williams has written in his opening statement, a simple rectifier will do.
I tried using germanium diodes and then schottkies (BAT48) but at +/- 40 Volts they are behaving very badly, so I am using 1N4148. It is not perfect, at low voltages there is a small error (say 300mV) which then increases for higher voltages.
And it shows peak voltage not RMS. Is there an easy way to convert peak voltage to RMS and display on a meter ?
I tried using germanium diodes and then schottkies (BAT48) but at +/- 40 Volts they are behaving very badly, so I am using 1N4148. It is not perfect, at low voltages there is a small error (say 300mV) which then increases for higher voltages.
And it shows peak voltage not RMS. Is there an easy way to convert peak voltage to RMS and display on a meter ?
Attachments
......I tried using germanium diodes and then schottkies (BAT48) but at +/- 40 Volts they are behaving very badly, so I am using 1N4148. It is not perfect, at low voltages there is a small error (say 300mV) which then increases for higher voltages.
And it shows peak voltage not RMS. Is there an easy way to convert peak voltage to RMS and display on a meter ?
With a 50v-peak signal, the diode will have to withstand 100v during negative half cycles, due to the +50v still being on the cap, hence your disappointment with the BAT48, only rated at 40v piv repetitive. Even 1N4148 is right on its limit here at 100v piv repetitive.
The diode across the cap isn't really doing anything.
Unless the error needs to be less, don't worry about it, or allow for it in your readings, it's quite predictable. As for RMS from peak, as it's a sinusoidal waveform, just divide by square root of 2, ie 1.41 etc. If you have to have it on the meter, you could build a voltage doubler, followed by a suitably scaled voltage divider of two resistors, (don't forget to allow for your meter resistance) but frankly, why bother, just use your calculator as you work, much easier.
Mmm....duh.....yes.
Irrefutable proof, as if it were needed, that I am indeed brain dead.
Akis, do you know the input impedance of the meter you're using, on the voltage range(s) you're using? In order to reduce rather than increase the ripple on the cap, you could make the meter's resistance the lower one of the two.
I use some ultra cheap DMMs, input impedance 1M on the voltage ranges, (I measured it just to check that the spec was correct). In order not to load circuits when I measure voltages in them, I have a 9M resistance that I can connect in series with it, giving a total of 10M. Effectively, it's a 10x probe, with banana plugs & sockets rather than bnc.
The 9M is composed of: 4M7 + 3M3 + 1M0
The rectifier itself loads the circuit under test (the capacitor is effectively a short to start with), and I just added a 1K resistor in series to stop it from triggering the current limit on my PSU, but still it loads the circuit with 1K load even for a short duration.
I take your point about the 9M in series with the voltmeter.
I take your point about the 9M in series with the voltmeter.
You may want to track down a Leader LMV-181A, -185A AC voltmeter:
They have an input range of 1mV to 300V, a frequency range of 5Hz to 1MHz, and are accurate to +/- 2% of full scale (or better if calibrated for a particular frequency/range. Also available is a 0-1V output that represents 0V to the the selected full scale, very convenient for remote monitoring or logging.
Always on eBay for $35 to $100 ($50-$60 will get you a good one). I have the owner's and service manuals if needed. There is also a 2-channel model, the LMV-186A.
They have an input range of 1mV to 300V, a frequency range of 5Hz to 1MHz, and are accurate to +/- 2% of full scale (or better if calibrated for a particular frequency/range. Also available is a 0-1V output that represents 0V to the the selected full scale, very convenient for remote monitoring or logging.
Always on eBay for $35 to $100 ($50-$60 will get you a good one). I have the owner's and service manuals if needed. There is also a 2-channel model, the LMV-186A.
Last edited:
Did you mean Fluke 8920? (or 8921 or 8922?)
Most multimeters are only designed to go to 100 KHz (the cheap RadioShack ones not even that). Look on eBay for an "AC Voltmeter" You will find many options. This is a very good value: BALLATINE 323 TRUE RMS AC VOLTMETER | eBay simply because its not well known. Or this HP Hewlett Packard 400E AC Voltmeter RMS Volt Meter | eBay
For AC current I would use a current transformer or a current probe like this HP / Agilent 456A AC Current Probe - 1mA to 1mV | eBay
it all depends on what you are trying to measure.
yes, i have an 8920A which was purchased for well under $100, but now they fetch multiples of this. i also have a 3403C.
I have been using the scope and a simple rectifier circuit (on a breadboard) with very good results. I am only measuring at 200 KHz for the time being while I am building a device - it is not a long term thing, although it would be great to have a proper bench AC Voltmeter .
Those ebay offers are for the US - I will have a look for UK ones.
Those ebay offers are for the US - I will have a look for UK ones.
Sad to say, it's often cheaper to buy a piece of gear on US ebay and pay the shipping than it is to buy it from the UK. Plus there's much more gear actually on US ebay in the first place.
There is a lot to be said for improvising test gear to suit one's immediate needs where possible. If you find yourself breadboarding the same, or a similar circuit over and over, build it, put it in a case and benefit from it permanently.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.