Another Thread About Using a Digital DC Voltmeter For Monitoring AC Track Voltage

A few weeks ago I acquired a miniature digital readout PC based AC/DC voltmeter for $25.00. Works very nicely.

BOB WALKER posted:

A few weeks ago I acquired a miniature digital readout PC based AC/DC voltmeter for $25.00. Works very nicely.

Bob; can we see it please?

Rod: How well does this scale through the range of possible AC voltages? Does it follow in a linear fashion? How many different voltages did you test?

I have no problem with the circuit. Looks good to me!

You'll find that any of these circuits are not all that accurate, especially for non-sinusoidal waveforms.  I went through this a few years ago when I was building a track voltage monitoring car.  I ended up finally finding AC meters, but even those don't do well with chopped waveform transformers.  However, they tracked a lot better than the DC meters I tried to use with rectification.

Leo, John; point taken on the tracking accuracy. I would expect this setup to be reasonably accurate for use in a fixed voltage command scenario in the 16-20 vac range. But in a variable voltage conventional situation I am sure the accuracy trails off below about 12 volts, as the fixed 1.4 volt loss of the bridge diode takes a larger percentage off the rectified DC voltage. We shall see.

I plan to do a little more testing nevertheless on the full range from 6 to 20 vac and just see how bad it gets. More later. 

Rod

I found this one that’s within .5  volts. I made the meter car just for a project. The red and black wires on the car are track power to the meter, with a diode to the AC to DC converter. The red from the converter powers the led meter at 12 volts. Didn’t need the DC ground; works without it.  

In this OGR thread, I show a similar approach using a 3-wire eBay voltmeter.

Note that when using the 2-wire the measured voltage is the same as the meter's power supply voltage.  In the 3-wire meters, the measured voltage is a separate input and typically has an input resistance of about 100K Ohms.  So this may be nit picking, but consider displaying 8.8 vs. 11.1 Volts.  Two 8's represent 14 "on" LED segments in the 7-segment digit.  OTOH, three 1's represent only 6 "on" segments.  That is, going through a 10K trimpot, a variation in LED display current of just 1 mA can have a whopping effect on the measured voltage.

In this different long OGR thread, I showed how to modify the AC digital meters that require 80V AC or 100V AC to operate at O-gauge AC voltages.  It is a tedious modification which I do not recommend.

In one of both of the linked threads, it is inevitable you fall down the rabbit-hole of true/pure sine vs. chopped sine train transformers with regards to RMS vs. averaging measurements.

When you have chopped waveforms, the voltages go way off with the simple rectification procedure.

FWIW, my voltmeters required an isolated supply, so I found a small isolated DC-DC regulator to power them from track power, hence there is no varying load as far as the measurement is concerned.

FWIW, eBay: 143361459558 has AC ammeters and voltmeters in all the desirable ranges, I haven't seen those lately.  For $6.99, they're a decent deal.  These use a separate isolated DC supply, this also makes them likely to be more accurate in varying loads.

You would need some kind of averaging circuit, integrate the rectified waveform over the cycle and then scale appropriately.

Stan’s modified 80vac vm sounds like fun though.

gunrunnerjohn posted:

You'll find that any of these circuits are not all that accurate, especially for non-sinusoidal waveforms.  I went through this a few years ago when I was building a track voltage monitoring car.  I ended up finally finding AC meters, but even those don't do well with chopped waveform transformers.  However, they tracked a lot better than the DC meters I tried to use with rectification.

I made a distance car once, calibrated it in feet. Useless but interesting to measure how much track you have.

I have a whole bunch of distance measuring cars.  They're otherwise named MTH PS/2 and PS/3 locomotives.   Just set the trip meter to zero and run the route.  Stop and read the trip meter...

gunrunnerjohn posted:

I have a whole bunch of distance measuring cars.  They're otherwise named MTH PS/2 and PS/3 locomotives.   Just set the trip meter to zero and run the route.  Stop and read the trip meter...

Ahh...kind of neat. I don’t have any MTH locos. Didn’t know that.

Today I did a quasi-scientific comparison of AC versus indicated AC (DC really) using the circuit in the opening post. It's actually better than I thought it would be:

Note that I was limited to only 1/10 of a volt resolution, due to the el cheapo DMM I had to use for the AC measurements. So that is kind of a limitation for sure. I might search around and see if I can scare up a better meter for the AC. The DC meter is a 6000 count and resolves to 1/100 of a volt.

Also these results were made using a pure sine wave old school adjustable AC transformer. So for sure they are way better than what you would see with a chopped wave transformer. I don't have an example of the latter here with me in AZ.

John, that meter looks pretty neat for what it can do. May have to order up a couple of those and try them out instead. 

The pure sine wave results makes sense, it's the chopped waveform that really screws up the results.

Rod

With a maximum variance of only 0.3V from your DMM, your circuit is accurate enough for its intended use:  monitoring track voltage.  I'm running command using Z-1000 and PH-180 bricks, and a Z-4000, (all sine wave) so this should work for me.   Here is a list of sine wave train transformers.  Mike Regan did a detailed video of chopped vs sine wave transformers here.

Bob

Sine wave Transformers:

All Lionel and AF post war transformers

Lionel Powerhouse 135 and 180

MTH Z-1000, Z-750, and Z-500 (brick only - no controller)

MTH Z-4000

MRC Pure Power AC

Right of Way 2020

For pure sine wave, that's correct, it's pretty easy to get fairly accurate readings.  The chopped waveform transformers are much more difficult to get decent readings from, even with a good meter.