Power Monitor on load side of SCR Power Controller

[EC Dan]

I'm interested in measuring demand (average) voltage, current, apparent power, and power factor on the load side of an SCR power controller (Watlow DT-348) feeding a delta-connected 480VAC three-phase heater. I need all three legs measured independently, which is why I can't use the SCR power controller's own diagnostics. I'm looking at an AB Power Monitor 500. I understand that power monitors can have issues with drive outputs due to voltage spikes and/or harmonics, so we would use a zero-crossing or burst-firing method of control which as I understand it minimizes those concerns. Does anyone see a reason why this wouldn't work? If the demand calculation is averaging over the duty cycle of the firing period, I believe I should get what I want. I did talk to Rockwell about this, and they say it is not recommended, but I'm not sure they fully understood the application. I will continue engaging with them, but wanted to get a second opinion.

 

[zbang]

The usual questions-
Why can't you put the same monitor on the line side of the controller?
Is this for troubleshooting? Unless the connections change, the % load on each leg should never change and you can verify that they're evenly loaded with a simple clamp-on meter. It would be an uncommon situation where all three legs need to be monitored independently.
Assuming purely resistive heaters, measuring PF will be irrelevant (it'll always be 1.0), so apparent power = actual power.
With burst-firing you won't have the harmonics of phase-angle firing but the averaging period will have to be longer; I didn't read enough into the AB manual to see how that'll work.

 

[steve66]

I'm assuming your using 60 hz. line voltage?

But in general, if its 60 hz., and the meter is a true RMS meter, I don't think you will have much trouble, unless you need super accurate measurements.

I see AB also has a "protective device" accessory. I wonder what that is.

 

[EC Dan]

The usual questions-
Why can't you put the same monitor on the line side of the controller?
Is this for troubleshooting? Unless the connections change, the % load on each leg should never change and you can verify that they're evenly loaded with a simple clamp-on meter. It would be an uncommon situation where all three legs need to be monitored independently.
Assuming purely resistive heaters, measuring PF will be irrelevant (it'll always be 1.0), so apparent power = actual power.
With burst-firing you won't have the harmonics of phase-angle firing but the averaging period will have to be longer; I didn't read enough into the AB manual to see how that'll work.

This is for an R&D application for a proprietary heater technology, so we can't assume balanced load over time or unity power factor (specifically we're looking for any capacitive effects). I thought about line side measurements, but we wouldn't get average voltage in that case. I'm ok with a long averaging period, but what I'm seeing in the description for burst firing says the duty cycle period is on the order of 16 cycles (ie. 50% duty cycle equates to 8 cycles on, 8 cycles off) so the averaging period may not need to be longer than 1 second anyway. The Power Monitor has configurable time intervals for DMD/AVG calculations of 1, 5, 10, 15, 20, or 30 minutes.

I'm assuming your using 60 hz. line voltage?

But in general, if its 60 hz., and the meter is a true RMS meter, I don't think you will have much trouble, unless you need super accurate measurements.

I see AB also has a "protective device" accessory. I wonder what that is.

Yes, supply voltage is 60 Hz and that will not change on the load side. The power monitor is a true RMS meter. Here is what is included in the protective device accessory (1400-PM-ACC):

 

[winnie]

I don't know the AB Power Monitor, so can't answer that part of your question.

But you can certainly measure what you want with a fairly simple DAQ system from any of a number of vendors.

You just need a high enough bandwidth and high enough rated voltage and current sensor for every line you want to monitor.

A couple of years back I needed to monitor power going to the input rectifier of a custom inverter. I used the LEM 210-s active hall effect current transducers and Isoblock voltage transducers to convert the measured values to +-10V signals, and fed these into the existing DAQ hardware (some overpriced National Instruments system)

Current transducer:

LF 210-S | Digi-Key Electronics

Order today, ships today. LF 210-S – Current Sensor 200A 1 Channel Hall Effect, Closed Loop Bidirectional Module, Single Pass Through from LEM USA Inc.. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

Voltage transducer:

ISOBLOCK V-1C (1000V 10V) | Digi-Key Electronics

Order today, ships today. ISOBLOCK V-1C (1000V 10V) – Voltage Transducer ±1000V Input ±10V Output 9 ~ 28VDC DIN Rail from Verivolt LLC. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

I like Measurement Computing for basic DAQ systems

MCC USB-1808 Series

Shop now for USB test and measurement devices, FPGA development boards, programming solutions and educational products.

www.mccdaq.com

Of course the above are examples of what I used, you would need to select devices appropriate to your voltage and current. Also the above transducers require power supply and interface components. These are simple and easy to supply, but other manufacturers can supply complete systems, eg. clamp on current probes with +- 10V outputs or differential voltage probes.

Jon

 

[synchro]

If you can afford it, it's likely that the following instrument will meet your needs when used with appropriate voltage and current probes:

https://www.keysight.com/us/en/prod...er-analyzer-4-channels-3-phase-ac.html#Extend

https://www.keysight.com/us/en/assets/7018-04657/data-sheets/5992-0324.pdf

An RMS measurement inherently includes some averaging due to the "Mean" in Root Mean Square, and so each measurement is taken over some time interval. Depending on where the measurement intervals are taken by the AB Power Monitor with a sequence of burst waveforms, measurement errors may result. "Averaging" might help reduce such errors depending on how it's implemented. If you already have or can borrow an AB Power Monitor, then you could program the SCR power controller for various burst modes and durations, continuous mode, etc. and compare the measured results to see if they make sense vs. what you would expect.

This is for an R&D application for a proprietary heater technology, so we can't assume balanced load over time or unity power factor (specifically we're looking for any capacitive effects). I thought about line side measurements, but we wouldn't get average voltage in that case.

Are you concerned about power factor because of the penalty that power companies will charge the user? If so, the power factor from the current drawn would have to be determined relative to the voltage waveform that's supplied by the power company, and not relative to the voltage after an SCR controller. The delay before an SCR turns on within a half cycle of a 60Hz waveform will cause the load current to lag the supplied input voltage, and result in a lagging displacement power factor. A "distortion power factor" would also include harmonics in the current that are caused by the SCR operation.

When you mention "capacitive effects" are you concerned about possibly higher peak currents when the SCRs turn on because of capacitance in the load? Or is it just the effect on power factor that you're considering?

I think you may need to make voltage measurements on both sides of the SCR power controller to get all the information you need.

 

[zbang]

I'd want to test the AB device with the controller feeding a known purely resistive load to see how it performs with the different modes, then maybe again with a known inductor thrown in the line to "lag" the PF.

 

[Jraef]

.If the demand calculation is averaging over the duty cycle of the firing period, I believe I should get what I want.

There is a way to set the "Filtering Coefficient"(Parameter FILTER CO) in the PM500. Matching that up to the duty cycle in the variable time base pulse firing scheme is not going to be easy, but I think if you can set it out long enough, it should work. Read this paper on the challenges, it describes their solution (using different transducers), but in the beginning, does a good job of explaining the challenge and makes some useful recommendations.

In the PM500 instructions, look up the FILTER CO for their description of adding that filter.

 

[EC Dan]

Are you concerned about power factor because of the penalty that power companies will charge the user? If so, the power factor from the current drawn would have to be determined relative to the voltage waveform that's supplied by the power company, and not relative to the voltage after an SCR controller. The delay before an SCR turns on within a half cycle of a 60Hz waveform will cause the load current to lag the supplied input voltage, and result in a lagging displacement power factor. A "distortion power factor" would also include harmonics in the current that are caused by the SCR operation.

When you mention "capacitive effects" are you concerned about possibly higher peak currents when the SCRs turn on because of capacitance in the load? Or is it just the effect on power factor that you're considering?

I think you may need to make voltage measurements on both sides of the SCR power controller to get all the information you need.

The theoretical capacitive effects would be due to where and how the heaters are installed, basically the presence of large energized surfaces close to large grounded metal surfaces. This is purely for parameterization of the technology, not for any direct economic reasons. You mentioned that a displacement power factor would appear if firing within the half cycle, but for zero crossing is this still true? What if the heater introduced it's own intrinsic (capacitive) power factor, such that the SCR fires at the voltage zero crossing, but not the current zero crossing. Would the SCR firing cause an additional displacement/distortion in the power factor? Sounds like it could result in current harmonics as you say. Let's assume that we install the power meter on the line side and sacrifice average voltage measurements, which I don't really see a having much value since the heater is only seeing system voltage or zero voltage anyway.

There is a way to set the "Filtering Coefficient"(Parameter FILTER CO) in the PM500. Matching that up to the duty cycle in the variable time base pulse firing scheme is not going to be easy, but I think if you can set it out long enough, it should work. Read this paper on the challenges, it describes their solution (using different transducers), but in the beginning, does a good job of explaining the challenge and makes some useful recommendations.

In the PM500 instructions, look up the FILTER CO for their description of adding that filter.

I'm ok with long averaging periods. The minimum in the PM 500 is 1 minute anyway, which would certainly qualify as 'long' based on that paper which concludes 5 seconds is suitable for a 3 second firing period. If I'm understanding the paper correctly, they switched to Hall effect sensors because the half-cycle control method results in a DC voltage, but with full-cycle control as with the Watlow, this shouldn't be a concern, correct?

I don't know the AB Power Monitor, so can't answer that part of your question.

But you can certainly measure what you want with a fairly simple DAQ system from any of a number of vendors.

You just need a high enough bandwidth and high enough rated voltage and current sensor for every line you want to monitor.

A couple of years back I needed to monitor power going to the input rectifier of a custom inverter. I used the LEM 210-s active hall effect current transducers and Isoblock voltage transducers to convert the measured values to +-10V signals, and fed these into the existing DAQ hardware (some overpriced National Instruments system)

Current transducer:

LF 210-S | Digi-Key Electronics

Order today, ships today. LF 210-S – Current Sensor 200A 1 Channel Hall Effect, Closed Loop Bidirectional Module, Single Pass Through from LEM USA Inc.. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

Voltage transducer:

ISOBLOCK V-1C (1000V 10V) | Digi-Key Electronics

Order today, ships today. ISOBLOCK V-1C (1000V 10V) – Voltage Transducer ±1000V Input ±10V Output 9 ~ 28VDC DIN Rail from Verivolt LLC. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

I like Measurement Computing for basic DAQ systems

MCC USB-1808 Series

Shop now for USB test and measurement devices, FPGA development boards, programming solutions and educational products.

www.mccdaq.com

Of course the above are examples of what I used, you would need to select devices appropriate to your voltage and current. Also the above transducers require power supply and interface components. These are simple and easy to supply, but other manufacturers can supply complete systems, eg. clamp on current probes with +- 10V outputs or differential voltage probes.

Jon

Thanks for this information. Hopefully we can find an off-the-shelf full contained solution but this gives me a good starting point for a potential custom solution.

 

[synchro]

I'm interested in measuring demand (average) voltage, current, apparent power, and power factor on the load side of an SCR power controller (Watlow DT-348) feeding a delta-connected 480VAC three-phase heater.

The theoretical capacitive effects would be due to where and how the heaters are installed, basically the presence of large energized surfaces close to large grounded metal surfaces. This is purely for parameterization of the technology, not for any direct economic reasons. You mentioned that a displacement power factor would appear if firing within the half cycle, but for zero crossing is this still true? What if the heater introduced it's own intrinsic (capacitive) power factor, such that the SCR fires at the voltage zero crossing, but not the current zero crossing. Would the SCR firing cause an additional displacement/distortion in the power factor? Sounds like it could result in current harmonics as you say. Let's assume that we install the power meter on the line side and sacrifice average voltage measurements, which I don't really see a having much value since the heater is only seeing system voltage or zero voltage anyway.

I think it might be wise to use an oscilloscope with appropriate probes to capture voltage and current waveforms when different firing approaches are used with the SCR power controller. That way you can see whether there's any unusual behavior beyond what would be expected for a pure resistive load (for example, any current peaking due to load capacitance, ringing, etc.). If there's nothing significant like that, then you can proceed with more confidence that you're not missing something when determining performance metrics and parameters on the heater system.

I need all three legs measured independently, which is why I can't use the SCR power controller's own diagnostics.

Do you intend to measure the individual phase currents passing through the loads that span across terminals of the delta, or just the three line currents feeding the delta?