I can monitor my HTQ Series compressor current, but it doesn’t correlate realistically with the reported total Outdoor Unit power consumption - what I see is kW = (0.42 x amps) + 0.15 (at R^2 > 0.9 so a reasonable fit).
I can believe the offset 0.15 is the non-compressor consumption (evaporator fan is approx 100W, so electronics a plausible 50W), but if I apply Volts = Watts / Amps to the measured kW (less the 0.15kW overhead), this would suggest about 420V (versus the actual 240V) - a ratio of 1.72.
This is a long shot, but I notice from the Outdoor Unit wiring diagram:
that the compressor is 3 phase, and wondered whether the controller current sensor could possibly be located on just one phase.
If so, I could divide the calculated voltage by sqrt(3), which would then match my actual voltage.
I’m no electrical/control engineer (as you can tell) but this ~1.7 ratio is fairly consistent over all compressor speeds, so if my theory is not correct, can anybody else come up with an alternative explanation?
Two things come to mind. The first is that there will be some reactive power, so you can’t just do P=U.I but have to multiply with the power factor cos(phi), the cosine of the phase delay between voltage and current. Secondly, if indeed only one phase is monitored and we assume that the other phases carry the same current and have the same power factor, then the factor sqrt(3) is indeed correct, as P=sqrt(3).U.I.cos(phi) for a balanced 3-phase load. This is approximately true for my Vaillant heatpump, with the exception that L1 also supplies power to the pump and hence there’s a bit more power on that line. If you observe a rather consistent sqrt(3) factor then this seems like a good empirical approach.
Thanks @Andre_K.
I seem to recall that the PF of heat pump inverters (and the smallish motors we get on them) is close to unity, meaning that W = V x A is a fair approximation. Maybe?
And that isn’t the full story if there are harmonics present in both the voltage and current waveforms.
P = V.I.cos(φ) (or in German P = U.I.cos(φ) ) only applies to a perfect sine wave, or to each harmonic individually. Clearly, if the sum of the harmonics is insignificant, they can be ignored. For an inverter, this probably will not be a valid simplification, so expect an error.
Good point, though I guess the heat pump will have some power factor correction to mitigate this. But you’re absolutely right and the RCDs used for inverter devices have to be sensitive to these harmonics as well as far as I know.
The Samsung manual shows reactor connections on the Inverter PBA, so I guess that’s exactly what they have done to reduce harmonics.
But why would they put a CT on just one phase? I’ll put it down in my “Samsung Mysteries” book (if there’s any room left in it…)
And mine is. If the load is balanced across the three phases - as it should be if it’s just the motor (the c.t. is on the inverter output) or the inverter together with the motor (the c.t is on the inverter input, which is more likely) then there’s no need to measure the current on all three phases, because it’s going to be the same on each.
It’s quite normal practice to measure only one phase and multiply by 3 when you know the load is balanced or nearly so. But if loss of a phase represents a danger, then the protection instruments must detect the imbalance even if you don’t care for the measurement.
That’s a measurement of my whole installation (outside + inside unit + controller). Pump is on L1, probably some other stuff as well, but it’s relatively well balanced.
I have made some actual measurements for cos phi of my single phase 3,6 kilowatt electrical power 8 kW heating power HP. The modulation % for compressor is from min 16% to 100%. For lower values of modulation cos phi is almost 0,8 and reaches 0,99 and 1 when consuption and modulation is close to maximum values. For standby actual cos phi is zero. Supply voltage for my Fujitsu compressor is close to 370 volts.
Those are very much the sorts of numbers I’d expect. The inverter designers will want the best possible power factor, but there’s a limit to what’s achievable over a wide range of operating conditions.
Sorry to be so stupid, @Robert.Wall but should that be sqrt(3)?
(The only elec eng I was exposed to was 50 years ago, and not too much lodged even then…)
The total power is 3 × the power on each individual phase. Whether you need the √3 or not depends on which voltage and current you are measuring - is it a 3-wire or a 4-wire system; is the load connected star or delta, so is the voltage you measure √3 × the individual load voltage (star connected - the load voltage is between line and star point, which might be the neutral, when you can only measure line to line); or is the current you measure √3 × the current in each individual load (the load is delta connected, and you can’t measure just one side of the delta)?
Problem is, the only arrangement data I have is the sketch I included with the initial post above. I see what looks like a 3-wire star arrangement, but don’t know where the current sensor is or how it is connected.
I’d like to be able to predict the ASHP total power consumption from the limited data available from my controller monitor (which doesn’t include power consumption). I had a quick look at the mathematics of inverters but quickly got over my head (modulation index etc ).
So a simple question if I may (based on some steady state readings today):
If I know my DC Link voltage (182V) and my “Compressor Current” (2.96A) when the compressor inverter was running at 23.9Hz, whilst the indicated total ASHP power input (not from the controller monitor) was 1.43kW (of which I’d guess 0.15kW was non-compressor consumers), is this enough information to derive an overall correction factor that I might apply at other speeds/compressor outputs - effectively a calibration curve for my compressor?
Probably the only way to get an accurate curve will be measure motor current and voltage (and even that won’t be the shaft power) and record that against the quantities you know - which means some temporary monitoring and then try to fool the heat pump into working over its full operating range.
I’ll put it down in my “Samsung Mysteries” book (if there’s any room left in it…)