mini split power observations

ionized_gwJuly 13, 2014

I started looking at the power/current draw of my mini splits last week.I was curious about the start-up current because when I operated them on a portable genset, I did not observe any sudden loading/bogging of the genset. In fact, the current starts low and rises to a max very slowly over a minute or two and then backs off by a noticeable amount after a few minutes. They seemed curiously high for the larger of my two compressor units.

Yesterday. I went further. I poked around in the service manual. There is a section with charts to be compared to readings in "Emergency cool" and "Emergency heat" modes that seems to take out lots of adjustments that the microprocessor can make, simplifying observations. Inverter operation is fixed frequency and fans are on high For each of the outdoor units covered (4), every indoor unit has a pair of graphs.

Both graphs for each indoor unit have outdoor ambient temp on the x axis (65 - 115 for cooling graphs). There is a series of lines, each at a different indoor DB temp 86,80, 75, and 70) for the pressure graph. On the current graph, there is only one line. From these I guess you are supposed to get expected currents and pressures on the Y axes.

I have two MXZ2b20NA outdoor units. One has two 9000 btu indoor units and the other a 9000 and a 6000. These all gave me between 2.5 and 2.7A which is probably within error on the graphs, but possibly a little low.

The MXZ3b24NA has three 6000btu indoor units. The currents I read for that were 3.7 to 3.9 amps which would appear high compared to the previously-given readings. Consulting the graphs for that outdoor unit, they are in line with the data given.

Why should this unit be that much less efficient? I was pretty disturbed. It was only just now that I noticed that the emergency cool mode puts the smaller compressor units (20,000 btu) into opertion at a constant 40 Hz and the larger ones (24,000; 30,000 and 36,000 btu) at constant 24 Hz.

Note that heating measurements are made at slightly different, 48 and 32 Hz. The larger compressor system may be less efficient at those settings, but when allowed to operate freely, it is probably closer to the smaller ones. I'll have to go back and measure both larger and smaller while they are running comperable indoor "loads" in their "unleashed" rather then constrained settings.

Thank you for reporting this comment. Undo
fsq4cw

Just out of curiosity, are you taking your measurements with a True RMS reading ammeter that has a wide frequency response range and can properly read complex waveforms with lots of THD?

These inverters may be introducing lots of distortion and hash on the AC line. Non True RMS ammeters will not be able to provide accurate readings under all operating conditions.

Just askingâ¦

SR

    Bookmark   July 13, 2014 at 10:47PM
Thank you for reporting this comment. Undo
ionized_gw

That is a very good question. Thanks for asking. I should have thought of that. No, I am not using a good meter at all. I am using a really cheap clamp-on meter. I repeat, really cheap. I checked its performance relative to a multimeter ammeter using a resistive load (toaster). It seemed OK. It is hard to justify an expensive meter for a hobby. Maybe I can borrow a better one. At the disconnect, It would be relatively easy to introduce a conventional ammeter. Would that work better or worse with distortion?

Some tables listing the electrical characteristics for these units have a column heading "P/F (%)". I assume that is power factor. I know what that is supposed to mean in practical terms, but I've thought little how the rating is determined. Is it for steady-state operation or does it include start-up, for example?

Fsq4cw, would the "distortion and hash" be higher during start-up or at speed? The "P/F" of these units is 95 with one indoor unit running and 97 with 2 or 3. Those numbers would indicate very little distortion wouldn't they?

    Bookmark   July 14, 2014 at 10:12AM
Thank you for reporting this comment. Undo
fsq4cw

Your toaster may not be a purely resistive load if itâÂÂs a coiled wire. That would make it an inductive load but not necessarily a complex reactive load as your mini split likely is.

Dispite all the âÂÂtheoryâÂÂ, there is no substitute for the proper instrument.

A Fluke True RMS clamp on meter should do or perhaps the instrument linked to below.

SR

Here is a link that might be useful: Fluke 345 - Power Quality Clamp-on Meter

    Bookmark   July 20, 2014 at 8:59PM
Thank you for reporting this comment. Undo
weedmeister

Power factor is not a measure of distortion. It is a measure of the phase relationship between current and voltage.

    Bookmark   July 21, 2014 at 1:43PM
Thank you for reporting this comment. Undo
ionized_gw

If the phase relationship between current and voltage is not in line, isn't that distortion? It may not be the only kind of distortion, but isn't it distortion and distortion of very common type and relatively high magnitude?

    Bookmark   July 22, 2014 at 8:05PM
Sign Up to comment
More Discussions
Manual J -> S ->D
JSHR: 0.81 MJ8 Tons: 3.03 SqFt/Ton: 1667 CFM/SqFt:...
Brian S
Greenspeed, is this setup ideal/ worth it?
I am building up on my house (cape style) in Long Island...
David
What in the world is this thing? (Water furnace) And I need a new one?
Hi. So my new condo has an ancient thing I've been...
halgreene
Are mini-splits a bad idea.
I have lived in both Europe and the Middle East and...
bry911
Baseboard heater thermostats
We are looking to install (2) 120v baseboard heaters...
stephja007
People viewed this after searching for:
© 2015 Houzz Inc. Houzz® The new way to design your home™