Hack inverter air to run on DC?

[Monomial]

I am currently investigating how to set up a building to run off grid that needs 24 hour refrigeration and air conditioning. The plan is to use solar panels and a large bank of batteries. As part of the design I am trying to reduce electrical load to the bare minimum, but I also have to use only cheap components that are readily available....meaning no imports of fancy solar equipment.

 

For the air and refrigerator, it seems stupid to me to convert the DC from the batteries to AC just so that I can run it to the inverter air conditioner which will then immediately convert it back into DC to create the variable frequency wave for the compressor.  It seems that would cost at least 10% in efficiency and probably more. It stands to reason that it would make alot more sense if I could simply use a DC voltage from the solar panels/battery bank that I could plug directly into the inverter bypassing both the lossy AC inverter and the initial AC rectifier. The problem is that I don't know what the optimal DC voltage is.  I assume that the inverter is going to have some kind of switching power supply which can take a large range of DC input voltages in and give out a constant 3 phase waveform at the right amplitude and frequency. But what would the minimum practical DC voltage be if I wanted to hack the inverter air?

 

I have checked on the internet but there does not seem to be much information about this. Does anyone know anything about how these are designed in practice, and is it possible to do what I suggest? If so, what would be the nominal voltage to target in this system? I think anything up to 96 volts would be practical to generate.  Would a typical home inverter air conditioner run on 96 volts DC (bypassing the AC rectifier stage of course)? Does anyone know what they require?

 

Thank you for any assistance, and if anyone has ever tried something like this previously please let me know the results.

 

[Crossy]

Never tried this with an A/C but I have had SMPS's running on DC, the principle is similar.

BUT

You need to be looking at a DC voltage around 300-350 VDC (rectified mains voltage).

 

Hunt on the net for inverter A/C service manuals to find schematics.

 

The PF correction circuitry may reduce the intermediate DC voltage somewhat, but I doubt it will be as low as 96V.

 

Why not run a higher battery voltage, most solar panels are good for string voltages up to 500V?

 

EDIT But a US standard A/C may have a more manageable intermediate voltage.

 

EDIT 2 Do be fully aware that the energy involved, even at 96V, is going to be significant, this kit is as capable of killing you as the mains supply (probably more-so as it's DC). Definitely not something to mess about with if not qualified.

[Wirejerker]

Be aware DC does not behave same as AC so your electrical components (switches, circuit breakers etc) must be rated for DC or you risk serious consequences. 

 

9 hours ago, Crossy said:

Definitely not something to mess about with if not qualified.

 

[Crossy]

54 minutes ago, Wirejerker said:

Be aware DC does not behave same as AC so your electrical components (switches, circuit breakers etc) must be rated for DC or you risk serious consequences.

 

Indeed WJ, I was coming to that if he insisted on proceeding

 

DC rated breakers, switches and the like (look at the solar energy sites).

[RichCor]

I was going to suggest looking into marine equipment, but that suggestion is never cost effective.

 

There are DC driven air conditioners being marketed. Some even designed to run off solar derived power.

Take a look at AliExpress: DC powered air conditioner

 

 

[al007]

I lived and travelled on my own 60ft trawler motor yacht for some 12 yrs which I also rebuilt

 

I had 12 volt DC, 24 volt DC 32volt DC, 110 volt AC and 220 volt ac with some 50cycle and some 60 cycle, I learnt a lot the hard way or good way through experience

 

I had airconditioning and refrigeration and deep freezes and ice-makers, and vacuum heads, some refrigeration 12 volt, but later switched to 110 volt, all water cooled

 

I had two  large main engines and two generators (I did not run generators a lot of the time) large battery banks and a 3kw Sine wave inverter inverter, which was also used for battery charging, which was three stage

 

the inverter produced large amounts of heat

 

Air conditioning and refrigeration is difficult to run on inverters, it also has very high start up loads as the compressors struggle to start, soft starts helped improve this, the temporary start up load can be many many times the running load

 

My battery banks were very large

 

In the end I accepted to run the a/c I needed a generator, some of the refrigeration I was able to run on the inverter if the main engines were running and the two high output alternators running

 

I replaced my batteries many times, I used 6 volt batteries from scrubbing machines

 

I will willingly talk if you wish and give you my tel no, I live KhonKaen send me PM

 

Good luck, my suggestion would be use the grid for refrigeration and air-conditioning, and solar power for the lighter loads, I will be interested to watch your progress

 

Some additional thoughts say you run a generator when you use the air-conditioning at the same time if you have holding plate refrigeration (reservoirs of brine that cool very cold and will even be good in deep freeze) the refrigeration can be cooling at the same time, Generally 4/5 hours a day of generator will give enough to drag down the temperatures for deep freeze and refrigerators

 

 

Edited July 10, 2017 by al007

[Monomial]

16 hours ago, al007 said:

 

Good luck, my suggestion would be use the grid for refrigeration and air-conditioning, and solar power for the lighter loads, I will be interested to watch your progress

 

Some additional thoughts say you run a generator when you use the air-conditioning at the same time if you have holding plate refrigeration (reservoirs of brine that cool very cold and will even be good in deep freeze) the refrigeration can be cooling at the same time, Generally 4/5 hours a day of generator will give enough to drag down the temperatures for deep freeze and refrigerators

 

 

Hi al007,

 

Can you expand on this at all? What do you mean by "holding plate refrigeration?".  I'm looking at a 50kW design right now, and nobody is happy with the cost.

 

In particular, there is no electric grid where I am talking about, so that option is out. I have a room that I need to keep at a constant 15 degrees Celsius 24/7 with controlled humidity for an industrial process.  I can't have the temperature cycling between extremes. I can use a generator for emergencies, but running a generator is nearly 3 times the expected long term cost of solar.

 

The newer inverter air conditioners and refrigerators don't have the startup rush problem you mention. The compressor is essentially always running when it is on, but the frequency is varied. Instead of the compressor being a binary on/off, it is more of an analog signal that can slowly ramp up during the day and slow down during the night. I would never try and do this on a conventional air conditioner.

 

But because this will be such a large part of the load, optimizing it is very important. Squeezing 10% efficiency out of this section means the capital investment is reduced by about that same amount. But I still need to use units that are readily available and recognized by your average technician, easily replaceable with spare parts and such. Bypassing the rectifier at the power input where the AC is converted back into DC might void the warranty, but fundamentally it is still the same AC that technicians see every day. If it breaks they will be able to fix it.

[Monomial]

On 7/7/2017 at 4:14 PM, Crossy said:

 

Indeed WJ, I was coming to that if he insisted on proceeding

 

DC rated breakers, switches and the like (look at the solar energy sites).

Thanks Crossy.

 

Yes, I am still actively investigating this.

 

But don't worry about details like safety. I am going to have a Cambodian tech do the installation. They don't worry about breakers. Works just fine without it. :-) 

 

One of the reasons I was trying to keep the voltage down to 96 volts was precisely because of the issues you mention. A -48v rail and a +48 v rail can still be reliably run on inexpensive 220VAC rated switches. Going beyond this raises lots of problems.

 

[Monomial]

On 7/7/2017 at 5:24 AM, Crossy said:

Never tried this with an A/C but I have had SMPS's running on DC, the principle is similar.

BUT

You need to be looking at a DC voltage around 300-350 VDC (rectified mains voltage).

 

Hunt on the net for inverter A/C service manuals to find schematics.

 

The PF correction circuitry may reduce the intermediate DC voltage somewhat, but I doubt it will be as low as 96V.

 

Why not run a higher battery voltage, most solar panels are good for string voltages up to 500V?

 

 

I wanted to take a minute to comment on this, because I am thinking something very different, and I'd like to know if I have misunderstood or if you know something that I don't about how these things are put together. So let me explain what I am thinking and you can tell me if I am wrong.

 

First, I am imagining that at the input stage the 220VAC waveform is fully rectified into a ripply DC waveform at a mean of around 200 VDC.  I agree that the mains voltage peaks at around 350v, but after going through the bridge rectifier and the input cap it should only be about a 40 Vpp waveform at a bias of around 200-210 VDC (the drop from the RMS voltage being due to the voltage drop by the diodes in the bridge rectifier).

 

I suspect this will then be fed into a PWM buck converter which will take a wide range of input voltages. Usually the acceptable voltage can be from about twice the RMS AC voltage, to allow for voltage spikes, all the way down to only a few percent above the output voltage. The PWM power supply controller accomplishes this by charging a bulk capacitor through switching the input current on/off in a square wave. Thus, a high input voltage would result in a duty cycle of only a few percent to charge the capacitor, while in the limit a low input voltage just barely enough will result in a duty cycle of nearly 100%.  This is a simple model of a switching power supply without PFC.  I imagine this is what cheap air conditioners will use.

 

The output voltage from the PWM should then be a stable DC voltage at approximately the RMS voltage of the compressor times root 2. There will then be a series of transistors used to convert the DC voltage into a true sine wave at the frequency needed to drive the compressor. Thus, my thinking is that what I really need to know is the voltage of the compressor that is used in an inverter air conditioner/refrigerator, and as long as my DC input voltage is well above this value plus 40% , then it should be fine. I am somewhat surprised at the figure of 350 volts. I was actually hoping for 48, and willing to settle for 96. My thinking is that there is simply the input voltage, the PWM controller, and the output DC which is at the voltage of the compressor (times the square root of 2). I am not sure what other intermediate voltages there might be. My other thought is that if they are going to build a circuit like they will try and keep the voltages low to reduce component costs. Unless the compressor becomes alot cheaper at higher voltages this would be the optimal strategy.

 

If I can figure out the voltage of the compressor then I can probably assume that a DC voltage about 20% more than the peak voltage should be sufficient. I just don't have an inverter air conditioner that I can tear apart to look at. I was hoping someone here might have tried something like this before.

 

Am I incorrect in how I envision the system would work?

 

 

[Crossy]

@Monomial you obviously have significant technical knowledge (I admit I was trying to frighten you off initially). But I think we are on a hiding to nothing here.

 

A couple of things I've found.

 

 

 

Note that we have active PF control. The brains of the whole thing is the IRMCF312 which takes input from the incoming mains to determine when to fire the front end IGBT. I'm not sure how it will react to not having the mains triggers (could be buried in the data sheet somewhere) it may shut down completely.

 

Data sheet here datasheet(1).pdf I have a feeling it's going to be way too smart for us mortals to defeat.

 

This is way more complex than the SMPS's I was running on DC (they had no micro-controllers, relatively simple analogue beasties).

 

Probably best to get one of the designed for DC units.

 

 

 

 

[al007]

25 minutes ago, Crossy said:

@Monomial you obviously have significant technical knowledge (I admit I was trying to frighten you off initially). But I think we are on a hiding to nothing here.

 

I would like to echo what Crossy said above

 

When I rebuilt my boat it was 20 yrs ago and technology has changed very substantially, and air conditioning would despite trying even with smaller units only run on generators

 

it was also easy to fry large expensive (6000$us) battery banks

 

However do not necessarily be put off, I think maybe you are on the wrong forum

 

You need young up to date electrical engineers advice who are familiar with large solar installations, I am guessing you are maybe talking in excess of 50 solar panels, and back up generator of say 20/30 kw

 

Do not be put off, but to achieve you aims maybe you have to have a very deep pocket

 

It sounds to me you are trying to build a facility for some secret process, remember the solar panels will still be seen from the air

 

In any case good luck, with money and effort in todays world I am sure you can achieve your goals

[Crossy]

4 minutes ago, al007 said:

It sounds to me you are trying to build a facility for some secret process, remember the solar panels will still be seen from the air

 

I must admit that the logic of building a facility that requires tight environmental control in a hot location where there is no mains power does intrigue me somewhat.

[al007]

9 hours ago, Crossy said:

 

I must admit that the logic of building a facility that requires tight environmental control in a hot location where there is no mains power does intrigue me somewhat.

I have been puzzling this all day and I enjoy the challenge

 

Further clues are using cambodian electrician, who may not use breakers, why the cambodian they are maybe not the worlds best

 

 

On 7/7/2017 at 2:03 AM, Monomial said:

but I also have to use only cheap components that are readily available....meaning no imports of fancy solar equipment.

 

So we are trying to do something very state of the art and high tech, but limited on funding , this is neither logical, nor practical

 

On 7/7/2017 at 2:03 AM, Monomial said:

I have checked on the internet but there does not seem to be much information about this.

Again top leading edge projects do not rely on the Internet

 

The above aside

 

I have no doubt there is a lot of information on solar power and storing energy into batteries

 

Efficiency is important but the primary requirement appears refrigeration and air conditioning in a silent hidden location, so initially put efficiency on the back burner if ten to 15% loss just add 10 to 15% more solar panels, thats easy keep it simple

 

Do not at this stage decide battery voltages and configurations, it can all be done later when the equipment is specified

 

Contact some of the major air conditioning manufacturers and tell them you need working equipment than can be run from batteries with inverters, or high DC currents, see what is available, especially with low start up loads

 

Do the same on refrigeration

 

When these two are done then config the battery bank, and that is also simple

 

This looks like an attempt to build some sort of high tech manufacturing in a hidden location, but then in Colombian drug production is done in the jungles with no clean and controlled environments

 

It can not be a hidden location for some top government official or drug lord as they would not worry about costs

 

Is this the topic for a university thesis for students and only academic

 

I think most of us at TV have failed, and not surprisingly

 

The OP was last seen on line some 15 hrs ago have we scared him off, or maybe he is doing a site visit where there is no communication or electricity

 

Are there roads to ship all this equipment in 

 

Have fun and thanks for the challenge

 

Another Thought, could it be a secret hospital doing plastic surgery for underworld bosses

[KittenKong]

My money is on cheese.

[al007]

12 minutes ago, KittenKong said:

My money is on cheese.

Smells correct to me,

 

Also fits low budget

 

I love good cheese, in very short supply here

 

OP still must be on site when i checked activity

 

However good goats cheese can be done without refrigeration or a/c

 

For me jury sent back for further consideration

 

 

[al007]

Monomial still on site visit

 

I hope he has not been inverted

Edited July 13, 2017 by al007

[al007]

Monomial, I trust you are Ok

 

I hope your Cambodian electrician has not with his lack of breakers electrocuted you

 

Let us know you are Ok

 

I continue to be intrigued on what you are building