inverter powered central heating

[Stephen]

Hello,

A few weeks ago I was asking here about powering my boiler from an inverter during a power cut. I have given it a go and here are my experiences and questions:

The boiler was fed by a fused connection unit. A poster, sorry I can't remember who, suggested that I replace the FCU with an unswitched socket and plug. He said that this fulfilled the requirements for isolation because the plug could be pulled out of the socket to isolate the boiler. Is the "unswitched" part important? I ask because I didn't have a spare unswitched socket so used a switched one, but I can easily go and buy an unswitched socket if regulations dictate. I'm trying to think of reasons why a switched socket might not be allowed. Does the switch only switch the live or is it a double pole switch? If single pole, I guess the switch cannot be used to isolate if it only switches one side but then a switched socket would still allow the plug to be removed for full isolation.

The boiler is oil-fired; does that makes a difference? An oil-fired boiler uses a pump to suck the oil. Does a gas boiler have a pump or does it just use the pressure in the supply pipe? The boiler manufacturer (Grant) said that the boiler could be used with *any* type of inverter. The boiler is connected to a Horstamn programmer and Horstman valves, as advertised in the Screwfix catalogue. I did contact Horstamn about using inverters with their products but never got a reply. The CH has a Wilo pump and sadly, Wilo did not reply to my questions either.

When connected to the modified sine wave inverter I had two problems:

1. the controller has a back light dot-matrix LCD. When the inverter was used the back light and the LCD flickered. Also some vertical lines appeared on the LCD. Clearly the display did not like the waveform from the inverter but this seemed to be just cosmetic; the switching functions worked as intended.

1. the boiler buzzed a lot. I think it was the boiler rather than the circulating pump but I will double check. I thought it was the oil pump or fan (or both). Whereas (1) was cosmetic and did not alarm me, I'm not so sure about this noise. Is it anything to worry about? Is it causing increased wear on whatever is making the noise? It will be only for occasional use but even so, should I upgrade to a pure sine wave inverter? The problem is these cost ten times as much!

I did also try to use lights powered by the inverter and both traditional filament and CFL bulbs seemed to work without any noticeable problems. One old CFL buzzed but new ones did not.

Thanks, Stephen.

[Dave Osborne]

No, I think the OP might have been concerned that you could accidentally switch the switch off and not notice (apart, I suppose from it being cold and there being no hot water).

No pump in a gas boiler.

The boiler

The controller does not have a mains transformer in it. This would normally act as a low-pass filter for the mains and takes out some of the high frequency crap you get with an inverter.

As it is, the controller itself takes very little power and this power is derived directly from the mains in a very crude arrangement. If you want to eliminate the display problems, then you need to insert an isolating transformer between the inverter and the controller. This should do the trick:

( )

Buy a bigger isolating trandformer and power the whole heating system via the transformer! You'll need to get a fix on the total current demand (i.e. boiler on, oil pump running, valves operating all at the same time) in order to get a transformer of a suitable size. Don't buy a transformer too big, cos the inverter might not be able to handle the inrush.

[Dave Liquorice]

Only to prevent "accidental" switching off or mistaken use as an isolation switch. You are obviously aware of what is required for that. I think you can get double pole switched sockets but that doesn't avoid the "accidental" switch off by the phantom switch twidler.

More power required for an oil boiler than gas. I think our oil boiler wants about 150W for the pump/blower the circulation pumps about 30W each, controller I'd not worry about in power terms. So you are looking at an invertor that can provide at least 300W continusly but with a fair bit of "reserve" to start the pump/blower motor.

Things are obviously not overly happy on the invertor you have. Does it have a high enough power rating? Is it a "quality" brand or a something something cheapy from China?

[Stephen]

Thanks for the fast reply!

I thought so.

I was going to ask whether I could fit something to smooth out the wave but then I thought that if a solution existed, surely the manufacturer would have built that in; obviously not.

I hadn't realised that an isolating transformer would be the solution. I have always thought of them as being for bathroom shaver sockets, not as filters. Is it something to do with inductance?

I have had another go and with the boiler switched off and the circulating pump on there is a very faint hum from the pump; quieter than the hum from a fridge for example. You really have to put your ear near to the pump to hear it. The louder buzz comes from the boiler once that is switched on. It's hard to describe how loud that is. It's like the buzz from a failing fluorescent tube. It's loud enough to notice when you walk in the room, louder than a fridge or freezer.

Whilst I can live with the display, it would be nice to quieten the hum. My worry was whether the hum was causing either damage to the pump or causing it to pump less efficiently. I ran it for a few minutes and the rads have got warm.

I think we worked out that the inverter needed to be rated about 200W continuous with 1kW surge. What is inrush? Is that similar to the start-up surge current in a motor? How would I calculate the inrush of a transformer?

Thanks, Stephen.

[meow2222]

someone needs to chill

both those devices will have issues with MSW. The programmer likely uses a CR supply, and these tend to die on MSW waveforms as the resistor grossly overheats. Pumps run hotter and less efficienctly on MSW, but they do run.

trivial

MSW waveforms contain lots of harmonics, so anything prone to hum or buzz will buzz many times louder. That's life.

The suggestion to use an iso tf as a filter strikes me as odd. These would only filter out frequencies so high that the inverter probably filters them out already. And if it doesnt, they wont really have a significant effect.

For low power items like the programmer, just a series resistor would help to reduce the crap and reduce the power diss of the inbuilt resistor in the CR supply.

For things like pumps, fans etc, I wouldnt worry about them. They'll just run a bit noisy. If youre as panicky as you seem you want want to check theyre not running frying hot (ie faulty but still operating) , if they were the MSW just might tip them over the edge.

NT

[Dave Osborne]

Yes, just like a start-up surge current in a motor.

How would I calculate the inrush of

You don't need to work out the inrush of the transformer. Rather, you need to work out the load which your heating system places on the inverter and specify the smallest suitable transformer (e.g. 200-300VA by the sound of it).

One of the things you could do to test the theory is to get two yellow

110V site transformers and wire them back-to-back to make a 230V double-isolating transformer. You can also get 230V isolating transformers (usually in a blue box, rather than yellow). However, such transformers will be between 750VA and 3kVA and your inverter might not be happy with the inrush.

[Stephen]

300W continuous, 1kW peak. No brand name but when I looked, I couldn't find anyone selling branded inverters. Do they exist and what are good names? I think the inverter is ok, it's just that the modified sine wave is not suitable.

[Andrew Gabriel]

It's not in the wiring regs, but I believe it's a requirment CORGI (as was) imposed on gas installers, and it was so they were not tempted to treat the switch as valid isolation for servicing/maintenance. Yes, you can get double pole switched sockets, but there's no way to know that's what's been fitted.

I've done this. The pump makes more noise due to harmonic components. These will also result in more magnetic losses, but as the pump is well cooled by the hot water (strange though this may sound), I didn't worry about the extra heat.

This might be more of a concern with some other components, where the higher power dissipation might not be so well handled. In my gas boiler, that's just the gas valve.

[Stephen]

I wasn't in a panic, I just wanted to be sure that I was following the regs and to understand why the regs required what they do.

Sorry, I'm not sure what CR is short for? I was thinking something regulation but since you mention resistors is it just resistor-capacitor?

That's what I thought at first. I can live with it flickering a bit but if you are telling me that the controller power supply will fail on the MSW then it ceases to be trivial as controllers aren't cheap!

I can live with the noise but, like you said above, I wonder about the reduced efficiency. The power wasted on humming is power that should be spent pumping. Will it have a noticeable effect?

Any suggestions on what resistor I should use? It's a shame Horstmann never replied.

Not panicky; just curious. I can't tell if the pump is running hot as it's always hot, it's pumping hot water ;)

I'm happy for the pump to run noisy as long as it is not causing premature wear and not loosing too much efficiency. OTOH I don't want to spend a hundred pounds on a new controller.

Thanks, Stephen.

[Stephen]

It looks like the jury's out about the isolating transformer. I was going to ask about power factors. If its 200-300VA and if that is roughly equivalent to 200-300W then it sounds fine. Would it be able to cope with any start up surge? The RS web site said something about testing 4kV for 6s, so I should think so.

Thanks again, Stephen.

[Dave Osborne]

Ok, your call.

I was

Ah, you're confusing voltage and current. The 4kV for 6 seconds thing is about proofing the insulation resistance of the transformer and has nothing to do with current surge. The transformer will absolutely cope with the current surge.

Transformers are not 100% efficient. Generally speaking the larger they are, the more efficient they are. Different types are more or less efficient (toroidal more efficient than laminated for example). A

200-300VA toroidal transofrmer is going to be 95% efficient. Power factor is not a significant issue. I would go with "200-300W is roughtly equivalent to 200-300VA" as a no-brainer in this situation.

HTH, Rumble

[Dave Osborne]

Good call. Nanny state at work, then ;-)

[Steve Firth]

Nikkai is a decent brand.

Yes, I use a square wave inverter (calling it MSW is a bad joke, it's a square wave with lumps) in a barn to provide lighting from solar panels/batteries. I've tried it with various electronic items and it killed a Targus PSU in seconds. I really wouldn't use one with any sort of switched mode PSU.

Besides true sinewave inverters and generators with sine-wave output can be had for reasonable prices. I'd suggest getting hold of a two-stroke

600W generator for about £50. It will keep your CH running until power is restored. Or if you want to do a decent job of it get a diesel generator and run it on the CH oil.

You'll get a 600! pure sine wave inverter for £50 if you look around (eBay).

[meow2222]

Yes. The capacitor acts as a dropper, with the resistor used to limit peak currents. MSW waveforms contain way more hf content, making the current such a supply draws very peaky. This causes the power diss in resistors in such supplies to go way up, they regularly die quickly.

relax. Its not worth spending =A350+ improving the efficiency of a small load thats rarely used.

I dont know the power rating of the thing, if its LCD lets guess at

3VA. i=3D 3/240 =3D 12mA

New products will cope fine with a 20v drop. 20v 12mA =3D 1.7k So a 1.5k resistor would be good. Make it a 3 watter.

If your controller ratings are different...

NT

[meow2222]

SMPSUs are inherently suited to MSW, as the smpsu input contains a rectifier and reservoir capacitor(s). However some have a resistor in the way to limit peak current, and this will generally fry on MSW. The solution is to fit a much more powerful resistor in its place and it should all run fine. More like 10w instead of 0.5w.

NT

[meow2222]

That would get you 5-10% less than 240v, which would work, but it wouldnt achieve anything. Mains tranformers wont filter anything out below 1kHz, which is a frequency above the bulk of the harmonics causing the issues. You could make it work by adding a capacitor across the tf secondary, forming an RC filter with the transformer's copper resistance and the cap, but its an inefficienct and overcostly/ bulky solution.

NT

[Dave Osborne]

You may well be right, but then again, RF Crap coming from the inverter may be affecting the microcontroller in the CH controller and this would be filtered out by a tranny. I'm no expert on the frequency response of

50Hz transformers or the waveform(s) of various types of inverters, so I'm not going to argue against your point, but under the circumstances, I would be in favour of galvanic isolation and if it were me I *would* cobble something together and see if made any difference.

Cheers, Rumble

[Steve Firth]

Yes, as usual from you that's an authoritative sounding answer that's useless. Perhaps you can explain how one opens a typical SMPS supplied for laptop (or in the case of the Targus general) use in order to fit a "more powerful resistor"? The Targus unit in question was about 4mm thick and ultrasonically welded.

[tony sayer]

Might he be better off getting a better inverter?..

[tony sayer]

Have you looked at the crap thats -on- the mains these days?..