How much power is shunted.

[seanchan00]

Hi Guys,

I am using two homebrew shunt regulators made from the schematic in the design by Chris Greacen in Homepower magazine. I am wondering each time the shunt turns on at the preset voltage eg 14 volts, what determines the power shunted. The Shunt Load or the power produced by the alternator. From what I can see from the circuit design, it seems to be the shunt load. If I use 100 watt shunt load then each time it turns on the full 100 watts is consumed by the shunt load till the voltage drops below 14 volts. If the alternator produces more than 100 watts the excess still goes into the battery. If less than 100 watts then the battery tops up. This way it is not a true shunt but an overvoltage switch that switches on more power usage when the voltage rises above 14 volts. Am I right?

SeanChan.

[jimovonz]

Right. So if you size your dump load to use all the power that your turbine makes just before furling you should be right. If your dumping while your turbine is at anything less than full output then the dump load will just cycle on and off as the voltage varies.

[Flux]

Yes, the shunt resistor should be capable of shunting the full power at furling point.

When the voltage exceeds the set point then a current equivalent to the full load set by the resistor is shunted until the comparator voltage falls below the set point. If the power is only just over that which the battery can absorb then the discharge pulses will be infrequent. As the input power rises the discharge pulses will happen more often so that the mean discharge current is equal to that required to maintain constant battery volts.

Batteries see strings of variable width discharge pulses more or less as though it was a constant current from a linear regulator maintaining constant terminal volts.

If the cycling of load on and off is done at very low frequencies then there may be some case for worry about it acting as a series of charge/ discharge cycles and reducing battery life, but at higher frequencies it just seems to be working within the surface charge and batteries don't mind. There is in fact evidence that at least for flooded cells this is better than a true constant charge without ripple.

[ghurd]

The Homepower circuit is designed for 2 cycles per minute, max.

It stays on 30 seconds each time the battery voltage reaches the set point.

SeanChan-

I have a 'better' circuit, I believe. Certainly simpler. Not yet ready for everyone, but soon.

I can email it to you if you like.

G-

[hydrosun]

I've been using a modified Chris Graeson circuit for over 10 years. I  changed the timing to  1/1000 of a second. This is fast enough to keep the battery voltage pretty steady.  I lower the set point in the summer to manually compensate for temperature. At my house I have three of these circuits, one for the hot water tank, one for the heater in the kitchen, and a longer 3 minute cycle to control a refrigeration compressor heat pump in the bathroom.  I change the setpoints to change the priority.  I also have a mx-60  auxillary relay controlling some fets to turn on an additional heater under the computer desk.  The home made circuits work well but don't automatically go to float or do temperature compensation.  For other people I put in Trace C-35 for 12 and 24 volt systems and  C-40 for 48 volt systems.  I did make some homemade circuits for the first few hydro systems I put in for others  but  I don't want to have to make repairs when something goes wrong. For my own system I can easily  diagnose and make repairs on such a simple circuit that I made myself  Chris

[seanchan00]

Hi Ghurd,

I would like that very much. Thanks a lot. My e-mail is seanchan00@lycos.com. 00 are numbers zero not alphabet o.

[seanchan00]

Hi Flux,

You always has a much deeper technical description of the situation and I had to digest it slowly, reading a few times before I think I understand your message fully. In fact if a posting don't have a rsponse from you I seldom open it and I usually open a posting just to read your opinion as I always learn something useful.

I have a nagging worry when I set my whole system to work ideally ie put in a shunt load equivalent to the power produced at/just before furling. That of course is near impossible for me in the first place as my wind is rather marginal which I am sure you remember and my batteries are old and quite low in AH capacity. In fact I have never seen the wind generator near furling point in the last two years of using it:) My worry is: what happens when the shunt regulator malfunctions? Will it leave the shunt load on and drain my battery flat thereby killing it? My experience playing around with them has been when the regulator fails it seems always to or malfunction with the shunt regulator set point turned on permanently. I just repaired (with your advice) such a situation by changing the LM 723 and Ne 555 in the failed one and now it is working again. My system is remote(I live 18km away) and I won't see the malfunction till the battery is destroyed. Will I be able to modify the regulating mode by putting a 20amp diode between the regulator and the battery? That way if the regulator ICs eg Lm 723 or NE 555 ever shorts for any reason then the regulator will only be able to shunt what is produced by the alternator and not drain my battery?

Alternator - Shunt regulator - 20amp diode - Battery

Question is the regulator will be subjected to repeated voltage variation from zero to battery voltage then down and up again depending on the wind. Will such a situation take the wind out of the ICs in this regulator thereby burning it out?

[commanda]

what happens when the shunt regulator malfunctions?

The answer to this problem is to use an exclusive-or gate.

One input goes to the Gate of the Fet.

Other input goes to the drain.

In normal operation, if the gate voltage is low, the fet is off and the drain voltage is high.

When the load is on, the gate voltage will be high, and the drain voltage will be low.

If the fet fails (open or short) or the dump load fails open, the output of the exclusive-or gate will go high. Connect this to some sort of alarm device with a mute switch.

This will cover most failure modes.

This technique is used in the dump load controller in my diary.

Amanda

[Flux]

I agree that most conventional controllers do run the risk of failing and running the battery flat. If it is in a remote place then this is even more of an issue.

Amanda's solution should be fine for normal use where the alarm can be noticed. For a remote site the failure signal would need to activate a relay to disconnect the dump resistor.

Your idea of a series diode is perfectly practical and if you use a schottky you will not notice the extra loss that it causes. For the Chris Greacen circuit you would probably be best to power the circuit and take the reference from the battery as before, but connect the positive end of the dump resistor to the windmill side of the series diode.

I think you may run into trouble if you power the circuit from the input to the diode, I have tried this with my boost converters and strange things happen.

Flux

[rossw]

If you have a "bang-bang" controller as some here have called it, why not try this instead.

Use the controller to pull in a decent double-pole relay connected BEFORE the bridge rectifier, that is, on the "raw AC" side.

When the battery volts get high enough to trip the dumpload, the load is applied to the output of the MILL ONLY, and NOT the batteries.

If the mill is making only a small amount of power, thats all that can be shunted.

A failure of the controller cannot flatten your batteries, the worst it can do is stall your mill so the batteries cannot charge (but you'd see no amps flowing and know why quickly enough)

Can be done with a minimum of rework of your existing setup, although it would require one or two more loads (resistors).

Happy to knock up a sketch of the circuit if you need it.

[seanchan00]

Thanks Flux,

As usual you have solved my problem. Why did'nt I think of connecting the positive lead of the shunt load to a series schottky diode before the regulator. Makes perfect sense

I remember when I asked you a year ago whether I can use an old battery as a dump load and you expressed reservations as to whether the FET will be able to switch off because of the voltage it is now esposed to? Can I overcome this with a series diode too placing the diode just before the battery acting as load?

Thanks again.

SeanChan.

[fungus]

I dont quite think it is. The 555 is set up so that it provides a timing period of around 1.1 seconds, with the off time varying with input. I would like to test your design, see if it works well. Email me at the above address.

Thanks,

[ghurd]

Yes, 1.1 seconds.

My fault.  I'm playing with one I changed to operate a motor for 30 seconds.

Sorry for any confusion I may have caused.

G-

[VinceB]

Hey Ghurd,

I would certainly like to see that schematic of yours.

My e-mail is vincevideos@hotmail.com

Thanks!

Vincent