Connecting dc generators in series

[Amine Bouabidi]

Hi,
I want to connect multiple small dc generators in series to increase the power , it is possible ? What are the problems that i might face while doing it ?

 

[zoki85]

It's possible, but problems with overvolting or/and overloading some of the machines are very possible too. Must be done with care, with machines of a same kind,voltage and aproximately same rated power. I don't recommend it.

 

[Amine Bouabidi]

Well i won't be having the same voltage because i'll be using wind turbines. Is there any safer way to connect many of them ( series or parallel ) ?

 

[zoki85]

DC generators are connected in parallel. If their voltage differs significantly, converters are used for parallel work.

 

[sophiecentaur]

DC generators are connected in parallel. If their voltage differs significantly, converters are used for parallel work.

I'm not sure that is likely in many cases. If the two output voltages are not identical then current will flow between them or, if standoff diodes are used, the lower voltage generator could provide no current. Generators are a bit more forgiving than batteries, I realize, but parallel batteries give huge problems unless they are very well matched in type and charge level.
Generators in series are more likely to behave well but, of course, the load must be suitable for operating at a voltage equal to the sums of the two generator outputs.

However you connect them, you really need to have good (sophisticated) regulation of the two. In the case of wind generators, the problem is even worse.

 

[zoki85]

I'm not sure that is likely in many cases. If the two output voltages are not identical then current will flow between them or, if standoff diodes are used, the lower voltage generator could provide no current. Generators are a bit more forgiving than batteries, I realize, but parallel batteries give huge problems unless they are very well matched in type and charge level.
Generators in series are more likely to behave well but, of course, the load must be suitable for operating at a voltage equal to the sums of the two generator outputs.

However you connect them, you really need to have good (sophisticated) regulation of the two. In the case of wind generators, the problem is even worse.

DC series wound generators with different emfs can be run in parallel. Equalizing conducting bar is put between armatures of the machines so that circulating current due to difference in emfs doesn't affect the field winding. It gets confined to the armature circuit and the equalizing bar. Main drawback is the efficiency loss.

 

[sophiecentaur]

DC series wound generators with different emfs can be run in parallel. Equalizing conducting bar is put between armatures of the machines so that circulating current due to difference in emfs doesn't affect the field winding. It gets confined to the armature circuit and the equalizing bar. Main drawback is the efficiency loss.

That's interesting. There are so many different approaches to this sort of problem - how is that any better than just using standoff diodes which would prevent current flowing into the lower voltage generator? (or is this 'historical', before they had massive diodes available?)
However, the OP is about small generators (permanent magnet probably) so this may just be an interesting diversion.
I would imaging that the best, modern approach to this problem would be (if the scale of the operation justified the cost) to use DC voltage converters to regulate the outputs to give the required output volts whenever the wind will provide enough power. This could be arranged to provide load sharing, which could obtain optimum power from both generators, when a simple system would only use the one that's, providing the higher V. I should imagine the system would also involve batteries (?).

 

[zoki85]

That's interesting. There are so many different approaches to this sort of problem - how is that any better than just using standoff diodes which would prevent current flowing into the lower voltage generator? (or is this 'historical', before they had massive diodes available?)

Not sure what do you mean by just using diodes there. If you prevent this way equalizing currents, than the voltages won't be equalized. The voltages must be equalized becouse the generators run simultaneously in parallel and are connected to the same bus. If the bus voltage stays higher than the lower voltage generator's output voltage, than this DC generator can't give away power to the network at all (so there's no point in running parallel combination: only higher voltage gen. wil give away the power).
Needless to say, converters are the best way to go today when DC generators run in parallel.

 

[sophiecentaur]

I just meant that if the output emf from one generator is higher than the emf from the other, current will flow back through the one with lower emf - a serious waste of Power. If you just put forward biased diodes in series with each output, you can be sure that the only current flowing will be from the generator with the higher emf. Apart for a 0.7V (or so) voltage drop you will lose no more energy. In both cases, the 'slower' generator will supply no Power but that's the consequence of using a system that's too simple.

A bit of cleverness will allow you to do very much better (up to almost twice the Power output!)

If an appropriate load is used, I should say that Series Connection could in fact be a more suitable answer to this particular problem.

 

[zoki85]

If an appropriate load is used, I should say that Series Connection could in fact be a more suitable answer to this particular problem.

Same machines, relatively small output voltage, small load, and I agree. Otherwise, I don't :D

 

[sophiecentaur]

I don't even think the voltage value is necessarily relevant, actually. The generators are "small" in the OP and that means (??) low power. If the load is heating then its resistance and operating voltage would be chosen to be appropriate for all generators at maximum output. Your objection about 'size' could certainly be relevant if we were talking about ten 240V generators, delivering several kW each and we do sometimes get such suggestions on PF. I would stay well away if my next door neighbour were contemplating anything like that, though.

 

[NascentOxygen]

Well i won't be having the same voltage because i'll be using wind turbines. Is there any safer way to connect many of them ( series or parallel ) ?

Is it possible that with a series connection, when a couple are out of service then you won't be able to use any if your converter is relying on a minimal combined voltage?

 

[Amine Bouabidi]

First, i would like to thank you all for your answers, and considering that the load in my case will be small, i'll go with series op. Question : what type of converter ( or regulator) should i use before attacking the batteries ?

 

[zoki85]

First, i would like to thank you all for your answers, and considering that the load in my case will be small, i'll go with series op. Question : what type of converter ( or regulator) should i use before attacking the batteries ?

If input voltages vary in a wide range, than a buck-boost converter type. For an apropriate choice, parameters to start with are: V_{in_min}, V_{in_max}, V_out

 

[sophiecentaur]

Series connection gives the same problem that the old domestic electricity distribution had - requiring continuity throughout but it wouldn't be hard to arrange a relay system to short out any non functioning generators. It would mean that a single regulator could deal with the charging current in the batteries (OK if all the batteries are of the same type and age) . But there are so many combinations, possible. One regulator per turbine would be a possibility which would give redundancy but it would be hard to decide how to make the charging most effective as you could miss out on some of the capacity at times.
You would need to do a lot of detailed sums, based on the actual equipment you plan to use, if you want an optimal solution. The definition of an Engineer is a guy who can do, for £10, what anyone else can do for £50.