AVR problem for 3-phase AC generator

[paddy lynch]

Hi,

I have posted this under the heading "electric motors" because I believe it is the closest topic I can find on this forum.
The problem I have is this:
I have a tri-phase AC generator,(6 KVA) with a busted AVR.
These AVR's cost a lot of $, and are very good at going "belly up", so I believe I should use a different approach to the field coil regulation.

I have mechanically added a 12 V car alternator to try and solve the problem, but the no-load - load difference is too big, and the voltage of said
car alternator is too low (in spite of the manual of the AC gen telling me it was 12 V)

So, I need to sense the voltage coming out of the generator constantly and crank-up or crank down the field coil voltage accordingly.
In the commercial version, this is done on one phase only,(with a dodgy triac circuit) hence the fact that one can't load these generators asymmetrically.
For now a similar solution would do.

I can either take the supply for the field coil from the generator directly (preferred) or from a separate field coil generator.

Understanding that this panel will/can't or shouldn't supply me with a "here-copy-this-solution" I would very much appreciate your
advise and insights in this problem.
The original AVR was covered in a black kit, that can't be removed, so there was no way I could study the design.
The Net provided some insights but it seems that proper diagrams of the AVR's are a "closely guarded secret"!
So, if you have a diagram of such a devise, even if it would only be suitable for a certain brand or type or mono phase AC generator
or any other wisdom regarding this circuit, I would very much appreciate your input.

many thanks!
Paddy

[mik3]

You can step down the output voltage of the generator with a transformer and rectify it. Then use an AVR (or whatever microcontroller you like) to measure the peak value of the rectified voltage and make the appropriate adjustment to the field voltage.

[retched]

mik3-
The AVR he speaks of isn't a uC, its a Automatic or Adjustable Voltage Regulator, I believe.

Paddy lynch-
Did you remove the diode-bridge from the alternator? If not, you will only be getting single phase converted to DC.

The three phase from the alternator can be kept as such by, after removing the diode bridge, running the 3 wires out from there. As for voltage regulation after that point, that will stay in phase individually, I am inexperienced.

Am I correct in my assumption so far? You are looking for 3phase regulation?

You want to sense the voltage and in turn adjust the stator voltage to maintain levels on the output, correct?

[eblc1388]

Originally Posted by paddy lynch

In the commercial version, this is done on one phase only,(with a dodgy triac circuit) hence the fact that one can't load these generators asymmetrically.

No AVR in the world can regulate individual phase voltage of a generator, if user load the phases with different loading.

Remember there is only one aspect user can change to control the output voltage of all three phases, i.e. excitation, via the field winding.

And why do you have the impression that commercial AVR circuits are dodgy circuit?

[paddy lynch]

Thank you all for your input:
The reason why I am convinced that the commercial AVR's are crap (pardon the French) is twofold.
First of all I have busted three of them, while treating them like fragile babies! and making sure all the phases were
loaded equally with ridiculously tiny tolerances!

Secondly the manufacturer of said generators, has now moved to permanent magnets in the generators (under 10 KVA)
just to get rid of the angry (future) customers!
This was done after moving the AVR out of their designated spot, because they couldn't stand any substantial T° increase
or vibration (on behalf of the manufacturer)

No, the AVR is no µ controller but indeed an Automatic Voltage Controller

yes I understand that it can never sense on phase an regulate the load for all three in a different way! (as there is only one field!)

The (tri phase) rectifier was included on the board that I have thrown a way (3 times) so, yes it has been removed.

I am aware of the fact that I am going to have to make an AVR, but at least if I have made it myself I can choose components/a design
that is a lot more tolerant that the commercial stuff that costs me over € 100 a pop!(and gets me nowhere)

So a circuit diagram of one(any one) would really help, to get an idea how the commercial vendors tackle this problem.
I have some ideas, but those combined with the knowledge of the standard circuits might get me there.

So far the reasoning I made is this:
-rectify the output (3-phase rectification)
-this voltage is a measure for the AC output
-this voltage is converted to a variable (V to f) AC signal
-the frequency of this signal will determine the output of a power supply (fed by the generated AC)
-this PSU will feed the field coils

further
I might incorporate a µ controller to regulate the whole process and to control some protective
circuitry.
In this way, I might not have to replace the thing every two weeks!


Paddy

[retched]

It may help us if you would tell us the load(s) being incurred.

What are you connecting to each of the three phases?

[paddy lynch]

The load on each of these phases is 1500 Watt, non capacitive, non inductive

Paddy

[eblc1388]

How is it possible that a commercial genset manufacturer can have its AVR failed three times each within such short two weeks period?

I would certainly complain to the manager and demand investigation and compensation for all my monetary losses. Send the failed AVR back for their examination too.

Tell them if nothing is forthcoming from them you will name their product and give details of the failures on a public forum.

[retched]

You should be able to hook standard breakers to the lines to keep the damage from occurring. You treat it as a standard home wiring after that.

If your loads are not changing, I don't see why you would need anything more than a transformer. The genset could be run at full power and the items would use what they need. The breakers would keep excess power from being pulled from each phase, protecting the rest of the unit.

Are your loads AC?

[paddy lynch]

doesn't solve my problem, and although it might make me feel a bit better, it still leaves me with no AC!

I do need more then a transformer, as there is control circuitry that switches loads, so the load is always well balanced, and that can't stand the higher AC when the load is switched of (or not on yet)

Also sometimes a similar load (in nature ) of 3 times a 1000 watt might be used.

So basically it has to do the complete "works of a commercial AVR"

Paddy