Synchronous Generator AVR Replacement

I can't answer all of your question fully, because I am not expert in how to size your AVR. But I know how the generator works, so I will answer your questions based on that:

1) Yes the regulating field should be in DC current, AC is never applied only in case of a brush-type exciter which is rectified by the wheeling diodes to DC eventually.

2) No idea, but I know that the AVR feeds voltage into the exciter and the current is generated based on the impedance of the exciter winding. So if you inject current instead, the voltage generated will be based on the impedance of the exciter. So if you want to know, measure the impedance of the exciter winding and calculate it.

3) Well, the damage can be on the AVR side and the exciter side, the exciter windings could be damaged from the high current.

4) Well of course it can be possible to supply the field from a DC battery which is called field flashing. But I am not sure about the value of the DC battery you need to use, 24V might be enough to build-up your output voltage during generator start-up. Your method of sizing could work, but I don't recommend it. I would say bring a vendor or the manufacturer and let him work on it, you don't want to start something and end up burning your exciter and that would be costly.

From the referenced catalog:

"The output voltage of the exciter (4) is regulated by two separate excitation windings. One of these windings (5) is supplied from teh compound transformer (8) and the additional resistor (9), while the other winding (6) is supplied from the voltage regulator (11). Compound transformer ensures compensation of the voltage dip while additional resistor ensures safe and reliable self-excitation. Direct regulation of the main generator output voltage is ensured by its own excitation winding (2) and rotating rectifier (3) of the exciter. The voltage regulator (11) ensures regulation of output voltage to the exciter to kip [sic] output votage of the generator constant and irrespective of the load."

This basically describes the function of the AVR. Unfortunately, the diagram and information in the document you referenced is insufficient to determine what that voltage should actually be. Apparently, the whole system relies on residual magnetism to get it up to output voltage from standstill, but it is hard to say. Did you not recieve any maintenance or operator's manual with the equipment?

I suspect you would be risking more damage to the exciter windings if you tried to fit a third-party AVR without knowing what the specifications on the original equipment are. Have you contacted the manufacturer in Yugoslavia?

Hi bibo, Cwarner, thanks for taking the time to take a look at the post and answer, I appreciate it very much.

@bibo - thanks for the reply and for the clarification. Regarding the current I seem to remember I read somewhere that for a single-stage exciter the ratio between generator rated current and exciter's output current should be somewhere in between 1:20 and 1:30. So for a 80 Amps of the generator, the exciter should have 4-5 Amps output at most. Is this correct, or I've just misunderstood what I've read? Based on this I'm keen to think that a 10 Amps AVR should suffice?

@Cwarner - I've contacted Sever, even moreso since I also live in former Yugoslavia (Bosnia, Sever is located in Serbia). The problem is, as most factories in ex-Yu Sever is pretty much going down fast (they're down to 300 workers, while they had ~3000 some 20 years ago). The older engineers have mostly been fired, and the one I know and I'm chasing after is on vacation at the moment. The additional problem might be that the whole genset was manufactured for the army, so I wouldn't be surprised if it turns out there's no documentation at the factory, or it's missing, or something similar. I don't have any documentation except for a general wiring diagram which has no information about the voltage level of the AVR output. The AVR has two sets of inputs: one is from the generator's terminal box (3x380 it would seem), and the other set has three connections marked 3x38V (not marked AC nor DC, and not shown on the diagram where do they come from).

One more thing...since the compound winding of the exciter is powered from a PCB that's still there, and working, and the generator picks up to rated voltage and frequency when started and no load is connected I could measure the voltage on the compound winding's. Does anybody know could this voltage possibly be used for determining the voltage of the regulation winding?

Best regards, Miki.

It seems the ratio you provided is more of a rule of thumb, but I can't really help you with it, because my knowledge is based on practical experience from working with generators and I never designed one. But I will look into this because it is interesting and I will come back to you about it.

Regarding the PCB that is connected to your exciter.. I think it is a comparator which compares the voltage obtained from the output terminal with a fixed value and then send gate signals accordingly to switch a DC-DC thyristor converter to control your excitation voltage in accordance to that fixed value (it could be varied with a potentiometer). This is just an assumption.

I think this PCB is all your generator originally needed and the actual concept of an AVR was not implemented just this comparator PCB module and that explains why you are able to reach the rated voltage in startup. So I guess now what you are trying to do by adding an AVR is actually a new modification on your generator and you need to remove that PCB when you able to determine and buy the new AVR (as a whole set). One way to determine your AVR ratings I believe is well explained by our friend RAMConsult in his comment below.

I would like to also add something, pay great attention about loading up your generator and how it reacts to the added load, does your voltage collapses when you add a huge load or it collapses when you start adding your load slightly. Also what type of load connected to the generator. Maybe your generator is tripping due to huge motor startup, because startup current drops the voltage down causing the generator to trip. In this case even if you bought a new AVR, it may not solve this problem and you could have just solved it by simply adding auxiliary circuit to boost the excitation current during this situation. All I am saying is for the time being since you have this PCB connected, try to monitor your generator, see how it reacts to different situation and assess your outcome. You may not need to go through the hustle of buying a new AVR.

Let us know what do you think.

Your best bet is to isolate both the compound and regulating windings so that you measure their resistances. If they are both the same than you can safely assume that whatever you measure on the compound also holds for the regulating winding. However since the compound winding supplies the base level of excitation while the regulating winding provides incremental excitation I suspect that the regulating winding has a higher resistance. In either case Ohm's law tells you the maximum current for the voltage applied.

If I were there I would use a continuously variable DC power supply with current control/limiting whose output is completely isolated from the ground. Get the system up to speed and record the no-load terminal voltage, start your dc supply at zero and slowly increase the voltage/current and see what effect it has on the terminal voltage (don't go more than 5% over the base). Back the DC supply back down and incrementally load the generator, then slowly add the DC supply so that you can see how the terminal changes versus the current injected. You should get a nice linear relationship between the DC parameters and the generator terminal voltage at various load points.

Your biggest danger is not the harming of the regulating winding, but instead is overexciting the generator if the load suddenly goes from full load to zero because your DC supply will not know this and continue to supply enough now extra current to overvoltage the generator. However if you take the maximum current from the 5% test and set that as your current limit on the DC supply you should be ok.

One more thought, make sure that you first check that the regulating winding is not grounded or damaged. The reason that the AVR module was removed may be that the winding might have been damaged by the original operators.

Let us know what you find.

Your item 4) is possible and can help you determine the current flow that the AVR will need to handle for different loads, especially for the maximum. You will need to round it up.

Regarding voltage of the AVR: It only needs to be able to work with the Main voltage available from the main output. Thus, if the AVR is connected between 2 phases, at 380 or 400 or 440 V, then that is the voltage of the AVR(minimum).

The Current: the AVR will output a voltage from 0 (+) to a maximum available from the main output, in order to drive enough current into the exciter regulating field winding to produce the required set output voltage up to the full load of the generator. Therefore, if the AVR is OK and the Field winding is sound, there is no need to worry about the current. In general, this current will be in the region of 10% of the main output current: in your case, In= 79.4A, take 80A, then the AVR should be able to handle ~8A to 10A output, without worrying about the exciter winding. it will self regulate as stated before. You should not have an AVR less than 10A capability output. The voltage, as said before, will be you generator output voltage, either Phase to phase or Phase to Neutral if single phase. You have a 3 phase 400V generator. Check how the AVR is connected (or should be connected) on the input.

I hope this will be clear.

A bit behind on the posts, sorry guys. And thank You all for taking an interest, and finding the time to help.

@Bibo...

Thanx again for taking the time and interest. Please post if you find anything about generator/exciter current ratio if you find the time.

As for the PCB present, I'm pretty sure it's a rectifier (if you take a look at the brochure I've linked in my original post it's the part between the compound transformer and the compound winding). Anyway, as I mentioned I do have one part of the wiring diagram, I'll scan it and upload it tomorrow. I'm still pretty sure I need an AVR, but as I said, this is the first generator I've ever looked at , so I'm a total newbie :)

@ RAMConsult

Thank You very much. I am planning to take the generator for a few days and test it thoroughly, but I'm trying to find out as much as I can prior to this to avoid 2 things...

1) I'd like to collect as much knowledge as I can so I can avoid the possibility of kicking myself a few days after I buy it/return it because I didn't try/test something;

2) I wouldn't like to burn or damage something in the process :)

Anyway, You mentioned a lot of guidelines for testing and I appreciate it very much. Will definitely print them out and use them during testing.

@LAA Lucke

Many thanx to You too. Mostly everything clear. Just to clarify the procedure step by step once more:

1) Start the generator with no load, voltage build up, regulating winding not powered;

2) Connect a small load; output voltage drops, I feed a small voltage to the regulation winding in order to get the desired output;

3) Repeat step 2 to full load;

4) Take all the measurements i need in the process;

5) Now start decreasing the voltage fed to the regulating winding step by step. First decrease the voltage, then disconnect one of the loads in order to avoid overexcitation;

6) Repeat 5 until I shut it down;

I just think we have a small misunderstanding regarding my concerns about voltage. The voltage supplied to the AVR is not an issue - I plan to buy a new one and I'll take a look at the datasheet to make sure what range can it take in. I was referring to the maximum voltage the AVR can generate on the output, since the one I was looking at has a max. 100V DC output (I'll leave a link at the end of the post). Anyway, if I do the tests as suggested and measure the full load current of the regulation winding I suppose I can measure the impedance of the winding and apply the Ohm's law to determine the max. voltage across the winding?

Finally, here's the link to the AVR I was looking at:

http://beltramecse.com/english/dettaglio_categorie_prodotto.php?id_parent=5&parent=5&id_cat=6&offset=0&id_prodotto=4

Best regards,

Miki.

After looking at the submitted drawings and brochures, and in reply:

The Measuring of the Voltage output from the AVR: Your method will be OK to give you an approximate idea of the voltage. But you could use a Voltmeter directly and measure it. It is a DC voltage but might have some ripples, therefore, take the peak values. Also, because of the wave shape, the Impedence is not a pure resistance to measure as such.

Also, as an approximate gauge to whether your 100V will be enough to drive 8 to 10A maximum into the field winding of the exciter, if the pure winding resistance is less than 10Ω it will be OK. The AVR will adjust the voltage output to drive enough current so that the main field will be strong enough to generate the required voltage at the specific load. Your worry will be to have enough output voltage to drive the exciter. A higher voltage rating for the AVR is not detrimental. the 100V rating looks very good to me.

The Compound field (since this was mentioned ...) is mainly to compensate, quickly, for sudden load variations. Its source of power is normally extracted from the main load current flow! therefore, at the start of operation, there will be no current flowing into the main load (negligeable to 0), and thus no voltage on the secondary of the compound transformer. This transformer will become active when the load start to draw significant current.

The AVR normally takes care to send all the generated power from the remanent magnetism directly into the exciter regulating winding. If the resistor (9) is really available on your set, this will also help.( but I am not sure if this resistor is connected on your unit?). Its value ??? its use is a little strange to me: At the start, you need all the juice you can get to chain reaction the buildup of the main field, and later give control to the AVR. If this resistor has a very small value for this purpose, then it will be a problem when the full voltage is achieved, but if it has a relatively high value then it will delay the buildup .... hence the AVR will be more effective ...

After checking the AVR data sheet it looks like it will work fine, but I noticed t doesn't have any Droop adjustment. This won't matter if you are not attached to any other generation on your system, and if you are you can always use the Cross-Current Compensation that hopefully is still part of your generator. Once you figure out the resistance of your regulating winding you'll know whether the 100V limitation will have any effect, but I doubt you will need to go above it.

Yes I believe this AVR is good and it has the basic control features and specially over voltage and under frequency protection.. But from it's drawings you need to remove the PCB because this AVR has its own rectifier. So all you need to do is to confirm the numbers and you are good to go.. Never run your generator with any other generators because this AVR doesn't have droop control as RAMConsult mentioned..

I will look into the generator/exciter current ratio and I will let you know.

Take care

Now You've got me a bit confused...I was thinking of leaving the PCB for the compound winding + installing the AVR for the excitation winding...

Anyway, I'll try to upload the diagram I have tomorrow morning (it's around midnight in Bosnia) so I believe some things might be clarified additionally.

As for droop control - is that something about synchronization - having multiple generators connected in parallel? If that's it, You don't have to worry, I wasn't even remotely planning to do that, I'm way in over may head with the AVR :)

Regards, Miki.

I see I am sorry I guess I missed few points about your generator set, I didn't realize your PCB is actually connected to a compound winding and not the excitation winding.. I personally haven't seen an AC generator with a compound winding but I know compound winding is used in DC generators.. Anyway, a diagram will be really helpful..

They are two separate systems and both are required for regulated operation. The compound winding provides the bulk of the current, it is unlikely that the AVR could handle the necessary current to get the generator to rated voltage.

Don't be confused, it's just a matter of inconsistent nomenclature. As used here the compound winding is a feed-forward loop that provides the base level of self-excitation; i.e., at start up it uses the residual magnetism in the rotor to build up (or compound) the terminal voltage, so that the small voltage gets rectified and fed back into the compound winding which leads to more field current which leads to more voltage, etc until the rated terminal voltage is reached.

The regulating winding should be thought of as the fine tuning control. It is a feed-back mechanism that reacts opposite to what the terminal voltage is doing; as the voltage rises above the desired setpoint it will provide less (and in some systems, negative or bucking) current into the regulating winding, if the voltage falls below the setpoint then more (positive or boost) current will be produced. The rotating exciter windings don't care, they simply see the net or resultant magnetic field and feed into the rotating diodes for conversion to DC for the main field.

So don't disconnect anything associated with the compound module, just add the AVR module, but be forewarned, any polarity errors on the wiring will result in very obvious misoperation.

Ok that is quite interesting, I guess generators can be build in my different ways.. The generators that depends on residual magnetism that I have seen uses the same winding as the exciter regulatory winding, No separate (compound) winding exists.

Thanks for the explanation, I wasn't aware of such system.

Here's the only document I have. I've translated the parts I think are important and added some remarks.

Hope it helps.

http://www.mediafire.com/view/?7nhfua2ayimfypx

One thought though...could it be that the three inputs on the AVR (TRG-05) connected from the generator's terminal box are sensing inputs, and the 3x38V inputs are power inputs of the AVR? In this case I suppose that the voltage output of the AVR is limited to 38V max?

Thanks for posting your drawings.. TRG-05 is indeed your missing AVR which was an IGBT controller, and it seems it is taking a voltage of 3 phase 38V from an external source as a power supply but that doesn't mean it is the max required voltage.. What I can see is that your excitation voltage is taking directly from the generator's terminal U,W & V.. It goes to L3 (7,4 & 1) and you can notice that there are resistors in series (r7, r8 & r9) to drop the voltage down to the required voltage level before it is rectified to DC and goes to the compound winding.. So you can expect similar thing for your new AVR, the AVR will take the same input from L3 (7, 4 & 1) which is equal to the terminal voltage and use some series resistance to drop it to the required level before rectifying it and controlling it. Is the value of the dropped voltage in the PCB equals to that of the new AVR? maybe if the compound winding's impedance is equal to the impedance of the regulating winding, or maybe not because I am afraid the AVR's purpose in this case is only to supply small voltages to the regulating winding to adjust the output voltage (unlike the PCB) which will be a good news for you because you will know that the required voltage is much lower for that in the PCB (I am not quite sure about this point I am just suspecting):

If this is true, here what I suggest you need to do:

1) Run your generator at no-load with full output voltage

2) Measure the voltages at L5 (2, 6 & 10) to see how much the AC voltage had dropped from the rated terminal voltage before it is rectified.

This way if the PCB measured voltages at L5 is 100V little more or less you are then in the safe side to buy that AVR. Otherwise, I am not sure what your next step would be. But if it works, once you buy the AVR, if its input circuit can only take a voltage of 100V then you may use a voltage transformer or a series resistors to drop it to that level. I hope I got that right, wait and see what RAMConsult will say about this.

One more thing which I didn't get from the drawing, what is the purpose of the transformer KD and the terminal box L9, I just don't see how they are used in the process.