View attachment 6975

View attachment 6976

I got it to work If anyone needs a good look at those 5 color band resistors I can email them to you. I am working on hand drawing the PCB with the print over it.

 

You have hijacked an ancient thread.

 

hello vks_foe , need some help body... I need schematic of AVR for 10kva 220vac single phase generator... few months ago just being interested about GENSET
to power a household of mine...

 

Plz
Sent me resister colour code

 

Plz
Sent me resister colour code

Is google so hard to use?

 

This is a AVR circuit ive reverse engineered & rebuilt which was faulty on a generator. It may be of some help.

7 years after your post, i wanna ask you some explanations about your circuit plz.
I have a 2.5kw single phase 220v gen (chinese) with a burned AVR.
i want build a simple AVR (am just an electronic hobbist) and i wanna understand some basics of your schema.
Actually, the gen work with the full rectified voltage from the exciter winding directly applied to the brushs springs of the rotor (90 Vdc).
The feed winding (~30Vac is not used actually)
I know this is not a solution, that's why i wanna build you AVR.
------
if i understand your schema, the transistor is biased via the secnd stage voltage divider (Zener + 1k)
is the mosfet is always ON and increase with the increase of the feed voltage?
The gate of mosfet is biased at 10V via zener diode
why -rotor is connected to the base of transistor?

thank you in advance.

Apology for my english, is just my third lang

 

Im not an engineer so im not realy able to describe mutch to you. But the regulator uses the 30V dc as a reference to control the Rotor voltage. The R2 & C2 are as far as I know a feed back circuit that causes the circuit to oscilate ( Crude PWM circuit)

 

Thank you for your reply Sir.
any other precision will be usefull for me.
did you already build this AVR and get it working smothly?
What is the frequency resonance of RC ?
If this is a resonance, how the duty cycle is affected by the feed voltage ?

good day for all

 

This circuit does work as it what the commercial AVR unit that's on generators. I pulled one apart & copied out the circuit. They were used up to 5Kv generators. They are available relatively cheep on Ebay.

 

Many Thanks for your reply and help Sir.
I will design and analyse how the circuit work on proteus.
Am from North Africa, eBay dont deliver tout my country
I wanna understand circuit logics before building it.

 

Hello @xing tao,

I want to buid the same type of AVR circuit as my original one has just died. I'll try to explain how this circuit works.

Once you have the exciter voltage available (90VDC in the diagram above), the power Mosfet's gate is biased through the R6 (15K/2W) resistor and the voltage limiter - the D2 zenner diode (10V). At this point, the Mosfet turns ON and the field (rotor) coil is energised.

When the feedback voltage on "top" (cathode) of the D3 zenner diode reaches 6.2V+0.7V, the BJT will start to saturate and it will shunt the Mosfet gate (hence the Mosfet will be switched OFF).

While the Mosfet is in OFF state, the field coil current keeps running through D1 (the freewheeling diode).

The adjustable voltage divider (R3 - R5 - VR1) is used to set the generator output voltage.


PS: Sorry for these bare explanations - english is not my native language.

 

Oh, I forgot about the R2 (10K) - C2 (3nF) circuit. It introduce some sort of hysteresis to the Mosfet gate control signal.

When the Mosfet is conducting (ON state), the voltage across C2 is kept to 0V (the Mosfet drain is almost at the ground level). When BJT starts conducting, the Mosfet is turned OFF thus its drain voltage rises to 90VDC + 0.7VDC (D1). At this point, we have another voltage divider (R2 - R4) which are keeping the BJT in conduction mode (hence keeping the Mosfet in OFF state) even if the feedback voltage is bellow 6.2V + 0.7V.

Now C2 will start to charge through R2 to the point when the R2-R4 voltage divider will no longer keep the BJT biased.

When the Mosfet is turning ON again (due to feedback voltage), its drain falls to ground hence the BJT is further turned off by the "reversed" voltage divider (R4 - R2) and the voltage across C2.

That's it, the R2-C2 time constant is very important for a proper circuit operation.

 

Thanks you so much for your explanations so helpfull.. i will take some Time to re read your reply and see if i Can all this ...
Let me know plz if your future homemade avr work After you building it.

Thank you again for your interest for my post however the topic IS so old

 

You're welcome. I'm going to buid the AVR in the next few days (I've just ordered the Mosfet and the high voltage electrolytic capacitor).

Btw, my generator has different feedback and exciter voltages available (around 17V and 110V respectively). Anyway, you can choose some slighty different values for R3 - R5 - VR1 to overcome this situation.

I hope it's not a problem if I post some related links for a similar circuit (they use a BJT as a power switch though).

http://portable.generatorguide.net/avr.html

https://ludens.cl/Electron/AVR/AVR.html

Anyway, I still prefer the Mosfet variant as it's more efficient (and it's using fewer parts too).

 

Here's another discussion about the AVR circuits:

http://cr4.globalspec.com/thread/67923/how-to-build-an-avr-for-a-three-phase-generator?pg=2&Pg=2

but, as far as I read, they don't use any positive feedback circuit (the R2-C2 circuit mentioned above) hence their Mosfet it's not working in switching mode (and that's the reason of the increased power dissipation).

 

Oh, I forgot about the R2 (10K) - C2 (3nF) circuit. It introduce some sort of hysteresis to the Mosfet gate control signal.

When the Mosfet is conducting (ON state), the voltage across C2 is kept to 0V (the Mosfet drain is almost at the ground level). When BJT starts conducting, the Mosfet is turned OFF thus its drain voltage rises to 90VDC + 0.7VDC (D1). At this point, we have another voltage divider (R2 - R4) which are keeping the BJT in conduction mode (hence keeping the Mosfet in OFF state) even if the feedback voltage is bellow 6.2V + 0.7V.

Now C2 will start to charge through R2 to the point when the R2-R4 voltage divider will no longer keep the BJT biased.

When the Mosfet is turning ON again (due to feedback voltage), its drain falls to ground hence the BJT is further turned off by the "reversed" voltage divider (R4 - R2) and the voltage across C2.

That's it, the R2-C2 time constant is very important for a proper circuit operation.

Sir How much voltage shoud i expect on the base of bjt?

 

If you refer to the reverse polarity base-emitter voltage, this situation occurs when the Mosfet turns ON only.

At that moment, the voltage across C2 (90V maximum) is applied across the R4-R2 voltage divider, thus only 1/10 of this voltage is (reverse) applied to BJTs base-emitter jonction - so that's around 9V).

Actually, you don"t need a high voltage BJT, as its collector-emitter voltage is protected by the zenner diode (10V).

Of course, that don't take into account the parasitic spikes or any other transients (I'll try an in depth analyze of the whole circuit if you want).

 

Btw, a Vds rating of 100V for the Mosfet is much too low (you need one and a half or even double the 90V voltage).

 

I will salvage some components from a pc PSU and an unused UPS. I hope find some usable parts for this project. WhatsApp do you think?
Thank you again for your Links and your clear english explanations

 

Btw, a Vds rating of 100V for the Mosfet is much too low (you need one and a half or even double the 90V voltage).

Mosfet on diagramme irfs640 IS rated for 200v vds and about 30v vgs