Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

3-phase generator current control

Status
Not open for further replies.

jcganley

New Member
Greetings all,

I am working on a somewhat strange project in which I am planning to use the rectified output of a three phase engine-driven generator to charge a large bank of ultracapacitors (20.8 F total, full potential of 389 V, total ESR ~100 mΩ). I will be using a series delta-wound generator capable of producing 266 VAC at 3600 rpm (375 V DC when rectified).

The challenge is to make sure that the DC charging voltage is kept below 50 A at all times in order to avoid overloading the generator (a real problem when the capacitors are at a low charge state).

I had first considered using phase-angle fired SCR to limit the AC current before rectifying, although this approach is expensive:

**broken link removed**

I am now wondering about direct control of the alternator field windings by with proportional feedback control. Basically, this involves controlling the current to the exciter stator. Excitation at zero load is 300 mA, full load excitation is about 1 A. The exciter stator resistance is 16Ω, so I could apply a variable DC voltage to the exciter between 4.8 and 16 V, by referencing the drop across a shunt resistor:

**broken link removed**

So, if I had a 1 mΩ shunt and wanted to maintain 50 DC amps, I would want the DC controller to increase exciter current for measured shunt voltage drops of less than 50 mV, and would decrease it if it was greater than 50 mV.

I already have the shunt and bridge rectifier in hand, but I have no idea if there is a device to control the power in this way. I have looked at programmable DC supplies without much luck in finding the flexibility in application that I'd need. I would hope to find as inexpensive a solution as possible.

This is a bit outside my normal field so I apologize for any silly mistakes, and I appreciate any advice. Thanks!
 
Control of the field is how it's normally done. First take a look at Electronic Components Distributor | DigiKey Corp. | US Home Page and search for ACS750 and LM395T, because these would be the componets I'd start with. The sensor will provide isolation and filtering for 75A so that's done. The LM395T is a Ultr Reliable Power transistor. It has a VCE(Sat) of 2V, so you would need an 18V (2V above 16)and a 5V supply at aminimum. I used a 75A transducer so your not at the limit of a 50A transducer.

An Op amp to scale the current (Scale), a voltage reference IC (Ref) and another OP amp (control). That transistor needs a very minisule amount of current on the base and it's protected all around thermally and short circuit. The control OP amp will have the reference and the setpoint attached to its inputs arranged such that the transistor is modulated until the setpoints agree.

This is just a verbal description and not a design at this point. Take a look at the parts.Find an available one andlook at the datsheets.

I do think that one thing is missing from your conception though. You have to guarantee the field is OFF while the engine is cranking which means sometime after you have oil pressure.

I'm not sure how the RPM of the generator is controlled either.

Usually, there is another winding on the genertor that generates some control voltage from the residual magnetism and this winding creates the voltage for the field. Once the system is producing, those voltages and the availabe currents increase. This also might complicate the control.

If there is a battery, which there probably is, the extra winding may not be used, but something is controlling the frequency. DC to DC converters could be used to generate the power for the circuits to operate.

Remember that there are overvoltage and undervoltage, frequency, water temp,oil pressure that are all part of the system.
 
You can also approach it from a block point of viewand use commercil stuff or make what you need.

Example: You need a wayto proportional control. There are PID controllers that you can configure to operate in P (proportional mode) that can output an isolated 0-10V proportional signal. They can take inputs from a current transducer and some have lots of ranges and flexibility like Eurotherm temperature controllers. These are the "Holy Grail" as far as I'm concerned. Not sure if you can find a 24 VDC controller though.

Inhibit might be done with a relay in the output loop. The oil transducer typically switches to ground when there is low oil presssure and you can "wire or" this signal to operate two things by placing the cathodes of the two diodes at the sensor and one diode to the original controller and the other to yours or you can window detect the frequency of the generator. I think you need something that tells you the generator is up to speed and start an inhibit timer.

You can create a gizmo that takes a 3-10 V signal or a 0-20 mA signal and isoltes it and turns on the transistor that we spoke of, so you have your output.

You can conditon the current signal or try to find a DC current transducer that operates in the correct range and provide that to the controller.

Again, there are just some ideas.

You need a way to excite the field with an inhibit function and there might be anuberof ways to do it.
 
Just idle the generator down to start with and speed it up as the amp load drops.

As far as 20.9 Farads go thats still not much power in comparison to what a 240 volt 50 amp three phase generator is capable of putting out. The generator will see those capacitors charging as being a similar loading effect as what big electric motor creates when starting up under load. No big deal really just a quick power sag for a few seconds as they charge up and level off.

At most just use a large heating element, like one or more from a hot water heater, in series with them to limit the first inrush current.
 
Last edited:
I work it out that 50 A will cause a 20 Farad capacitor to rise at 2.5 V/s, so it will take around 3 minutes to charge the capacitors.

As Tcmtech said, the alternator might be perfectly happy doing that.

There is an approach that I think will be far simpler than a current control circuit. Alternators have a lot of inductance. The voltage that they produce is approximately proportional to the excitation. However, when there is a load, most of that voltage is lost in the winding inductance, so as more current is taken, much more excitation is needed to keep the voltage the same.

This is what the OP, jcganley, found, and it is why alternators nearly always need voltage regulators. When the excitation hits the maximum that the exciter can produce, all you have is a big, fixed voltage generated by the alternator, and the big fixed inductance of the windings. The maximum current is the ratio of the two, so is approximately proportional to the excitation current. If you limit the excitation current, you limit the output current. A few ohms in series with the exciter coil will reduce the maximum current.

Car alternators don't have current limiters in them, because their very design limits the current. If the output load current drops suddenly, the voltage drop due the the inductance of the windings disappears, leaving only the massive generated voltage, which is what causes load dump surges. If there is a battery there, it absorbs the current until the excitation has dropped.

You will have to supply your exciter from some other supply, or your problem with discharged capacitors will be that you can't charge them at all quickly. If the output voltage is low, and that supplies the exciter, there will be no excitation, so no current generated.
 
Is the generator putting out 50 amps at the DC voltage or 50 amps per phase I assumed it was 50 amps per phase? At 50 amps per phase that would give you a 150 amp DC capacity after rectification which would bring the capacitors up very quickly.

The initial inrush current would be high but would rapidly taper off as the capacitor bank charges up causing the overall load to drop back to within the normal working load capacity of the gen set in a very short time much like starting a heavily loaded electric motor does. The engine would likely pull down for a moment but the overall drain would still be in the acceptable limits of what any good gen set should handle.

If not a two stage charge system would also work. Initially the load current is limited by a large resistance until the capacitor bank achieves a high enough voltage and then the resistance is bypassed and the capacitors get charged all the way through a direct connection to the rectifiers. Its a common setup in many large inverter type power supplies that take the AC input and rectify it to DC then stabilize it with large banks of capacitors.
 
Thanks for the insight, all. The capacitor bank is used for storing energy which will be intermittently used, sometimes at 100 kW for a few seconds at a time, but mostly at rates below 15 kW. The engine driving the generator will be rated at around 20 kW (27 hp). Sorry about the current ambiguity; the goal is to keep DC current delivered to 50 A or less, so that the power level stays within the engine's rating.

The initial charging will call for high currents before the capacitors can accumulate charge, but at least that charging is done at low voltages (low power) which the engine might be able to deal with well. Also, the charging from 0-200 V or so is only done once; the capacitor bank is to be used swinging between about 375 V at a maximum down to about 220 V or so, in order to drive the load (3 phase AC motor, controller with 200 V LVC).

The alternator does have a regulator which can be tweaked a bit: (link)

I may be able to work with the voltage rise time control (STAB pot to adjust stability network, page 2) in order to keep output current in the right window. Or, as was mentioned, the engine may just be able to handle things as they are.

So Diver300, your suggestion would be to add a small amount of resistance to the DC feed to the exciter? That is easy enough to try.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top