current sensing

What kind of generator is that- it looks like the field coils and armature are fed in parallel from Batt(+) and the current limit opens the high side armature/field feed while placing resistance in series with the field coil low side return to GND. The low side of the armature is permanently connected to GND. It looks like the old time current regulators routed a portion of the field coil current through the armature current sensor magnet for hysteresis. Your circuit proposals are shunting the armature current away from the battery and doing nothing to limit generator output. Part of the current limit is to protect the generator against overcurrent and your scheme does nothing about that. It would be a shame to burn up that relic. You can use an LM185 to make a precision high side current sensor if you want to stay away from GMR or Hall Effect. But this isn't going to do you any good unless you use it attenuate that field coil current.

Reply to
Fred Bloggs
Loading thread data ...

Gee, maybe if you would have said it was for a BIKE in the first place I wouldn't have hurt your little feelings by asking a simple question.

I can read and know the difference between an alternator and a generator.

Got a real nice book on automotive generators (from 1925), by the way...

Reply to
Rick

Go monolithic reference, then even a cheap pot works with this: View in a fixed-width font such as Courier.

View in a fixed-width font such as Courier.

. .. .. .---------------------. R1 .. | | 1.24V - (Vled-Vbe)*-- .. | | R2 .. ----+-----[Rs]------+-------> I= ---------------------- .. | | | RS .. | | | .. | | | .. | [R1] | .. | LM185 | | .. --- | | .. / \\ -------------+ --- .. --- | \\ / led .. | | --- .. [R4] [R2] | .. | | | .. | | | .. +---------------' | .. | | .. | | .. | | .. >| | .. |-------------------+ .. /| | .. | | .. +---> to comp [R3] .. | | .. [Rc] | .. | | .. --- --- . ..

Reply to
Fred Bloggs

On the contrary, I didn't mean to ascribe any ignorant mistakes to you-just wasn't sure what else you had done, the application, or the way you were attempting to limit the current output....

Reply to
Rick

Confusing the temperature derivative equation on that last one: View in a fixed-width font such as Courier.

. .. .. .---------------------. R1 .. | | 1.24V - (Vled-Vbe)*----- .. | | R2+R1 .. ----+-----[Rs]------+-------> I= ---------------------- .. | | | Rs* R2/(R1+R2) .. | | | .. | | | .. | [R1] | .. | LM185 | | .. --- | | .. / \\ -------------+ --- .. --- | \\ / led .. | | --- .. [R4] [R2] | .. | | | .. | | | .. +---------------' | .. | | .. | | .. | | .. >| | .. |-------------------+ .. /| | .. | | .. +---> to comp [R3] .. | | .. [Rc] | .. | | .. --- --- . ..

Reply to
Fred Bloggs

Yes, it's normally a shunt generator with field resistance control as you describe. Could have up to 3 relays in the regulator-one for current regulation, one for voltage regulation, and a cutout. Some have a third brush on the armature to gradually increase the output as the generator speed increases.

Reply to
Rick

By the way, I wasn't questioning the need to regulate the current output, just the 10A value. That seemed a little low for an auto, but then I didn't know it was for a bike.

Reply to
Rick

Not sure what you are referencing here, but old automotive voltage regulators used 2 relays to regulate generator output: a voltage regulating relay which bypassed an additional series resistance in the field circuit when the voltage is below setpoint, a current limit relay which bypassed a separate additional series resistance in the field circuit when the output current is below the current limit. The field circuit is internal generator V+ to field winding to regulator (with variable resistance via regulating relays) to ground. Some also had a separate cutout relay to disconnect the field or possibly the entire generator V+ connection when the ignition is off or there is no generator output voltage, not part of the regulating ckt.

Reply to
Glen Walpert

Amazing what the old boys could do with a couple of relays, eh : )

The more important point Fred made is the various proposals so far do not limit total armature current, just current to the battery....

Reply to
Rick

Sorry, an LTP won't work. Wasn't thinking. But you could use two PNPs in a current mirror, with separater emitters as inputs.

Reply to
Derek Potter

I used the PNP approach with a low-voltage micropower opamp to grab the current on a 250V supply rail and translate it down to ground level. The transistor (an mpsa92) introduced about 1% error from its base current, matched by 1% tolerances in the three resistors used.

.. +250 --+--+-- Rsense --+----> To Circuit .. | | | .. Rc '----+------ | ---, .. +------ |----, | | .. | _|_ | | \\_|_ .. e\\| / -|--' | /_\\ .. PNP |--< | | | .. c/| \\__+|-----' | Vout = I Rsense * Rout/Rc .. | |____________| .. Rout | .. | | .. gnd ---+--------/\\/\\--------'

If I had used PMOS, I'd have been forced to use a higher-capacitance part and suffered from a slower response, given my puny opamp running from the +250V rail, off of a zener diode plus a resistor to ground.

The technique can be extended to higher voltages with cascode PNPs. Clearly one wants to keep the opamp current as low as possible, to limit the power dissipation in the R1 R2 zener resistors.

.. +600 --+--+-- Rsense --+----> To Circuit .. | | | .. Rc '----+------ | ---, .. +------ |----, | | .. | _|_ | | \\_|_ .. e\\| / -|--' | /_\\ .. PNP |--< | | | .. c/| \\__+|-----' | Vout = I Rsense * Rout/Rc .. | |____________| .. | | .. e\\| R1 .. |------------------+ .. c/| | .. | R2 .. Rout | .. | | .. gnd ---+--------------------'

--
 Thanks,
    - Win
Reply to
Winfield Hill

In article , Winfield Hill wrote: [...]

You can use a PNP if you flip the inputs of the op-amp and the op-amp output works up to the +V.

Another handy variation is to use a small MOSFET for the transistor. Supertex makes some small high voltage ones. The advantage of the MOSFET is that there is no bias current to introduce errors. The disadvantage is that you have to protect the gate against high voltages.

--
--
kensmith@rahul.net   forging knowledge
Reply to
Ken Smith

In article , Winfield Hill wrote: [....]

The LM301's (IIRC) inputs will work up to the positive rail. Its output doesn't like to but you can feed the base through a voltage divider. I used just about exactly the circuit above to bring a current measurement down from an 80V supply.

Either that or use a less puny op-amp and a resistor with fins.

That depends on where you live. Some places you want a hand warmer. :)

I'm going to modify your drawing a little to put in some stuff that people using it should consider adding:

R5 and R6 divide down the drive to the PNP. If there is a massive over current, it protects the PNPs from also being over currented.

R3 limits the currents that can flow into the input pin of the op-amp.

R4 Balances the drop on R3 due to bias current and also allows the capacitor C1 to be added if desired.

C1 slows down the frequency responce and can help with stability problems. Also if the 600V goes to a DC motor, there can be a lot of RF brush noise on it. C1 prevents this from getting to the input of th eop-amp.

C2 should actually be near the op-amp. It lowers the supply impedance at high frequencies.

--
--
kensmith@rahul.net   forging knowledge
Reply to
Ken Smith

There's a movie? Boy, talk about the ultimate geek flick! ;-P ;-P ;-)

Reply to
Rich Grise, but drunk

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.