opamps coming off the rail

May 16, 2013 27 Replies
S

Don't know about fast, but clean:

LT1678 - Dual/Quad Low Noise, Rail-to-Rail, Precision Op Amps Features

Rail-to-Rail Input and Output 100% Tested Low Voltage Noise: 3.9nV/rtHz Typ at 1kHz 5.5nV/rtHz Max at 1kHz Single Supply Operation from 2.5V to 36V Offset Voltage: 100µV Max Low Input Bias Current: 20nA Max High AVOL: 3V/µV Min, RL = 10k High CMRR: 100dB Min High PSRR: 106dB Min Gain Bandwidth Product: 20MHz Operating Temperature Range: -40°C to 85°C Matching Specifications No Phase Inversion

Vote
Twitter Facebook LinkedIn
M

Are you saying the LM324 really is an op amp?

Vote
Twitter Facebook LinkedIn
T

Mirror2.JPG

What kind of gate capacitance are you talking about? 1us settling implies a pretty high bandwidth to the circuit -- you're going to need a low output-impedance op-amp (and possibly a high-frequency high-power op- amp) to get any kind of bandwidth.

(I've done this once with IGBTs. 'taint easy when the power levels go up and the speed remains high).

Tim Wescott Control system and signal processing consulting www.wescottdesign.com
Vote
Twitter Facebook LinkedIn
J

The IXYS part has something like 26 nF Cgs, astonishing. That seems to be the price you pay for a huge SOAR.

1us settling

For the initial application we could live with maybe a 5 us current risetime at

150 amps. But you're right, we'll need a hefty amount of gate drive. Something in the ballpark of 50 mA might do, which leads me back to LM8261 as a candidate. Something beefier would be nice for a standard-product pulser where bragging rights matter.

An opamp and a pair of complementary emitter followers would work, but again we'd need a negative supply. Maybe. Maybe not.

formatting link

Microseconds aren't high speed!

John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom timing and laser controllers Photonics and fiberoptic TTL data links VME analog, thermocouple, LVDT, synchro, tachometer Multichannel arbitrary waveform generators
Vote
Twitter Facebook LinkedIn
J

candidate.

It'll be interesting to see if there exist values which will make it both stable _and_ fast.

And the other issue I mentioned, gate charge plateau, remains ignored in this thread. ...Jim Thompson

| James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Vote
Twitter Facebook LinkedIn
T

Current_Sources/

Big_Sink_1.JPG

It's not just raw current drive to fill the capacitance that you need at your desired bandwidth -- it's a low-enough impedance in the op-amp output so you don't have to cripple the loop keeping it stable against the FET gate capacitance.

(One of those amps designed for a capacitive load may help -- but it may just get really, really slow).

Your circuit, as shown, is going to have terrible crossover distortion, of a sort that's going to make for a high output impedance in the transition from charging to discharging the gate. That, in turn, will get you oscillation or a terribly slow loop.

You might be able to live with a resistor-loaded emitter follower, and just accept that you'll be burning up some watts to keep your output impedance and parts count low.

Controlling 150A at a 1MHz bandwidth is high speed.

Tim Wescott Control system and signal processing consulting www.wescottdesign.com
Vote
Twitter Facebook LinkedIn
J

I've done this a zillion times, and it's not hard to get it to work. All I need is the right opamp.

One of my guys found the AD8565, which looks good. 250 mA peak output from an SC70!

Not really. Gate drive is managable. Cap ESL and fet source inductances don't matter at numbers like 10 to 100 amps/usec. Numbers in the amps per ns are "fast", where every nH makes volts of drop.

John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
Vote
Twitter Facebook LinkedIn
L

mp

nd

V
k

ax

a
e

go

reminded me of this:

formatting link

-D.pdf

describes using a sensefets kelvin connection to "bypass" the source inductance

-Lasse

Vote
Twitter Facebook LinkedIn

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required