EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: DanInvents on June 20, 2024, 12:14:52 pm
-
Hi!
For an experiment I'm interested in building a power amplifier that allows me to generate a sine wave signal that is 2A peak-to-peak and up to 30 kHz in frequency. I have been considering using an LM675 for this purpose and I found a suggested implementation for a current source from the datasheet. There they use a sense resistor, but the power dissipated over it is considerable and I would like to reduce that.
I found this post https://electronics.stackexchange.com/questions/666372/how-does-this-current-sink-work-figure-15-lm675 (https://electronics.stackexchange.com/questions/666372/how-does-this-current-sink-work-figure-15-lm675) where they derive the current output as a function of input voltage. I think that if I would use a current sense amplifier, I could use a 0.1 ohm shunt together with 100 gain to have a 1V to 1A (see attached file).
Am I correct in my assumptions?
Thanks!
-
You want a two amp peak but at what voltage? What is the expected impedance of your load?
-
The load is a wire whose resistance is approximately 1 ohm. The drive voltage would be 2 volts peak to peak to generate a couple of amps.
-
The LM675 can indeed be used to drive a current by using a sense resistor in the feedback loop.
The power dissipated in the sense resistor can be significant, especially at higher currents. Reducing the resistance of the sense resistor reduces this power dissipation but requires a way to accurately measure small voltage drops.
-
You said you want a 2 A sine wave (peak? rms?).
The circuit you show will just clip the sine signal to a certain current level.
More information needed.
Do you need current drive or is voltage drive with level control also OK?
It's very diffuse.
-
TS design looks like a Howland current source, which is a typical solution for AC current source. The only mistake I see - it seems opamp inputs are confused (inverting with non-inverting).
But sinse LM675 is not a precision opamp there will be some flaw with DC accuracy, like +-100 mA DC bias current through a load.
If TS need excellent DC precision then you have to use a precision opamp with a buffer (the same LM675 can be buffer too).
I guess LM675 need somefhing like +-9...+-12 VDC supply for this exact load.
You may use two complementary MOSFETs as a buffer too (with some crosstalk distortion).
-
The late great Jim W tells you how.
Op Amp Booster pt1 - LINEAR TECH ampslab.com/RECOMMMENDED%20READ/OPAMP%20BOOSTER/LINEAR%20TECH%20OPAMP%20BOOSTER%20STAGE.pdf
Op Amp Booster pt2 - LINEAR TECH ampslab.com/RECOMMMENDED%20READ/OPAMP%20BOOSTER/LINEAR%20TECH%20OPAMP%20BOOSTER%20STAGE_2.pdf
-
Can you afford wasting a little bit of power?
Then I´d just take a conventional Audio-Amplifier IC like the TDA7379. If the resistance of your 1Ohm-wire is loading the TDA too much, I´d put a 3Ohm-resistor in series. Then you have a load resistance of 4Ohm like a standard speaker. The resistive part of the loads also makes the circuit quite stable, the IC only has to regulate the voltage and the current sets itself with little chance of oscillation or instability.
2A at 4Ohm are 8V and 16W - that´s absolutely in the range of these IC´s. 30kHz is a little outside of the audio-range, but usually these class AB-amplifiers are good for more than 50kHz.
-
No, you need a loss of 10, for VIN=1V, RS = 0.1 and IOUT = 1A.
Since it's AC, can't you use a transformer?
R5 is only for DC stability.
R6 is the AC current sense resistor.
Use a low ESR capacitor. If a non-polarised capacitor with a high enough value is expensive, use two aluminium electro lytic capacitors back-to-back. Use low ESR capacitors. You might need to measure the ESR and adjust R6 accordingly.
[attachimg=1]
-
I bought a complete build amplifier from alie with the OPA541.
bridged the AC coupling input capacitors (no AC coupling at the output)
Adjusted the first opamp amplification so the total amplification is 10x.
So if I set my sig gen at 1V DC the output is 10V DC
If i set it to a 2V square my output is 20V square. According the TI datasheet the output is max 10A peak.
The below unit is the unit I bought, there are a lot of the same units on the interweb
https://nl.aliexpress.com/item/1005002346215365.html
P.S. there is now also a OPA549 that is more beefy
Benno
-
@BennoG That´s exactly what I had in mind. For frequencies below 50kHz this is really cheap and astonishingly okay regarding noise, distortion, overload protection, ...
-
@BennoG That´s exactly what I had in mind. For frequencies below 50kHz this is really cheap and astonishingly okay regarding noise, distortion, overload protection, ...
Bear in mind it has a constant voltage output, whilst the circuit you posted at the start of the thread is constant current. If constant voltage is suitable, then go for it. In fact an ordinary class AB audio amplifier will do.
-
In fact an ordinary class AB audio amplifier will do.
I agree. An car stereo amp can do that job - eg old Sony XPlöd or similar
-
Bear in mind it has a constant voltage output, whilst the circuit you posted at the start of the thread is constant current. If constant voltage is suitable, then go for it. In fact an ordinary class AB audio amplifier will do.
That´s why I proposed switching a small resistor in series to the consumer wire. The higher the linear resistive part of the load is, the more a voltage source will also be a current source.
Of course if you need quick current regulation - e.g. for driving laser diodes - then it´s not working. But for a wire at a fixed frequency it should do.
-
use a 100W hifi amp
-
They make opamps that can output up to ~10A, usually like a TO-220 style for heatsinking but longer and with more pins.
But they can be bit expensive. simple to use though.
-
How about a hearing aid Induction Loop driver? Almost designed for the job, if it can handle the frequency
-
The late great Jim W tells you how.
Op Amp Booster pt1 - LINEAR TECH ampslab.com/RECOMMMENDED%20READ/OPAMP%20BOOSTER/LINEAR%20TECH%20OPAMP%20BOOSTER%20STAGE.pdf
Op Amp Booster pt2 - LINEAR TECH ampslab.com/RECOMMMENDED%20READ/OPAMP%20BOOSTER/LINEAR%20TECH%20OPAMP%20BOOSTER%20STAGE_2.pdf
Following Terry Bites suggestion I simulated in LTSpice the circuit present in Fig.3. After some tweeking (my LTSpice does not include Darlington transistors) I got it to work and the results look great. The current is 180 degrees out of phase (blue trace, the phase is not important in my application) with respect to the input voltage (green trace) and I can adjust the voltage so that there is one-to-one correspondance with a one-ohm load (1A corresponds to 1V input voltage). The author claims that the harmonic distortion is 0.05% an at least by naked eye the graph looks like a perfect sine wave. Neat! Thank you very much!