0
\$\begingroup\$

For an experiment on photosynthesis, I would like to amplify a 0-10V 100Hz to 1µHz (~a day) sine wave from a function generator to drive a 24V 3A LED module.

I already did that with a square signal and a MOSFET as amplification system and everything worked fine.

For the sine wave, I can't use MOSFET, so I was looking at OPamp but I read in some various places (here for example : Very low frequency instrumentation amplifier) that, as far as I understood, classical OPamp based systems were not well suited for low frequency and DC amplification.

I also thought to an other option which could use a higher PWM base frequency and then modulate the duty cycle to simulate the sine wave.

Would you have any advice to guide my search ? Thank you !

\$\endgroup\$
2
  • \$\begingroup\$ You can use MOSFET with PWM. It's pulse width modulation, so modylate with your sine wave. Nice way to do it is by filtering out the PWM with inductor, but in some range even that is not required. \$\endgroup\$
    – user76844
    Commented Apr 27, 2018 at 16:56
  • \$\begingroup\$ sounds like a darlington bjt would fit the bill. if i want to dim without pwm, i use a darlington to create a voltage follower (series pass). This regulates pretty well for lighting applications, relying on the raw high HFE and a known load instead of an opamp. be sure to heatsink. \$\endgroup\$
    – dandavis
    Commented Apr 28, 2018 at 10:02

3 Answers 3

1
\$\begingroup\$

Classical op-amps work just fine down to DC. However the input noise does increase at low frequencies. At sufficiently low frequency, we stop calling it 'noise' and start calling it 'drift'.

If you want an amplifier with the lowest possible noise down to DC, then use a chopper-stabilised op-amp. This employs an auto-calibration loop round two amplifiers internally, to null out the drift as it happens.

However, it's quite unusual to have an application that is so sensitive to drift that you need to go to the expense of this type of amplifier. A good compromise between the el-cheapo amplifier and the chopper stabilised is a 'low drift' amplifier with a specified input drift and input offset voltage tempco.

\$\endgroup\$
1
\$\begingroup\$

For an LED the op-amp will work perfectly down to 'DC'.

Your problems, if any, will likely have to do with the power supply and thermal considerations. Linear control of a 72W load could require the amplifier to dissipate quite a bit of power, depending on your supply voltage and what the LED I vs. Vf curve looks like. It's definitely not going to be a small chip, it will be something with a large heat sink and perhaps a fan.

You mention a sine wave, which is bipolar. To state the obvious, LEDs only emit light, they can't emit dark or suck light, so you will, at best, be dealing with a sine wave + offset voltage so the sum stays greater than or equal to zero.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks ! Briefly why working with a LED is ok for the op-amp ? And about the power, I could find a module with less consumption but heat management is not a problem for this experiment. Do you think this link may be useful for my application ? dmcinfo.com/latest-thinking/blog/id/9462/… \$\endgroup\$
    – Remy F
    Commented Apr 28, 2018 at 9:18
0
\$\begingroup\$

The ideal way is to use a MOSFET controlled boost regulator with current sense feedback . With PFM and PWM the dynamic range of current intensity can be controlled by the resulting out filtered DC current limited only by the linearity of the emitters and of course the plant. For further accuracy a photo diode array feedback could be used at the desired target.

Synthesis of the current control input is easily done commercially with 10~100% range but a special design can achieve 0 to 100%

It is irrelevant to discuss Op Amps here since there are many ways to achieve this after specs are defined.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.