RF detector with my comparator

[neazoi]

Thanks a lot Dana.
And how about keeping this LED at constant brightness independent of the VCC?
A constant current source?
Can this one be used for the purpose instead of the anode resistor?

 

[danadakk]

Here is what solution would look like with ATTINY85 -

Regards, Dana.

--- Updated Dec 8, 2020 ---

Thanks a lot Dana.
And how about keeping this LED at constant brightness independent of the VCC?
A constant current source?
Can this one be used for the purpose instead of the anode resistor?

Personally I would use a processor, sample a current sense R in series with
LED, and drive the series string with a onboard PWM whose duty cycle I change
in code to implement a brightness control loop.

But yes another bunch of discretes to implement a current source OK.

But your implementation no, there is a direct function of Iled and Vcc -

Regards, Dana.

 

[danadakk]

Simple current regulator diodes -

https://www.diodes.com/assets/Datasheets/AL5809.pdf

You can also use LDO regulators in constant current mode (note the 2 in4148's useds to sim a led) -

Regards, Dana.

--- Updated Dec 9, 2020 ---

The reg I show in prior post is not LDO, just adjustable. LDO a better choice as
it will improve compliance range of current source.


Regards, Dana.

 

[betwixt]

I think we are talking about uA constant current sources here. A 1.5A rated regulator is 'overkill' to say the least and the current at its ADJ pin alone is probably more than needed.

Possibly a JFET or even a constant current diode would be a suitable device to use.

Brian.

 

[danadakk]

No he was referring earlier to a LED drive current source.

So and adj or LDO in a TO92 package or sot would do. If high power
LED then TO220 or DPAK...


Regards, Dana.

 

[avisitor]

Hi neazoi,

Say buddy, to bias the q2 at 90mV always, maybe you could make q2 input a non-inverting summing input kind of thing, or if that can't work, use a non-inverting summing amplifier at the q2 input, and bias the circuit to this permanent 90mV you need with a resistive divider from V+ as one of the summing amp inputs to generate the summing amp permanent output to keep q2 biased?