Photodiode Amplifier Sanity Check

[Chet T16]

I have tried to reverse engineer a circuit from an old through hole design and reuse it on my SMD board, i made a stupid layout mistake when drawing the circuit so it didn't work, i think i've now corrected it but because i'm copying the design without any understanding I can't verify that it right before i waste more money on another batch of useless pcbs 

The circuit is an IR photodiode receiver, from what i have seen in testing when the receiver gets a pulse the output gives a 100us pulse. The actual received pulse is 40us but i've verified the 100us output when the input is between 40 and 200us.

I have drawn the circuit on Falstad and it seems to work correctly but then the "wrong" circuit seemed to as well!

Questions:
1. Does this look like a reasonable circuit to do the above or have i made any obvious mistakes.
2. What determines the 100us pulse? I assume some RC combination
3. I cannot understand what is happening with the 330k and 1.5k resistors and how the emitter of one and base of the other are connected, whats going on there?

Here is the circuit and falstad link, apologies that the layout is backwards. The switch and resistor is the emulate the current flow through the photodiodes as the are not available in the simulator


Thank you!

[jwet]

It looks semi-reasonable but not knowing much about the application, its hard to say.

The timing is driven mostly by the charge formation in the photodiode driving essentially a charge amp- the first stage. Decreasing the 100 pF would increase gain some probably, this RC is lower than required.  The other coupling C's form high passes, basically just DC coupling.

The 330k and 1.5k are feedback around the transistor pairs.  The first Q is a common emitter amp with a gain around 200.  The second is a common emitter with feedback taken from its emitter through the 330k to make the amp act as transimpedance amps, a charge amp in the first stage.  This connection makes the second stage look like an emitter follower- IIRC, this is shunt fed feedback.

I don't understand the switch with the Bias R- maybe to decrease sensitivity to dc light swamping?

The output stage with the 1M collector looks wierd.  Can you probe the DC nodes of the real circuit.

[jwet]

I found a reasonable explanation online- its actually called shunt-series.  Its a current amplifier essentially as I thought.

[Chet T16]

It looks semi-reasonable but not knowing much about the application, its hard to say.

The timing is driven mostly by the charge formation in the photodiode driving essentially a charge amp- the first stage. Decreasing the 100 pF would increase gain some probably, this RC is lower than required.  The other coupling C's form high passes, basically just DC coupling.

The 330k and 1.5k are feedback around the transistor pairs.  The first Q is a common emitter amp with a gain around 200.  The second is a common emitter with feedback taken from its emitter through the 330k to make the amp act as transimpedance amps, a charge amp in the first stage.  This connection makes the second stage look like an emitter follower- IIRC, this is shunt fed feedback.

I don't understand the switch with the Bias R- maybe to decrease sensitivity to dc light swamping?

The output stage with the 1M collector looks wierd.  Can you probe the DC nodes of the real circuit.

Thanks for this, and the follow up video.

I mentioned the switch in the first post, this was just so i could emulate some current flow into the circuit in place of the photodiodes.

I took a look at the circuit i had and the new one i'd drawn and it I realised i was just missing a single connection (its amazing how moving things around can change perception of differences!) so i soldered it in and the circuit started working as it should