Bootstrapping a photodiode

Pablo Beret

Other Parts Discussed in Thread: LM6171, LMC6082

Hi,

I'm trying to design the first stage (transimpedance amplifier) of a radation detector with a "common" photodiode.

Sytem details:

The goal of the first stage is to have a pulse when some kind of radiation impact the photodiode. This will produce a 15/20 nA - 50 us pulse.
Photodiode: Vishay BPW34 (70 pF - 2 nA dark current in photovoltaic mode (0V bias))
Necessary bandwidth: 20 kHz.

The transimpedance amplifier seems to be ok with low frecuencies. I've simulated a 3 photodiode array with C1 (capacitance = 210 pF), IS1 (dark current = 6 nA) and IG1 (a 15 nA - 50 us pulse).

Naturally, the capacitance of the photodiode begins to cause problems with frequency. As is indicated in most technical literature, the photodiode can be bootstrap with a voltage follower, improving the bandwidth. But when I've tried it I begin to have problems.

This is, for example, the AC gain with a LM6171 as voltage follower:

What I'm doing wrong?

Thanks!Transimpedancia 01.TSC

over 8 years ago

0 Pablo Beret over 8 years ago

Prodigy 45 points

There is a mistake in the first AC gain image. Should be this one:

0 Lothar Kowalik over 8 years ago in reply to Pablo Beret

Prodigy 90 points

Hi,

the main problem is the too large dimensioned feedback resistor - and the capacitances of the photodiode and the feedback capacitor.

This both capacitors with her low impedances do dominate the gain response and not the feedback resistor.

Try to use only 1MOhm (or 100kOhm) and a voltage follower with gain of 10 (or 100) behind the TIA-stage.

The result should loook much better.

Best Regards from Berlin

Lothar Kowalik

0 Hooman Hashemi over 8 years ago in reply to Lothar Kowalik

TI__Mastermind 22825 points

Hello Pablo and Lothar,
I see that you are running the LM6171 from 9V single supply (VEE tied to ground). The input common voltage range of the LM6171 (VCM) does not include the VEE (or V-) rail. VCM requires at least 1.3V headroom to V- and to V+.

Therefore, you would need to change the LM6171 VEE pin to at least -2V supply or more negative, to have the LM6171 in its linear region when input is 0V (as is the case the way you've tied it to the inverting input of the LMC6082).

Regards,
Hooman

0 Pablo Beret over 8 years ago in reply to Hooman Hashemi

Prodigy 45 points

Thanks Lothar and Hooman for the help. I got the point.

Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

Amplifiers

Amplifiers forum

Bootstrapping a photodiode