1. Home
  2. Electronics Design
  3. bandwith limitations in transimpedance amplifier

bandwith limitations in transimpedance amplifier

Hello

There are already a lot of topics about transimpedance amplifier with a lot of references and I'm now waiting for the Graeme's book about photodiode amplifier. But there is one issue that is not so discussed : the bias voltage.

I'm using a custom bidimensional PSD with 2 anodes and 2 cathodes. The typ. value of the resistance between 1 anode and 1 cathode is 150 Ohm (which is very very low compared to photodiode resistance : 100meg or giga ohms) the capacitance between 1 anode and 1 cathode is 1,5 nF (which is also very big). The output current is about 100=B5 amp and its bandwidth is around 4 MHz.

I decided to use the OPA657 for my transimpedance amplifier. I have a feed back resistor of 1.8K with a parallel capacitance of 12 pf. This configuration gives me not so bad results : ~3 MHz bandwidth (simulated with PSpice). But my problem is that by addind a bias voltage of few volts on the non inverting input, both simulation and real measurements showed a reduced bandwidth to ~100kHz. I think this is due to the small detector resistance but I'm not so experienced to say more about.... Any ideas ? by the way the max positiv voltage of the opa657 is 2,5V, do you know similar product that could achieve a bias around 10 or 15 V ?

Thanks you electronicians!

Reply to
popi
Loading thread data ...

Check out figure 3 in the data sheet, on page 11

formatting link

which shows you how to put 15V bias on the photodiode while working within the common mode range of the OPA657 (+1.8V to -3.3V).

--=20 Bill Sloman, Nijmegen

Reply to
bill.sloman

I can't use this schematics because each cathode and anode of my psd are linked to the inverting input of an opa657. Thus I have 4 channels. And in my case can only be applied at the non inverting input which have some limitations

snipped-for-privacy@ieee.org a =E9crit :

Reply to
popi

Use the low-supply-voltage part - it's better, but with boostrapped supply voltages. Then diff amp the outputs.

--
 Thanks,
    - Win
Reply to
Winfield Hill

wait a sec, some of these numgbers don't seem to be in the right ballpark:

1,500 picofarads? Wow, that's huge. Are you sure that's correct? Guess it could be if these are large-area diodes.

How did you choose the 12pF? And does that include the stray capacitance?

Reply to
Ancient_Hacker

Yea this is really huge but the sensor is 45mm*45mm which is also huge! Calculations of the capacitance is very well documented every where and spice simulations helped me to choose 12pF to get the flatest frequency response. I can't get a high gain because it's linked to the GWB by this ration in my case : f-3db=3Dsqrt(GWB/(2.Pi.Gain.2*C1)) with C1=3DCs+Cin(from amp)

that gives GWB =3D 4.6*10^5 * Gain for f-3db =3D 5MHz

with a gain of 1000 i already need an op amp with a 450 MHz GWB. My question is these calculations are always made assuming gain wait a sec, some of these numgbers don't seem to be in the right

f=2E

Reply to
popi

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.