OPA847: opa847

BAOLIN CHEN

Part Number: OPA847

hello everyone,

i designed a transimpedance amplifier circuit using opa847, with the cd=30pf， rf=1k

when the cf=0pf, the fft of the output noise seems to have a high q,

while i tuning cf to 0.3pf， the q seems to decrease and the circuit is stable, but the q stills seems to be higher than 0.707 for the fft of the noise is not close to butterworth response.

in theory ,i can still increase the cf and the circuit will still be stable.

but when i tunes the cf to 0.4pf， the circuit becomes unstable,some other peaks in fft of the output noise .

could anyone tell me how to why the circuit becomes unstable before it achieve the butterworth response ?

over 1 year ago

0 BAOLIN CHEN over 1 year ago

Prodigy 120 points

0 Michael Steffes over 1 year ago in reply to BAOLIN CHEN

Guru 47485 points

Cf=1.7pF will compensate this to Q=0.7 with F-3dB at about 132MHz.

Be careful how you probe this,

The noise will always peak with the noise gain shape,

0 kai klaas69 over 1 year ago in reply to BAOLIN CHEN

Guru 140200 points

Hi Baolin,

for all Cf between 0fF and 400fF your circuit is not stable:

baolin_opa847_3.TSC

Kai

0 BAOLIN CHEN over 1 year ago in reply to kai klaas69

Prodigy 120 points

hi ,Kai

maybe the parastic capacitance on the PCB pads of cf is large than 1pf？

but i compute the parstice cap using c=er*s/4pi*k*d; c<=0.1pf;

i wonder where does the extend 0.9pf come from ?

baolin

0 BAOLIN CHEN over 1 year ago in reply to Michael Steffes

Prodigy 120 points

hi ,steffes

will there be a possibility that the parastic capacitance on the PCB pads of cf is large than 1pf？

i compute the parstice cap using c=er*s/4pi*k*d; and gets that parastic cap<=0.1pf;

i wonder where does the extend 0.9pf may come from ?

baolin

0 kai klaas69 over 1 year ago in reply to BAOLIN CHEN

Guru 140200 points

Hi Baolin,

if you have a 4-layer PCB made of epoxy with a distance of 140µm between the top layer and the second layer, then 2mm x 2mm copper above the second plane gives a stray capacitance of about 1pF.

Please note that you must not directly touch any signal pin of OPAmp with the scope probe. Michael already mentioned this. Always insert a suited isolation resistor first. Or use the 50R technique to connect the output of OPAmp to the 50R input of scope (50R series termination resistance at the output of OPAmp + 50R cable + 50R input of scope).

Kai

0 BAOLIN CHEN over 1 year ago in reply to kai klaas69

Prodigy 120 points

kai，

thanks .

indeed,i have a 2-layer PCB made of epoxy with a distance of 1.60mm between the top layer and the bottle layer,and i use the 50R technique to connect the output of OPAmp to the 50R or 1M input of scope.