Other Parts Discussed in Thread: OPA695, OPA846
sch1:
sch2:
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
hi, kai
the first stage in your sch is not a tia,have you miss some components?
whatever,the output is the same as my sch1,i have test that circuits,which is not very good for its freq response and noise perfoemence.
Kai is just showing your 50ohm are not to ground, but series output to a matched load,
The 1st stage is just not drawn out completely,
I would suggest a series 50ohm into the 2nd stage amp V+ input after you tap off the matched line, this isolates the input C from the 1st stage,
Hi Baolin,
and eventually you overload the OPA847 (100R in parallel to 220R or 330R) and might want to use an additional cable driver?
Kai
hi,michael
i want tia output ac couple with the second noninverting amplifier stage , after the seriers 50ohm resister,can i add a CR filter circuit next to the second sstage input?
baolin
hi ,kai,
100R is overloaded for opa847?never heard about this thing. i just want to test both the signal of the first stage and the second stage individually,which i mean is that when i test the first stage ,i connect p1 to osc,while choose the match 50ohm resistor of osc .
baolin
hi ,kai
another question.
when i add Cpd1 and Cpd2,the tia output become unstable.
but i remove the Cpd1 and Cpd2,the tia output become stable .
So is there any wrong in bias circuit in sch1?
best wishes,
baolin
Hey Baolin, the details here are getting confused, can you attach a TINA file showing what you have now?
hi,michael,
i draw it in tina, choosing eq circuit to take place of the PIN detector. when i finish the sch, i choose ac analysis, a irregular circuit report appears. wish you can use it .
TIA_Schematic.TSCbaolin
So there was a miswire in the negative bias to the diodes, fixed that - the response to the output of the OPA847 looks fine, not sure this is the shape you want.Those diodes divider bias networks are causing that lower frequency shift and reducing the midband gain to below 330ohms. But it looks stable,
You don't need to use a JFET part for the 2nd stage, Another OPA847 or OPA846 would do - or even back to the CFA OPA695
Michael,
in actual,
i need a 200Mhz bandwidth in first stage , but it seems like there is only 10M bandwidth?
And the bias circuit is right?
baolin
Thanks Kai for pointing out those two supply bias connections where missing, here that is corrected and it looks a lot more reasonable, if peaking badly,
Here is a simple solution claiming 290Mhz F-3dB, but the actual high F noise gain is too low for the OPA847 and it is still peaking 4.5dB. You don't want to match the C on the non-inverting bias current cancellation R (1nF here)
Might be able to get this to work with the lower minimum gain OPA846,
Yes, that is a lot more stable and still pretty flat through 200Mhz, I am only looking at the Zt output right now,
and this last file,
hi ,michael,
what does this mean? "You don't want to match the C on the non-inverting bias current cancellation R (1nF here)"
according to the datasheets,the C on the non-inverting bias current cancellation R is usually 100nf ,OR parallel with 100pf.
is there a method to choose the value of C?
thanks
baolin
So you only put a matching R on the V+ input to reduce output DC offset - you are AC coupled, so not sure why are doing that - but, in any case, those caps are only there on the V+ node to reduce the noise contribution of the resistor -
Zt is short for transimpedance
if i remove the C on the V+ input,the Zt output will appear a sustain oscillation,i don't know why .so i choose to solder the cap on the pcb.
thanks for your patience and reply!
baolin.
Really, that is a little disturbing - essentially, the OPA847 is on the edge of oscillation where you don't want it to be, the OPA846 would be safer.
One issue I also just noticed in your sim file, you show +/-15V supplies on the OPA847 - hope not on the board as this part is +/-6V max.
hello,michael
another question about the tia input,i wanna ac couple for that,330Ohm is a satured gain for the dc part of the input light.
baolin
Hi Baolin,
AC coupling at the input is no good idea, because the midpoint of the two photodiodes would float away, if the two photodiodes see different light intensities. You need the midpoint of the two photodiodes sitting at virtual ground and for that they must see the virtual ground generated by the DC coupled TIA. You can do AC coupling at the output of TIA, but not at the input.
Kai
Kai has a good point, but actualy there are too many unresolved issues here to really know where to go next. If you have to reject a large DC current, you can always use a servo loop on the 1st stage, but
1. what supplies are you physically using.
2. Servo loops inject their own low frequency noise, easy to handle, but a concern,
3. why does that small cap on the V+ node show stability in your apparent bench testing, has to be something with the total apparent impedance at LG xover for that 1s stage, but again right now the OPA847 is maybe not the right part, either change it to the OPA846 or if you can't stability issues need attention.
hello,michael
what is "servo loop"? a technology ?or a IC?how to achieve it in my circuit? ''servo loop''can mean many things.
the opa847 is working on +_5V state,explaining for the mistake in .tsc above.
about the small cap on the V+ node showing stability in your apparent bench testing, i have no idea, just according to the datasheets,when i solder it ,the oscillation in tia's output disappear.
baolin
I am working from this background information for transimpedance, from which I developed all those datasheet examples for the OP847, OP846, etc. etc. etc.
3125.Transimpedance design flow using high speed op amps.pptx
hi ,michael
thanks for your source,it's so nice of you.And i still feel confused about something.eg:
in this sch,series c1,c2 consists of the Capacitance of detector part of full tia circuit ? how can? if the cap of detector is 20pf,
should i choose c1=c2=40pf?
btw,if i test the tia in this way,it means ac couple right?
baolin
Hi Baolin,
by "servo loop" the use of a "DC servo" is meant.
Here's a nice explanation (for audio applications, though):
https://sound-au.com/articles/dc-servo.htm
Kai