Hi Jari,
The E2E editing tool is changed two days ago, but I think that certain things are made easier. Since the tool is new, there may be some learning curve because of the new menu system.
To insert an image file, you may cut (e.g. snipping tool) and paste an image directly into the inquiry text. You may also use Insert command below and click image/video/file tab to upload a file.
Can you send us the schematic that you are working with and see if we are able to simulate via Tina first? I do not work with hspice tool, so we may not be able to reply certain issues that you encountered with the tool.
Best,
Raymond
I'll check... This looks working well with paste.
TRAN simulation result:
NoiseTRAN simulation result:
And a lot of case where the noisetran stops with wrong or too small timestep error.
The issue that TRAN & NOISETRAN results are quite much different, could point for some error here (maybe timestep related, etc).
--
Jari
The schematics as above is quite simple, I guess this could be published here openly.
Main issues is NOISETRAN-case where it fails to operate, or maybe is it possible that it really would oscillate that case?
But that could be modeling failure that affects noise simulations in the spice model...
The used capacitor values may have been different than here in the earlier simulation cases.
Hi Jari,
LMP7721's supply voltage, Vs (V+ - V-) is rated up to 5.5Vdc. Enclosed is an example of TIA circuit from 1nA to 100nA, since I am not clear that the photodiode's parameters are. If you are able to provide us the TIA's application requirements, I can help you to simulate one via Tina.
Our pspice model engineer is not familiar enough in supporting hspice issues. Enclosed is the link of Tina-TI simulation tool in our website and it can be downloaded for free.
https://www.ti.com/tool/TINA-TI#downloads
Best,
Raymond
Ok. I might check if I get it working with the Tina.
However the system has dual boot for Windows10 + Linux dualboot and mainly I'm operating with LinuxMint20.1
I simulated it with 5.5V supply (hspice) now result is as below:
So still in hspice case the transient noise has strange oscillating and different voltage results.
I would rather use the hspice, because it works here in linux mode.
Does the TINA have transient noise option?
Alse diode model is for hspice, and it can not be translated to the Tina:
So the diode would be missing in the simulation if I use Tina.
As shown below, the different model blocks should be combined into same simulation:
I should know how to get the diode model prepared for TINA or LMP7721 model prepared for hspice.
The light condition is not problem here, as this is for dark condition leakage measurement.
If I transfer the diode model to TINA, the results show Out = 0.000 V !
So I guess this still does not work properly... But why?
And there could not be found transient noise analysis (TNA) in TINA-simulator:
DC analysis case the output shows voltage of about 1.76V, but that seems wrong also and like if the diode current flowing from anode to cathode rather than c to a.
Hi Jari,
Enclosed is a simulation with photodiode current from 1pA to 1nA. The transfer function Vout = Rf * IG1, where 1pA <= IG1 <= 1nA.
With single supply rail and DC bias of 2.2V, Vout = 2.2V + Rf*IG1 = 2.2V + 1x10^9*1nA = 3.20V at IG1=1nA.
Here is the simulation.
Please modify the simulation for your photodiode sensor. Once you are getting the correct response, we can check the TIA stability.
BTW, what your sensor's BW required for the TIA application. The dominated pole in the above TIA circuit is reduced at approx. 159Hz currently.
Best,
Raymond
Can you recheck the simulation with feedback resistor value of 1T (not 1G)?
The measurement range would be 1f-1000f (about x1000 times smaller current that what you used).
It looks also temperature sweep case the result shows zero.
It could be also that the diode model for the nwdio is not 100% compatible here...
The current of R4 in below image is 'negative'.. this is strange.
Maybe this opamp could not handle such small current or what can be causing this?
Simulator accuracy settings wrong?
The below circuit seems to be going right:
I think this case is correct, the V(Out1)=4V is correct.
However if resistor value 1T is used the results does not seem to be what is requeseted:
This case of course if the opamp has low leakage the V(Out1) would be 4V as well.
But I could not find what actually is causing this...
Hi Jari,
Q: Maybe this opamp could not handle such small current or what can be causing this? The measurement range would be 1f-1000f (about x1000 times smaller current that what you used).
LMP7721 is unable to meet your 1fA-1pA TIA design requirements. LMP7721 is also the best fA amplifier we have, and we do not have lower Ibias op amp part.
The use of 1 TΩ Feedback resistor will not be feasible in TIA application. If you need to higher gain output, you have to use multiple stages to gain up the I-to-V conversion. Without knowing your application, I placed 1GΩ Feedback in the TIA circuit, which the gain is already high enough (You may modify the working circuit to increase the gains). You also have to consider op amp stability, BW and taken into account other environmental consideration. It is possible that your environmental noises are already exceeded 1fA under regular lab environmental conditions.
You may have a chance to use LMP7721 for 1pA or may be slightly lower in IV conversion. If you are able to lower operating temperature in LMP7721 and purge with low dew point gas (e.g. N2), you may be able to stabilize the Ibias current and improve the sensitivity of the part. However, it is going to be difficult to detect 1 fA of current because of the level, which is already exceeded the part's specification.
BTW, Digikey does sell 500GΩ resistor from Ohmite ($5.737/each). However, if you use the type of resistor, the BW of TIA will be very low, down to Hz or subHz range. However, the issues are that 1fA current range is lower than the LMP7721 can handle. If you still want to use LMP7721, I can put together a simulation, but I need to know your BW requirements.
LMP7721 up to 10pA E2E 03232021.TSC
I simulated the TIA from 1fA to 10pA for you with 100GΩ feedback resistor, but it does not mean that the simulation will work in the real part (maybe not in the 1-50fA range).. In fact, I already see some model issues with Vbias voltage from the plotted data. The BW is configured approx. 5Hz currently.
If you have other questions, please let us know.
Best,
Raymond.
