Photodiode falltime problem

[biff44]

Hi,

I don't know what you want me to explain.
A capacitor, a scope probe, a TIA, why a TIA improves fall time?

Everything already is explained. Do a forum search, do an internet search first.

If then still are questions you are welcome to ask a detailed question. Please refer to documents you have read, so we can discuss about it.

Klaus

Op-Amp Transimpedance Amplifier

A transimpedance amplifier (TIA) converts a current to a voltage and is often used with current-based sensors like photodiodes. It’s also a common building block that helps explain the performance and stability limits of many other op-amp circuits.

ultimateelectronicsbook.com

.

the input impedance of a transimpedance amplifier is NOT zero at high frequencies. so, where exactly do you expect that stored charge in the photodetector to go? to Disney world?

 

[KlausST]

input impedance of a transimpedance amplifier is NOT zero at high frequencies.

I agree, nobody said so.

Let's do some calculations.
What Opamp? What feedback resistor value? What frequency?

Klaus

 

[pooya_BA]

I want to output this photodiode to an analog to digital converter like ADS 820 and process this data by FPGA.
For work I need to have accurate falltime.
I work on phosphorescent materials and it is important for me to get accurate lifetime.
For this reason, instead of increasing the voltage, accuracy in the waveform is more important and I want to have an accurate waveform.
According to the simulations with the parts I can prepare, the best answer was the waveform.

What is the lowest voltage I can give the 8042 ad amplifier?

--- Updated Apr 1, 2021 ---

Op-Amp Transimpedance Amplifier

A transimpedance amplifier (TIA) converts a current to a voltage and is often used with current-based sensors like photodiodes. It’s also a common building block that helps explain the performance and stability limits of many other op-amp circuits.

ultimateelectronicsbook.com

.

the input impedance of a transimpedance amplifier is NOT zero at high frequencies. so, where exactly do you expect that stored charge in the photodetector to go? to Disney world?

I will read this article and answer you

 

[KlausST]

Hi,

Please give values.
Every body has a different idea what "accurate" means.

If you are interested in timing, then an ADC might not be the best solution.
An ADC is mainly used to measure amplitude.
A comparator and a timer_capture will be more accurate and with higher resolution.

Klaus

 

[pooya_BA]

Hi,

Please give values.
Every body has a different idea what "accurate" means.

If you are interested in timing, then an ADC might not be the best solution.
An ADC is mainly used to measure amplitude.
A comparator and a timer_capture will be more accurate and with higher resolution.

Klaus

I want to use opa 354, which works at a frequency of 250 MHz. And then I use ad 8042, which works at a frequency of 160 MHz, for further amplification.

I want to make the circuit according to this simulation that I did.

--- Updated Apr 1, 2021 ---

Hi,

Please give values.
Every body has a different idea what "accurate" means.

If you are interested in timing, then an ADC might not be the best solution.
An ADC is mainly used to measure amplitude.
A comparator and a timer_capture will be more accurate and with higher resolution.

Klaus

Yes, you are right, it is better to use a timer for time measurements.
But for my work I need to check the time and the voltage level at the same time.

 

[KlausST]

Hi,

Usually the requirement is determined by your idea, not by datasheets or parts.
It usualky is the other way round: You have a requirement for a timing measurement with A accuracy and R resolution, then you choose circuit and parts to meet this requirement.
For sure you are free to go your own way.

So your circuit seems to be fixed.
So your question just is:

What is the lowest voltage I can give the 8042 ad amplifier?

What voltage are you talking?
Supply voltage? At which pin?
Absolute input voltage (range)?
Operable input voltage (range) in circuit?
Minimum detectable (AC) voltage that can be processed? Which frequency/timing/waveform.

In either case one needs to consult the datasheet. Either you or one of us.

*****
I'd never use an ADC for timing measurement. To get good results you need a clean design, difficult PCB layout, expensive parts...and then you get a huge amount of data with low information and low resolution.
But maybe we just don't have full information what you really want to achieve.

Klaus

 

[pooya_BA]

What voltage are you talking?
Supply voltage? At which pin?
Absolute input voltage (range)?
Operable input voltage (range) in circuit?
Minimum detectable (AC) voltage that can be processed? Which frequency/timing/waveform.

In either case one needs to consult the datasheet. Either you or one of us.

Absolute input voltage
Operable input voltage (range) in circuit

 

[KlausST]

Hi,

Did you check the datasheet? What does it say?
Or you want us to do your job?

Klaus

 

[pooya_BA]

Hi,

Did you check the datasheet? What does it say?
Or you want us to do your job?

Klaus

hi
Yes, I checked. The maximum input voltage is reported to be 9 mV. I do not know what this data means.
What is the minimum detectable value of this amplifier?
Should I design this 9 mV based on the maximum output of the photodiode?

--- Updated Apr 1, 2021 ---

I really do not intend to take this time with you and the rest of my friends. I really do not know about amplifier parameters.
I searched the books but there was no explanation for this information.
I really do not know what is meant by "input compensation voltage"?

 

[KlausST]

Hi,

Reading the given datasheet in the ABSOLUTE MAXIMUM SPECIFICATION: (Table 4)
Input Voltage (Common Mode) ±VS ± 0.5 V
Differential Input Voltage ±3.4 V

A datasheet has sections:
* Absolute maximum ratings: limits. Stepping over those limits may kill the device.
* Recommended operation values / specification:

I really do not know about amplifier parameters.

You feel lost, lost in some details.

Developing electronics is like doing a travel. It needs a plan, an idea a destination. This is what I miss here.
I don't know about your goals. This makes it hard to give good assistance.

Comparing this thread with a travel.
You talk about Opamps, ADCs, datasheet details.
It's like talking about train, airplane, customs, safety scans, vaccinations. But nobody knows where your travel starts and where your destination is. Nobody knows whether a train is suitable or an airplane, maybe you need to use a ship.

After knowing the destination it makes more sense to talk about the details.

The target of your application surely is not building a hardware using a AD8042 (just to have it), nor doing high speed ADconversions.
You want to find out something, usually some data, some information, the result of an experiment...using measurements and calculations.
But for us it's difficult (impossible) to validate if the measurement makes sense or the use of an AD8042.

Maybe using the above mentioned circuit and devices make perfectly sense, maybe it's completely unsuitable.

Klaus

 

[pooya_BA]

In this work, LED light is radiated to materials that are able to emit phosphorescence, and then the Phosphorescence emission is received by the photodiode. In the figure below, you can see the general schematic of this work for calculating the radiation lifetime using the time-resolved phosphorescence intensity, which is the time dependent-voltage of the photodiode.
For my work, I have to read the voltage level from the photodiode. The higher the resolution, the better the result. I am writing some parameters that I have considered:
- Photodiode current output is 350 nA and maximum 510 nA.
- I want the resolution to be around 100 nanoseconds.
- I want the output of my amplifiers to be from 0 to 5 volts so that I can process it with an analog to digital converter.
- The total data acquisition time is 200 microseconds.

 

[KlausST]

Hi,

Very useful informations.

(Some pedantic notes: the main goal is to get emitted radiation timing. For this you don't read photodiode voltage (slow), but using a TIA photodiode current. The output of the TIA is voltage)

Calculation: if you want the TIA to generate 5V from 500nA, then the feedback resistor needs to be R = V / I = 10M.
I'd reduce it to get some margins. Maybe 4M7.
You need a careful PCB layout with proper GND plane for the whole circuit!

I'd realy try it as one stage at first.
It's likely that we need a second gain stage in future.

Use 0805 SMD resistor (maybe 0603) and short traces to get low series impedance.
Consider a compensating cap (0805 or 0603, COG/NP0) in the range of 1pF ... 10pF in parallel with the feedback resistor.
Try 3pF at first.

Do some tests on the TIA output with a scope. Use fast pulses from a LED directly on the photodiode. You will need some optical filter to get the desired 500nA.

Mind that the usual circuit TIA output has a DC offset and is inverted.
Maybe it is at 4.5V when dark and goes down to 0.5V when iluminated.

So far so good. Take some photos of your circuit and scope screen.

***********
For a two stage system maybe 20k feedback + inverting gain stage with gain of -200. Again: careful PCB layout is essential.
Don't forget power supply decoupling capacitors at each supply pin of each IC. Traces as short as possible.

Indeed this will be a challenge. Not very suitable for an electronics newbie.

Klaus

 

[pooya_BA]

Hi,

Very useful informations.

(Some pedantic notes: the main goal is to get emitted radiation timing. For this you don't read photodiode voltage (slow), but using a TIA photodiode current. The output of the TIA is voltage)

Calculation: if you want the TIA to generate 5V from 500nA, then the feedback resistor needs to be R = V / I = 10M.
I'd reduce it to get some margins. Maybe 4M7.
You need a careful PCB layout with proper GND plane for the whole circuit!

I'd realy try it as one stage at first.
It's likely that we need a second gain stage in future.

Use 0805 SMD resistor (maybe 0603) and short traces to get low series impedance.
Consider a compensating cap (0805 or 0603, COG/NP0) in the range of 1pF ... 10pF in parallel with the feedback resistor.
Try 3pF at first.

Do some tests on the TIA output with a scope. Use fast pulses from a LED directly on the photodiode. You will need some optical filter to get the desired 500nA.

Mind that the usual circuit TIA output has a DC offset and is inverted.
Maybe it is at 4.5V when dark and goes down to 0.5V when iluminated.

So far so good. Take some photos of your circuit and scope screen.

***********
For a two stage system maybe 20k feedback + inverting gain stage with gain of -200. Again: careful PCB layout is essential.
Don't forget power supply decoupling capacitors at each supply pin of each IC. Traces as short as possible.

Indeed this will be a challenge. Not very suitable for an electronics newbie.

Klaus

Hi
Thanks for the good tips
I simulated the circuit and I want to know your opinion about this circuit.
This simulation was created by LTspice software.
This circuit has 3 amplifiers. The general plan of this circuit can be seen in the image below.

The first amplifier is responsible for converting current to voltage. This amplifier converts a 510 nA pulse of 1 microsecond into an 8 mV pulse, here from a 15 k resistor.
I used it with a 1.5 pf capacitor, which can be seen in the diagram below.

In the second amplifier, the output pulse of the second amplifier is amplified again and reaches a voltage of 92 mV. The diagram below shows the output of the second amplifier.

In the third amplifier, this pulse reaches 4 volts.
The value of 4 volts is chosen because the analog to digital converter can receive up to 4 volts.
A filter is also placed between each amplifier.

 

[KlausST]

Hi,

If you use a couple of gain stages in series, the the first ones should have the higher gain.
I'd not use this setup: gain 10 followed by gain 50

********
I recommended a high value feedback TIA as first test.
I don't see the results. Either you did not follow my recommendation or you were not satisfied with the result. I don't know.

If multiple stages are necessary ... I recommended a single inverting gain stage.
Again you did not follow my recommendation without any comment. You will have your reasons.

****
You ask about my opinion...

I don't use this circuit, so it's more interesting what you think about the results.
If you are satisfied, then go for it.
If you are not satisfied then tell us why.

Klaus

 

[pooya_BA]

Hi
I did not use invert pulse because I wanted to apply invert bias voltage to the photodiode, and most of the circuits I have seen in articles were similar to this.
I did the simulation again with inverting bias voltage and designed the first amplifier with invert gain.

In this form, it is the first amplifier obtained in invert.

Then in this part I inverting the amplification

In the third amplification, I reached the desired voltage with a further increase.

The reason I do not consider a higher gain for the first amplification is that the waveform changes drastically, as shown in the image below.

Please let me know why the first amplifier need to be with inverted gain .

pooya

 

[KlausST]

Hi,

An inverting stage usually generates less noise,
(And with low gain it generates less distortion and is more stable)

The Cfilter at the inverting stage does not work this way. If you want to generate a filter then put the Cfilter across the feedback.

You add two voltage gain stages (I did not recommend this .... you will have your reasons) ... but then the argument about polarity does not count anymore. Two inverting are non inverting.
For further digital processing positive or negative polarity makes no difference anyway. (It just makes a difference in your imagination of the waveform)

You are free to do what you want ... but if you want us to validate your design, then let us know your ideas.

We don't know what amplifier you use as TIA.
If you want high gain bandwidth and low noise you could try so called "current feedback Opamps"

Klaus

 

[pooya_BA]

hi
Thanks for your tips and good information.
I removed the filter and it had a very bad effect on the waveform as you can see in the image below. I use OPA 354 amplifier for TIA.

As you can see in the picture, there was an error removing the second output filter.

Changing the C filter in the feedback did not show a good result.

My reason for the two-step increase was that when I use a boost step, falltime increases dramatically.
With a two-stage increase, falltime increases by about 20 nanoseconds, but in a single-phase gain, an increase of about 100 nanoseconds was observed.

pooya

 

[KlausST]

Hi please reread my post.

You modified the TIA capacitor... I did not recommend this.
You removed the input resistor of the inverting amplifiers... I did not recommend this.

Klaus

 

[pooya_BA]

hi
I have made these changes.
Please comment.

 

[KlausST]

Hi,

Please comment.

What comment do you expect?

You followed some of my recommendations some not.
This is how a designer should work. Looking for informations / recommendations and improve them to design a circuit "on his own".

I can´t say if it meets your expectations or not.

Klaus