LMH5401 for transimpedance amplifier

Kangbaek Kim

Other Parts Discussed in Thread: LMH5401, OPA857, OPA856, LMH6629, OPA659, OPA656

Hello,

I am trying to design a TIA based on LMH5401, but there is no information regarding the current noise. Can you share a plot for current noise in frequency domain for different input capacitance? Typically the current noise increases proportionally to the frequency, so it may dominate a noise floor at 500 MHz, which is the frequency I am interested.

If you were to design a photodiode TIA that has BW > 600 MHz, which op amp would you use to achieve the lowest input referred current noise? FYI, my PD capacitance is ~ 1.5pF and has series resistance of 10 Ohm.

I am looking for input noise less than 5pA/rtHz.

Thanks,

Kangbaek

over 8 years ago

0 Loren Siebert 1 over 8 years ago

TI__Genius 13935 points

Hello Kangbaek,

The current noise should be pretty flat at 3.5pA/rt Hz up to at least 1GHz. I do not have any plots with respect to frequency.

Do you want to use the LMH5401 with a differential or single ended input? Do you require differential output?

I have not designed any TIA circuits with the LMH5401 yet. We do have dedicated TIA amplifiers. If you give some more specifics I may be able to recommend a better amplifier.

Regards,

Loren

0 Kangbaek Kim over 8 years ago in reply to Loren Siebert 1

Intellectual 320 points

Hello Loren,

I need single input and single output.
There are two sets of narrow band signals I am interested in to measure:
- Fc = 225 MHz with 50MHz BW
- Fc = 500 Mhz with 100 MHz (or can be less) BW
My goal is making input referred current noise about or less than 5pA/rtHz.

Did you mean dedicated TIA amplifiers like OPA857?

0 Loren Siebert 1 over 8 years ago in reply to Kangbaek Kim

TI__Genius 13935 points

Hello Kangbaek,
I think the OPA856 is faster than the OPA857. But yes, one of the dedicated TIA amplifiers may work better.
Regards,Loren

0 Loren Siebert 1 over 8 years ago in reply to Kangbaek Kim

TI__Genius 13935 points

I looked at our online selector tool and the LMH6629 is our fastest TIA and has 2.6pA/ rt Hz current noise.
THe OPA device is the OPA656 or OPA659
www.ti.com/.../Transimpedance-Amplifiers-products.page

0 Samir Cherian over 8 years ago in reply to Loren Siebert 1

TI__Genius 12631 points

Hello Kim,
How much transimpedance gain are you looking to have? 500MHz is going to be a challenge, especially if the gain requirement is high. What is the end-application here? Is the signal CW (like a sinusoid) or is it a pulse signal?
-Samir

0 Kangbaek Kim over 8 years ago in reply to Samir Cherian

Intellectual 320 points

Hi Samir,

The more the better to achieve low input referred current noise.
I have multi tone sinusoidal signals.

Kangbaek

0 Kangbaek Kim over 8 years ago in reply to Samir Cherian

Intellectual 320 points

Samir,

I need a large gain to achieve low input referred current noise - eventually it will be limited by op amp's current noise.
That is why I was looking at LMH5401 which has a large GBP.
I use multi tones sinusoidal signals within the BW. And then measure the phase delay less than mRad.

Kangbaek

0 Samir Cherian over 8 years ago in reply to Kangbaek Kim

TI__Genius 12631 points

Hello Kangbaek,

I just ran some numbers....for a 600MHz BW, using the LMH5401, you can get around 2kOhm of transimpedance gain. Is that amount of gain sufficient for your application? You will have to be very careful with your layout and can only have around 150fF of total feedback capacitance from parasitics on your PCB at this level.

Attached is a TINA spice simulation of a potential circuit. Let me know your thoughts

-SamirLMH5401_TIA.TSC

0 Kangbaek Kim over 8 years ago in reply to Samir Cherian

Intellectual 320 points

Hello Samir,

Thanks for taking your moment to calculate the performance.

Your calculation appeared to nicely reach 5 pA/rtHz, but without photodiode.

If you put 1.1 pf before LMH5401, this will increase the input noise to 8.4 pA/rtHz because the input noise of LMH5410 is proportional to the input capacitance.

And 8pA/rtHz is where I am stuck right now using any MMICs or op-amp amplifiers in the market.

Thanks,

Kangbaek

0 Samir Cherian over 8 years ago in reply to Kangbaek Kim

TI__Genius 12631 points

Hello Kangbaek,
You are correct...increasing the input capacitance will certainly cause noise peaking (due to the amplifiers voltage noise) as a function of frequency. One way of reducing this noise peaking is to reduce the feedback resistance, however this is also going to reduce your system SNR so no real benefit. If your signal is repetitive, you could use averaging to lower the noise floor.
-Samir

0 Kangbaek Kim over 8 years ago in reply to Samir Cherian

Intellectual 320 points

Samir,

There is a trade off in averaging. Speed vs noise.
I think I am already averaging enough to increase SNR now.
As the noise level is dominated by the op-amp, I will keep my eyes on any future release on TI op-amps.

Thanks,
Kangbaek

Amplifiers

Amplifiers forum

LMH5401 for transimpedance amplifier