0
\$\begingroup\$

I'm trying to replicate a Silicon Photomultiplier array multiplexing scheme by SensL. The link to their application note is found here:

http://sensl.com/downloads/ds/TN-Signal_Driven_Multiplexing_Method.pdf

According to the datasheet of their MicroFJ-60035-TSV silicon photomultiplier, the fast output signal is 150 mV, so some amplification is necessary. In their paper, they used a set of Minicircuits amplifiers in order to amplify the multiplexed output signal, the MiniCircuits ZX60-43 and ZFL-1000LN+ external amplifiers. They were connected in series in order to provide a larger gain of ~200. Links to the datasheet are below:

https://www.minicircuits.com/pdfs/ZX60-43+.pdf https://www.minicircuits.com/pdfs/ZFL-1000LN+.pdf

I never worked with external amplifiers such as these before, and I'm looking for a way to swap them out for smaller IC amplifiers that can be put on the same board as the SiPM array. Going from larger amplifiers such as these to ICs, how are they different, and how can I choose the best IC (or pair of ICs) to provide the same gain? Why are external amplifiers such as these used in the first place rather than ICs? Also, given how they're both broadband amplifiers, how would they be put in series? It feels that the amplifier order matters, but I'm not sure how.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ If you want to replicate the performance of those amplifiers it's not the gain you have to worry about, it's the bandwidth. Unless you have serious microwave design experience, (and this question tells us you don't), forget about ICs and stick with these modules. Otherwise, a very long and very expensive time from now, you will have serious microwave design experience. \$\endgroup\$
    – user16324
    Commented Dec 18, 2017 at 19:11
  • \$\begingroup\$ @BrianDrummond Thank you for your reply. Yes, I do not have any experience in microwave designs; I was hoping that it would be possible to replicate the series with ICs. Looking at the potential 4 GHz amplifiers on Digikey, they seem to just be differential amplifiers, and I have no experience with that. \$\endgroup\$
    – BestQualityVacuum
    Commented Dec 18, 2017 at 20:01

1 Answer 1

Reset to default
1
\$\begingroup\$

The reasons that external amplifiers were used are:

1) Simplicity/cost of the Eval board

2) Wide range of possible SiPm signals - under high pulse illumination SiPm signal, even the 'fast' one might be very large (talking ~1V)

To preserve the bandwidth = pulse shape of the 'fast' SiPm of 6mm size, you need roughly around 300 MHz or so. It is a bit on the limit for the opamps, it is easier to use RF-amplifiers ( the 'fast' is anyway AC coupled by nature with few MHz low freq cutoff ).

You can use e.g. BGA2818 or similar, however one has to be RF aware when routing the board (which is not that difficult). You can perhaps acquire some Eval boards for them first and solder bigger coupling caps there to improve low freq cutoff.

Also, you should have a look at boards like this.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for your reply. Previously, my team used a single MFJ60035 SiPM in a transimpedance configuration (properly biased and a feedback configuration of 235 ohms in parallel with 3 pF cap) with the OPA656, the AD8014, and the AD8001 (not all at once, but independent circuits to test the amplifiers). The AD8014 had fast speeds, but it couldn't preserve the shape even though it has a larger bandwidth than the other two, while the other two amplifiers did at the cost of speed. Assuming the same circuit for each SiPM, what would cause loss of the shape? Is it possible to have too large of a BW? \$\endgroup\$
    – BestQualityVacuum
    Commented Mar 6, 2018 at 17:36
  • \$\begingroup\$ Using current feedback opamps (AD8014 and AD8001 are current feedback type) in transimpedance configuration is a bit tricky - they are effectively common base on the inverting input. That means their inverting input is of low impedance, therefore the overall gain is not completely defined by just feedback as in usual transimpedance case. \$\endgroup\$
    – Nikolay
    Commented May 22, 2018 at 13:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.