Welcome to the e2e forums,

May I assume that there is no information in the input and output levels. In other words if the input was a square wave of 0.1V to 3.2V, would the correct output be a square wave with 0V and 10V levels? 

What is the maximum frequency of the input?

What are the characteristic of the output load. For example '10k resistance to ground'.

Hello Ronald, thanks so much for your reply.

The input is a signal that comes from an Arduino due, it is a square wave where the maximum amplitud will be changed from 0.5 to 3.2V. The frequency can vary from 10 Hz to 1 kHz.

The output needs to be amplified so when the max input amplitude is 3.2V, the max output amplitude is 10V. The output load is 10 kilo-ohms. It will be fed to a TREK 2210 HV amplifier (characteristics copied below, also at http://www.trekinc.com/pdf/2210-Sales.pdf)

That is a fast rise time power amplifier. Therefore the preamplifier to get up to 10V will also need to have a fast slew rate. OPA810 is a good choice for this application. It will need a positive power supply above 10V. To get more detailed assistance with using the OPA810, start a new post with that part number selected.

Thank you very much Ron for your help.

I am looking for a board that is already soldered and ready to use. Is there a ready to use evaluation board with the OPA810? Or maybe a different chip?

Do you think any of these would work for my application?

LMH6881EVAL, THS4601EVM, or LMH6881EVAL/NOPB

I understand now. The EVM that TI has will require soldering and having additional components available to get the desired gain. It would be good to find an amplifier that has a potentiometer (pot) adjustment to set the gain (turn a knob). There are lots of these available but they are meant for audio and will not work well because the DC component matters here. 

For a solder free 'ready to go' solution, I suggest trying the Arduino forum (or try it again).  There should be an adjustable DC amplifier that is used by the hobby-ist there. The edge rate might not be great, but that might not matter in your use case.

Hi Ron, thanks again.

I have been searching for ready to go solutions without much success because our signal requires a bandwidth from DC to 9GHz.  I may try to put together the board and its components (or find help for getting it done).  I used to be an electrical engineer in a different century at another place, and am now working for a small research lab at a university. I appreciate your help very much. 

I am considering then purchasing two EVMs for my project.

It consists of:

Photodetector (10 to 400 mV) -> EVM 1  -> Arduino due analog input (3.2V max) ->Arduino due analog output (3.2V max) -> EVM 2 preamplifier (10V max)-> TREK 2210 HV amplifier 

EVM1: I asked on a separate post about using the LMH32401RGTEVM, which seems would work and comes mostly assembled.

EVM2: It seems that either the LMH6881EVAL or the  THS4601EVM are assembled for the most part. Would any of these work for my application? 

Just in case this is useful info for you, the detectors we are using are: PDA8GS (https://www.thorlabs.com/thorproduct.cfm?partnumber=PDA8GS), and DET08CFC  ( https://www.thorlabs.com/thorproduct.cfm?partnumber=DET08CFC#ad-image-0)

Thanks again for your help.

Hi,

the 9GHz bandwidth was a typo?

Kai

Hi Kai, thanks for your reply!

No it is not a typo.

There is a snapshot of our detector specifications pasted below.  

This is the configuration I would like to try, if I purchase the two EVMs from TI:

Photodetector   (https://www.thorlabs.com/thorproduct.cfm?partnumber=PDA8GS) --> EVM 1 (  LMH32401RGTEVM) -> Arduino due analog input (3.2V max) ->Arduino due analog output (3.2V max) -> EVM 2  LMH6881EVAL or the  THS4601EVM preamplifier)-> TREK 2210 HV amplifier (http://www.trekinc.com/pdf/2210-Sales.pdf)

I would like to know your view on the feasibility of this setup if I purchase EVM1 and EVM2 from TI. 

Thanks again for your help. Kind regards

The good news is that all three EVM are supported by the same high speed amplifier team. I will transfer this thread to that team. 9 GHz requires special techniques. I do not think 9GHz was considered in the Arduino design. 

Hello,

I have a few questions in regards to this design:

1) Is the 9GHz requirement in regard to these EVMs? The LMH32401 only has a 450MHz bandwidth at the lower gain configuration. 

2) What is the sampling rate on this Arduino? Will it able to read the high frequency signals of the LMH32401? 

3) In regards to EVM 2, the reason for using high frequency amplifier is because of the slew rate of the TREK 2210 HV amplifier? Would a comparator work for EVM 2? 

In regards to an EVM for the OPA810, while the current EVM for the board is a blank board to be populated, there is an EVM for the OPA2810 (the dual version of this device). You will probably still have to change the resistors on that EVM for your desired gain.

Best,

Hasan Babiker

Thank you Hasan and Ronald for your replies and help.

The fast response required in EVM1 is because the signal from the photodetector has sharp spikes, with a duration of a fraction of a millisecond. Also within a fraction of a millisecond, after a couple of comparisons operations, we need to send a trigger signal to the HV amplifier.

Because of your comments, I have been looking at our current system, implemented with an Arduino, which I inherited from a person who left the lab. You are very right in questioning the speed of the Arduino. It may be worth to re-do the system again with a faster microprocessor board, maybe one such the Rasberry, and get the right EVMs. I don't have any experience yet with either Arduinos or Rasberry, but I can write programs so I think I can deal with them in time. The Rasberry does not have any ADC or DACs but it still seems can do much better. 

Hi Aeg,

the Arduinos I know have the ATMEGA328 microcontroller.

Kai

Thank you Hasan.

I have one more question.

Is there an evaluation module for a fast amplifier, analogous to the  LMH32401 , but with a gain larger than 10? It would be used to amplify our photodetector signal which has a maximum amplitude of 100 mV.

Hello,

Can you clarify is your input a voltage or a current from photodiode source? The LMH32401 is meant to be used in a transimpedance configuration with a gain of 2k or 20k. 

If your input is a voltage can you confirm what gain and bandwidth you are looking to achieve? Are there any design constraints (noise, dc precision, power consumption, etc.)?

Best,

Hasan Babiker

Hello Hasan, 

I am not sure I understood all your questions but will try to reply the best I can by sending the details of the setup. Sorry for the long reply.

1- Detector and signal 

We are using two different photodetectors, both with a pulse response in the order of 300 picoseconds.

The  PDA8GS Thorlabs amplified photodetector  (https://www.thorlabs.com/thorproduct.cfm?partnumber=PDA8GS) has already a transimpedance amplifier.

We also have the biased photodetector  DET08CFC  as a backup, ( https://www.thorlabs.com/thorproduct.cfm?partnumber=DET08CFC#ad-image-0), this one I understand does not have a transimpedance amplifier. It has a shunt resistance of 10 MΩ.  

When we use the PDA8GS, the signal is a  DC-coupled signal with a spikey shape, with a baseline around 35 mV, and upwards and downwards spikes, with a relative amplitude from the baseline of +-20 mV, and a width at half maximum of about 400 microseconds. We will use a cable with SMA connectors, with a capacitance of 96 pF/m  (https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_ID=2888) to connect the detector the LMH32401RGTEVM .

2- The  LMH32401IRGT Evaluation Module  says that it comes with a jumper that can be used to get a 10 V/V gain. Since our signal is ~40 mVpp, by selecting the 20k jumper we would get 400 mVpp. Is there another EVM that would give us a higher gain, so we get a signal in the order of a couple of Volts?

3- On the  LMH32401IRGT data sheet, page 2, Table 1 - It says the max output voltage swing (VCC = 3.3 V, 100-Ω load) is 5 Vpp. How can the 5Vpp be achieved with this board?

Thanks again for all your help.