This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMP7721: Gamma radiation sensor

Part Number: LMP7721
Other Parts Discussed in Thread: OPA4350

Hello,

I have inherited this radiation detection board design at work. None of the original designers are with us and very little documentation exists. The design has been through several revisions and still 30% of the boards do not work as desired, or they work at our facility when we test them but do not work at customer's site. 

A PIN diode (PDB-C156-ND) is used as the sensor, followed by an LMP7721 as the first stage transimpedance amplifier. I first suspected layout and noise issues, because there were no guard rings on the board. The whole circuit is enclosed in an aluminum shield, but no guard rings. As I investigated more, now I suspect the feedback capacitor and resistors values as well. 

This stage is followed by a band-pass filter. Of course I found cut-off frequency settings to be a little off and as I changed the values performance improved significantly. 

Though I still cannot figure out how previous designers of this board calculated capacitor and resistor values for LMP7721, and I need a second opinion with that since my own calculations yield different values.

You can see the LMP7721 part of the circuit below, where;

C19 = 0.6pF, R3 = 10Mohm, R17 = 100Kohm & R18 = 49.9Kohm

Using equation 8 from the datasheet (page 19) and assuming that;

Cs = Cin + Cj + Cstray, where Cin is LMP7721 input capacitance (11pF), Cj is junction capacitance of photodiode (10pF) and assuming 3pF for any stray capacitance. I have Cf ~= 0.4pF. Where if I Do not change R17 & R18, then C19 should be 0.8pF, or if I change R17 & R18 to 50K and 10K, C19 nan be 2pF.

So I have two questions;

1- Do you think the current values will work fine and I should just leave them the way they are, or shall I replace them with any of these configurations I mentioned above?

2- Do you suggest that I change the layout and add a guard ring for LMP7721?

Thanks for your help,

May Varza

  • Hi May,

    For now, let's assume the design is okay because about 70 % of them appear to be working as needed.

    Please explain what "30% of the boards do not work as desired, or they work at our facility when we test them but do not work at customer's site" in terms of how the working, and non-working, boards behave?

    The LMP7721 has ultra low input bias current and it is very easy to corrupt the performance of the amplifier if extraordinary efforts are not in place to assure an ultra clean PC board and component environment. If there is solder flux contamination, or finger oils on the PC board, or across the components, the operation of the circuit can be severely degraded.

    What PC board cleaning do you have in place to assure the complete removal of solder flux and any other contaminants that are present from the board assembly and handling operations? We use water soluble flux for our evaluation board soldering and then run the PC board through two separate ultrasonic cleanings using deioinzed water.

    Regards, Thomas
    Precision Amplifiers Applications Engineering
  • Hi Thomas,

    As far as desired functionality, the rest of the board consists of a band-pass amplifier, a comparator and a final stage low pass filter with a long time constant. In presence of radiation final output should read 5V and with no radiation 0V.

    As for cleaning, I am not sure what procedure is followed. I have to wait for our mechanical engineer to come back from vacation, since he is the one who has the information.

    I can share the complete schematic is necessary.

    Thanks,

    May

  • Hi May,

    Okay - I understand the basic transimpedance function of the LMP7721 in you application, but it is not clear how the ~30 % are failing. Can you describe how the failing devices are behaving, or what they are failing to do?

    Please do provide the portion of your schematic where the LMP7721 and the filter that follows it are shown. Include the power supply, input signal, output load, etc, information. I have ran some simulations on the LMP7721 circuit you provided here on the E2E and have confirmed the viability of the design. I did have to make an assumption about the Vref_Pre level, but even if it is different from my assumed value the circuit performance should be similar.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas,

    Thank you for validating the design. I have attached the complete schematic.

    Please take a look at what I was told about what defines failure with these boards:

    "The following is a description of a standard test and defines a failure:

    • Sensor is plugged in to test setup

    • Turn on power supply o Normal operation: multimeter shows voltage spike to ~5V, then decay to ~20 mV (“background” value)

    • Move sensor near (less than 6”) radiation source o Normal operation: multimeter shows values between 0 and >200 mV, depending on activity amount and exact distance to sensor o Failure: multimeter value does not change from background value 


    These failures are surmised to be due to a failure in the diode which detects the radiation, although this has not been definitively proven. There is currently no established QC protocol for the radiation sensor and board before or during assembly. The only QC check the radiation sensors go through is the end-point check: a validation that the sensor is responsive to radiation after assembly (described above).

    Most failed sensors are caught during this end-point check. However, there are instances of radiation sensors failing in the field after having previously been shown to function correctly at SOFIE. The reasons for these failures are not well characterized. In addition, during the assembly and testing process we have also noticed boards failing. These failures almost always result in a short to ground from the 12V input."

    I also know that a brass shield is soldered to the board to cover the photodiode. This shield is soldered to the completely assembled board, here at our facility. Could the heat from this procedure damage LMP7721?

    Another thing I noticed is that, it is mentioned above that in normal operation multimeter connected to output shows 0-200mV in presence of radiation, which is not what I measure. I use a function generator in series with a 1Mohm resistor to create input current pulses to feed to -IN on LMP7721. I would really appreciate if you could point me into the right direction in case I am doing anything wrong.

    BTW, C17 is marked as DNI on BOM!

    Thanks,

    MaySOF-1861031103 - RADAFERevG_LOW.PDF

  • Hi May,

    Reading through the information you provided it appaears that the particular component failure that occurs hasn't been established, although it suggest its is the PIN photodiode. You also mention, "These failures almost always result in a short to ground from the 12V input." Is that short from the VIN input to GND in the Precision Power Supplies, or another connection in the circuit? If the LMP7721 or OPA4350 were to short between their V+ and V- supply pins that would occur after the TPS7A4901 regulator and its output would be pulled down to an unexpected voltage level. I don't know if that would result in a short indication from the 12 V line to GND.

    Have you taken one of the failed boards and tried replacing the PIN photodiode, or the LMP7721, and then retesting? Note that if the diode isn't operating that the LMP7721 VOUT_PRE voltage level should be about equal to the Vref_pre applied to the non-inverting input.

    Certainly excessive temperature can damage semiconductors. The LMP7721 datasheet provides information regarding soldering the devices on a PC board; Infrared or Convection (20 sec) 235 °C MAX, Wave Soldering Lead Temp. (10 sec) 260 °C MAX. If the procedure for soldering the brass shield on the PC board subjects the LMP7721 and the other semiconductors to temperature/dwell conditions that exceed that specified for the LMP7721 they might be damaged. Each semiconductor device is different and will have different heat tolerance. You may want to review the specifications and procedure used for that soldering process.

    Regards, Thomas
    Precision Amplifiers Applications Engineering