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.

Isolator For Photodiode Operation in Photo-voltaic/conduction mode

Other Parts Discussed in Thread: TMUX7219, TMUX6219, TPS2121

Hello All,

I am looking for Digital Isolator, my application is to run the photo-diode in photovoltaic and photoconduction mode.

The photo-diode will be connected to the SMP connector, the shield of the SMP connector will be connected to supply for photovoltaic mode and to ground for photoconduction mode.

The signal line of the SMP connector will be connected to the amplifier.

Can you please suggest which isolator would be best suitable for my applications?

Thank you,

Ravikumar

  • Hi Ravikumar,

    Thank you for posting to E2E! It seems like you might be looking for high-side and low-side power switches instead of digital isolators, so I am moving this thread to the appropriate forum.

    Please let us know if you have any additional questions meanwhile.


    Respectfully,
    Manuel Chavez

  • Hello Manuel,

    Thank you for the reply.

    The reply to this post will be visible here, correct?

    Thank you,

    Ravikumar

  • Hi Ravikumar,

    You're welcome! And yes, you are correct: the reply from our Power Switches team will be visible on this thread.


    Thank you,
    Manuel Chavez

  • Hello Manuel,

    Thank you for the update.

    Why do you suggest power switches here, instead of isolators or mux?

    Any specific reason which I am overlooking.

    Why I choose isolators?

    Because, power switches might have leakage ratings, compare to isloators.

    Thank you,

    Ravikumar

  • Hi Ravikumar,

    TI's digital isolators are designed to isolate two different voltage domains, protecting devices on each side from high voltage transients and ground shifts in the other. Our isolation devices maintain an isolation barrier that prevents unwanted voltages and currents from crossing while allowing digital signals to "pass across," but the isolation barrier is not shorted for communication or power transmission.

    From the diagram shown, there is a single power supply rail (12V) and a single GND potential that need to be directly connected to the SMP connector's shield at different times, and it seems like controllable switches are a simple solution to achieve this. A multiplexer may also work if you prefer, however as you noted, both have higher leakage currents than digital isolators because they are meant to join two circuits, where isolators must not.


    Thank you,
    Manuel Chavez

  • Hello Manuel,

    Thank you for the detailed explanation.

    I understand your point, but I do not understand this statement,

    both have higher leakage currents than digital isolators because they are meant to join two circuits, where isolators must not.

    Because in OFF state either, the leakage ratings are unwanted, in my case the shield of the SMP will be either +12V or at GND, so I need full isolation between these 2 power plans.

    Please correct me, if I overlooked something.

    Thank you,

    Ravikumar Jain

  • Hello,

    After looking at the MUX catalog, I found TMUX7219 or TMUX6219, to be satisfactory.

    Can TMUX7219 or TMUX6219 be operated in the following scheme?

    Thank you,

    Ravikumar Jain

  • Hi Ravikumar,

    You're welcome! Regarding your question on leakage currents, switch devices specify leakage current values as they are a non-ideality. Leakage currents are a small parasitic made from realizing the switching device. Even though they are present when the switch is OFF and they are undesirable, the switch will still disconnect both sides of the circuit when it is OFF.

    Since isolators do not actually join two circuits, they do not have leakage currents on the same order as switches. However, the complete isolation of the power supply rail, SMP connector shield, and GND means the SMP shield would not ever be directly connected to either the 12V supply or GND, as shown in your diagram. Isolators are meant to fully separate circuits and only allow signal or power transfer.

    Engineers from our Multiplexers team will follow up on this thread since you mentioned the TMUXx219 devices Slight smile Thanks for your follow-up posts, and please let us know if you have additional comments or questions!


    Thank you,
    Manuel Chavez

  • Hello Manuel,

    Thank you for the detailed information.

    I'll be waiting for the response.

    For this thread, this is the only query I have.Slight smile

    Thank you

    Ravikumar Jain

  • Hi Ravikumar,

    I am from the Multiplexers and Protection Devices team. 

    The TMUX7219 should not be operated in a way in which the voltage on the I/O pins exceeds the power supplies. It doesn't really work as a high-side to low side transfer. I don't think we have something in our mux portfolio that can meet the needs of this application. 

    If you have any other questions please let me know!

    Best,

    Parker Dodson

  • Hello Parker,

    Can you suggest what can be the suitable solution for this...

    Thank you,

    Ravikumar

  • Hello Parker,

    If I supply the MUX with +/-12V, in that case, it should work, correct?

    Thank you,

    Ravikumar Jain

  • Hi Ravikumar,

    We do not have a solution from MPD I have moved the thread to power switches to see if they have any solutions they should answer soon.

    Best,

    Parker Dodson

  • Hello Parker,

    Thank you, let see what they have...maybe finally if nothing works Ill for the below configuration...

    Hope this works, whats your opinion Parker?

    Thank you,

    Ravikumar Jain

  • Hi Ravikumar,

    From a Power Switch perspective, our devices are built to handle DC current and do not have the switching speed, leakage specifications or control that our signal switch portfolio has. From your explanation, I do not think we have anything that fits your requires in the Power Switch portfolio. I am moving this back to the signal forum so that your follow up questions can be answered. 

    Thanks,

    Shreyas

  • Hi Ravi,

    Sorry for the delay - 

    This configuration is fine w.r.t. to the voltages being passed through the device - the one thing that you must ensure is that the maximum current is respected - please see below:

    Since you are operating at +/-12V the current limit will be between single supply 12V and +/-15V supplies but in general the current limit will be closer to the +/-15V case. 

    If the voltage at the D pin is +/-10V when the switch connects S2 to D there will be a current of  +/- 10/R_on which will be ~2.5 / 2.6 Ohms which will cause a current surge that could damage the device.  The current through the switch should be within ratings to prevent damage to the switch.

    If you have any other questions please let me know.

    Best,

    Parker Dodson

  • Hello Parker,

    Thank you for the information.

    The signal is the voltage signal, moreover, it's for IV measurements, the signal will changing gradually from -10V to +10V.

    About inrush current, how this can be solved? do you have any suggestions for this problem?

    I checked about the load switch, which can be employed between +/-10 V and S1, or a simple LC-pi network.

    Thank you,

    Ravikumar Jain

  • Hi Ravi,

    The voltage range is fine even if it is ramping as described. 

    To limit the current through the switch a current limiting resistor would need to be used. At max temp (125C) the max current is going to be ~100mA through the switch (this is worst case scenario). That means that if the switch is switched when D is at +/-10V the resistance to ground seen by the D pin needs to be at least 100 ohms so I'd advise the addition of at least a 100 Ohm resistor. This will slow down how Fast D goes from +/-10V to 0V as it is adding resistance and the current will be limited. 

    If the current limiting resistor is not possible then the best solution is  going to be discrete because our signal multiplexers aren't good low-side switches due to them having lower current limits and building a switch out of MOSFETS with higher current limits is going to solve the design problem you are having - I don't have any particular suggestions on a discrete implementation but its looking like there isn't a good solution in IC form at this time. 

    Best,

    Parker Dodson

  • Hello Parker,

    Thank you for the detailed explanation and solution.

    What I understand is below attached, please correct me if it's not according to your explanation.

    Thank you,

    Ravikumar Jain

  • Hi Ravi,

    Yes that is correct. 

    For this application that is where you would most likely have to place the resistor. However this doesn't fully protect in situations where when D closes to S2 there could exist a brief amount of time where a current transient could go through the device and damage it - however this isn't as large of a concern with a 100 Ohm resistor as this is still going to act as a pretty strong pull-down. 

    However if the capacitance is kept low this should prevent the IC from becoming damaged when D is switched to S2. The 100 Ohm resistor should keep the current through the switch to +/-100mA which is the worst case current performance (at T = 125C). Increasing the resistance will give a bit more margin but 100 Ohms should be a good value and I'd try to go with a tolerance of 1% to ensure that 100mA isn't in worst case conditions. 

    Best,

    Parker Dodson

  • Hello Parker,

    Thank you for the explanation again.

    While designing a PCB layout I'll try to keep the device as close as possible to the connector, so that the capacitance concern will be minimum. I think that shouldn't be a big problem in this case with a 100ohm resistor.

    Yes, I'll use the 100ohm with 1% tolerance. 

    Thank you,

    Ravikumar Jain

  • Hi Ravi,

    I am sorry - I seemed to have misjudged the in-rush current due the capacitance. 

    I don't think this part will work because there is a a capacitance at the output of the mux that is ~148pF. There will be an inrush current over the switch that will most likely violate the max spec's and while this pulse is short it may be large enough to damage the switch. I'd advise against using this part.  Our signal switches aren't good for this type of application - this is typically a power switch application but we also seem to not have anything that matches this applications needs in our power switches portfolio as well. At this time I think a discrete solution is a better approach as we don't have an IC that can support this application.

    Best,

    Parker Dodson

  • Hello Parker,

    Thank you for the clarification.
    Can you please suggest some strategies?

    Thank you,

    Ravikumar Jain

  • Hello Parker,

    The problem is inrush current from D to S2, so would this strategy work?

    Thank you,

    Ravikumar Jain

  • Hi Ravikumar,

    TPS2121 cannot handle negative voltage. This device is a 2:1 powerMUX that takes in two DC power input and then passes it to one output.

    It is built for (comparatively) low speed applications and not for transmitting signals.

    I don't believe a PowerMUX will help your application here.

    Unfortunately, it looks like we do not have the device you are looking for. The different teams will look into the best way to fill this gap.

    Thanks,

    Shreyas

  • Hello Shreyas,

    Thank you for the comments.

    In my application I have DC powers only, 1 input will have +/-10V power and the other will have GND.

    For example:

    Input1 Input2 Output

    Sel

    +5 GND +5

    1

    -6 GND -6

    1

    -10V GND GND

    0

    +10 GND +10V

    1

    Thank you,

    Ravi

  • Hi Ravi,

    As mentioned, TPS2121 cannot handle negative voltage. 

    Thanks,

    Shreyas

  • Hello Shreyas,

    I understand, something else....or how it can be resolved?

    Thank you,

    Ravi

  • Hi Ravi,

    I think you will have to use a discrete implementation that involve FETs and gate drivers. TI does offer gate drivers but I recommend looking at them and starting a new thread if you have any questions on the specific gate driver part. This will ensure you get the fastest support possible.

    Thanks,

    Shreyas

  • Hello Shreyas,

    Thank for the suggestion.

    Does TI have some start examples for the same?

    Thank you,

    Ravi

  • Hi Ravi,

    Please reference this overview page for more information about gate drivers: https://www.ti.com/power-management/gate-drivers/overview.html

    Thanks,

    Shreyas