• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 514 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

OPA4991: OPA4991

HH1992
Prodigy 120 points

Part Number: OPA4991

Tool/software:

Hi,


I'm using this circuit as an overvoltage protection circuit. I've tested it, and it works. However, there's an issue: when a load is connected, the gate voltage drops to around 3V(still Mosfet is on ). When I remove the load, the gate voltage switches between 0V and 10V as expected.

Why does the gate voltage drop when the load is connected? How can I solve this issue?

Thanks

  • 0
    TI__Guru* 77501 points

    Hi Gholam,

    HH1992 said:
    when a load is connected, the gate voltage drops to around 3V(still Mosfet is on ). When I remove the load, the gate voltage switches between 0V and 10V as expected.

    When the gate voltage is turn on, it should be measured at approx. 10Vdc. However, when the Mosfet is on, your 48Vdc may droop significantly because it is sourcing approx. 48V/10 = 4.8Adc. Somehow it affects your 10V supply rail. Your 10V supply rail is likely generated from 48Vdc voltage source. My guess is that your 10Vdc is no longer 10Vdc when 48Vdc is sourcing 4.8A or mosfet is ON. Also please add LPF filters at the input and increase the bypass capacitor C44 from 100nF to 1uF or 2.2uF. 

    As it is configured, Vin- > Vin+ and hence the TP13 should go low when the comparator is turned on (likely a typo in the schematic, should be Vin+ > Vin- ). Please make sure that 48Vdc is stiff in voltage and does not droop when current is sourcing when load is ON, where 48V*5A = 240W power supply. You probably need at least 300W 48Vdc or greater power supply for the application. 

    If you have other questions, please let me know. 

    Best,

    Raymond

  • 0 in reply to Raymond Zhang1
    Prodigy 120 points

    Hi Raymond,
    I've checked the 10V supply rail, and it remains mostly stable when the MOSFET is ON. Could something else be affecting the gate voltage?
    Thanks

  • 0 in reply to HH1992
    Prodigy 120 points

    I've checked the gate using an oscilloscope, and this is the result when I used a power supply and also another high-current supply.

      
    3A power  supply                                                                                                       10A Power Supply                 
         

  • 0 in reply to HH1992
    TI__Guru* 77501 points

    Hi Gholam,

    The issues are your 48Vdc and 10Vdc. Please filter out the noises from 48Vdc and 10Vdc right at the pin5 of IC6 op amp.

    The 10Vdc @3A power supply is terrible with high frequency ripple voltage (48Vdc is likely a switcher, if it is 10Vdc is regulated from 48Vdc). Please add 100-1000uF large capacitor next to the 48Vdc, and at 10Vdc rail next to IC6's pin5, place 2.2uF-10uF low ESR bypass capacitor to decouple the high frequency noise, see the image below. 

    I assumed that this plot is shown 10Vdc with 3A power supply at pin5 of IC6 op amp. What is the triangle waveform measured from?

    You need to attenuate the high frequency noise first before you are able to get conclusive results. When you measure the 10Vdc and 48Vdc, please limit the scope's BW to 20MHz or lower, As it is right now, DC + high frequency noise is shown on the scope capture. 

    48Vdc/10 ohm > 3A and your power supply is NOT rated for the load application. Please change R31 from 10ohm to >16ohm, say 20ohm or 25 ohm range for the 3A  rated power supply. 

    Best,

    Raymond

  • 0 in reply to Raymond Zhang1
    Prodigy 120 points

    Hi Raymond,

    I think I didn't explain it clearly earlier, or the images weren't good. Both images actually show the gate voltages. The noisy one was measured while the 48V was supplied by a 3A power supply. The other one (the triangular waveform) was measured using another power supply with a higher current rating.

    I measured 10V at the gate, and it was clean—no issues there.

    Also, without any load connected, the MOSFET switches without any issue, and the gate voltage is 10V. However, when I connect the load, the gate voltage changes to a triangular shape, with a peak of only 5V.

  • 0 in reply to HH1992
    TI__Guru* 77501 points

    Hi Gholam, 

    HH1992 said:
    However, when I connect the load, the gate voltage changes to a triangular shape, with a peak of only 5V.

    This is very common characteristics in poorly designed switching power supply. The ripple voltage is relative low when there is no load or light load. With a heavy load, the switching noise is increased and so is the regulated ripple voltage and output current. 

    You have to find a way to filter out the ripple voltage from the setup. You can tell if your switching power supply is any good and doing its job in %regulation. You can select one of your load, say 10ohm load (>250W @48Vdc - power resistors ). Connect the 10ohm load to the switcher and monitor the output voltage of 48Vdc. The switcher's output voltage will droop significantly, if the switcher is only able to source 3A. By your load requirements, the Rdson should be in mOhm, so 48Vdc should be applied across the load or 10ohm, which it will source approx. 3.8A (depending on R's tolerance in ohm's law). 

    If 48Vdc switcher droops significantly, you have to find a better power supply. The 10Vdc configured for OPA4991 has to be stable as well. If the 10Vdc is doing similar things, the OPA4991's output will alter and doing strange things when applying saturated gate voltage (close to 10Vdc between gate and source). 

    You have to resolve both regulated supply first before you attempt to control the op amp's voltage. Otherwise you are unable to find the root causes in your setup. 

    Best,

    Raymond