Post Go back to editing

Reliability issues

Thread Summary

The user experienced issues with the LTC7000 MOSFET driver not fully enhancing the MOSFETs, resulting in only half the intended gate voltage. The final answer suggests adding a diode to the TS pin and considering current-sharing ballasting or using a larger single FET. The accompanying answer shows a scope trace indicating improved performance with a diode, suggesting a flyback diode is necessary for better undershoot protection.
AI Generated Content
Category: Hardware
Product Number: LTC7000A

I have been experiencing issues with the LTC7000 mosfet driver. It is used as a static switch driver in a failsafe system for depowering an electronic speed controller. What happens is that over some extended use the driver will no longer fully enhance the MOSFET, only getting about half the intended gate voltage. The connected ESC is driving a moderately sized motor load at 11-13A and about 8V. So far i am suspecting that it might be because we lack a diode at the TS pin, but please do tell me if there are other obvious issues.

Attached are pictures of the schematic we used, as well as the PCB design:

/resized-image/__size/640x480/__key/communityserver-discussions-components-files/330/pastedimage1783422417790v2.png

/resized-image/__size/640x480/__key/communityserver-discussions-components-files/330/pastedimage1783422605608v3.png

/resized-image/__size/640x480/__key/communityserver-discussions-components-files/330/pastedimage1783422656809v4.png

  • Hi Tormod,
    Yes I think what you noted regarding missing the diode on TS might be part of the issue.
    Another might be since you are using 2 FETs where these might require some type of ballasting to share the current. Or replace with a larger single FET.

    Best Regard,
    Chris
     

  • Here is a scope trace of the TS pin with and without a diode. This is the moment a large motor is switched off by the static switch. it seems that the undershoot is better with a diode, but it could be better with a purpose selected diode(the one i used was just a random one we had laying around). So the lack of a flyback diode was the most likely cause.

  • Hi Tormod,
    Thanks for sharing the scope shot overlay. 
    I agree with you on this. 
    How about the BAT52 as that's one I've noted in many applications for protection?
    If planning on a PCB redesign, make the  traces on both FET's the exact same geometry to the motor. The PCB trace will help with ballasting since there's some small resistance. 
    But I think at this current level you are OK as it does appear to be the missing diode causing this issue.  

    Chris

  • Ill order up your recommended diode, and once i have some new tests done ill get back with a reply here. thank you for the help!