Actions
Forum Thread Details
  • Replies 5 replies
  • Subscribers 56 subscribers
  • Views 985 views
  • Users 0 members are here
Related

Triac protection against short-circuit

femtotech

Hi!

I have a board with the basic triac driver circuits as described at page 8 (fig. 14) of MOC3052:

 

https://www.fairchildsemi.com/datasheets/MO/MOC3052M.pdf

 

The triac is a BTA204S-600D (600V 4A).

The load is a small electric motor (less than 250W) but there are no snubbers because the leakage is not acceptable for the customer.

 

I tested a resistive load up to 750W with no problems. The circuits works well also with lighter inductive loads, about 100W.

With the original one (250W) after few commutations the triac goes in short-circuit - I guess due to the back-EMF when disabling the triac.

 

I tried with a very big triac (25A, 250A peak) and there are no problems, but of course it's not a good solution.

I also tried with a TVS (275V) to clamp the VAC across the line (between L and N, and between T1 and T2) with no chance.

Also a snubber between L and N does nothing.

 

Any idea?

Parents

  • The spike injects current into the control terminal latching it back on.  The TVS has to change state in order to protect, the triac turns on in the meantime.  A tiny snubber might get the thing to work.  You are using a very-sensitive gate triac which is contrary to your interests.  Being easy to turn on means that it is easy to turn on spuriously.  Back-to-back SCRs are the way the hip kids are doin' it, nowadays.

  • in reply to D_Hersey

    Thank you for your answer.

    Ok, the spikes spuriously turn on the triac, but why it breaks? I mean, perhaps the spikes are so high they burn the control terminal? Because the triac appears in short-circuit between T1 and T2 when removed from the board.

    About the snubbers, I tried:

     

    - across L+N (across the load device) with no effect

     

    - across T1 and T2: it works, but the small leakage is not accepted by the customer, because there is a little amount of current on the device even when the triac is off (due to the C impedance, of course)

     

    Do you suggest a different position?

    Or another triac I could use for such a load (they are in DPAK).

  • in reply to femtotech

    You could try a choke in series with the triac.  You'd normally fit one to stop noise from the triac conducting out, but it might do it.

    Out of interest, why is the client objecting to leakage.  Is it for a 'green' or medical product?  There is permitted LN (&E) leakage for medical.

    There is an interesting note on the datasheet on type approvals.  As part of the CE approval (assuming your device will be used in the EU), your circuit may be subjected to surges, so it looks like you might need the snubber.  Depends which parts of the EMI standard / target market the product is for.

     

    This is a possible explanation for device destruction.  The motor emf, which triggers the triac conduction, could occur when the mains is at say -230V, and the EMF at say +1000V.  Way outside the triac operating region.

Reply
  • in reply to femtotech

    You could try a choke in series with the triac.  You'd normally fit one to stop noise from the triac conducting out, but it might do it.

    Out of interest, why is the client objecting to leakage.  Is it for a 'green' or medical product?  There is permitted LN (&E) leakage for medical.

    There is an interesting note on the datasheet on type approvals.  As part of the CE approval (assuming your device will be used in the EU), your circuit may be subjected to surges, so it looks like you might need the snubber.  Depends which parts of the EMI standard / target market the product is for.

     

    This is a possible explanation for device destruction.  The motor emf, which triggers the triac conduction, could occur when the mains is at say -230V, and the EMF at say +1000V.  Way outside the triac operating region.

Children
No Data