Jump to content


Photo

Gate-driver Recommendations For Phase Control Of Medium Ss


  • Please log in to reply
9 replies to this topic

#1 bookevg_ekra

bookevg_ekra

    Junior Member

  • Full Member
  • PipPip
  • 7 posts

Posted 11 July 2007 - 01:26 PM

When ss starts work the firing angle is 150-165 degree. If duration of firing angle is equal 120 degree then 100mA(the leakage current)*1500V = 150W = it's to lead to device failure. What are ways to solve this problem?

#2 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,601 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 11 July 2007 - 06:56 PM

Hello bookevg_ekra

Where does the 100mA leakage current come from?
SCRs do not usually have appreciable leakage current. I have never experienced a leakage current so high.

Best regards,

#3 bookevg_ekra

bookevg_ekra

    Junior Member

  • Full Member
  • PipPip
  • 7 posts

Posted 12 July 2007 - 05:10 AM

Look to ABB's application note:
Gate-drive Recommendations for Phase Control ...

#4 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,601 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 12 July 2007 - 08:36 AM

Can you give a link to this?

#5 bookevg_ekra

bookevg_ekra

    Junior Member

  • Full Member
  • PipPip
  • 7 posts

Posted 12 July 2007 - 12:38 PM

You must use to search: Gate-drive Recommendations for Phase Control ...
here
Word by word
When the thyristor is in its reverse-blocking state, there is no risk of triggering neither by a positive nor by a
negative gate-current. However, applying reverse anode-voltage and positive gate-current simultaneously will
lead to an increased leakage current. A reverse-biased thyristor will act as a transistor with a gain in the
range of 0.1 0.5. A gate current of 1 A can cause an increase in the leakage current of some hundreds of
mAs even at room temperature. This leakage current can, in certain circumstances, lead to device failure and
should be avoided. An example of the transistor action is seen in Fig. 10.
Fig. 10 Leakage current increase due to gate current for 5STP 25L5200 at VR = 1500 V, IG = 2 A (channel 2), Tj = 100 C, Ir = 200mA/div (channel cool.gif, t = 20 μs/div


#6 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,601 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 12 July 2007 - 07:41 PM

Hello book_ekra

Thank you for the link.
Now I understand your question.

Yes, if you have a reverse bias on the SCR, and a high forward current on the gate, it acts as a very low gain transistor and you will have an increase in the current through the anode. If this was sustained for a period of time at a high voltage and a relatively high reverse current, then the power dissipation would certainly lead to an increase in the junction temperature of the die and could lead to premature failure if this power dissipation was not taken into account.

I assume that you are considering the effect of the reverse current on SCRs in a soft starter as your post is in the soft starter section, so in that case, you have a reverse parallel SCR across the SCR under analysis.
If you are applying the gate signal to both SCRs at the same time, then you would only be applying a Gate signal when you are expecting one of the SCRs to be turned ON. In this case, the actual period when Both SCRs are turned OFF and gate current is ON is virtualy zero so there is no problem.

If you are independantly controlling the SCRs, and have a large overhang on the gate signals to each SCR, then this is certainly a factor that needs to be considered. Remember that the period is only while Both the SCRs are not conducting. The average gate current can be reduced by the use of picket fence firing with control over both the current amplitude and the mark-space ratio of the pulses. Reducing the overhang to the minimum required will also reduce the negative current.

If you had full gate current (hard fire) for the whole negative cycle on the SCR, then at a conduction angle of 150 degrees, you have an OFF time of 30 degrees so you will have high reverse power dissipation for 16% of the cycle. The average power dissipated in your quoted case would be around 10W which is not so bad.

Best regards,

#7 jraef

jraef

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 683 posts
  • Gender:Male
  • Location:USA, California

Posted 13 July 2007 - 03:56 AM

Another resource for you...

http://www.geocities...54/scrsfire.pdf
"He's not dead, he's just pinin' for the fjords!"

#8 bookevg_ekra

bookevg_ekra

    Junior Member

  • Full Member
  • PipPip
  • 7 posts

Posted 13 July 2007 - 11:01 AM

1.When soft-starter starts to work the initial firing angle (delay angle or alpha) is about 120 degree. It's discontinuous conduction mode. If duration of firing angle is equal 120 degree the thyristor is in its reverse-blocking state about 60 degree. So 60/360*0.1*1500=25W
2.When soft-starter starts to work the initial firing angle (delay angle or alpha) is about 150 degree. It's discontinuous conduction mode. If duration of firing angle is equal 120 degree the thyristor is in its reverse-blocking state about 90 degree. So 90/360*0.1*1500=35,5W
Then firing angle shift to 15-20 degree (alpha minimum) when motor is speeding-up.

#9 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,601 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 15 July 2007 - 04:40 AM

Hello bookevg_ekra

1.
QUOTE
If duration of firing angle is equal 120 degree the thyristor is in its reverse-blocking state about 60 degree. So 60/360*0.1*1500=25W

Yes this is true if you are applying gate current through all the reverse blocking period. I many cases, the two SCRs are controlled equally so there would be no gate current during this period. If there was, the forward SCR would be turned ON and there would be no reverse voltage to block.
Additionally, the SCR is blocking a sinewave, the average voltage is a function of the non conduction angle and the shape of the sinewave. The chances are, that if you have a peak voltage of 1500 volts, the average during the non conduction period may be less than half of this.

2. You apply the gate current when you want the SCR to turn ON. I do not see why you would apply the gate signal for all of the reverse blocking non conduction period. Part of it maybe, but not all.

Another point to remember, with the figures that you are quoting, you are talking about a high power SCR which in normal operation is going to dissipate much more that say the 35 watts in reverse, so the heatsinking etc must be easily able to absorb this power level.

Low power SCRs have a much lower gate current and reverse blocking power dissipation.

Best regards,

#10 bookevg_ekra

bookevg_ekra

    Junior Member

  • Full Member
  • PipPip
  • 7 posts

Posted 17 July 2007 - 09:38 AM

Thanks for answers.
Our developing of medium voltage soft-starter is completed. We ship and installation our soft-starters.
Now we will start to develop load-commutated invertor (static frequency converter) for medium voltage synchronous motors.





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users