Assymmetric outputs in half-bridge - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 25th August 2005, 08:33 PM   #21
Pierre is offline Pierre  France
Banned
 
Join Date: Nov 2004
Location: Paris
It is true that IR2110 are tricky to use. I have a long experience using them in Class-D amplifiers (have a look at "Mosfet reliability in Class-D amplifiers" thread, or something like that, under "Class-D" forum).
First, a lot of mosfets blew up instantaneously just like in this supply, with no apparent reason. I rearranged the layout paying great attention to the gate loop and used more rugged mosfets (moved from OnSemi to IRF) and since them I have had no problems, and I have tested them quite a bit!

All that experience has been applied here, so my layout, at least in that respects to the IR2110 chip and gates should be ok. Anyway, I was cautious enough to add a double footprint in order to add a gate drive transformer in the same board, that I will test soon (I am fed up with so much mosfet and drivers changes!)

This way even the SG3525 chip seems to have died.

If the problem was excessive Vgs voltage due to resonance, what about adding 15V zeners in parallel with Vgs? I know they are not recommended for high frequency, but at 80 KHz...

But if it is due to spurious turn on causing shoot-through, that's a different story...

BTW: Danko, my bootstrap capacitor is 10uF, should be enough. Gate resistors are now 15 ohm.
  Reply With Quote
Old 26th August 2005, 06:01 AM   #22
Pierre is offline Pierre  France
Banned
 
Join Date: Nov 2004
Location: Paris
I have changed the IR2110 to a SD250-4 Coilcraft gate-drive transformer. I still haven't soldered my mosfets, but with 1nF capacitor load, the Vgs waveform looks clean (bipolar, +/-15V amplitude, and a rise time from 0 to 15V of around 350nS, I assume that this will produce much faster rise time in the mosfets, as the Vth level is reached in around 50-80ns)

My mosfets are rated at +/-30v Vgs voltage, so they shouldn't be damaged by the +/-15V bipolar wavefrom... I will cross my fingers.

This afternoon I will test all and tell you how it goes. At least I won't have to change my chips once a day during my testing phase.

Pierre.
  Reply With Quote
Old 26th August 2005, 09:37 PM   #23
Pierre is offline Pierre  France
Banned
 
Join Date: Nov 2004
Location: Paris
More about my progress....
I exchanged the IR2110 by a GDT (Coilcraft sd250-4). The figure shows the waveform of Vgs. It is clean although a bit slow for my taste, and there are a peak that may trigger the mosfet (although the other will be off so I don't think that's an issue).

I have been able to obtain 375W with this and observe a good switching waveform. I haven't tried more but I am sure it will do well. I am almost sure that the mosfets heat up more quickly than with the IR2110 driver, I suppose that the rise/fall times are slower, so
I will try to add a NPN/PNP buffer for the SG3525 to speed up things a bit.

I have finally decreased sw. freq. to 40KHz, and now they become less hot, so switching losses must be dominating.

Unfortunately, when I have soldered the transformer and everything ok, I have done an stupid mistake and something has exploded and broken (but not the mosfets) But as soon as I have repaired all, I will try to add regulation.
Attached Images
File Type: jpg gate_waveform_80khz.jpg (27.0 KB, 268 views)
  Reply With Quote
Old 26th August 2005, 09:41 PM   #24
Pierre is offline Pierre  France
Banned
 
Join Date: Nov 2004
Location: Paris
And this is the switching waveform with 200W load at 80KHz. It still shows discontinuous mode.
Attached Images
File Type: jpg primary_voltage_200w_80khz_gdt.jpg (29.1 KB, 256 views)
  Reply With Quote
Old 26th August 2005, 09:53 PM   #25
Danko is offline Danko  Hungary
diyAudio Member
 
Join Date: Apr 2004
Location: Hungary
How would the waveform look like in continous mode?
__________________
Best regards, Danko
  Reply With Quote
Old 26th August 2005, 09:54 PM   #26
diyAudio Member
 
carvinguy's Avatar
 
Join Date: Jan 2005
Location: So. California, USA
Yes, the SG3525 doesn't have a whole lot of drive current available, so rise times are slower than what you'd like because it takes longer to get charge to the gates. The IR part has 2 amps available for each device compared to 400mA for the SG part. I think your results will be much better after adding the additional driver circuitry.

Matt.
  Reply With Quote
Old 27th August 2005, 04:05 AM   #27
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
The transformer gate drive waveform shows that the transformer is saturating and that leakage inductance is already limiting MOSFET turn-off speed, so it can't be improved. There may be also some cross conduction due to the leakage inductance spike.

You can't just drive MOSFET gates directly from a transformer and expect low losses, this only works with IGBTs that have input capacitances almost an order of magnitude lower than MOSFETs, and with bipolar transistors that are current driven.

It would be a good idea to add some kind of gate buffering, at least for turn-off. Don't try to load the transformer with a lower gate resistor to get faster turn-off since this will only make the leakage inductance spike issue worse.
  Reply With Quote
Old 27th August 2005, 07:50 AM   #28
Pierre is offline Pierre  France
Banned
 
Join Date: Nov 2004
Location: Paris
Eva, you are right in that last affirmation: I already tried to lower the gate resistor. Now is 15 ohm but I tried up to 0 ohm, and yes, it switches a bit faster but the spike rises, so I didn't try with the mains switched on for security.
So, from what you are saying, if I keep the mosfets, I should look for another transformer, right? Can you suggest any commercial model that could work, having a look at the specifications of SD250-4?

Thanks!
  Reply With Quote
Old 29th August 2005, 10:50 AM   #29
Pierre is offline Pierre  France
Banned
 
Join Date: Nov 2004
Location: Paris
mmmm.
Have been thinking about the GDT a little bit more. Eva, how do you know that it is saturated?
It is rated at 375V-us, so as I drive it in a bipolar fashion, (+/-12V), at 40 or even 80 KHz, there shouldn't be any problem, right?
If it is still usable at that frequency, how can I speed it up to minimize losses?

Another question: can I simply substitute the mosfets by IGBTs (properly rated and fast enough for that frequency, of course)?

Thanks!
  Reply With Quote
Old 29th August 2005, 01:28 PM   #30
mzzj is offline mzzj  Finland
diyAudio Member
 
Join Date: Jul 2005
Location: 65N 25E
Quote:
Originally posted by Pierre
mmmm.
Have been thinking about the GDT a little bit more. Eva, how do you know that it is saturated?
It is rated at 375V-us, so as I drive it in a bipolar fashion, (+/-12V), at 40 or even 80 KHz, there shouldn't be any problem, right?
If it is still usable at that frequency, how can I speed it up to minimize losses?

Another question: can I simply substitute the mosfets by IGBTs (properly rated and fast enough for that frequency, of course)?

Thanks!
From drooping dc level on your gate pulses. 10 different choices also why it is happening, I dont know your exact part values and scematics so difficult to say. Do you use dc-decoupling cap on GDT primary? too small dc-decoupling cap or too small chip vcc cap can cause similar dc pulse drooping as core saturation.

I dont see any tricks to speed it up, lower leakage tranny is only option if your layout is already reasonablyy good. I have had good results with DIY gate drive toroids, sub 50ns risetimes to 6nF load with minimal ringing. Sorry, cant recommend any off-the-self products.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Full Bridge X Half Bridge: What's The Best? CHACALPOWERS Power Supplies 8 18th July 2008 04:45 PM
Half Bridge Reference maverick_123 Class D 0 7th December 2007 02:06 AM
half-bridge VS. full-bridge gearheadgene Class D 3 25th May 2007 08:49 PM
The difference of Push-pull,Half-bridge,Full-bridge digi01 Chip Amps 0 8th September 2006 02:10 AM
Whats the difference between full bridge and half-bridge SMPS ? skaara Class D 6 3rd February 2005 07:23 PM


New To Site? Need Help?

All times are GMT. The time now is 11:06 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2