Class-D Amp with IRS2092 - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

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 3rd February 2008, 09:11 PM   #21
mag is offline mag  Switzerland
diyAudio Member
 
Join Date: Jan 2008
Location: Chiasso
Attached the drain waveforms
The scale is 10V/div 2us/div.
I see lots of ringing there. Is my transformer the source of the problem?

What do you think about those drain waveforms?

thank you
Attached Images
File Type: jpg drain_2us_10v.jpg (64.2 KB, 5719 views)
  Reply With Quote
Old 4th February 2008, 09:54 PM   #22
mag is offline mag  Switzerland
diyAudio Member
 
Join Date: Jan 2008
Location: Chiasso
Probably I have found the root cause of the mosfets heat up.
The huge gate to source capacitance of the IRFB4110 (about 10nF each device) creates a very high current peak trough the shunt when the gate is charged. My gate driver is rated 9A and I have a 3.3ohm series resistor to the gates.

I have tried to increase the value that resistor up to 10ohms, the voltage spikes on the shunt reduces from about 2V to 0.9V.

I have even tried to replace these mosfets with the IRFZ24N (I know that it is not enough for this application but I have only this part available now). The IRFZ24N has a gate capacitance of about 370pF and the voltage spikes are even more reduced. The gain and drain waveforms are much better now with the IRFZ24N.

Do you think I should change the SMPS mosfets?
The same mosfet I use for the class-D can do the job (IRFB4227)

thank you
  Reply With Quote
Old 4th February 2008, 10:34 PM   #23
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
Ouch! You have still a lot to learn.

The 1uF capacitor directly in parallel with the shunt is a completely crazy idea as a means to compensate inductance. I told you to use a RC filter. What do you get if you place R and L in parallel???

A 9A gate driver is also a crazy idea for your application. The gates have to be charged much slower than that. You could even drive them directly from a SG3525 with cleverly used ferrite beads.

Go back to the IRFB4110 and find a reasonable gate drive speed and dead time.

You are going to need an oscilloscope with at least 40Mhz bandwidth and digital storage mode comes very handy too.

Next time you take oscilloscope pictures, adjust the image to show the edges of the waveforms because nothing interesting happens during the flat portions.
__________________
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 4th February 2008, 11:27 PM   #24
mag is offline mag  Switzerland
diyAudio Member
 
Join Date: Jan 2008
Location: Chiasso
No the 1uF is not in parallel with the shunt.
There is a low pass filter with R=12ohm and C=1uF (see attached schematics). The other 22ohm resistor is used as a voltage divider. The current limits trips on on the SG3526 when the voltage is > 100mV. I can not use higher values resistor because the input of the SG3526 is intended for a low impedance load (<50ohms). The R=12ohm and C=1uF create a pole at 13.2kHz, more or less the same frequency of the shunt zero (15.9kHz for 20nH and 0.002ohm).

Probably this week I can access to a 100MHz BW oscilloscope at my job. I will try to capture the waveform with that scope.

Do you think that the SG3526 outputs are strong enough to drive two IRFB4110 fast enough at 110kHz? (the Qg is 150nC).

I will try to remove the driver and see what happens. For the ferrite bead you mean using it in place of the gate resistor?
Is it used just to avoid possible oscillations of the mosfet or it has some other purposes too?

The dead time you see in the waveforms is huge, I know this. It was just to be sure that no cross-conduction appears

thank you
Attached Files
File Type: pdf lpf_schematics.pdf (19.8 KB, 2129 views)
  Reply With Quote
Old 4th February 2008, 11:46 PM   #25
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
Then the LPF is OK.

Concerning the gate drive, leave the 9A ICs in use but add a diode in parallel with a resistor in series with the existing gate resistors so that you can set independent turn-on and turn-off "speeds".

3.3 ohms is probably producing a way too fast turn-on and a barely fast enough turn-off, 10 to 50 ohms may improve things a lot, but keeping the 3.3 ohms for turn off. Try to aim for cross-over times in the 100ns to 200ns range.

Ferrite beads as gate drive "ballasts" have a series of unique advantages. You may want to use the search engine of the forum to find and read my comments on them. Anyway, the best way to learn about them is to try them in various sizes and study the resulting waveforms.
__________________
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 5th February 2008, 10:57 AM   #26
mag is offline mag  Switzerland
diyAudio Member
 
Join Date: Jan 2008
Location: Chiasso
Just to be sure you mean to try a circuit as the one shown in the
attached schematics.

I think that the diode must be a schottky type with at least 1A continous current. I have a bunch of BYV10-40 schottky diodes, can they do the job?

thanks
Attached Files
File Type: pdf gate_drive.pdf (17.2 KB, 1781 views)
  Reply With Quote
Old 5th February 2008, 11:02 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
Yes, you got it right

The diode is not required to be schottky.
__________________
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 5th February 2008, 11:03 AM   #28
mag is offline mag  Switzerland
diyAudio Member
 
Join Date: Jan 2008
Location: Chiasso
can I use a standard 1n4004 or it is too slow?
othewise I need to find our some fast diodes because I don't have them available now
  Reply With Quote
Old 5th February 2008, 11:09 AM   #29
IVX is offline IVX  Russian Federation
diyAudio Member
 
IVX's Avatar
 
Join Date: Jul 2003
Location: Everywhere (Buddhist's context)
4004 too slow, 1N4148 or similar typically used.
__________________
Best regards,
E1.
  Reply With Quote
Old 5th February 2008, 11:16 AM   #30
mag is offline mag  Switzerland
diyAudio Member
 
Join Date: Jan 2008
Location: Chiasso
4148? with only 100mA average current and 500mA peak I think it will blow up with such current...
I will search for a fast diode rated at least 1A continuous
  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
IRS2092 customizable DarkOne Class D 55 30th July 2014 11:15 PM
IRS2092 Amp Design: Please help! Devendra Class D 11 17th March 2011 10:10 AM
Another amp with IRS2092 mag Class D 23 18th June 2010 08:25 PM
IRS2092 Amplifier zox2003 Class D 22 31st August 2009 10:49 PM
Any suggestion for IRS2092 ja2dhc Class D 4 2nd August 2008 11:53 AM


New To Site? Need Help?

All times are GMT. The time now is 08:12 PM.


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