diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Class D (http://www.diyaudio.com/forums/class-d/)
-   -   PROBLEMS WITH IRS2092?????? (http://www.diyaudio.com/forums/class-d/219373-problems-irs2092.html)

sweetperfume 9th September 2012 08:33 AM

PROBLEMS WITH IRS2092??????
 
i bought a kit from sure electronics it worked well....then i thought to take the voltage readings so that i can make sureness when i build my own one if it is correct or not.....i took all the readings but when i went to take the reading of vcc mistakenly the probe touched to VS pin but dint short to vcc and the amp turned dead....i tested this thing on one side of the board but the other side also was off after that dont understand why is it so.....

NEED HELP,,,,,any feedback would be highly appreciated

Tekko 9th September 2012 10:41 AM

You blew it due to a short between two pins due to a probe slip. To get it working again, you need to replace the gate driver and the output fets.

sweetperfume 9th September 2012 08:03 PM

Quote:

Originally Posted by Tekko (Post 3157911)
You blew it due to a short between two pins due to a probe slip. To get it working again, you need to replace the gate driver and the output fets.

one thing for sure the probes were not shorted....and i found that on ref kit the bias voltages of both channels are joint thru a jumper....and was not knowing that ...made me blow another fet.....how much RMS output can be pushed out from IRFI4019

nigelwright7557 10th September 2012 01:52 AM

I found the 2092 to be very touchy about scope probes.
I touched a single pin and blew it up.
If you want to monitor voltages switch the amp off first. Let PSU discharge then connect your probe securely then turn on.
I use a 150r resistor on a croc clip to discharge the supply caps completely.
I also blew up a couple of 2092's by inserting them with undischarged supply capacitors.

I would check the mosfets too as they sometimes go with the 2092.
I have had numerous D-S shorts but also once got a G-S short.

sweetperfume 10th September 2012 06:55 PM

Quote:

Originally Posted by nigelwright7557 (Post 3158707)
I found the 2092 to be very touchy about scope probes.
I touched a single pin and blew it up.
If you want to monitor voltages switch the amp off first. Let PSU discharge then connect your probe securely then turn on.
I use a 150r resistor on a croc clip to discharge the supply caps completely.
I also blew up a couple of 2092's by inserting them with undischarged supply capacitors.

I would check the mosfets too as they sometimes go with the 2092.
I have had numerous D-S shorts but also once got a G-S short.

Thanks nigel

sweetperfume 10th September 2012 07:04 PM

Nigel
Thanks nigel...i think I got it...whatever I need to do with attachment of probes is while the supply is off and the caps are fully discharged...can I monitor the hi and lo voltages with same manner....and can the OCP works well if gates are buffered and output is lets say 1000watts will OCP work very well as it does with low voltages and output without damaging the IC???....and buddy I also have 100's of fets damaged in that manner;)

nigelwright7557 12th September 2012 12:29 AM

Quote:

Originally Posted by sweetperfume (Post 3159657)
Nigel
Thanks nigel...i think I got it...whatever I need to do with attachment of probes is while the supply is off and the caps are fully discharged...can I monitor the hi and lo voltages with same manner....and can the OCP works well if gates are buffered and output is lets say 1000watts will OCP work very well as it does with low voltages and output without damaging the IC???....and buddy I also have 100's of fets damaged in that manner;)

Yes you can look at HI and LO outputs using that method.

I found I had to scale back the OCP threshold to stop the amp cuting out on high power.
Buffering the gates will speed up switch on of mosfets and stop OCP being so sensitive to slow edges.

Buffering the gates will allow driving more mosfets or more capacitive gate mosfets. The gate buffering takes the load off the 2092.
I use TC4420's for gate buffering.

You might find you have a problem with turn off thumps or siren type power down noises when the power supply gets really big. If you can live with that then fine but I couldnt and added a small PIC based circuit to detect low VCC and then turn off CSD using an opto coupler. I used to be a PIC consultant so use them a lot but you migh tget away with a comparator and opto coupler. I also found I couldnt drive CSD direct from the opto-coupler, I had to use a 390r is series with it to stop blowing up the 2092.

sweetperfume 13th September 2012 08:30 PM

Quote:

Originally Posted by nigelwright7557 (Post 3161359)
Yes you can look at HI and LO outputs using that method.

I found I had to scale back the OCP threshold to stop the amp cuting out on high power.
Buffering the gates will speed up switch on of mosfets and stop OCP being so sensitive to slow edges.

Buffering the gates will allow driving more mosfets or more capacitive gate mosfets. The gate buffering takes the load off the 2092.
I use TC4420's for gate buffering.

You might find you have a problem with turn off thumps or siren type power down noises when the power supply gets really big. If you can live with that then fine but I couldnt and added a small PIC based circuit to detect low VCC and then turn off CSD using an opto coupler. I used to be a PIC consultant so use them a lot but you migh tget away with a comparator and opto coupler. I also found I couldnt drive CSD direct from the opto-coupler, I had to use a 390r is series with it to stop blowing up the 2092.

thanks .....it was a great information for me....what do u think how much pairs should i use for getting 1000watts @ 4ohms load with what voltage......i'm getting too humbled as i'm unable to find IRFB4227 around anywhere......can u suggest any other mosfets which i can use for the amp!!!!!!

nigelwright7557 13th September 2012 08:39 PM

I would suggest 2 to 3 pairs for 1000 watts peak into 4 ohms.
Around 63 volts will give you 1000 watts into 4 ohms.

You will get irfb4227 from farnell, RS and Future.

sweetperfume 13th September 2012 08:55 PM

Quote:

Originally Posted by nigelwright7557 (Post 3163883)
I would suggest 2 to 3 pairs for 1000 watts peak into 4 ohms.
Around 63 volts will give you 100 watts into 4 ohms.

You will get irfb4227 from farnell, RS and Future.

63volts will give 100watts????????????????//.....buddy i need to know about 1000watts


All times are GMT. The time now is 07:42 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