+15V pre-regulate for classD amp

[opor]

From this circuit,it is +15V pre-regulate for classD amp . R114B and R115B are calculated base on -B=-50V . So if -B =-90V what should R114B and R115B be? And how to calculate?

[opor]

Sorry ,here is circuit.

[mag]

Hi opor,
For R115 is easy but for R114 it depens on how much current you need from the 15V.

with R115=15k and -B=-50V we can calculate the current in the zener diode.

Iz=(50V-15V)/15k=2.3mA.

Now let's keep 2.3mA in the zener diode and calculate R115 to have Iz=2.3mA at -B=-90V.

R115=(90V-15V)/2.3mA=33kohm.

R114 helps the transistor to dissipate the power, start with something like
2.2kohm and check the temperature rise of the resistor and the transistor.

Is this regulator used for the VCC supply of the IRS2092 IC?

Supposing that current needed from VCC is 25mA you will have the following
figures from power dissipation:

1: -B=50V Iout=25mA

P_R114=25mA*25mA*1kohm=0.625W (1W resistor is needed)
P_trans=Vce*Ic=(50V-25mA*1kohm-15V)*25mA=0.25W

2: -B=90V Iout=25mA

P_R114=25mA*25mA*2.2kohm=1.37W (3W resistor is needed)
P_trans=Vce*Ic=(90V-25mA*2.2kohm-15V)*25mA=0.5W

These calculations applies if Iout=25mA if it is more you need to select a proper value for R114.

Btw: if it is the IRS2092, don't even try to power it up at +/-90V, it will blow up. You need at least some totem poles drivers to drive the mosfets.

[opor]

Dear Mag,
Do you have any idea ,is this schematic work or not ? It is hard to get IRFB4020 .If I use IRFP250 instead . Which component I have to change?

[mag]

Hi opor,
IRFP250 has much higher Qg compared to IRFB4020 (123nC vs 18nC).
On top of that is slower and also its body diode is slower.
Regarding conduction losses they are more or less equivalent (80mohm vs. 75mohm).

Looking at your schematics I have serious doubt about getting 700W rms at +/- 90Vdc on 4ohm load. A single pair of IRFB4020 (or even worse IRFP250) are not enough due to the high conduction losses.

On top of that driving the IRFP250 with it 123nC of gate charge needs for sure some external totem pole drivers in order to switch it reasonably fast. The drivers integrated in the IRS2092 are quite weak.

I suggest you to go on IRFB4227 which is the new generation of IRFB4020 with reduced Rds_on (19.7mohm) and reasonably fast switching characteristics.

I have tried to go to +/-50V with IRS2092 + IRFB4227 and I could easily get more than 250Wrms into 4ohm. I have tried also to connect two of these modules in bridge and I could get around 1kW into 4ohm with +/-50V rails.

Then after changing some components I have tried to increase the rails at
+/-80V and everything blow up... The IRS2092 has not enough drive current to switch the IRFB4227 reasonably fast when running at +/- 80V the swiching losses are huge and there is the possibility of cross conduction of the two mosfets.

You have to add two external totem poles or two mos drivers rated at at least 4A if you want to go at +/-90V with IRS2092. If you want to stay on IRFP250 use 2 of them in parallel and care at the driver.

[Cristi]

regardless of the MOS-FET's which you use or if there will be used a totem-pole drivers or not, do not not exceed +-75V for supply rails or +-85 V in BTL mode. the voltage spikes and pumping effect will kill the ic.

[opor]

Hi Mag,

How about IRF640 and IRF540? Is it good for this design?