LM4651 Boomer

[DjLeco]

Hi,I received today 4 pieces of National Semiconductor LM4651 ClassD Boomer.
I have some questions about it.

Looking on his PDF,the mazimum power supply voltage is simetrical at +/- 22 (44) volts.

On my schematic, the final mosfets is supplied simetrical at +/- 60 (120) volts.
First question is wich is the maxim simetrically voltage supply (+XX/-XX volts)that I can use to the final mosfets,without destroying the LM4651
Par example supplying The Lm4651 at +/-20 volts,can I supply the final stage mosfets at +/- 80 (160)volts?
On this example using Mosfets with 200 Vds and 30 Amperes.
The second question is about the (L) inductance from outputs.

[mgm2000ro]

SALUT
http://www.national.com/appinfo/audi...19,133,00.html
Nu merge am incercat si eu treaba asta ....cauta in posturile mele vechi.
Is not orching

[DjLeco]

Now other questions about the output inductors.

I tried(for testing inductance value only) to make those inductances from 3 different type of cores,the number of windings(spires) was 10 turns on each.

First core was from GECALLOY,the code is G1W 88 S540,is an R50 core, made from metallic powder.
After 10 turns the inductance was around 12 uH (microhenryes).

The second core was an Kaskhe R40 core made from ferrite.
After 10 turns the inductance was HUGE, around 1,2mH (milihenryes).

The third core is unknown for me but I saw 2 pieces same model in an ClassD JBL A1201. their code is T822 L56uH (uH is like microhenries) made from ferrite too.
And surprise, after 10 turns,the inductance was almost the same like gecalloy core, around 12uH (microhenries)
Npw, the final question is:

Wich core is better for the output inductors, Gecalloy metallic powder with 35-50 turns according to inductande that I need, or Kaskhe feritte core with much less turns 2-8 turns, according to the inductance that I need?

I will attach some pictures of the cores too

[DjLeco]

First picture

[DjLeco]

third

[DjLeco]

4-th

[DjLeco]

5-th

[DjLeco]

6-th

[DjLeco]

7-th

[DjLeco]

last