All Hitachi Lateral Fet amplifier for DIY described by Paul Kemble

[wahab]

Originally Posted by donpetru

I said saturate faster to the same input signal comparative with case when the voltage Uds would be more than 10V. For example, my project DP400 requires the input of a voltage 1.8Vrms to charge nominal power. If you use the J-FET with a voltage less than 10V Uds, it can not ensure a proper trip to other transistors saturation voltage, so will tend to increase the THD close to the nominal power, at a slightly higher value in the circuit input and overall the whole amplifier. [/COLOR]

This amp use negative feedback so the Jfets need only 100mV
variation in their drain voltage to correctly make the VAS swinging
the full rail voltage, hence the latter provid most of the gain.

That s about 1% of the 9.6V VGS provided by a 10V zener
in the cascode, wich will be hardly a limiting factor.

[donpetru]

If you choose 2SK170GR then every J-FET will be polarized at a current of approximately 3 mA, which means a total current of 6mA for CCS on the input. In this case, change the values ​​of R1 and R2 as follows:

R2 = (0.8 * Uzener_D2 - 0.7) / I_vas, where I_vas = 0.006 A.
R2 = (0.8 * 15 - 0.7) / 0.006 = 1833 Ohm - you can choose 1k8

And the value of R1 will be:

R1 = Vcss / I_css = 10 / 0.006 = 1666 Ohm - choose a default value of 1K5 or 1k6.

Again, attention to the current input circuit; 6mA current is quite large - using a popular expression - can gather a lot of noise but worth trying.
Let us know here on the experimental stage of this amplifier, some opinions, after the first version of the prototype. Then will apply improvements that will be necessary. Good luck.

Later edit: @wahab, I understand your point of view. It's a good point.

[gaborbela]

Originally Posted by
Again, attention to the current input circuit; 6mA current is quite large - using a popular expression - can gather a lot of noise but worth trying.
Let us know here on the experimental stage of this amplifier, some opinions, after the first version of the prototype. Then will apply improvements that will be necessary. Good luck.

[B

Later edit:[/B] @wahab, I understand your point of view. It's a good point.

Hello

donpetru
Thank you once more!
GR IDSS rating 2.2 to 6mA, I will pick the lowest matched 2PC I have.
I upgraded the circuit after the advised value
I still need some value for start for D2, for R3 if I remember well you wrote 4K7 is wrong value
Do I keep HZ15V or go back to 10V?

@wahab, I understand your point of view. It's a good point

I added a input capacitor, I will start with 4.7uF after the test (I'll hear the amp (sound)tone) I will decide about that.
I will work on the layout so I can start testing what we have now.
I'll buy the missing parts (resistors and diodes) tomorrow.
Thank you for the help to all of you!

Greetings Gabor

[donpetru]

R3 = 2k4, as I said here:
http://www.diyaudio.com/forums/solid...le-naim-5.html
In that post I wrote wrong only this part:

Let's say you choose 2.5mA:
R3 = (0.8 * Uzener_D2 - 0.7) / I_vas
R3 = (0.8 * 15 - 0.7) / 0.0025 = 4520 Ohmi or you can choose standard value 4k7.

Corrected should look like this:

Let's say you choose 2.5mA:
R2 = (0.8 * Uzener_D2 - 0.7) / I_vas
R2 = (0.8 * 15 - 0.7) / 0.0025 = 4520 Ohmi or you can choose standard value 4k7.

So, it was only mistake notation R3 ---> R2.
Now, replace the part above from quote with this:

R2 = (0.8 * Uzener_D2 - 0.7) / I_vas, where I_vas = 0.006 A.
R2 = (0.8 * 15 - 0.7) / 0.006 = 1833 Ohm - you can choose 1k8

Concerning D2, I quote below what I said in a previous post when I determined R3. So, diode D2 wil be 15V zener.

Current in VAS is determined by mathematical formula:
I_vas = 0.8 * Uzener_D2 / R3
or, whether, it requires a current of 5mA then we find the value of R3:
R3 = 0.8 * Uzener_D2 / I_vas
Suppose you chose a zener diode of 15V when R3 is:
R3 = 0.8 * 15 / 0.005 = 2400 Ohm or 2K4

[gaborbela]

Thank you donpetru!!!

Upgraded circuit

Greetings

[gaborbela]

Hello

Just finished the layout for the amplifier.
So soon I can start test her.
I draw with my computer paint program so I use color code for the parts.
Red- mosfet, JFet, transistor
Green- resistors
Blue- capacitors
Grey- trimmers
Yellow- jumper
Purple- diodes
So here is it if someone interested, the layout not tested yet but I drew more than a dozen of layout like these, 99% work at first.

Greetings Gabor

[gaborbela]

Hello
Here is the all black layout.
The right size 139 X 73 mm.
Greetings

[donpetru]

Some transistors (eg. VAS circuit) will need a heat sink. Do not forget that.

[wahab]

Before building this amp , a few points that matters if black smoke
is not the goal...

The biaising trimmer should be about 500R , as already pointed ,
if you use Renesas/Hitachi laterals wich have low Vth.

It seems that there is two critical flaws in the schematic.

The first is that the VAS current source as well as the differential current
source voltage reference , D2 , is fed by one leg of the differential VAS ,
wich is surely a mistake in the schematic.

Second , the differential fet current source has a diode , D1 , in serial with its
gate that is placed the wrong side.

[ashok]

T7, T8,T9 could be bipolar too .
Maybe just keep the input transistors and power output stage as fet's.
Would it make an 'audible' difference ? It certainly would be easier to build !