Need help to simulate my MOSFET Amplifier using TINA

[udhay_cit]

I don't have much experience with Audio amplifier design. I found one 25W MOSFET audio amplifier circuit. 25 Watt MosFet Audio Amplifier - RED - Page2
I like to start this design with simulation. My TINA simulation get clipping at output.
The circuit is proven & working but i don't know what is wrong with the simulation...
Facts and Figures - RED

Regards
Udhay

 

[Mooly]

I don't use TINA but the amp seems to have a voltage gain of around 70 (R7/R6) and runs on 66 volts .

So applying 2 volts peak to peak as an input will clip on those rails. Try reducing the input voltage down to say 200mv peak to peak initially and see what happens.

 

[udhay_cit]

Mooly,
Thanks for the reply. Yes i got the output for 200mV & 300mV Input. Is it Gain=70 too high? What will be the recommended input voltage or how much gain i can set?

Regards
Udhay

 

[udhay_cit]

My simulation waveforms....

 

[Mooly]

The gain of 70 is fine and the amp and its compensation and so on will have been designed around that figure. Although you could alter the value of R6 (increasing lowers the gain) I wouldn't do so in practice. It just means the amp has a high input sensitivity and as it says, no need for a preamp.

 

[Mooly]

Not to sure about running HEXFET's with a simple resistive bias generator (the preset pot) as it will be very prone to thermal runaway. Normally a transistor vbe multiplier type circuit is used.

If you build the amp for real be very aware of this

 

[udhay_cit]

PCB Placement

Can i use this for 5.1 Surround? Can i connect it into my DVD player or PC sound card? (Sorry i don't know the p-p signal Output voltage) without clip?...

 

[Mooly]

Can i use this for 5.1 Surround? Can i connect it into my DVD player or PC sound card? (Sorry i don't know the p-p signal Output voltage) without clip?...

Yes you can. The high sensitivity should work with most source components.

Just remember what I said about the bias stability though

 

[magnoman]

udhay

I dont understand why the output waveforms from post #4 look as they do.

Are you sure the feedback resistor (R6, 33K) is actually connected to the output node?

Thanks
-Antonio

 

[Mooly]

I threw the circuit into LTspice. This is at the point of clipping into 8 ohms.

 

[udhay_cit]

yes R6 is connected to the output node....

I think i selected a wrong simulation tool. I try to simulate the circuit using LTspice....

Meanwhile can anyone verify my PCB routing. This PCB is designed only for compactness...

Regards
Udhay

 

[Mooly]

Here's the LTspice file to get you started. You need to put the .text file in the same folder as the main file as this contains the transistor models.

 

[Boscoe]

This circuit will blow up. No thermal run away compensation...

 

[Ouroboros]

The TINA sim works perfectly, but with an Iq of the output stage at only 2mA (with the 50% default on the pot), the distortion for a 200mV (peak) input signal is high at 0.16%. (see attached TINA transient run).

 

[Ouroboros]

Sorry, at 50% setting of the pot, the Iq of the output stage is only 75uA!

Setting the pot at 42% gives an Iq of 150mA, but as others have said, the quiescent varies wildly with temperature. This is to be expected when using VMOS FETs with no Source resistors. A proper temperature compensated bias is needed at least.