Anybody willing to simulate ?

[ingenieus]

I've drawn a blank with SPICE models for the BD829 and BD830. Can I substitute them with something else like the MJE243 and MJE253? The fT is a lot lower though.

[wahab]

High frequency THD is awfull , wich is a consequence of the "design choices" ,
mainly the low differential current.

As spoted by Godfrey the CCS resistor should be reduced to 2.2K while
the 33K must be also reduced to 3.3K or suppressed.
In the meantime the current mirror emitter resistor should be reduced
to 220R.

Of course the amp become way faster with theses mods , so the
compensation has to be adapted , best is to suppress the shunt network
consisting of 27R + 470pf , replace the 22pF cap by a 100pF and replace
the serial 1K resistor by a short.

Also , a zobel output network should be added.

Below is the simulated THD at 10KHZ 25W RMS 8R.

[wahab]

Originally Posted by ingenieus

I've drawn a blank with SPICE models for the BD829 and BD830. Can I substitute them with something else like the MJE243 and MJE253? The fT is a lot lower though.

The 829/830 are in fact the 100V version of the BD139/140.

[sakis]

mje 243 looks fine but only slower

[sakis]

ok wahab ... thanks for the advice so far will look in to it once more and get back

[ingenieus]

I found the missing models for the BD829/BD830. Here they are...

.MODEL BD829 NPN (IS=9.530E-0019 NF=714M NR=714M RE=170M RC=1 RB=10 VAF=40 VAR=20 ISE=236F ISC=236F ISS=0 NE=1.5 NC=1.5 NS=1 BF=74.9 BR=5 IKF=878M IKR=878M CJC=131P CJE=131P CJS=0 VJC=515M VJE=515M VJS=750M MJC=330M MJE=330M MJS=0 TF=637P TR=82.8N EG=1.11 KF=0 AF=1)

.MODEL BD830 PNP (IS=1.130E-0018 NF=714M NR=714M RE=159M RC=1 RB=10 VAF=40 VAR=20 ISE=2.79F ISC=2.79F ISS=0 NE=1.24 NC=1.24 NS=1 BF=106 BR=5 IKF=525M IKR=525M CJC=131P CJE=131P CJS=0 VJC=515M VJE=515M VJS=750M MJC=330M MJE=330M MJS=0 TF=2.12N TR=276N EG=1.11 KF=0 AF=1)

Google/Yahoo was of no help. Don't ask me where I got them, hehehe.

[ingenieus]

Here are the results that I got:

THD for 25W RMS into 8 ohm at 1kHz is 0.005640%.

[sakis]

now ...wahab obviously we are missing something here .

Ingenious ...thanks very much for your time and effort did you see something else that troubles you ?

kind regards sakis

[discrete]

Originally Posted by sakis

now ...wahab obviously we are missing something here .

10kHz vs. 1kHz THD?

[ingenieus]

I am afraid that the slew rate is rather abysmal. Even at 6kHz, the waveform is already badly distorted. At higher frequencies the output is a triangle wave, for a sine wave input.

I left out the 330pF cap by accident, but the results remain the same if I put it in.