Ticle Zen Amp

[kasey197]

Cool stuff - congrats!

I think you don't need the caps on the outputs - just connect the load to the junction of v2 n v1 like post 11.... wonder if you tried that and had probs?

For the thermal tracking n bias, easiest is to use lateral fets ( so no thermal tracking needed) else for the parts you are using, the pass f5 article shows an elegant solution using a thermistor.

 

[tom42107]

Thanks kasey197. I didn't try hooking the speaker to V2/V1. For some reason it made me nervous, not sure why, but I will try it. And thanks for the F5 tip.

Tom

 

[padamiecki]

Ok, so I just went ahead and built the schematic in post #11. I used IRF510 and IRF9510. No transformer input. Schematic below. I changed some of the resistor values based on what I had and added the two caps because I couldn't figure out how to hook up the speaker without them. I admit it, I got a lot to learn. I used a capacitance multiplier power supply that I've built for an upcoming JLH build. I started with about +-10v and ran up the volume. Out came music. Sounded pretty good. Fets started to get warm so I shut it down, screwed them to a heatsink and tried again. This time +- 12v. Again, music sounding good. I started to worry about thermal runaway (which I got from the padamiecki post) so I shut it down. That's where I stopped. Now I need to figure out how to implement Vbe multiplier or diodes like padamiecki recommends. It did sound good though.
Tom

Hi
just back from hollidays
indeed, you have to put vbe multiplier instead of R6,7
I will post the proposition later/tomorrow
cheers

 

[tom42107]

Thanks padamiecki. A visual would be really helpful for me.

Tom

 

[padamiecki]

Thanks padamiecki. A visual would be really helpful for me.

Tom

this is my proposition:
BD139 on the heatsink
R6 = 120k + 50k precision pot in series
play with R1 to set offset (2 ohm 1-2W noninductive resistors in parallel to set ~0.6 ohm)
do not expect it will sound better than std EFPP
the only one advantage is that you do not have to use insulators during mounting trans on the heatsink.
forgive to rubbish original thread with this proposition, I should set another thread...
in the next post I will send ltspice file and this will do...

 

[padamiecki]

fourier looks worse...

 

[padamiecki]

but this one with laterals should work from scratch (play with R1 to right offset):
laterals have negative temperature coeff, should be not thermal runaway.
fourier 15V pp looks amazing