Andrea Ciuffoli's power follower

[SteveG]

http://www.geocities.com/ResearchTriangle/8231/

Does anyone know if there is a reason not to use many paralleled FETs in this design to achieve less dissipation per device (and spreading the dissipation out over many heatsinks)?
Steve

 

[GRollins]

Parallel devices is cool. Just keep in mind that some point the cumulative capacitance from the gates will start to bog down whatever you're using to drive the output stage. High frequency response will suffer.

Grey

 

[jam]

It could be argued that too many paralled devices could 'fuzzy' up the sound. This is probably more true of tube amplifiers.
Comments anyone?

Jam

 

[SteveG]

Yeah, I was looking at that. Assuming each device is 2200pf, with 6 devices I come up with needing an output impedance of about 120 ohms to drive it to 100Khz. I was thinking of using a follower to drive all the other followers. What do you think?
STeve

 

[jam]

Steve,

This is an interesting problem. The follower would get you the extra bandwith but when I built the Pass A-75 I reduced the number of output devices by half and got better bandwidth and a more focused sound to boot.

Jam

 

[SteveG]

how much power dissipation are you able to get out of a single IRF240? It would be a trade off I guess. It seems that the design is already bandwidth challenged, or at least requires a lot of preamp muscle. I was think of making a cascoded/current sourced (I'll let Jam and Grey fight that one out) balanced line stage to drive it with, probably integrating it into one chassis/channel- more of an input stage than a preamp really(don't know if I already stated that). I would like to keep it to two stages for the sake of simplicity. My problem is that the heatsinks I have are not that great- 2.2c/w. I will be using them in groups, hoping to get this thing cool enough to push enough bias through for 20 watts. So, where do I compromise- More devices in the output with a driver, or less devices w/no driver and less power? My speakers are about 92.5db- 1w, but I do like to turn it up a little. I eventually will move these to just tweeter duty.
Steve

 

[Jocko Homo]

Maybe you could provide a direct link to the circuit in question. Instead of home page filled with gratuitous images. Some of us have slow dial-up accounts that charge out the wazoo when you exceed their menial monthly usage limits.

Jocko

 

[jam]

Steve,

Your speakers seem pretty efficient. I would reduce the number of output devices and up the bias.( No driver)

Jam

 

[paulb]

Maybe you could provide a direct link to the circuit in question.

I think this is it:
http://www.geocities.com/ResearchTriangle/8231/my/Follower_99c.gif

 

[SteveG]

Sorry jocko... took the page off of the word document. I'll be more careful next time.

 

[tiefbassuebertr]

http://www.geocities.com/ResearchTriangle/8231/

Does anyone know if there is a reason not to use many paralleled FETs in this design to achieve less dissipation per device (and spreading the dissipation out over many heatsinks)?
Steve

All links are death in the meantime (10 years old).
here the currently URLs to this excellent project:
http://www.audiodesignguide.com/my/Follower_99c.gif
check also out the associated links under the headline
Single Ended Pure Class A Power Follower Amplifier
about the URL
Andrea Ciuffoli - Home Page
THis topology as "Front-End" could provide an additional sound enhancement:
http://www.diyaudio.com/forums/solid-state/226535-unity-gain-solid-state-buffer.html

 

[tiefbassuebertr]

Parallel devices is cool. Just keep in mind that some point the cumulative capacitance from the gates will start to bog down whatever you're using to drive the output stage. High frequency response will suffer.

Grey

For me the most interesting version for low Cin is those from post 27 about
http://www.diyaudio.com/forums/soli...uras-mosfet-power-follower-3.html#post3804254
(Ladislav Bunta's advanced circuit).