N-channel amp.

Status
Not open for further replies.
N-channel-sound...

I have been comparing the N-channel with amps from
Dynamic Precision, http://www.dynamicprecision.no and Electrocompaniet, http://www.electrocompaniet.no, (and an old Adcom), It seems to be a bit more forward and "pushy" compared w. the master of analytic, (Dynamic Precision), and has a quick, deep and hard bass. And, just as important, I have not singled out any "flaws", and is a joy to use. Not even a problem w. pops or noice when turning on or off.

My advice (for what it's worth) TRY IT!

Arne K
NORWAY

BTW. I use AudioVector S6 speakers, http://www.audiovector.dk,
front end: modified Marantz CD-14, Monarchy Audio DAC, Oracle Delphi + other modified or diy-units.....
 
Re: Less power?

mik said:
Can this amplifier give less output power with lower supply rail or there is another way to do this because I don't so much power and since everybody telling nothing but good about this amp it is worth to givit a try.:xeye:

CAN THIS AMP VORK ON A 2 X 90 V ??? CAN IT BE MODIFIED VOR THAT VOLTAGE ?
 
Re: Less power?

mik said:
Can this amplifier give less output power with lower supply rail or there is another way to do this because I don't so much power and since everybody telling nothing but good about this amp it is worth to givit a try.:xeye:

CAN THIS AMP VORK ON A 2 X 90 V ??? CAN IT BE MODIFIED VOR THAT VOLTAGE ?
 
Re: MTW35N15E Mosfet in N-Channel amp???

2002tii said:
Is it OK to use MTW35N15E mosfet`s in this amp?
What will be the difference comparing to the IRFP240?
I have a lot of this mosfet (840pieces) and would like to use it to something (like an ampliifer).
Datasheet:
http://www.onsemi.com/pub/Collateral/MTW35N15E-D.PDF

If I don't remember wrong you should choose FET's with Rdson as high as possible (for your power needs). Less than 0,3-0,5 ohms is pure waste. Think decibels in power loss when you calculate. 1 or 2 dB's doesn't matter.

Low Rdson FET's have large caps and can be hard to get working in a source follower (self oscillation). This FET have max 5 nF (ugh!) as Cgs and up to 300 pF as Cgd. The old Hitachi from stone age had less than a factor 10 at all values!

As a conclusion: You may succeed but not without testing and development.
 
The gate C in the N-channel design has little to do with it Objective performance, I have used N-channel devices with Gate C of 7nf with great results.
It is a fact that this design requires Mosfets with a high forward transconductance. Mosfets such as the Hitachi 2sk1058 Lateral type. Just won't work as they have a siemens rating of 1

Types as the MTW35N15E have a rating of 18. Which is very high.
Will have no trouble working in this design.
Its is also very important that the supply voltage with the existing design not exceed 75 volts.
At least not without resigning the thermal tracking circuit as the Gate threshold voltage on these types of mosfets drop as the supply voltage increases closer to 100 volts.
Having said all of this peranders is correct in saying high Rson devices will be more thermally stable in designs such as the n-channel design.

Cheers

Anthony Holton
www.aussieamplifiers.com
 
: O )
 

Attachments

  • nchannel.jpg
    nchannel.jpg
    75.6 KB · Views: 1,129
Thanks for your answer Anthony.

I thought I read somewhere on your web site that one could use any industrial Mosfet.
Not just my bad memory then 🙂

Can I compansate for the low Rdson by using bias resistors with higher resistance?
Or do I just have to use bigger heatsinks???

Or doesn`t this help at all?

I`m not 100% sure how an thermally unstable amp behaves, and what I can do to cure it...
 
I have 2 problems so far.

One my amd gets hot with no load or any signal input why. i have a new heat sink on it with 1/4 base with 40 finns on it and 3" on all directions.
2 when my amp is on no input but a speaker atached there is a buzz. when i hook up my preamp i get a faint hiss and a litle bit of humm

my transformer is a toroidal 300va 2 x 50v
and 2 x 15,000mf caps @125v

ANy sujestions.
 
Status
Not open for further replies.