N-channel amp.

[LBHajdu]

Sorry for not writing back to you sooner. I just got a new job at Brookhaven National Laboratory.

Ratz,
I don’t see any way to send attachments by way of the e-mail system in this site. However if you send me an e-mail at LBHajdu@optonline.net I will reply and attach this file for you.

The http://www.aussieamplifiers.com has been somewhat stagnant for the last 2 months. I can’t say I blame him. Sometime I have time for DIY and sometimes I don’t. Let’s hope that there will be updates soon.

Has anyone had the good fortune of comparing the N channel Amp with Holton’s Symmetrical MOSFET Output Amplifier. I noticed that the new amps he is producing are mostly derived from the original Symmetrical MOSFET Output Amplifier.

[Workhorse]

Hi Anthony Holton,

have you checked your n-channel amp for the cross-conduction, one of my friends build it and found out it to be in cross-conduction when there is no load presented at the output, but input signal is present.....

regards,
K a n w a r

[The Saint]

Greetings

Cross conduction as far as I am aware is only a problem in high speed switching circuits, such as SMPS.

In Linear circuits such as analog audio amplifiers we actually want it to happen to reduce crossover distortion.

Unless I am misunderstanding what cross conduction is, that is.
I would suggest that your friend closely match the output stage devices and try again.

Also the performance of the N Channel design requires a very well laid out PCB.

Anyway If I have missed the boat completely on cross conduction I am happy to be corrected. no doudt I will be hearing from some of you.

[Workhorse]

Saint hello,

Cross-Conduction is totally different from cross-over distortion, and you simply donot understands or know the concept is something amazed me lot.....i canot even believe it that you are unfamiliar of this concept , though you are a mosfet amp designer....

In your n-channel amp which is a modified version of Bengt Olson's design...the upper and lower output N- channel mosfets cross-conduct the current, i.e. there is improper turn-off of mosfets....

In Class-AB, the upper mosfet conducts and lower one is partially off[though it is little "on" by the virtue of biasing]and then the lower one turn on and upper one gets partially off ,
but the situation is different in your amp ..when the upper one conducts the driver n-channel doesnt turn off the lower mosfet so well and it also conducts and the whole current conducted through upper mosfets is then sinked by lower mosfet and thus causing catastrophic dissipiation of heat....the turn-off time of the mosfets is very slow as compared to turn-on time and this is because the n-channel driver doesn't drives the output mosfets fast enough so that the cross-conduction could be avoided...

Secondly..Cross-conduction has nothing to do with the layout of PCB's[as far as in linear amps but in switching amps they does matter] its an inherent flaw in the design itself....

We also manufacture N-channel mosfet amps which never cross-conduct even at 200KHZ....

If you redesign your driver stage and implement seperate drivers for both upper and lower mosfets then the cross-conduction could be eliminated....

I have also seen your 1KW amp which is operated from +-110VDC..which dissipiates lots of heat and its efficiency is poor because the output is not rail to rail ..i.e. the max output voltage never gets even close to the supply rail...because there is very high rail loss created by the driver mosfets and output stage mosfets---4+4 = 8 volts of rail loss causing very heavy dissipiation when about 16 amperes of contnuous current flows and that turns out to be around 120 to 200 watts of worst case wastage of power into heat......if you run your Frontend + driver stage at +-110 and your output stage at about +-100V ..your efficiency must boost and the dissipiation also decreases alot....

regards,
K a n w a r

[The Saint]

This reply is typical of the type of misunderstanding that goes on in this kind of forum and it is why I rarely respond to such posts.

I had a lot more to sayand reply to, But I don't think I will bother

[Workhorse]

Thanks Saint for diverting your precious meditation time for this forum..meditate alot and seek nirvana in audio

sometimes when people are out of answers..these types of comments from them are usual in nature....

K a n w a r

[Lars Clausen]

I agree with Kanwar.

Cross conduction is something you absolutely dont want!

You can say the difference between BIAS current and cross conduction is that BIAS current is stable in the whole frequency range of the amplifier, while cross conduction tends to grow with the frequency, and potentially kill the amplifier.

In a class D amplifier you are dealing with cross conduction in the few nS scale, making it necessary to take this effect very seroiusly, and put a lot of maticulous work into getting rid of the cross conduction. Anthony is right that in a linear amplifier, things don't go quite as fast, but that only means it is much easier to avoid cross conduction. Not that it should be overlooked as kind of a perk for lower crossover distortion.

So M50C: Take cross conduction seriously, or it will most probably kill your amplifier one day.

I am pretty sure Kanwar has a very good grip on building N channel amplifiers, but still i wait to see the sch's of his designs I mean the good ones.

All the best from

Lars

[Workhorse]

hello Lars Clausen,

good to see you after after so many days,

cross-conduction is dreaded in amplifier and Mr.Anthony has difficulty in accepting the truth about his designs whether its cross-conduction or rail-loss.....

You have the Zeta and i have the AT--EZ!

If i reveal my excellent commercial designs here then from very next day i would be on the roads

my prototype for diying is going to be ready and i would post it ..So stay tuned for .............

i appreciate your comments.....lars....your Zeta is only suffering from rail-loss but nothing else....

regards,
K a n w a r

[The Saint]

I wasn't out of answers.

I just don't like having strips taken off me, so you can feel better about yourself.

and I wasn't saying cross conduction wasn't a problem or it should be ignored.

se'ya guys

I don't think I want to play anymore..........

[Workhorse]

The Forum is full of experts like Lars, Graham, MikeB, and many other gentlemen and its not a game to play , its an entity to prove....