Various things are possible in life… not all are very realistic. For normal and “composed” user that probably won’t happen ever. Very similar amp just with less output devices and on slightly lower supply has been tested for hours without problems. 15034/35 are very robust units quite capable to survive anything that those tiny Zeners might impose on them. Anyway it’s quite easy to add appropriate resistors between their emitters and cathodes of Zeners. I am designing amps for normal use, not for abuse. None of my amps are idiot-proof and I am not planning to start designing such even if that category of users prevails one day. Of course, other designers will follow their own way and approach.
"15034/35 are very robust units quite capable to survive anything that those tiny Zeners might impose on them."
You mis-understand the failure mode.
When driven into full output into a low impedance load the gate voltage is clamped by the zener. The zener goes open from the high current from the driver transistor. The gates on the outputs now burn up.
I have had to rebuild Hafler amplifiers for this very reason.
You mis-understand the failure mode.
When driven into full output into a low impedance load the gate voltage is clamped by the zener. The zener goes open from the high current from the driver transistor. The gates on the outputs now burn up.
I have had to rebuild Hafler amplifiers for this very reason.
Zeners almost always fail shorted (or semi-shorted with low resistance ). In these circumstances, the MJE1503x will be left driving the output alone and will fail too. Note that the resistance of a shorted small zener may be high enough to allow the MJE1503x to produce enough Vgs to blow the output MOSFET before they blow themselves up. In any case, the MJE1503x will fail shorted and will cause *all* the output devices to blow too.
An active limiter for Vgs is required, and also a current limiting stage for the VAS.
An active limiter for Vgs is required, and also a current limiting stage for the VAS.
Driving Vertical N-channel Mosfets through the same DRIVER STAGE which is meant for Bipolar Output Devices is the worst mistake committed unknowingly by some people.
The zener placement on BORAOMEGA's design is yet another silly approach. In that way zeners willnot only destroy outputdevices but drivers also, when subjected to overdrive. Also the baxandall diode used for CFB has no effect with Mosfet output stage as already pointed out by DJK.
Zeners must be connected directly to gate-source terminals of each mosfet individually. In this way gate resistor will eventually limit the zenercurrent to safe limit hence will also protect zener from getting smoke+ added bonus of safety of mosfets from gate voltage overdrive along with drivers.
The zener placement on BORAOMEGA's design is yet another silly approach. In that way zeners willnot only destroy outputdevices but drivers also, when subjected to overdrive. Also the baxandall diode used for CFB has no effect with Mosfet output stage as already pointed out by DJK.
Zeners must be connected directly to gate-source terminals of each mosfet individually. In this way gate resistor will eventually limit the zenercurrent to safe limit hence will also protect zener from getting smoke+ added bonus of safety of mosfets from gate voltage overdrive along with drivers.
designers ......should
also keep in mind that all amplifiers some time will suffer some ammount of abuse .....
i can bring a million of examples ....one of the most common : kids at home .....very close to apreamplifier ....even closer to the volume control ....resulting a couple of JBL E60 northbridge ...... brand new cost 900 euros ....blown to mars ...... together with a few mosfets some fuses ......
nice amp ( was ) very clean very powerfull ....very straight.... very simple ..... but no protection .....( And i am not talking about output protection )
Latest days i go for less power more quality and a guaranty SOA ....
also keep in mind that all amplifiers some time will suffer some ammount of abuse .....
i can bring a million of examples ....one of the most common : kids at home .....very close to apreamplifier ....even closer to the volume control ....resulting a couple of JBL E60 northbridge ...... brand new cost 900 euros ....blown to mars ...... together with a few mosfets some fuses ......
nice amp ( was ) very clean very powerfull ....very straight.... very simple ..... but no protection .....( And i am not talking about output protection )
Latest days i go for less power more quality and a guaranty SOA ....
Driving Vertical N-channel Mosfets through the same DRIVER STAGE which is meant for Bipolar Output Devices is the worst mistake committed unknowingly by some people.
Designing a Mosfet amp like it was a BJT amp is a good way to discover that you will not like Mosfets, then you will become one of those here at DIY who criticizes the use of Mosfets because they don't know how to use them properly.
A gate and a base are not the same thing..........
Sorry but I think I have been missunderstood. I was asked by my nephew to find a schematic of a high power mosfet amplifer having (possibly) only n-channel mosfets because it's what he has at home. He wants to make experiance and noise at the moment. Since the schematic I found was unpritable, I just asked to have it clear.
Regarding amplifier design, I know very well the difference between bjt, mosget and valve design.
Sorry again to have been missunderstood.
Bye
abiondia
Regarding amplifier design, I know very well the difference between bjt, mosget and valve design.
Sorry again to have been missunderstood.
Bye
abiondia
Designing a Mosfet amp like it was a BJT amp is a good way to discover that you will not like Mosfets, then you will become one of those here at DIY who criticizes the use of Mosfets because they don't know how to use them properly.
Well now, why dont you tell us how to drive a mosfet then, I find no reason to not drive them with a emitter folower or even 2 emitter folowers as in a triple.
OK and I repeate :
I was just serching a schematic of a non clear one found because my nephvew asked for it ( and I still do'nt have it).
I absolutely did'nt want to be a part of a kind of an "audio social network".
Allthough I found the anwered quite interesting, but technically "speaking" they are quite a nonsense.
I still hope that all this it due a big misunderstanding.
Regards
abiondia
I was just serching a schematic of a non clear one found because my nephvew asked for it ( and I still do'nt have it).
I absolutely did'nt want to be a part of a kind of an "audio social network".
Allthough I found the anwered quite interesting, but technically "speaking" they are quite a nonsense.
I still hope that all this it due a big misunderstanding.
Regards
abiondia
Inefficiency!
Class B output stages are fairly inefficient when used in PA. If pushed hard, say with compressed bass signals worst case efficiency is about 50%
So to produce 2000watts you have to dissapate anything up to 1000w of heat
This is one reason why high power amplifiers use multiple split voltage rails.
The amplifier either switches the output stage to the higher voltage when required (class G), or uses tiers of output devices between the various voltage rails (class H).
This greatly increases the efficiency of the amplifier.
Also the tiers of output devices can use lower voltage parts.
Overall what I am trying to say is, if you want to modify your original amplifier to use Quasi topology then fine. But this does not mean that you can then just scale up the design. I think, even if you could get it to be stable, you would quickly run in to thermal problems as soon as you tried to use it at full power!
I am curious to know what amplifiers you know of that use +/- 150V rails in a Class AB amplifier?
Class B output stages are fairly inefficient when used in PA. If pushed hard, say with compressed bass signals worst case efficiency is about 50%
So to produce 2000watts you have to dissapate anything up to 1000w of heat
This is one reason why high power amplifiers use multiple split voltage rails.
The amplifier either switches the output stage to the higher voltage when required (class G), or uses tiers of output devices between the various voltage rails (class H).
This greatly increases the efficiency of the amplifier.
Also the tiers of output devices can use lower voltage parts.
Overall what I am trying to say is, if you want to modify your original amplifier to use Quasi topology then fine. But this does not mean that you can then just scale up the design. I think, even if you could get it to be stable, you would quickly run in to thermal problems as soon as you tried to use it at full power!
That is a good way, maybe even with a totem pole for each device (there is usually 3-4V Vgsth for a typical vertical fet), with two series zeners (opposing polarity) from gate to source on each transistor as close to the device as possible. Minimize lead length to minimize the effect on RF stability. A gate stopper resistor along with a gate Zobel filter is a nice way to help control this instability. PCB layout and minimizing certain track inductance is very important here as well. Address the non-linearity between Vgs and Gm by some sort of feedback/feedforward EC.
This is just a few points I can think of just off the top of my head. It can become more complicated than a typical BJT layout with the BCE wires to the output transistors strung loosely across the whole project. Simulations generally do not take these important factors into account.
That is a good way, maybe even with a totem pole for each device (there is usually 3-4V Vgsth for a typical vertical fet), with two series zeners (opposing polarity) from gate to source on each transistor as close to the device as possible. Minimize lead length to minimize the effect on RF stability. A gate stopper resistor along with a gate Zobel filter is a nice way to help control this instability. PCB layout and minimizing certain track inductance is very important here as well. Address the non-linearity between Vgs and Gm by some sort of feedback/feedforward EC.
This is just a few points I can think of just off the top of my head. It can become more complicated than a typical BJT layout with the BCE wires to the output transistors strung loosely across the whole project. Simulations generally do not take these important factors into account.
Amen to that, my prefered way is totem pole btw but more complex, in a triple configuration it works mighty good though, and i find it more stable than with bjt otputs.
I am curious to know what amplifiers you know of that use +/- 150V rails in a Class AB amplifier?
Class B output stages are fairly inefficient when used in PA. If pushed hard, say with compressed bass signals worst case efficiency is about 50%
So to produce 2000watts you have to dissapate anything up to 1000w of heat
This is one reason why high power amplifiers use multiple split voltage rails.
The amplifier either switches the output stage to the higher voltage when required (class G), or uses tiers of output devices between the various voltage rails (class H).
This greatly increases the efficiency of the amplifier.
Also the tiers of output devices can use lower voltage parts.
Overall what I am trying to say is, if you want to modify your original amplifier to use Quasi topology then fine. But this does not mean that you can then just scale up the design. I think, even if you could get it to be stable, you would quickly run in to thermal problems as soon as you tried to use it at full power!
this is all academic since the post is of course very dead but for the fun of it.... ( presumably together with the original poster that he is very happy now with his brand new behringer amplifier
) None of the "chinese amplifiers with 2sa1943-5200 " and rails of 150+150 volt ( ???? ) works on class AB most if not all of them work in class G or H
then again so many nice people add their 2 coins in the post some of them with local advice and some of them with complete schematics
finally my advice is simple as that ...to any of schematics or design ideas simply FACE IT !!!!! no Amplifier above 100W if designed for PA use will last if each and every division of the amp is propelrly protected .... Ltp ,vas , drivers , output stage ...
I find Dr Bora's sayings the most correct :"i dont design idiot proof amplifiers"
take my word for it a PA amplifier has to be designed as Idiot proof
AndrewT's opinion on the same subject is as always correct and its based on soa plots of tranistors and the logic that 1+1 makes 2 fair enough but not cost effective ...
my 2 coins .....
This Thread is obsolete ...
but Quasi NMOS Amplifier Design is excellent sounding N Channel Mosfet Amp
have built 3 of them
@ SAKIS or Dr Bora's sayings "i dont design idiot proof amplifiers"
I dont like amplifier without protection like crowbar, VI limiter, Output Protection
my speakers to expensive to use with amplifiers without protection
but Quasi NMOS Amplifier Design is excellent sounding N Channel Mosfet Amp
have built 3 of them
@ SAKIS or Dr Bora's sayings "i dont design idiot proof amplifiers"
I dont like amplifier without protection like crowbar, VI limiter, Output Protection
my speakers to expensive to use with amplifiers without protection
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.