skaara said:
It's non-linear in a different way. IME, the biggest non-linearity problem with BJTs is how they act around 0V on the base. The slew rate on the output voltage doesn't always match the input voltage right around that point (similar to how the output of a diode is when you ramp the voltage from 0V to 10V). The biggest issue (again, IME) with FETs (and tubes) is transconductance, where the output voltage may follow the input voltage well, but the output current is changing at a non-constant rate.
Those may not be either components biggest problem, but that's what I don't like about each, based on my experiences with them.
I don't like to get into arguments about Mosfets vs Bipolar
because they both have their strengths and weaknesses.
Having said that, it is my experience that people arguing
the superior linearity of BJT's over Mosfets are using cases
where either the bias or the source impedance is low. For
example, the typical big Bipolar device has about 1/3 the
distortion of a Mosfet as a follower - if the current isn't
very high and if the source impedance is just a few ohms.
As the source impedance and the current rise, the Mosfet
will equal and ultimately surpass the Bipolar.
The argument is a red herring anyway, since we have already
decided that we don't base everything on THD numbers
anyway.
because they both have their strengths and weaknesses.
Having said that, it is my experience that people arguing
the superior linearity of BJT's over Mosfets are using cases
where either the bias or the source impedance is low. For
example, the typical big Bipolar device has about 1/3 the
distortion of a Mosfet as a follower - if the current isn't
very high and if the source impedance is just a few ohms.
As the source impedance and the current rise, the Mosfet
will equal and ultimately surpass the Bipolar.
The argument is a red herring anyway, since we have already
decided that we don't base everything on THD numbers
anyway.
Actually!!
Well, this is really a little conservative. If you use one of NP's designs you will gain some efficiency due to the active current source, but even then you are at 300W / channel for 100W output. So, now you need 1.5 kVA in transformer and 0.05 C/W in heatsinks.
"Standard" SE Class A will probably be around 20% efficiency, so you will be up to 500W/ch ..... ouch!
Does anyone have any push-pull schematics at 100W .... this might be the only chance
And here comes the "broken record" bit ...... exactly why do you need 100W ??
cheers, mark
PS: If price, heat and space are not a concern ...... build Aleph2 monoblocks !!
Well, this is really a little conservative. If you use one of NP's designs you will gain some efficiency due to the active current source, but even then you are at 300W / channel for 100W output. So, now you need 1.5 kVA in transformer and 0.05 C/W in heatsinks.
"Standard" SE Class A will probably be around 20% efficiency, so you will be up to 500W/ch ..... ouch!
Does anyone have any push-pull schematics at 100W .... this might be the only chance
And here comes the "broken record" bit ...... exactly why do you need 100W ??
cheers, mark
PS: If price, heat and space are not a concern ...... build Aleph2 monoblocks !!
Hope it is not too late to suggest:
http://www.borbelyaudio.com/
Erno Borbely is some one who has made many contributions to Audio DIY.
Have a look.
Aud_Mot
http://www.borbelyaudio.com/
Erno Borbely is some one who has made many contributions to Audio DIY.
Have a look.
Aud_Mot
you can find a free pdf version from Mark Finnis. http://www.users.bigpond.net.au/mefinnis/passlabs/
he has very cool boards and projects
he has very cool boards and projects
What about the plate on http://web.vip.hr/pcb-design.vip/ ?
It's better looking, don't you think?
thx,
DJG
It's better looking, don't you think?
thx,
DJG
I have provided positive images which I used for home PCB manufacture. No question the PCBs I made do not compare to professional grade milled / laser-drilled / silk-screened PCBs.
If the later is what you want then the above website is really very good indeed. My only thought about these PCBs is the close proximity of the FETs and heat dissipation. Other than that they look absolutely fantastic.
One small advantage of DIY, and this sort of community, is the willingness of people to help each other. So, if someone piped-up and said, "hey, we have a supply of great heatsinks, but the PCB needs the devices to be at 'x' mm spacing" ..... then you would likely find an appropriate TIFF appear in the not too distant future!
cheers, mark
If the later is what you want then the above website is really very good indeed. My only thought about these PCBs is the close proximity of the FETs and heat dissipation. Other than that they look absolutely fantastic.
One small advantage of DIY, and this sort of community, is the willingness of people to help each other. So, if someone piped-up and said, "hey, we have a supply of great heatsinks, but the PCB needs the devices to be at 'x' mm spacing" ..... then you would likely find an appropriate TIFF appear in the not too distant future!
cheers, mark
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.