60V? The power dissipation is too much if I want to run 750mA quiescent current!!! Why do you need over 150W per channel? Normal music at home, you never see over 3W, say peak at 30W!!! I am sure I run less than 2W in my listening.
Seriously, I know some of you guys want 500W amp. But you either have to give up Class A at any normal listening or you have to have air conditioning to cool the room!!!
I even consider reduce down to 30V rail then increase the quiescent current and get more wattage in Class A. What's wrong with that?
Thanks
Seriously, I know some of you guys want 500W amp. But you either have to give up Class A at any normal listening or you have to have air conditioning to cool the room!!!
I even consider reduce down to 30V rail then increase the quiescent current and get more wattage in Class A. What's wrong with that?
Thanks
Last edited:
That's another reason I want to keep the voltage low. I can use much cheaper cap also.
I see opinions of using higher rail. But what is the disadvantage of using a lower rail? I really want to hear the real reason, not just everybody else using higher voltage.
lower rail = lower output swing at Hi-Fi speaker ( Re ~ 6 ohms typical) unless you like car stuff.
lower volts = lower efficiency or more copper wire and switching devices for a fixed power. related to I2R losses. power conversion.
don't forget energy stored in a cap is a V squared function! so there's that too
lower volts = lower efficiency or more copper wire and switching devices for a fixed power. related to I2R losses. power conversion.
don't forget energy stored in a cap is a V squared function! so there's that too
Last edited:
Hi,
This is all over the place. 10V C to B to avoid a pole ? How and Why ?
Output stage needs careful design to optimise class A as high current
optimal aB or heavily biased AB, the latter needs more heatsinking
and of course because of that lower rails voltages will help somewhat.
Optimum bias, heatsinking, and compensation are not simple.
rgds, sreten.
This is all over the place. 10V C to B to avoid a pole ? How and Why ?
Output stage needs careful design to optimise class A as high current
optimal aB or heavily biased AB, the latter needs more heatsinking
and of course because of that lower rails voltages will help somewhat.
Optimum bias, heatsinking, and compensation are not simple.
rgds, sreten.
Last edited:
then hurry up and buy it....com'on and build something....and post pictures will you? we are eager to see you build something after all the talking....😉
nothing wrong with this as long as you start building instead of arguing ad infinitum....
But I already calculated and you pretty much varified 68W for 8 ohm. I calculated 112W for 4 ohm speaker. That's not enough?
The amp is specifically designed for my pair of speaker that is 4 ohm. I don't have a choice. I just have to use more caps and bigger wires for the extra current.
yeah build it, there's nothing wrong cold transformer and hot heat sinks.
Bridging amps are a lot of fun next time. build 2 extra channels in advance
Bridging amps are a lot of fun next time. build 2 extra channels in advance
Last edited:
class A amps needs psu traffo that is bigger compared to classAB or
classB if you want them cool...this has been my experience thus far...
no amount for formula will get you the right size,
the best you will ever get is an engineering guess....
classB if you want them cool...this has been my experience thus far...
no amount for formula will get you the right size,
the best you will ever get is an engineering guess....
Split the "job" up ....
Ideal situation would be class AB (even under-biased) for LF ( a sub ? ) .
X-over distortion would matter less here.
Heavy bias AB or go full class A for >100Hz.
200+ W for the sub , 50W X 2 (class A) for the full range.
OS
Ideal situation would be class AB (even under-biased) for LF ( a sub ? ) .
X-over distortion would matter less here.
Heavy bias AB or go full class A for >100Hz.
200+ W for the sub , 50W X 2 (class A) for the full range.
OS
I have a sub, so I don't have to deal with that ( thank God)!!!
You guys are scaring me, I was so sure I want the 40V rail. Now I start questioning myself!!!! I have been so busy searching transistors, chassis and everything. My head is spinning.
Last edited:
Alan has 4 ohm speakers hence he is designing for a nominal 4 ohm load. Makes sense to me.
Alan, for what it's worth, I think that 625VA unit will work a treat.
Alan, for what it's worth, I think that 625VA unit will work a treat.
This is really for home use, 99.9% of the time is at conversation level. High power is of no importance for me. I am not showcasing the amp or use it anywhere else. From my reading, keeping the amp in Class A in my listening level is more important than the peak level.
I was even considering a 25V 25V transformer to get 35V rail so I can get more quiescent current to get a little more power in Class A. But I chicken out and go with 40V.
I can still remove the artificial limit I set and run to 35V peak. So the peak current is 8.75A and W = 0.5 X 35V X 8.75A=153W. That's a lot of power for 4 ohm speaker. For 8 ohm, peak current is 4.375A and W = 0.5 X 35 X 4.375 = 76.6W.
I was even considering a 25V 25V transformer to get 35V rail so I can get more quiescent current to get a little more power in Class A. But I chicken out and go with 40V.
I can still remove the artificial limit I set and run to 35V peak. So the peak current is 8.75A and W = 0.5 X 35V X 8.75A=153W. That's a lot of power for 4 ohm speaker. For 8 ohm, peak current is 4.375A and W = 0.5 X 35 X 4.375 = 76.6W.
Then 10W is oodles for your needs. Why not begin by constructing a simple, very low cost amplifier to remove the obvious doubts and misunderstandings you have about audio power. Then you can plan your magnum opus with confidence - your own decisions rather than polling opinions for every choice that presents itself.
Just an observation ....
Have a OEM "run of the mill" amp , occasionally engaging - never breathtaking.
Have a overpowered , "psycho ".... over-designed muscle amp - ALWAYS
effortless , engaging , breathtaking.
Even a few watts (average) of digital material can call upon 100's of watts (short term).
An amp that has a high upper limit handles this effortlessly. (and sounds better).
Decide to push the limits (short term 800W peaks) , beyond "breathtaking". (scary).
You can physically fit a 200w unit in the same space as a 50W unit. Pack it
tighter power supply wise and make sure your C/W ratings are up to the output
stage.
PS - you also can rest assured that a few glasses of wine won't blow your amp 😀 .
You don't have to use all that power all the time - why buy a Ferrari if the speed
limit is 65mph ?
OS
Well, even with 40V rail, I get 150WRMS if I don't artificially limit to 30V peak. That is high power by any measure. The YBA amp that push me into searching for a better amp was only 70W per channel into 8ohm. So I won't say high power is everything. There are many people swear by single ended tube amps that only have 3 to 5W output, they sure sell for a lot of money and I listened to a Cary very loud in the showroom. It was loud and was one of the prettiest sounding amp I ever heard.
I can tell you I got quite an improvement from my Acurus by increasing the quiescent current from 50mA to 200mA total. Problem is the amp is 80V rail and the amp gets really hot already. That gives me the reason to lower the rail and increase the current. My plan is 3 to 4 stages and 250mA each. So I get 750mA or 1A quiescent current respectively.
I am still waiting for Mr. Cordell to respond to the question about gm doubling. I still think if I lower the emitter resistor to 0.1ohm, I can run at 250mA and get the optimal crossover distortion characteristics.
I can tell you I got quite an improvement from my Acurus by increasing the quiescent current from 50mA to 200mA total. Problem is the amp is 80V rail and the amp gets really hot already. That gives me the reason to lower the rail and increase the current. My plan is 3 to 4 stages and 250mA each. So I get 750mA or 1A quiescent current respectively.
I am still waiting for Mr. Cordell to respond to the question about gm doubling. I still think if I lower the emitter resistor to 0.1ohm, I can run at 250mA and get the optimal crossover distortion characteristics.
Last edited:
indeed there are many examples of lousy class AB esp at higher power, it all comes down to the many gotcha's or the little details that are so very easy to get wrong by experts as well as novices. These are after all is said and done, commutating power conversion circuits in lots of aspects. pure class A the power current is constant so not switching. easy for DIY.
Maybe you should investigate proven designs not modified high bias class AB.
KRELL or PASS designs?
Maybe you should investigate proven designs not modified high bias class AB.
KRELL or PASS designs?
Last edited:
A normal low-bias class AB amplifier is ~70% efficent, to put out 100W it will need 140W in.
A normal amplifier power supply will have the rectifiers feeding large filter caps, a brute force pseudo regulated supply if you will.
The power factor on this will be ~.7, so you will need a 200VA trafo to output 140W.
So we need a 400VA trafo for a stereo 100W amplifier.
Highly compressed rock/rap music has a 50% duty cycle, but a real speaker may look like a 45* load. Call it a wash.
Rule-of-thumb = 2VA for 1W out.
I have built as low as 1VA for 1W out. It sounded fine until driven into clipping (no one ever does this do they?). Going to 2VA for 1W made the amplifier sound much more 'solid'. Going above 2VA for 1W gave diminishing return.
Running a 50mA regulated supply for the front end of the amplifier did far more than going above 2VA for 1W, and cost far less.
Amplifiers I have owned that were built like this:
McIntosh (series regulator for front end)
Electrocompaniet (series regulator for front end)
Rowland (shunt regulator for front end)
Can you go bigger? Sure. My 120W/channel McIntosh MC2120 has a 600VA transformer with regulated supplies for the voltage gain stages. That follows my calculations too, but for 100% duty cycle. Of course it sounds like a brick house in the bass, not to mention weighing 58 lbs.
A normal amplifier power supply will have the rectifiers feeding large filter caps, a brute force pseudo regulated supply if you will.
The power factor on this will be ~.7, so you will need a 200VA trafo to output 140W.
So we need a 400VA trafo for a stereo 100W amplifier.
Highly compressed rock/rap music has a 50% duty cycle, but a real speaker may look like a 45* load. Call it a wash.
Rule-of-thumb = 2VA for 1W out.
I have built as low as 1VA for 1W out. It sounded fine until driven into clipping (no one ever does this do they?). Going to 2VA for 1W made the amplifier sound much more 'solid'. Going above 2VA for 1W gave diminishing return.
Running a 50mA regulated supply for the front end of the amplifier did far more than going above 2VA for 1W, and cost far less.
Amplifiers I have owned that were built like this:
McIntosh (series regulator for front end)
Electrocompaniet (series regulator for front end)
Rowland (shunt regulator for front end)
Can you go bigger? Sure. My 120W/channel McIntosh MC2120 has a 600VA transformer with regulated supplies for the voltage gain stages. That follows my calculations too, but for 100% duty cycle. Of course it sounds like a brick house in the bass, not to mention weighing 58 lbs.
Last edited:
- Status
- Not open for further replies.