I'm begining to think this project isn't even worth it.... although I did just order more transformers, and today went and bought a (hopefully) suitable case for it....
I've spent $155 so far... that includes a rack mount case, 1250VA of toroidals, and 75,200uF of caps.... but the toroidals are so big that I'm not really left with enough room for the heatsinking... I don't want to have a seperate PSU to the power amp either.
finding suitable heatsinks will be a pita.... and output devices for the amp will set me back another $150 or so.... so far, everything has been cheap.... maybe someone has some surplus mosfets or transistors I could buy??? LATER!!!!
I've spent $155 so far... that includes a rack mount case, 1250VA of toroidals, and 75,200uF of caps.... but the toroidals are so big that I'm not really left with enough room for the heatsinking... I don't want to have a seperate PSU to the power amp either.
finding suitable heatsinks will be a pita.... and output devices for the amp will set me back another $150 or so.... so far, everything has been cheap.... maybe someone has some surplus mosfets or transistors I could buy??? LATER!!!!
Hi,
your 50v rails do not seem too low.
I ran 625va toroid 4% regulation 35vac into 2 Crimson 1704 boards. I got 110w into 8r both channels driven and 200w into 4r one channel driven. and the output stage only has one pair of 2n3773. Crimson has put in good protection that you never hear & I have never damaged them in 20 years & over 3 generations of boards.
Caps started as 2x4,700uF per rail per board.
Later amps run 3x 6,800uF per rail per board, bass strength and depth much improved. Bypass was mix of polycarbonate & polystyrene. input cap changed to 2x33uf solid tant back to back bypassed with 1uf polysomething, I,m getting too old in the grey matter.
regards Andrew T.
your 50v rails do not seem too low.
I ran 625va toroid 4% regulation 35vac into 2 Crimson 1704 boards. I got 110w into 8r both channels driven and 200w into 4r one channel driven. and the output stage only has one pair of 2n3773. Crimson has put in good protection that you never hear & I have never damaged them in 20 years & over 3 generations of boards.
Caps started as 2x4,700uF per rail per board.
Later amps run 3x 6,800uF per rail per board, bass strength and depth much improved. Bypass was mix of polycarbonate & polystyrene. input cap changed to 2x33uf solid tant back to back bypassed with 1uf polysomething, I,m getting too old in the grey matter.
regards Andrew T.
"see, I feel that having 1000va of transformers for 220watts of output is a TAD over rated "
Just shows what you know.
A capacitor input filter has a really poor power factor, even worse with large caps. If you keep the caps reasonable sized it might be 0.7. That means your 1KVA transformer can only provide 700 watts DC to the load.
A typical Class AB amplifier, with moderate bias, and an unregulated supply, will have an efficency of 60%.
AndrewT's amplifier put out 200W/4R, so 200W X 2 = 400W, 400W/60%=667W required from the supply.
This is assuming you are driving a resistor with a 0 degree phase angle. A real loudspeaker will look more like a 45* load, and an electrostat can come close to 90* load. A 90* load will require TWICE as much input power to the amplifier. If I remember correctly a 45* load would require 1.414 more watts.
Music MAY have a high peak-to-average ratio, then again it MAY NOT.
Modern dance music can measure as low as a 58% duty cycle, FM radio as low as 25%.
A 50% duty cycle is a wash for a 45* load, a 25% duty cycle is a wash for a 90* load.
My McIntosh is a well designed amplifier, it has a 600VA transformer for 120W X 2 at 4R.
If you do the math on this, it comes out exactly the same as the 1KVA for 200W X 2 example.
Do you have to make yours this big?
Of course not, and most commercial designs don't.
I have a HUGE pile of transformers removed from Denon, Yamaha, and Sony receivers for the thermo-switch in the transformer going open (I used to be an authorized service center for these brands).
Cost/performance ratio is a curve that generally has a steep knee in it. On the flat portion below the knee, a little bit of extra cash gives a lot of improvement. Above the knee, it takes a lot of extra cash to give a little bit of improvement. I like to be on the knee somewhere, how about you?
DIY is great, you can make any design choice you have a mind to.
Just shows what you know.
A capacitor input filter has a really poor power factor, even worse with large caps. If you keep the caps reasonable sized it might be 0.7. That means your 1KVA transformer can only provide 700 watts DC to the load.
A typical Class AB amplifier, with moderate bias, and an unregulated supply, will have an efficency of 60%.
AndrewT's amplifier put out 200W/4R, so 200W X 2 = 400W, 400W/60%=667W required from the supply.
This is assuming you are driving a resistor with a 0 degree phase angle. A real loudspeaker will look more like a 45* load, and an electrostat can come close to 90* load. A 90* load will require TWICE as much input power to the amplifier. If I remember correctly a 45* load would require 1.414 more watts.
Music MAY have a high peak-to-average ratio, then again it MAY NOT.
Modern dance music can measure as low as a 58% duty cycle, FM radio as low as 25%.
A 50% duty cycle is a wash for a 45* load, a 25% duty cycle is a wash for a 90* load.
My McIntosh is a well designed amplifier, it has a 600VA transformer for 120W X 2 at 4R.
If you do the math on this, it comes out exactly the same as the 1KVA for 200W X 2 example.
Do you have to make yours this big?
Of course not, and most commercial designs don't.
I have a HUGE pile of transformers removed from Denon, Yamaha, and Sony receivers for the thermo-switch in the transformer going open (I used to be an authorized service center for these brands).
Cost/performance ratio is a curve that generally has a steep knee in it. On the flat portion below the knee, a little bit of extra cash gives a lot of improvement. Above the knee, it takes a lot of extra cash to give a little bit of improvement. I like to be on the knee somewhere, how about you?
DIY is great, you can make any design choice you have a mind to.
wintermute said:hey skinnyseems you have a habit of putting the cart before the horse
next time I suggest deciding on an amp and then working out what components you need for the power supply
Tony.
I got the capacitors for free, does that count?
and I had chosen what amp I was going to make... I just changed my mind so I bought more transformers ok, I now have +- 63 volts at 1250va
lolanyone got any nice, simple transistor amp schematics that will run off this voltage?
wintermute said:my amp runs on 63V rails (actually it is supposed to be 66 but in reality is 63) maybe I should buy the toroidals and caps off you
I wouldn't recomend trying to make one, the circuit layout is apparently very sensitive!!!!
What made you change your mind on the amp??
Tony.
my transformers!!!!
they are hUgE!!!!!! expect this amp to weigh 25Kgs or so the amp I was going to make was like a 6 channel LM4780 amp..... but as if you would!!!
we'll see how it all goes.....
350W at 4R with ±70V indicates it has a healthy supply.
I would not expect that much normally.
An Adcom GFA555 is only 325W at 4R with ±77V and has 60,000µF of filter caps too.
And I am surprised to only see four pair of plastic outputs, normally you need to run 50%~100% more output devices when you run 150*C plastic vs 200*C rated TO3 outputs.
I would not expect that much normally.
An Adcom GFA555 is only 325W at 4R with ±77V and has 60,000µF of filter caps too.
And I am surprised to only see four pair of plastic outputs, normally you need to run 50%~100% more output devices when you run 150*C plastic vs 200*C rated TO3 outputs.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.