From the article
The gain of this front end is set at about X 6, or 15 dB. You can have higher gain simply by changing R203 and R204 to higher values. You can double the gain (to about 20 dB) by increasing these to 100 Kohm, and you can push it to 150 Kohm (about 23 dB) without changing the compensation capacitors.
The gain of this front end is set at about X 6, or 15 dB. You can have higher gain simply by changing R203 and R204 to higher values. You can double the gain (to about 20 dB) by increasing these to 100 Kohm, and you can push it to 150 Kohm (about 23 dB) without changing the compensation capacitors.
Thanks for that Tea-Bag.
I'll re-read the article in case I've missed anything else. Do you think there would be any change to the overall sonic characteristics of the amp?
Something I have come across when altering the impedence of interconnects (the more you twist the more focused the image for example and the less a sense of 'air') is their affect on the soundstage and something that is also mentioned in the Ayre K1 pre-amp user's manual. I quote, "Usually, a lower load impedence results in a more focused soundstage, with greater image solidity and less apparent distortion. A higher load impedence will give a more open sound with greater 'air' and 'life'."
Regards,
Chris
I'll re-read the article in case I've missed anything else. Do you think there would be any change to the overall sonic characteristics of the amp?
Something I have come across when altering the impedence of interconnects (the more you twist the more focused the image for example and the less a sense of 'air') is their affect on the soundstage and something that is also mentioned in the Ayre K1 pre-amp user's manual. I quote, "Usually, a lower load impedence results in a more focused soundstage, with greater image solidity and less apparent distortion. A higher load impedence will give a more open sound with greater 'air' and 'life'."
Regards,
Chris
Backbones:
I built my BA-2 with 150k in the feedback (R203, R204) so I could drive it from my Benchmark DAC without changing the Benchmark, haven't had any problems. Remember if you are running the +/- supplies at 25volts then the amp is only around 40watts at 8 ohm. This may be the difference you are seeing between the amps.
I built my BA-2 with 150k in the feedback (R203, R204) so I could drive it from my Benchmark DAC without changing the Benchmark, haven't had any problems. Remember if you are running the +/- supplies at 25volts then the amp is only around 40watts at 8 ohm. This may be the difference you are seeing between the amps.
I have several questions on POWER DISSIPATION
I read “Burning Amp 2 by Nelson Pass”, but …
Here are several statements:
Page 2
”As the typical Burning Amp will dissipate maybe 300 watts,…”
(I understand per 2ch)
Page 6
“Two channels of this amp will may draw as much as 300 watts…”
(It is almost equal to heat)
I understand, that Nelson describes 6 pairs per channel version with 250mA current bias. Later he mentions 1.5A bias.
Page 7
“The 250 mA setting results in dissipation of about 6 watts per transistor for the output stage shown.”
If we have 12 transistors per channel and 6W per transistor, there is 75W/ch (12x6W) or 150W/2ch or in other way we have here 1.5A bias per channel (6 pairs x 250mA=1.5A).
It looks right – 1.5A x (25V+25V)=75W or for single transistor 250mA x 25V=6.25W
Page 7 further
“If you have lots of heat sinking, you could consider twice as much bias, for 3 amps of bias per channel. This would be about 300 watts per chassis”
3A x 50V=150W per channel, 150Wx2=300W
So why then is there mentioned, that typical (I assume 1.5A/ch) dissipates 150W/ch or 300W per 2 channels?
May somebody explain whic calculation is correct?
Is a typical 1.5A bias per channel dissipates 150W or 75W per channel?
Thanks
I read “Burning Amp 2 by Nelson Pass”, but …
Here are several statements:
Page 2
”As the typical Burning Amp will dissipate maybe 300 watts,…”
(I understand per 2ch)
Page 6
“Two channels of this amp will may draw as much as 300 watts…”
(It is almost equal to heat)
I understand, that Nelson describes 6 pairs per channel version with 250mA current bias. Later he mentions 1.5A bias.
Page 7
“The 250 mA setting results in dissipation of about 6 watts per transistor for the output stage shown.”
If we have 12 transistors per channel and 6W per transistor, there is 75W/ch (12x6W) or 150W/2ch or in other way we have here 1.5A bias per channel (6 pairs x 250mA=1.5A).
It looks right – 1.5A x (25V+25V)=75W or for single transistor 250mA x 25V=6.25W
Page 7 further
“If you have lots of heat sinking, you could consider twice as much bias, for 3 amps of bias per channel. This would be about 300 watts per chassis”
3A x 50V=150W per channel, 150Wx2=300W
So why then is there mentioned, that typical (I assume 1.5A/ch) dissipates 150W/ch or 300W per 2 channels?
May somebody explain whic calculation is correct?
Is a typical 1.5A bias per channel dissipates 150W or 75W per channel?
Thanks
Your enquiry has shown a discrepancy.
your 6pr output stage with device bias set to 250mA gives a total bias current of 1.5A
+-25Vdc supply rails times 1.5A gives 75W of dissipation per channel.
Similarly 250mA times a single supply rail of 25Vdc gives a device dissipation of 6.25W.
12 devices @ 6.25W gives 75W of dissipation per channel.
BA2 is a push pull ClassA amp. The peak ClassA current must be <3Apk for a 1.5A total bias.
3Apk across 8r0 is 24Vpk indicating a maximum ClassA power output of ~30W into 8r0. (if the amp can deliver 24Vpk and 3Apk with supply rails @ +-25Vdc.)
your 6pr output stage with device bias set to 250mA gives a total bias current of 1.5A
+-25Vdc supply rails times 1.5A gives 75W of dissipation per channel.
Similarly 250mA times a single supply rail of 25Vdc gives a device dissipation of 6.25W.
12 devices @ 6.25W gives 75W of dissipation per channel.
BA2 is a push pull ClassA amp. The peak ClassA current must be <3Apk for a 1.5A total bias.
3Apk across 8r0 is 24Vpk indicating a maximum ClassA power output of ~30W into 8r0. (if the amp can deliver 24Vpk and 3Apk with supply rails @ +-25Vdc.)
AndrewT, thanks.
It means 300W transformer for 2 channels with 1.5A each should be enough, isn’t it?
Is it correct then, that 3A pk and 25V supply rails across 4 Ohms will deliver around 50W of class A?
By doubling of bias to 3A it may increase power output to 60w/8R and heat dissipation to 150w/ch.
It seems a recommended bias limit is 1.2A per transistor pair (30W/25V=1.2A) according the statement “If you have enough heat sinking, you can
consider as high as 30 watts per device” (page 11).
It converts to 3.5A/ch with 3 pr or 7A/ch with 6 pr and using appropriate heatsinks and transformers.
If we discuss a power output, there is mentioned that BA-2 above class A at certain power point is turning to Class B. This breaking point is defined by bias.
Page 8 “the push-pull output stage in BA-2 can be biased at lower current values and still achieve the same maximum output power. Of course it will do so by falling into Class B (also thought of as AB)”.
What is a power output in class AB (or B)?
What is defining it? Supply rail or number of transistors?
It means 300W transformer for 2 channels with 1.5A each should be enough, isn’t it?
Is it correct then, that 3A pk and 25V supply rails across 4 Ohms will deliver around 50W of class A?
By doubling of bias to 3A it may increase power output to 60w/8R and heat dissipation to 150w/ch.
It seems a recommended bias limit is 1.2A per transistor pair (30W/25V=1.2A) according the statement “If you have enough heat sinking, you can
consider as high as 30 watts per device” (page 11).
It converts to 3.5A/ch with 3 pr or 7A/ch with 6 pr and using appropriate heatsinks and transformers.
If we discuss a power output, there is mentioned that BA-2 above class A at certain power point is turning to Class B. This breaking point is defined by bias.
Page 8 “the push-pull output stage in BA-2 can be biased at lower current values and still achieve the same maximum output power. Of course it will do so by falling into Class B (also thought of as AB)”.
What is a power output in class AB (or B)?
What is defining it? Supply rail or number of transistors?
Nelson usually recommends twice the idle draw as the bare minimum. So if the amp is dissipating 150W then 2x150W=300VA as a bare minimum. Best to triple it and round up. 3 x 150 =450, so that becomes a 500VA transformer for stereo.
So just remember idle power dissipation x 2 as a bare minimum.
This is not the whole answer, but I am trying to keep it simple.
P=I^2 x R
No, Power is limited by V^2/R
Yes to 150
Yes 1.2A is ok with 25V rails
Shouln't that be 3.6A and 7.2A based on 1.2A each
P=V^2/R
Last edited:
I’ve asked a lot of questions regarding the transformers in the PSU and different kind’s usability. On my search for suitable transformers I’ve found a commercial PSU for a very reasonable price, so I bought it. If this PSU work, which I believe, my amp will become a bit less DIY. I’d like it to be as muck DIY as possible but on the other hand the level of DIY is never 100%. It isn’t me who has constructed the amp and will never bee. 
Do you think this PSU will work?
/Forsman
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Do you think this PSU will work?
/Forsman
yup
you can :
1. find schematic of it and see , or
2. figure by your self -
does xformer is with Center Tap or not .
say that's 36Vac xformer with center tap , and 25Vdc is made with 2 diode full wave rectifying ;
in that case - you can use Graetz to have +/-25Vdc ( in reality around +/-22Vdc with A class as load ....... so - considering that you have 2 of them - each channel can have own supply
contrary to that - if xformer have sole secondary - you must use both of them to manage +/- supply ; it's best to each secondary have own Graetz and tie them after Graetz to have symmetric PSU
look in any FW service or user manual and you'll see PSU schematic drawn simply enough to understand
you can :
1. find schematic of it and see , or
2. figure by your self -
does xformer is with Center Tap or not .
say that's 36Vac xformer with center tap , and 25Vdc is made with 2 diode full wave rectifying ;
in that case - you can use Graetz to have +/-25Vdc ( in reality around +/-22Vdc with A class as load ....... so - considering that you have 2 of them - each channel can have own supply
contrary to that - if xformer have sole secondary - you must use both of them to manage +/- supply ; it's best to each secondary have own Graetz and tie them after Graetz to have symmetric PSU
look in any FW service or user manual and you'll see PSU schematic drawn simply enough to understand
Hi Melon Head,
Thanks for your explanations.
Probably I am too fresh to this, but I still have question.
Power output in class A
P=I^2 x R
It means if we have 1.5A bias per channel - power output in class A is 18W or 9W for 8 and 4 ohms speakers (1.5A^2 x 8=18W).
In case of 3A bias BA-2 would deliver 72W/8R and 36W/4R (with power dissipation of 150W/ch and at least 600W transformer).
No, Power is limited by V^2/R
Yes to 150
Total power output limitation (in class AB) by supply voltage rails is 78W/8R is 156W/4R. (25V^2/8=78.125W)
I do not see why power output wouldn't be increased in class A if bias will be changed from 1.5A to 3A/ch. With 3A bias 72W/8R class A power output is below limitation of 78W (limited by 25V supply rails).
The output power would be limited with 3.5A bias for 8 Ohms speakers by 78W. But it wouldn't be limited for 4 ohms speakers (49W).
Is that all approximatelly correct?
Thanks for your explanations.
Probably I am too fresh to this, but I still have question.
Power output in class A
P=I^2 x R
It means if we have 1.5A bias per channel - power output in class A is 18W or 9W for 8 and 4 ohms speakers (1.5A^2 x 8=18W).
In case of 3A bias BA-2 would deliver 72W/8R and 36W/4R (with power dissipation of 150W/ch and at least 600W transformer).
No, Power is limited by V^2/R
Yes to 150
Total power output limitation (in class AB) by supply voltage rails is 78W/8R is 156W/4R. (25V^2/8=78.125W)
I do not see why power output wouldn't be increased in class A if bias will be changed from 1.5A to 3A/ch. With 3A bias 72W/8R class A power output is below limitation of 78W (limited by 25V supply rails).
The output power would be limited with 3.5A bias for 8 Ohms speakers by 78W. But it wouldn't be limited for 4 ohms speakers (49W).
Is that all approximatelly correct?
No, for push-pull output stage class A limit is 2 x bias
So for 1.5A bias P = (2x1.5)^2 x 8 = 72W pk = 36W avg
For 3A bias P = 6^2 x 8 = 288W pk = 144W avg
However the rails of 25V limit the max power to (25-3)^2/8 = 60W pk = 30W avg
Melon Head,
Thanks for the formulas, it is more clear now.
If I apply it for 4 Ohms speakers:
1.5A bias P=(2x1.5)^2x4=36W pk = 18W avg
3A bias P=(2x3)^2x4=144W pk = 72W
25V limit (25-3)^2/4=121W pk = 60W avg
Based on calculations above I have couple conclusions:
There is no point to use bias above 1.5A if speakers are 8 Ohms.
2.5-3A bias would be optimal for 4 Ohms.
Is that correct?
That would be a fair conclusion if speaker loads were purely resistive.
In practice though this is not the case. It is not uncommon for speakers of 8 Ohms nominal impedance to have impedance dips as low as 3Ohms.
Now that you have the full picture, I will give you my philosophy about all of this.
If more current sounds better then it is better, if less current sounds better then it is better. If you can't tell the difference, then you can't tell the difference.
I usually bias as high as my heat sinks will tolerate and within the safe operating area of the devices. However if I can't hear any difference then I am happy to lower the bias.
In practice though this is not the case. It is not uncommon for speakers of 8 Ohms nominal impedance to have impedance dips as low as 3Ohms.
Now that you have the full picture, I will give you my philosophy about all of this.
If more current sounds better then it is better, if less current sounds better then it is better. If you can't tell the difference, then you can't tell the difference.
I usually bias as high as my heat sinks will tolerate and within the safe operating area of the devices. However if I can't hear any difference then I am happy to lower the bias.
Last edited: