Hi,
I´m thinking about dimensioning of my Babelfish J while waiting for the boards to come.
Bought two heatsinks from Seifert with 0,125 K/W (according to their brochure), so they get rid of about 200 W/channel.
Could I go with 4 IRF 150 per channel with an 18V transformer (4 x 2 x 24V x 1A = 200 W?)? Do I have to change any values on the base- or daughter-boards … and would one 400 VA transformer per channel be a good choice?
Thanks.
Axel
I´m thinking about dimensioning of my Babelfish J while waiting for the boards to come.
Bought two heatsinks from Seifert with 0,125 K/W (according to their brochure), so they get rid of about 200 W/channel.
Could I go with 4 IRF 150 per channel with an 18V transformer (4 x 2 x 24V x 1A = 200 W?)? Do I have to change any values on the base- or daughter-boards … and would one 400 VA transformer per channel be a good choice?
Thanks.
Axel
Is there a thread describing the basic babelfish circuit or project description? I've searched but can't find one.
Hi,
I´m thinking about dimensioning of my Babelfish J while waiting for the boards to come.
Bought two heatsinks from Seifert with 0,125 K/W (according to their brochure), so they get rid of about 200 W/channel.
Could I go with 4 IRF 150 per channel with an 18V transformer (4 x 2 x 24V x 1A = 200 W?)? Do I have to change any values on the base- or daughter-boards … and would one 400 VA transformer per channel be a good choice?
Thanks.
Axel
first .... with 18Vac , resulting in something as 22V5dc rails , it's completely unnecessary going above 3A for Iq , just because you'll hit voltage threshold before you hit I threshold
further , when calculating dissipation , observe that current is flowing through transistor vertical pairs (from +rail to - rail ) , so if you have 3 vertical pairs (if you go for 4 mosfets per channel , why not taking another two ..... ) , you multiply intended Iq/pair with 3
so , let use 25W dissipation per mosfet as example , so 50W per vertical pair - that will give you 50Wx3=150W of heat per channel
(150W of power consumption ..... no less than 350VA xformer per channel and at least 66mF per rail)
for 50W/pair , rail to rail 45Vdc , that's something as 1A1 for Iq ..... giving 3A3 as Iq for channel
it'll result in 25W-ish rms power @8R (voltage threshold) , while 40W-ish @4R load (current threshold)
plenty , if you ask me .......
if you need more power - time is for higher rails , and wise choosing of Iq
so , what are size of room , which speakers and loudness of your listening ?
Obviosly didn´t understand your comments in post 21...
I have a TMR Standard II, 90dB 6Ohms in 35 qm, normal listening levels, so 25W could be enough:
That means 6 x IRF150 per channel, all resistor as in your schematics (R2, 4, 6 = 0,27R, R1, 3, 5= 180R, R31 ... R36= 0,27R)?
Increasing to 20Vac, the dissipation (+10%) still could be handled by the heatsinks?
Thanks.
I have a TMR Standard II, 90dB 6Ohms in 35 qm, normal listening levels, so 25W could be enough:
That means 6 x IRF150 per channel, all resistor as in your schematics (R2, 4, 6 = 0,27R, R1, 3, 5= 180R, R31 ... R36= 0,27R)?
Increasing to 20Vac, the dissipation (+10%) still could be handled by the heatsinks?
Thanks.
go with 20Vac secondaries , 3 pairs in output and you'll set Iq as high as heatsinks can handle
in your case - without going to most masculine dissipation per device , you can go with IRFP240 too , easily obtainable from member h_a , in case you already didn't bought quantity to match them by your self
in your case - without going to most masculine dissipation per device , you can go with IRFP240 too , easily obtainable from member h_a , in case you already didn't bought quantity to match them by your self
OK, did my homework studying this thread. Here is my newbie talk:
My goal is to get 50W @ 8ohm / ch
1. Rule of thumb for single-ended class A: the output power and dissipated power ratio is about 1:8. Eg. Aleph J has 25W output at 200W dissipation.
2. ZM's post #69: using FW/PL rail voltage 21.5Vdc (+/-21.5V so 43V rail-to-rail), Iq=4A using 1x IRF150/quadrant. Dissipation power=43V * 4A = 172W. Using 1/8 ratio, this will give 20W output.
Question: what is "quadrant"? I guess each channel must have two quadrants since ZM said "42Wrms/8r" in post #69.
3. In ZM's post #21, he said "30V rail - 2A Iq - 50W/mosfet".
Questions: this "/mosfet' means "/IRF150", right? so each "quadrant" has 2x IRF150, right? So that means daughter board is needed to put 2xIRF150 there, correct?
So using the example in Post #69 as a starting point. That means for each channel I need a big square main board (which has 2x mosfet mounting points at two corners). I will need two daughter board. Each daugther board holds two IRF150 and goes to the mosfet mounting point on the main board. I guess that is one "quadrant", since I got 2X IRF150 there, I had 4A Iq. If I use 21.5Vdc rail, I will have 172W dissipation and 20W output from each "quadrant" or daughter board. The total two daughter boards in one channel give me rought 43W @ 8ohm, correct?
For 40W average power, the 8ohm load needs peak volt: 18*1.414 = 25V , or peak current: 2.23*1.414 = 3.1A
Question: I am voltage limited? (25V>21.5V rail) or not yet (25V < 21.5 x 2)?
So if I want to reach 50W, I need to raise rail voltage?
Sorry for the long post and thanks for the help!!
My goal is to get 50W @ 8ohm / ch
1. Rule of thumb for single-ended class A: the output power and dissipated power ratio is about 1:8. Eg. Aleph J has 25W output at 200W dissipation.
2. ZM's post #69: using FW/PL rail voltage 21.5Vdc (+/-21.5V so 43V rail-to-rail), Iq=4A using 1x IRF150/quadrant. Dissipation power=43V * 4A = 172W. Using 1/8 ratio, this will give 20W output.
Question: what is "quadrant"? I guess each channel must have two quadrants since ZM said "42Wrms/8r" in post #69.
3. In ZM's post #21, he said "30V rail - 2A Iq - 50W/mosfet".
Questions: this "/mosfet' means "/IRF150", right? so each "quadrant" has 2x IRF150, right? So that means daughter board is needed to put 2xIRF150 there, correct?
So using the example in Post #69 as a starting point. That means for each channel I need a big square main board (which has 2x mosfet mounting points at two corners). I will need two daughter board. Each daugther board holds two IRF150 and goes to the mosfet mounting point on the main board. I guess that is one "quadrant", since I got 2X IRF150 there, I had 4A Iq. If I use 21.5Vdc rail, I will have 172W dissipation and 20W output from each "quadrant" or daughter board. The total two daughter boards in one channel give me rought 43W @ 8ohm, correct?
For 40W average power, the 8ohm load needs peak volt: 18*1.414 = 25V , or peak current: 2.23*1.414 = 3.1A
Question: I am voltage limited? (25V>21.5V rail) or not yet (25V < 21.5 x 2)?
So if I want to reach 50W, I need to raise rail voltage?
Sorry for the long post and thanks for the help!!
Last edited:
If you want 50W with some head room 34V rails will well and truly get you 50W.
Or you could settle for 30V rails (a bit under 50W)
Or you could settle for 30V rails (a bit under 50W)
Thanks for the conclusion! 😀
So I am voltage limited now?
I will also appreciate more comments from my questions, so I can learn to get estimation more accurately. 😱
So I am voltage limited now?
I will also appreciate more comments from my questions, so I can learn to get estimation more accurately. 😱
Last edited:
You want a supply capable of at least 30V dc on both rails.
P= V^2/R
=30^2/8
=900/8
= roughly 112W peak
= 56W avg into 8 Ohms
Except a 30V supply won't give you 30V at the speaker load you migh only get 27V at the load.
Repeat calculation with 27V for a bit of practice.
P= V^2/R
=30^2/8
=900/8
= roughly 112W peak
= 56W avg into 8 Ohms
Except a 30V supply won't give you 30V at the speaker load you migh only get 27V at the load.
Repeat calculation with 27V for a bit of practice.
Thanks for the math. I guess what I need more help is on the number of transistors and boards. 😀
Thanks for the math. I guess what I need more help is on the number of transistors and boards. 😀
I'll let Zen Mod fill you in on those details.
Generally speaking around 35W per device, but from memory Zen Mod uses double die devices. He is best person to advise, you might be lucky to push them to 50W each.
you Gentlemen were typing while I was sleeping , for change
I need first coffee now , before typing anything related to power and numbers ....... or Dukja could end managing smaller power station
I need first coffee now , before typing anything related to power and numbers ....... or Dukja could end managing smaller power station
Good morning, ZM! (maybe afternoon, 🙂 )
I forgot to mention my gear. I got a about 85dB speaker (New Philharmonitor from Dennis M.) and listening distance is about 7 feet. Room is about 15x30x7 feet. Usually my listening volume is less than 70dB. With soundtrack, classical music, and a few female vocal music, I don't think my SPL meter ever register more than 90dB. I use some SPL calculation template and it seems with 30W I can get around 95dB before my Aleph 3 clips. That is what I am running now.
So I figure that 50W BJ should let me run up almost 3dB. I just realized that how little amp watt can do with our nonlinear ear. 🙁 So what do you think the best watt route for me? thanks!!
I forgot to mention my gear. I got a about 85dB speaker (New Philharmonitor from Dennis M.) and listening distance is about 7 feet. Room is about 15x30x7 feet. Usually my listening volume is less than 70dB. With soundtrack, classical music, and a few female vocal music, I don't think my SPL meter ever register more than 90dB. I use some SPL calculation template and it seems with 30W I can get around 95dB before my Aleph 3 clips. That is what I am running now.
So I figure that 50W BJ should let me run up almost 3dB. I just realized that how little amp watt can do with our nonlinear ear. 🙁 So what do you think the best watt route for me? thanks!!
It is original.
ZM your memory failed you. 😀 (but you got too many people to remember🙂 )
There are plenty of photo in this thread that you helped me before:
http://www.diyaudio.com/forums/pass-labs/200231-aleph3-recap.html
ZM your memory failed you. 😀 (but you got too many people to remember🙂 )
There are plenty of photo in this thread that you helped me before:
http://www.diyaudio.com/forums/pass-labs/200231-aleph3-recap.html

meaning that Americans are equally dumb as Europeans , just in opposite direction
so , 6L6 , you're Babelfishing Babelfisherman
![]()
that was interesting thread .........
so - we can convert your Aleph 3 to Aleph J3 (simple to do) , and you can make somewhat stronger Aleph , so having two great amps in the house
😉
if you click on , you'll find SM for Aleph 5 , which is roughly what you're looking for , regarding power range
look at schematic and read the text
that means - 30-ish volts PSU rails , around 150W of heat per channel
it demands (speaking of Babelfish J V2.) main pcb and two daughter boards , for 3 pairs of mosfet outputs, all that speaking per channel
didn't looked on posts you mentioned above , but if I was writing about quadrants , no other way than it was strictly related to Aleph X , or some other Papamp , being symmetrical in construction ..... meaning - having four quadrants populated with mosfet outputs
that's not the case with these simple Aleph iterations
Hi Dukja,
Have you seen this thread? :
http://www.diyaudio.com/forums/mult...much-voltage-power-do-your-speakers-need.html
Assuming you are listening with a digital source, you can get an idea how much
voltage swing you are currently using, and hopefully how much larger an amp you
want.
In the event a larger Aleph is required, just build a new one and send your original
Aleph 3 to Zen Mod...for proper disposal. 🙂 🙂 🙂
Cheers,
Dennis
Have you seen this thread? :
http://www.diyaudio.com/forums/mult...much-voltage-power-do-your-speakers-need.html
Assuming you are listening with a digital source, you can get an idea how much
voltage swing you are currently using, and hopefully how much larger an amp you
want.
In the event a larger Aleph is required, just build a new one and send your original
Aleph 3 to Zen Mod...for proper disposal. 🙂 🙂 🙂
Cheers,
Dennis
- Home
- Amplifiers
- Pass Labs
- About possible Babelfish J interest