Yeah, I now classA idle bias is at max always, from idle to full blast
Im by no means an expert on such things
But anyway, I dont think you can just use idle consumption to calculate what kind of power is consumed at output terminals, due to various speaker loads
I dont think that idle consumption is the same as power supply needs in low impedance load
But those kind of calculations are really not my kind of thing
I will follow what Nelson recommends, 6A continious and 10A peak, each channel
2x 18Vac, 10A peak, equals 180VA at each secondary = 360VA(watt) total, each channel
I may have been wrong saying it would be minimum
Its just whats optimally needed, no less no more
With this you can probably run it at close to full blast, with just a small audible increase in distortion
With smaller supply distortion may increase before, but on that point Im way over my head
As said, less may work fine, if not loaded too hard
If someone wants something else, tailored for lighter loads, fine
I have done what I could to be very clear on this
F5 manual is quite clear on this
I dont think Im wrong
Im by no means an expert on such things
But anyway, I dont think you can just use idle consumption to calculate what kind of power is consumed at output terminals, due to various speaker loads
I dont think that idle consumption is the same as power supply needs in low impedance load
But those kind of calculations are really not my kind of thing
I will follow what Nelson recommends, 6A continious and 10A peak, each channel
2x 18Vac, 10A peak, equals 180VA at each secondary = 360VA(watt) total, each channel
I may have been wrong saying it would be minimum
Its just whats optimally needed, no less no more
With this you can probably run it at close to full blast, with just a small audible increase in distortion
With smaller supply distortion may increase before, but on that point Im way over my head
As said, less may work fine, if not loaded too hard
If someone wants something else, tailored for lighter loads, fine
I have done what I could to be very clear on this
F5 manual is quite clear on this
I dont think Im wrong
I think the big difference between Babowana's and Tinitus' calculations is:
B. starts from an average dc current value while T. starts from a peak value.
Isn't it?
Who is near from the right line?
Could you help us, Nelson?
Certainly many of us would like to better understand amplifiers transformer rating.
On the other hand we are into a "First watt" thread. How hard a F5 can be pushed is not her more interesting aspect.
Her sweetness when used in pure class A power makes her best signature.
For hard rock and techno, why not to add a class AB or classD bottom amp, and so, to keep the F5 for a more subtile task?
B. starts from an average dc current value while T. starts from a peak value.
Isn't it?
Who is near from the right line?
Could you help us, Nelson?
Certainly many of us would like to better understand amplifiers transformer rating.
On the other hand we are into a "First watt" thread. How hard a F5 can be pushed is not her more interesting aspect.
Her sweetness when used in pure class A power makes her best signature.
For hard rock and techno, why not to add a class AB or classD bottom amp, and so, to keep the F5 for a more subtile task?
bobodioulasso said:
I believe Babowanas calculation is as accurate as can be
But with no "safety" volume added
Its more like if I had calculated with continious consumption, and not used peak value
I use peak value, yes
But its the peak value given by Nelson
Safety volume/margin has been calculated already, I guess
Its just to use the 10A given, and no more
May be thats what you would call headroom
Sure, with smaller trafo you may have perfectly fine peak/headroom value, but maybe only up to about 15watt
But sure, a total peak situation may not happen so often
With sensitive speakers and not exstremely loud you would probably be fine with just 100watt trafos, in normal daily use
Im not arguing with that at all
But you could just as well say, hell Im only using say 10watt, why then a 25watt, its 15watt wasted
But OK
Theres nothing wrong in calculating for certain application, speakers etc
But I dont think such special dimensioning should be used for general trafo size suggestion here
Recommended/suggested trafo size hould be what Nelson has calculated to be opitimal for any kind of loads, no more no less
The optimal for ANY kind of loads
To conclude
Im not saying less than optimal wont work, cause smaller trafos will be fine in many situations, just not the exstreme
Like say, playing at full blast using all 25watt, and maybe into 3ohm
If you want to be able to draw absolute maximum power, you will need big "oversized" trafos, or you will get distorted transints
OK I will try to have a hack at this.
First of all there are 2 concerns here:
i) The first one is the idle heat dissipation 60Watts per channel.
From memory you need to double this (for class A) so that is 120VA per channel.
ii) The second one is that this amp goes into class B and according to the F5 manual, current limiting cuts in at 10Amps peak current per channel so 24V x 10A = 240W peak power = 120W rms power per channel so you need at least 120VA
Then which ever one is higher out of the two is the bare minimum. In this case the requirements between i) and ii) are identical (which is rarely the case) so you would need at least 120VA per channel or 240VA for stereo.
So round up 240VA to the next transformer size and you get 300VA.
However if you are like me you are likely to build another amp in the future, so to save yourself having to buying another higher rated transformer (eg you want to build the Aleph J) then you might find better value in buying a 500VA transformer.
First of all there are 2 concerns here:
i) The first one is the idle heat dissipation 60Watts per channel.
From memory you need to double this (for class A) so that is 120VA per channel.
ii) The second one is that this amp goes into class B and according to the F5 manual, current limiting cuts in at 10Amps peak current per channel so 24V x 10A = 240W peak power = 120W rms power per channel so you need at least 120VA
Then which ever one is higher out of the two is the bare minimum. In this case the requirements between i) and ii) are identical (which is rarely the case) so you would need at least 120VA per channel or 240VA for stereo.
So round up 240VA to the next transformer size and you get 300VA.
However if you are like me you are likely to build another amp in the future, so to save yourself having to buying another higher rated transformer (eg you want to build the Aleph J) then you might find better value in buying a 500VA transformer.
labjr said:
Wow, I second that motion! I'm nowhere near a Nelson Pass category in my understanding, but you guys are making my head hurt.
If you want to stay in Class A operation then you cannot exceed the bias current, it doesn't matter what load you are using. If you want to optimize for a greater load (4 ohms) then you have to bias it higher, but the fets are already around 30W dissipation at 1.3A.
When you move into class AB you start to turn things off and on, so your average dissipation goes down and your peak demands go up. Technically speaking, each channel pulls about 75W, so 150W constant is needed for 2 channels and the rule of thumb is 2 to 3X for a quiet and cool trafo. If you're going to run it way up into AB then you might want to go bigger, but why?
Just my two cents.
Bill
I built my F5 with 2x250va transfos because when reading Nelson's pdf, i understood 300va recommended per channel.( 6 amps continuous and 10 amps peak)
But now, i can read and read again the pdf i cannot understand wether he recommands 300 va for one only channel or for both.
Calculating the rating of a transfo is not that simple because, in a linear PSU, the current drawn by the load and the current flowing through the transfo secondary windings are quite differents.
But now, i can read and read again the pdf i cannot understand wether he recommands 300 va for one only channel or for both.
Calculating the rating of a transfo is not that simple because, in a linear PSU, the current drawn by the load and the current flowing through the transfo secondary windings are quite differents.
I guess I should finish my answer. You hear some people complaining about their trafo buzzing and that is what you get when you use more caps than the trafo can handle because of the peak charging current. Mr Pass has already done the math, if you use more caps then you need a bigger trafo.
Best, Bill
Best, Bill
I mean that a not negligible amount of power is lost into the rectification and filtering process , more than 30%.
go to fig2.
http://sound.westhost.com/power-supplies.htm#intro.
go to fig2.
http://sound.westhost.com/power-supplies.htm#intro.
No way! Let's not get wrapped up in PS design on this thread, it's just not helpful to most readers. You're speaking of the difference between VA and Watts, that's called Power Factor and involves phase angles, not a casual subject. Suffice it to say if you double or triple the trafo size, as has been done, you have that covered.
Best, Bill
Best, Bill
Looks like Papa got his own BlowTranny thread.
Why not use whatever pleases you.
Just don't waste dinero on a transformer with a VA size that's more than the output stage can handle.
Example from the middle ages : Metaxas Solitaire 150W Class AB stereo power amp.
1st generation :
16 x 150W power devices ( superduper 60/80MHz Toshiba 2SA1095/2SC2565 ring-emitters)
16 x 150W= 2400W.
Transformer=> 2000VA = less than 2400W.
2nd generation :
Toshiba stopped production of the lightspeed devices, so Metaxas switched over to 50-60MHz fT NEC 120W power transistors. (A1227/C2987)
16 times 120W = 1920W.
Consequently, the transformer was downsized to 1600W. (= 1/6th less than the output stage number, same as for the first series)
The F5 only has a single pair of 150W devices in the output stage.
Stereo : 4 x 150W = 600W
Anything over 600VA for a power transformer would be wastefull.
If you follow Costas Metaxas wet pinky rule : 5/6 times 600 = 500VA Max for the transformer of a stereo F5.
Just read the first few pages of this thread, the 6 Amp continuous and 10 Amp peak numbers will just pop out of the hat.
Why not use whatever pleases you.
Just don't waste dinero on a transformer with a VA size that's more than the output stage can handle.
Example from the middle ages : Metaxas Solitaire 150W Class AB stereo power amp.
1st generation :
16 x 150W power devices ( superduper 60/80MHz Toshiba 2SA1095/2SC2565 ring-emitters)
16 x 150W= 2400W.
Transformer=> 2000VA = less than 2400W.
2nd generation :
Toshiba stopped production of the lightspeed devices, so Metaxas switched over to 50-60MHz fT NEC 120W power transistors. (A1227/C2987)
16 times 120W = 1920W.
Consequently, the transformer was downsized to 1600W. (= 1/6th less than the output stage number, same as for the first series)
The F5 only has a single pair of 150W devices in the output stage.
Stereo : 4 x 150W = 600W
Anything over 600VA for a power transformer would be wastefull.
If you follow Costas Metaxas wet pinky rule : 5/6 times 600 = 500VA Max for the transformer of a stereo F5.
Just read the first few pages of this thread, the 6 Amp continuous and 10 Amp peak numbers will just pop out of the hat.
Indeed, for any reactive load operating in a steady state sinusoid...
There is Average power in Watts:
Reactive power in Volts Amps reactive (VAR)
And the Apparent Power which is the Root of the sum of the two above squared)
The angle between them is the power factor angle and is the phase difference between the voltage across and current through the load.
If you read Mr. Pass's articles, starting with the Zen series, he walks through the power requirement calculations.
Ryan
There is Average power in Watts:
Reactive power in Volts Amps reactive (VAR)
And the Apparent Power which is the Root of the sum of the two above squared)
The angle between them is the power factor angle and is the phase difference between the voltage across and current through the load.
If you read Mr. Pass's articles, starting with the Zen series, he walks through the power requirement calculations.
Ryan
I think somewhere back in this thread it mentions the transformer. If you look at the 6 moons review you can see it's a 300VA Plitron.
http://www.6moons.com/audioreviews/firstwatt7/f5_2.html
We listened to Tea-bag's F5 and it had a 300VA Plitron in it. From what I understand the Plitrons regulate well right up to the VA rating.
I think some have used a 400VA when using an Antek transformer just to be sure.
Were not making this up !
http://www.6moons.com/audioreviews/firstwatt7/f5_2.html
We listened to Tea-bag's F5 and it had a 300VA Plitron in it. From what I understand the Plitrons regulate well right up to the VA rating.
I think some have used a 400VA when using an Antek transformer just to be sure.
Were not making this up !
Nelson Pass said:Like you guys need help coming up with a power supply that
meets that description.
Poor babies...
This topic has been discussed earlyer, this was Nelson's answer.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=121228&perpage=25&pagenumber=9
post #212
bobodioulasso said:
Quote from F5 manual
"The power supply of the tested amplifier is +/-24V and should be rated at 6A continious duty, and more than 10A peak per channel"
Looks perfectly clear to me
Other than I just realised "MORE than 10A peak, pr channel"
Remember that both secondarys are summed to give total VA rating of trafo
Thats how trafo manufactorer does VA rating
We have to do it the same way, or we wont get what we expect
When looking at what is used in FirstWatt, its different
Nelson may conclude that the 300VA rating for a mono amp is good to use for normal stereo also
But you should not calculate backwards from that, and use only half of that for mono amps
You can use 300VA trafo fore a stereo amp
But also use 300VA for mono amps, or else theres no advantage of doing mono amps
Stereo amps is kind of a funny thing to me
You may not really need to have double VA rating
But fore tough loads we all know we want mono amps
Same thing with an integrated
You wont expect a stereo amp to have the same power as mono amps
There is also a reason why stereo amps are also made as "double mono"
Stereo amps are not twice a monoamp, and mono amps are not half a stereo amp
Let's assume we have been advised to use a 300VA 18Vac + 18Vac transformer for our ClassA power amplifier project.
The transformer manufacturer will stipulate that the transformer must be de-rated when feeding a capacitor input filter.
The usual de-rating factor is between 60% and 70%.
I'll use the 67% figure for this example.
The maximum voltage after the rectifier when the transformer is delivering maximum power is
~sqrt(2) * Vac -diode drop.
Our 18Vac secondary will give approximately 24Vpk, using a bridge rectifier on each secondary.
If mains ripple is low then that is the PSU output voltage.
We have available 67% of 300W @ +-24Vdc at the PSU output.
The maximum continuous current available is
Icont ~ 300 * 0.65 / [24+24] ~ 4.2Adc
Note the ratio between rated AC current and maximum continuous DC current (~ 2:1).
The transformer will run at the manufacturer's maximum rated temperature if one draws the maximum rated DC current.
De-rate by a further 50% to keep the transformer cool.
The cool running maximum DC current is ~25% of the AC rating.
So our 300VA, 8.3Aac transformer can supply 2.1Adc all day.
Well in excess of the 1.3A bias current of the F5.
A 6Aac transformer (225VA, 18+18Vac) can meet the 1.3A bias requirement.
11Aac (400VA 18+18Vac) will manage 2channels each @ 1.3A of Class A bias.
That is the minimum transformer size to meet the bias requirement alone.
The transformer manufacturer will stipulate that the transformer must be de-rated when feeding a capacitor input filter.
The usual de-rating factor is between 60% and 70%.
I'll use the 67% figure for this example.
The maximum voltage after the rectifier when the transformer is delivering maximum power is
~sqrt(2) * Vac -diode drop.
Our 18Vac secondary will give approximately 24Vpk, using a bridge rectifier on each secondary.
If mains ripple is low then that is the PSU output voltage.
We have available 67% of 300W @ +-24Vdc at the PSU output.
The maximum continuous current available is
Icont ~ 300 * 0.65 / [24+24] ~ 4.2Adc
Note the ratio between rated AC current and maximum continuous DC current (~ 2:1).
The transformer will run at the manufacturer's maximum rated temperature if one draws the maximum rated DC current.
De-rate by a further 50% to keep the transformer cool.
The cool running maximum DC current is ~25% of the AC rating.
So our 300VA, 8.3Aac transformer can supply 2.1Adc all day.
Well in excess of the 1.3A bias current of the F5.
A 6Aac transformer (225VA, 18+18Vac) can meet the 1.3A bias requirement.
11Aac (400VA 18+18Vac) will manage 2channels each @ 1.3A of Class A bias.
That is the minimum transformer size to meet the bias requirement alone.