Actually I need to start with a voltage first. Since the recommendation says a few volts above +/-22, and mauro used 25Vac, I think I'll go with 25Vac.
But beyond this point it's all theory and no concrete figures that can help calculate the required VA for the xfr. If I go by the 2:1 rule then for 80W then 160VA should be enough. Is this figure correct? A simple answer could be a yes or no or the VA value.
But beyond this point it's all theory and no concrete figures that can help calculate the required VA for the xfr. If I go by the 2:1 rule then for 80W then 160VA should be enough. Is this figure correct? A simple answer could be a yes or no or the VA value.
And to think we (just about every EE/audio techie) have been fighting this fight for four decades. It looks like even the FTC got it right after 25 years, but by that time the damage had been done:
http://www.n4lcd.com/RMS.pdf
If Mark Twain had lived in modern times he would have said "lies, damned lies, and audio power amplifier ratings."
Yes.
Var,
you are coming back questioning the "simple" rule of thumb and repeatedly asking for justification.
If you need justification, then do what I said way back
Did you take the time to look at Gootee?
Did you learn anything?
Did you do any research?
or are you just here trying to annoy us with the same questions and no input from yourself?
you are coming back questioning the "simple" rule of thumb and repeatedly asking for justification.
If you need justification, then do what I said way back
Did you take the time to look at Gootee?
Did you learn anything?
Did you do any research?
or are you just here trying to annoy us with the same questions and no input from yourself?
Last edited:
Apparently my questions have been sending the wrong signal. Yes I did some research and read the philosophy and there are several threads from Gootee so I'm still going through what I find is relevant. It's simple to multiply a given power output figure by a factor two given that there is consensus on the power output. 135W? 50W? 60W? In order to derive a figure I need to have some confirmed values which I can put in the formula. In fact all this while when I did the calculation I realized that my calculation in the initial question was wrong because of the formula (W = V*I not I^2*V). And even though I really appreciate all the answer and they are valuable, they came in without an understanding of the original query. I'm not designing a PSU. I'm just trying to get my choice right of several available ones.
benb's post has the answer I was looking for. +/-25Vac@160VA is good enough for myref dual-channel. This will hopefully get me up and running without 25yrs of research since my existing xfr went kaput. I also have learnt from your inputs and you've given me enough homework for a few days 🙂
I continue to search and research and learn with inputs from you and this great community out here 🙂 Apologies for the trouble.
benb's post has the answer I was looking for. +/-25Vac@160VA is good enough for myref dual-channel. This will hopefully get me up and running without 25yrs of research since my existing xfr went kaput. I also have learnt from your inputs and you've given me enough homework for a few days 🙂
I continue to search and research and learn with inputs from you and this great community out here 🙂 Apologies for the trouble.
25VAC gives +/-33V so 55W per channel in 8 Ohms and another 25W heat so 80W/ch*. In that case 160VA transformer will run both channels sine wave at full power all day. so with music you will have 6-18dB of headroom. Whether you consider that hitting the sweet spot or over designed is up to you.
* supply voltage may vary by half a volt depending upon the bridge being used.
* supply voltage may vary by half a volt depending upon the bridge being used.
Thanks billshurv. That means I could use the 225VA xfr without issues. Just to be on the safe side I thought I should go with a 300VA xfr (someone suggested in this thread to go with a higher VA to get a strong bass). I just don't want an overkill; that keeps things easy on the mains. And running 8 hrs works for me (since sometimes I forget to switch off the system and sleep to glory).
So how did you arrive at the 55W figure (LM3886 datasheet?)? And what formula did you use to get the 160VA figure? Just curious.
So how did you arrive at the 55W figure (LM3886 datasheet?)? And what formula did you use to get the 160VA figure? Just curious.
160VA will do a sine wave all day, which is the worst case signal you will find (crest factor of 3dB). worse case compressed rock will be 6dB.
55W was an interpolation off the data sheet, from where I also got the max dissipation for the rails that you have, add the 2 gets 80W peak power consumption per channel. Double that for stereo and you get 160VA.
300VA will not hurt, you just don't need that power. But companies like Krell have made us believe that huge iron is good. And nothing wrong with the warm glow you get from knowing that you have 5lbs of iron in there. Andrew will disagree of course!
LM3886 chip power amplifier power supply design. has a really good analysis of power supplies.
55W was an interpolation off the data sheet, from where I also got the max dissipation for the rails that you have, add the 2 gets 80W peak power consumption per channel. Double that for stereo and you get 160VA.
300VA will not hurt, you just don't need that power. But companies like Krell have made us believe that huge iron is good. And nothing wrong with the warm glow you get from knowing that you have 5lbs of iron in there. Andrew will disagree of course!
LM3886 chip power amplifier power supply design. has a really good analysis of power supplies.
I would consider peak and continuous conditions individually. You want the power supply to satisfy both for optimum power output capability.
power supply should satisfy the peak, not necessarily the transformer. That's what capacitors are for...
Two Members that are awake.
Average conditions define the heatsink temperatures/dissipations.
Transient conditions define the junction temperatures.
what speakers do you have and how loudly do you listen? Its easy to forget that the difference between 30W and 60W is barely audible due to our damned logarithmic ears.
yes for every 3db, the watts are doubled/halved, and the difference is barely noticeable unless the difference is 10db where in the sound perception doubles/halves?
I have 8Ohms and I listen full-blast when I exercise or have a coffee.
I also wanted to keep it safe for an occasional 4Ohm speaker.
I have 8Ohms and I listen full-blast when I exercise or have a coffee.
I also wanted to keep it safe for an occasional 4Ohm speaker.
8 ohms and what sensitivity? If you sit 4m away and have 86dB/W sensitivity with peak of 100dB (enough to damage your hearing long term) then we have 80dB/W for the pair at 4m so you need +20dBW or 100W. If you are only 2m away you can deafen yourself with 50W per channel.
20W will do the damage then. Anything more is headroom! so back to original question either transformer will work. Decide whether you prefer headroom or more current capability in case a 4 Ohm speaker gets plugged in for a headbang.
I'd go with more current. How much voltage should I go for? 22Vac? My primary speakers still will be 8Ohms@93db.
The concern is actually the damage that a high voltage can cause. 25-0-25@300VA can measure 27Vac unloaded. Could this damage the amp?
If I plug a 4Ohm speaker with low volume, will the amp be able to take it at 25-0-25@300VA? Sound output is not the issue. Safety is.
A higher supply voltage will help avoid clipping. And as per the datasheet, 25Vac still keeps us well below the absolute max supply voltage (84Vdc)
If I plug a 4Ohm speaker with low volume, will the amp be able to take it at 25-0-25@300VA? Sound output is not the issue. Safety is.
A higher supply voltage will help avoid clipping. And as per the datasheet, 25Vac still keeps us well below the absolute max supply voltage (84Vdc)