Some good reading here Challenges in Faithful Reproduction of Musical Peaks: A Power Amplifier Designer's Perspective Article By Simon Thacher Ph.D. Of Spectron Audio
Ordered 25-0-25@300VA. Thanks folks. It makes for an interesting discussion and the more I read the more inquisitive I become.
Ordered 25-0-25@300VA. Thanks folks. It makes for an interesting discussion and the more I read the more inquisitive I become.
Basically there are a host of things that effect the perceived dynamics. The speaker drivers are most limiting. Pretty hard to generalize other issues because each system have their own flaws in different ways, then you have interfacing issues. The MyRef is already quite an achievement. But in order to get everything right, it is necessary to tackle the last part of the analog chain. It is a pity that the effort is really tremendous to get a good team to integrate well.
I can certainly vouch for Thacher's 400W Grand Piano. On own recordings, I clip a 2x1000W amp with 90dB/2.83V 1m. speakers trying to achieve realistic piano in the living room.
But you need far less for bleeding ear levels for pop concert in the living room
The difference is dynamic range ... remember, Soft bits & Loud bits? Beethoven has dynamic range. A rock concert is ALL con belto.
It's ludicrous that after striving for decades in the last century for dynamic range, achieving it as a given this century .. that new recordings have effectively no dynamic range. Just look at some modern music (?!!) in Audacity.
The trend is getting worse. One of Bob Cordell's favourite tracks for dynamic range in its original CD issue, a Ricky Lee Jones song .. has far less dynamic range in this Millenium's re-issues.
There's loadsa good stuff recently in this thread about transformers for a given amp. But also some rubbish.
I'll just add that the VA rating isn't only based on thermal considerations. A good maker will also quote the regulation at that VA. This give the effective series resistance that you need to put into the LTspice sim that AndrewT posted.
If we take a commercial view, a good 2x50W amp may use a 160VA item but that would need to have good regulation. More than 250VA would be icing on the cake and start giving you problems with switch on surge.
For LM3886, be careful of too high voltages on 4R speakers. What the datasheet only hints at is elevated voltages + higher dissipation, together = more triggering of SPIKE protection. This sounds very nasty .. a snap crackle pop noise .. unlike the rough noise of plain clipping. AndrewT's recommendations are appropriate.
Thacher (and Otala & other pseudo gurus) is (are) also misinformed about speaker impedance and the loads they present to amps. For ALL practical speakers, the lowest point on the impedance curve is the most difficult.
There's only one situation where the speaker may draw more current than that represents and if you hit it, the speaker is toast.
Last edited:
Sad is that the vinyl re-issues actually have a greater dynamic range than the CD issues these days. I bought some Chopin on CD a few weeks back. Has about 35dB dynamic range on the solos, which listening suggest has about 18-20dB of limiting added.
I have a bruckner CD on BIS that is actually unlistenable in the car as it keeps vanishing under the road noise. For me that is a compromise I can live with.
Would it be that hard for the labels that issue dual layer CD/SACD to put the uncompressed version on the SACD. If you have gone to the trouble of getting a player that can do the audio, surely you deserve the 'reality' of it.
And will ANYONE release a recording of 'Zadok the priest' that gives any sense of the kick in the head you get when a full choir comes in at '11'.
/rant on dynamic range.
IME, a 60W amplifier working correctly into say, 90dB sensitivity speakers has no problem generating subjectively intense sound - grand piano is not a problem, if the system is working properly. This sort of setup will easily start your ears ringing within 5 minutes, at well less than maximum volume, if playing more recent, heavily compressed rock music with 100% clean reproduction.
Most amplifiers wimp out at higher volumes, hence they sound like more power is needed - whereas the truth is, they just need fixing ...
Most amplifiers wimp out at higher volumes, hence they sound like more power is needed - whereas the truth is, they just need fixing ...
I think then I won't be able to use 4Ohm speakers after all, except for short duration testing. I do have 12Vdc fans on the heatsinks (I'm using the CPU heatsinks on each LM3886). I could go with a 25-0-25@225VA if that will suffice and could work with 8Ohms without switch-on-surge. I don't really want to plan for an occasional 4Ohm. What's the right thing to do?
Look at the impedance curve for your speakers.
If the lowest point is 6R4 or more, you'll have no problem treating them as 8R speakers.
60W+60W into 90dB/W @ 1m speakers is about 7dB short of achieving 110dB on the transient peaks at a 2.5m listening distance.
That 7dB below a "loud enough" target, is translated into needing ~300+300W into 90dB speakers.
points to a target transient peak well above 110dB at the 2.5m listening distance, or a very big listening room.
Last edited:
Hmmm ... had a look around, 109dB seems to be a figure offered up by many audio websites as to what a piano can do - but is that a meaningful number? I can't find a reference to a "proper study" that has measured that value - and, are there any recordings that have captured that peak volume with zero compression, deliberately or otherwise, in them?
Put it this way - playing a solo piano recording at a gain a few clicks from maximum creates the sense of a real instrument playing at the distance of the speakers or a touch more, for me - when listening to the real thing, the latter usually sound softer, because 99.9% of the time the musician is not playing them with the foot flat to the floor ...
Typically in the US on "modern" circuits any xformer with less than 400va is safe for turn on surge. Above 400va you will need a CL60 or soft start.
25 volt secondaries will work with a 4 Ohm load just not to the same volume as an 8 Ohm speaker. It's not like you can NEVER use a 4 Ohm speaker, you just can't use it as loudly. The spike protection will kick in and protect the amp once limits are reached.
Every transformer sold to the public will be designed/manufactured/tested to be "safe" for the user.
As a result every transformer sold to the public "is safe for a turn on surge".
The transformer will not blow up.
The transformer will not be damaged.
But the fuse required between the Mains supply and the transformer Primary can be very different depending on how one decides to start it up.
A 300VA 115:x-0-x Vac transformer will blow a T3.1A fuse on start up.
It will probably require a T6A or T8A fuse to start up reliably over many years of repeated cold starts.
A 400VA 115:x-0-x Vax transformer will also blow a T3.1A fuse on start up.
It may require as high as T10A to restart from cold repeatedly over many years.
Put in a CL60 with a cold resistance of ~10r to 15r and it will restart from cold repeatedly with a T3.1A fuse. Some 400VA may require a T4A and CL60 for reliable start up over many years.
The CL60 limits the current during start up such that a close rated fuse does not blow on repeated cold start ups.
That is not a safety issue. It is simply about whether one wants to adopt a close rated fuse, or not and selecting a start up strategy to suit.
It makes no difference whether it is a 300VA, or >400VA transformer.
Both categories of transformer size are "safe for a turn on surge".
As a result every transformer sold to the public "is safe for a turn on surge".
The transformer will not blow up.
The transformer will not be damaged.
But the fuse required between the Mains supply and the transformer Primary can be very different depending on how one decides to start it up.
A 300VA 115:x-0-x Vac transformer will blow a T3.1A fuse on start up.
It will probably require a T6A or T8A fuse to start up reliably over many years of repeated cold starts.
A 400VA 115:x-0-x Vax transformer will also blow a T3.1A fuse on start up.
It may require as high as T10A to restart from cold repeatedly over many years.
Put in a CL60 with a cold resistance of ~10r to 15r and it will restart from cold repeatedly with a T3.1A fuse. Some 400VA may require a T4A and CL60 for reliable start up over many years.
The CL60 limits the current during start up such that a close rated fuse does not blow on repeated cold start ups.
That is not a safety issue. It is simply about whether one wants to adopt a close rated fuse, or not and selecting a start up strategy to suit.
It makes no difference whether it is a 300VA, or >400VA transformer.
Both categories of transformer size are "safe for a turn on surge".
Last edited:
Actually it points to the the Moonlight Sonata played by a good pianist on a Steinway in any room in which a Steinway is appropriate.
If the pianist was Ludwig himself (a notorious headbanger) you'd need even more.
It's very educational trying to record Beethoven piano stuff played as God intended. If you ever try this, your ideas about sound reproduction will be changed forever.
This Millenium, the gear to do this properly is finally available at low cost.
Got a link with more info on these CL60 devices?
Anyone tried them on a big 500VA toroid for a PA and measured the effect on clipping power?
A deaf guy used to play the piano really loud...No sh*t sherlock
My room is not sadly not even appropriate for a mini grand. small upright is all I could squeeze in