Iain Churches wrote:
>
> "Alan Rutlidge" <don't_spam_me_rutlidge@[EMAIL PROTECTED]
> wrote in message
> news:47e1c209$0$6869$5a62ac22@[EMAIL PROTECTED]
> >
> >
> > I haven't yet bothered measuring the peak power delivered to my
speakers
> > during either moderate or loud listening levels, however doing so
would
> > give a reasonable indication of what output power capability of the
toob
> > amplifier would be deemed reasonable taking into consideration the
wide
> > genre of music I listen to.
>
> It would be interesting to put a DVM with peak hold
> across the speakers, and work out the power level with
> V2/R. Alternatively, to get a fig with a generous margin
> you could set the system to play as loud as is comfortable
> and the replace with CD player with a generator at 2V
> and the speaker with a dummy load, and measure.
>
> >
> > The 2.5 ohm nominal Z of both the Apogees and the Jupiters does IMHO
pose
> > a bit of a challenge to the average toob amplifier, but taking into
> > account we aren't talking about your average toob amplifier and the
> > possibility the OP trannies can be wound to match a 2 - 2.5 ohm load
is
> > encouraging.
>
> They can indeed. The 4 Ohm winding on the C50 the
> pic of which you have referred to earlier) is actually
> 3.5 Ohms. So that is close. Let's see what the winders
> think about 2.5 Ohms.
There is no problem whatsoever for a skilled OPT winder to
make an OPT which has more than one load match and
a match to any low ohm value.
It all depends on who designs it, and who winds it.
Accountants do the design now, and winders have no skills...
Hammond makes the lowest common denominator type of OPT
and a visit to their website explains how they arrange
their windings to get 16, 8 & 4 ohms in all of their OPT.
Most Chinese OPT are worse, so a Hammond replacement is an improvement.
A visit to MY website will show you how to get any number of load
matches
and how to alter the arrangement of a larger number of secondary winding
sections to
maintain the same winding loss % regardless of load match chosen, and
maintain equal
current density in all individual separate windings, and to maintain the
same
amount of leakage inductance.
Nobody who builds amps commercially wants to go down this inconvenient
path because owners changing OPT
winding sections with solderable straps across terminals here and there
invariably make a mistake,
with a smoky result for music. In today's world of ignorant users,
this is a ****al solution.
But I get 400 hits a day at my site, so obviously diyers might have a
different view.
When I know nothing about how the amp is to be used, I supply with ONE
default 5.6 ohms nominal load.
SE amps are made so maximum PO is at 5 ohms.
This allows good current performance to 2 ohms, and good voltage
performance to where RL < 8 ohms,
which is nearly always the case with bass speakers and their resonant
peak.
With PP amps which have the option or ""benefit"" of AB action,
I make the amps so absolute max PO is at 3 ohms, so 2 x KT88 will give
about
60 watts at clipping into 3 ohms.
About 50 watts is possible at 5 ohms, and 35 watts into 8 ohms.
Many makers nor make the load where absolute max class AB occurs is at 8
ohms
when the 8 ohm outlet is used.
This is real ****tard design for hi-fi, because speakers Z falls
alarmingly below 8 ohms in many cases. A 4 ohm outlet is mandatory,
but even then the power is mainly AB with a 4 ohm load connected.
Many makers now insist they will the war of ****ing watts.
Distortion and tube cooking is barely mentioned in inadequate specs.
The modern buyers have no clue, so we have many makers producing what
are virtually PA amps as regards loading, but with an OPT with high BW
to allow only a very moderate amount of NFB.
Distortion is 10 times what it was in 1960.
The other method of output load match change is to provide an octal
socket
and with an octal plug which has TWO ways of plugging in, so that say 6
ohms can be had
with N turns on the OPT sec, and then for 3 ohms you rotate the plug
position
and you get 0.707 N turns.
The plug needs to be tied to the amp with a chain to prevent its loss.
This turn relation****p of 1 : 0.707 isn't very convenient for winders
and its
difficult to fit the turns on so you get full layers, so study my
website for how to sectionalise
secondary windings to get a useful variety of sections and very nice
outcomes.
Winders HATE putting in necessary taps within the width of any layer of
winding.
Designers cave in to demands of the low skilled winding workers
and try desperately to get by with having all taps if any at the
beginning or end of a layer
of wire, and as a result the performance is always compromised with
regard to losses
and HF response.
This desire for no taps within layers suits winding machines where
perhaps 20 bobbins are being wound by the same lathe from 20 spools of
wire.
For example, the 43% tap recommended for the UL screen connection is
because
43% is 3/7 of a 1/2 primary. Typically, an OPT winder is told by the
designer
to put on 14 total layers of anode primary.
So there are only 2 sections of secondary between 3 sections of primary.
So on goes 4 layers of P, with a tap for a screen, then a section of S,
then 3 layers of P, and a CT,
then 3 more layers of P, and another tap for a screen, then another S
section, and then 4 more layers
of P, and its done.
And its pretty ****ing awful.
But that's how most mediocre OPTs are wound which give high leakage
inductance,
and they are difficult to keep unconditionally stable with any amount of
global NFB,
AND to get wide HF extension.
I have to say I don't much like taps along windings myself.
But done properly with polyester sleeving on the wire in, and wire out
to make a tap, its OK.
I did wind one power tranny with 48 taps and one low voltage darn
winding developed a shorted turn
after I potted it. I'd used only heatshrink sleeving.
It heated up and analysis showed one turn of 1.25mm dia wire was
shorted,
and must have been running red hot. The tranny was a write off,
and I put it in the fireplace and cooked it with a wood fire for 3 hours
to easily extract the core for re-use.
>
> > As has been mentioned, my only concern might be the DF of a toob
> > amplifier. Having become somewhat accustomed to a tight and controlled
> > bass response from the ME850 HiCap will the toob amplifier live up to
> > these expectations?
>
> Yes. That is indeed something that needs to be carefully
> investigated. I have spoken with several people on the
> phone yesterday, and exchanged a couple of e-mails.
> So far, no real doubts have been raised.
> But the proof of the pudding, as they say......
The DF of a tube amp can be good as it ever NEEDS to be.
A typical Williamson of 1947 with KT66 had Rout = approx 0.2 ohms
because of the 20dB
of NFB, if the load was meant to be 4 ohms.
Using 300B in the same circuit and with the same amount of NFB
would give 0.1 ohms, because the Ra of a 300B is 1/2 the Ra of KT66 in
triode.
SS amps with 40dB of emitter follower NFB and 50dB of global NFB have
Rout in milliohms.
This does NOT sound any better than if you had 0.2 ohms, even if the
load was 2 ohms.
A DF of 10 is plenty.
>
> > The ME has the ability to confortably deliver 70A into
> > a suitably low Z load, implying its source impedance is extremely low,
> > hence the high damping factor even into a 2.5 ohm speaker
The current ability of an amp DOES NOT indicate low Rout.
A tube amp might produce a maxima of 0.5 peak amps at an output tube
anode.
In the class AB working of the amp, with an effective OPT turn ratio
of say 16:1, you get 8 peak amps possible at the output.
This does not indicate a low Rout.
The Rout is indicated by how much the output voltage changes when the
load
value is changed, so if Rout was say 5 ohms which is typical
for a UL amp without global NFB, then a 5 ohm load would halve the
output level
when it is connected.
But if there is global NFB arranged, then the amp output might only drop
from say 10V to 9V when a 5 ohm load is connected.
The V change = 1V, and the current change is from 0.0 amps with no load
to
4.5 amps with 5 ohm load, and Rout is calculated as
Rout = V change / I change = 1.0 / 4.5 = approx 0.23 ohms.
>
> That's impressive. It probably has 75% of the open loop
> gain used in NFB:-) A good PP tube amp with about
> 12-15dB NFB can manage Zo of 0.4 Ohms. Triode
> wired PPP is considerably lower. We shall see:-)
I am getting 0.32 ohms with a single 13EI, with 12dB CFB from the OPT
to 13E1 cathode, and only 8 dB of global NFB.
All music sounds well and the bass is well defined, giving the firm
foundation for the rest of the music.
Patrick Turner.
> >
> > Without any doubt, a new high powered toob amplifier would represent a
> > substantial investment and as such I would be far more comfortable in
> > coming to a decision to purchase after doing an extensive audition of
the
> > amplifier(s).
>
> I would rather you have a pair of monoblocs on
> extended load or a sale or return basis. Your
> evaluation is more im****tant to me than making
> a sale.
>
> Iain
|
