Re: Safe headphone amplifier
- From: terryS <tsanford@xxxxxxxxxxxxxxx>
- Date: Mon, 08 Oct 2007 11:06:57 -0700
On Oct 7, 12:21 am, "Ian Iveson" <IanIveson.h...@xxxxxxxxxxxxxxxx>
wrote:
A high fidelity valve amplifier should IMO
* have an output impedance of no more than one tenth of the
impedance of its load. This is an opinion I wish to pursue.
* properly load the output valve(s), so that a triode, for
example, should be loaded with at least 2.5 times it's anode
impedance. O.K., another rule of thumb, but I want to stick
to it.
Looking at available output transformers intended for
headphone amps, it is possible to make an amp for 70 ohm
phones that meets those criteria without using overall
negative feedback, but only just.
For example, there is a Sowter transformer offering 10k : 70
ohm. When driven as recommended, the resulting amplifier
provides a load for the driving cathode follower of 10k, and
has an output impedance of just under 7 ohms.
However, I want to drive Grado 'phones of 30 ohms. A little
calculation would reveal that the Sowter example above has
winding resistance, referred to the secondary, of about 4
ohms. It is possible that the same transformer with
secondary connected or wound for 30 ohms would just about
scrape home because the reflected primary resistance would
be reduced, but I doubt it.
I don't want to use a cathode follower either. If I use my
intended triode-connected 6CH6 with the transformer coupled
to the anode, as I prefer, I end up a few ohms too high.
Consequently, overall feedback seems a necessity. I don't
mind that per se, but it complicates the issue of soft
limiting.
I don't want to risk hurting my ears, so I want the output
stage to be limited. I want hi-fi up to a reasonable output
level, beyond which it should as gracefully as possible
compress any further excursion to within a safe limit. I
feel that this limiting should be done by the output stage
because that would protect me from signals generated by the
amp itself under fault conditions.
Feedback tends to harden limiting. If I arrange the output
stage to limit how I want it to, then feedback will make it
too sharp. If OTOH I make it so soft that it is right with
feedback, then the limit will introduce unwanted open-loop
distortion into the signal at a reasonable listening level.
Anyone got any ideas on how best to limit the output of my
headphone amp so I get max hi-fi without risking damaged
ears?
cheers, Ian
As a complete newcomer to this audio business!
And not quite sure what is meant by 'limit'? Assuming it is not some
'active' circuit that cuts in or cuts off peaks of audio;
it becomes s a matter of proper matching?
You mention that you prefer plate coupling not cathode follower. I'd
like to ask a question about CF in separate thread.
But it appears that you wish to reflect the correct impedance back
into the plate of the OP tube's with 30 ohm phones intead of 7 0hm
speakers?
Is it not possible to shunt the phones with some resistance. Pure non
reactive resistance in parallel withe phones.
Shunting with 30 ohms will immediately halve the impedance reflected
back through the OP transformer/s to the plate circuit/s? Assuming
stereo!
Shunting with 10 ohms will provide an approx. 7.5 ohm load!
But as said as a newbie to audio amps. etc. stand to be corrected.
Presumably we are talking millwatts of power here; even 5 volts across
30 ohms is less than a watt (peak)?
A 10 ohm resistor will dissipate approx. 3 times the audio power that
is turned into sound by the headphones? Couple of watts,
intermittent?
BTW don't pentodes and tetrodes have higher plate load? If soi could
the same transformer be used with the OP tube (the 6CH6 mentioned for
example) configured that way instead of triode connected?
Flame/blast away; willing to learn!
.
- References:
- Safe headphone amplifier
- From: Ian Iveson
- Safe headphone amplifier
- Prev by Date: Re: May I recommend . . . . ?
- Next by Date: Cyclothmia & England
- Previous by thread: Re: Safe headphone amplifier
- Next by thread: FA: 1930 era mono type 50PP amp
- Index(es):
Relevant Pages
|
