Menu
>Second, I too could have just sent people to one or more
>of my favorite websites, but I chose instead to write
>this all out in my own words. While you may dispute
>some of it, I challenge you (in the politest sense of
>the word) to put forth your reasons for calling me into
>question, preferably in your own words please.
I guess I did come across as rather short. I've been really swamped
lately with little or no free time... but anyway, here's a few
comments on your article:
--- In , "unlunf" <unlunf@y...> wrote:
>
> Hmmm, the whole idea of an article that is tech oriented,
> and completely off topic (surf music).... is this a good
> thing? So let's cut to the chase, and eliminate all the
> blathering. We'll also eliminate the math portion of our
> curriculum. <g> Since we started talking about impedance
> in the power amp section, we'll stay there for now. In a
> few paragraphs we'll deal with low signal level impedance.
>
> First, every circuit in electronics has two impedance
> factors, the input and the output. Some components
> in those circuits are impedance-neutral, such as resistors,
> diodes, batteries, etc. Others are very impedance critical,
> these would be capacitors, inductors such as transformers, etc.
> Impedance is frequency sensitive.
I think in the above paragraph you are saying that impedance is
related to reactance. But I'm not sure what you mean when you
say 'impedance critical'. To be accurate, impedance is comprised of
a real and imaginary part. Components such as resistors are
primarily 'real' only (resistance). While capacitors and inductors
are primarily 'imaginary' only (reactance). Impedance is used to
decribe something that contains both components. Amplifiers are
specified for output impedance because the load can be resistive or
reactive or anything in between. But this probably qualifies as 'nit-
picking'....
> Let's just take it for granted that a tube has fairly high
> input and output impedances. OTOH, a speaker's impedance
> is low. We use a transformer to convert from one to the
> other. So, what happens if we don't match the speaker to
> the transformer?
>
> Well, the transformer doesn't do anything in and of itself,
> it's not an active component. What comes in, goes right
> back out, after the conversion. If we hook up a 4 ohm
> speaker to a 4 ohm output transformer, then the tubes will
> be happy. If we hook up an 8 ohm speaker to that same
> transformer, what happens to the tubes? They 'see' twice
> the number of ohms of impedance. Is this a bad thing?
> Depends on the tube, but for the most part, no. Lots of
> tolerance is built into both tubes and transformers, and
> even most speakers can handle some mis-matching of
> impedance values.
> Some guidelines that have been accepted by the amp design
> and repair industries over the years:
>
> You can usually feed a lower output impedance into a higher
> speaker impedance without causing grief to the tubes, but
> there are limits. You shouldn't go over twice the rated
> number of ohms. Remember, as the ohms went up, the tubes
> tried to overcome that, and tried to draw more power from
> the power supply. Go too high, and the power supply will
> crap out on you. You'll end up with less power output at
> the jack, and probably a muddy sound.
As a rule of thumb, nothing greater than 2 to 1 mismatch is a good
one. However, with regard to what happens when you do mismatch the
load, when you put a higher impedance load on a tube amp, the tubes
see less of a load and the voltage swing on the plates goes up. The
tube tries to pull the same amount of current through the
transformer but since it is now a higher impedance, ohms law tells
us the voltage is going to increase. The problem is they will arc
internally if the output impedance, and thus voltage swing, gets too
high. The power supply is more than capable of providing everything
the tubes need to accomplish this.
> Pretty much the same holds true in reverse, but that way
> is a bit more dangerous. All the good amp designs, not
> just Fender, can handle a halving of the impedance (2 ohms
> instead of 4), and the tubes should survive. As has been
> pointed out before, always look at your tubes for awhile
> after doing this. If they start glowing red, STOP! The
> tube is not sufficiently loaded down (too few ohms), and
> is drawing too much power. This is something that can't
> be fixed by twiddling the bias, you'd have to reduce the
> power supply voltage (the B+). Moreover, the tubes will
> very likely clip in the second harmonic range, and that
> will pass right through the transformer and take the
> speaker out like a snuffed candle. Too much of this,
> and the transformer will also buy the farm.
This is simply not true. Higher output impedance is more dangerous
for a tube amp than a low impedance. (the opposite is true for solid
state) Lower the load on a tube amp and the gain goes down. The tube
will have to dissipate more of the power and consequently glow red
if the load is small enough and is used that way long enough. But
this is a much slower and less spectacular way of dieing than having
the tubes arc over.
I have no idea what you mean when you say it will cause a tube to
clip in the second harmonic range. I do know that when you drive the
tubes into clipping they generate lots of higher order harmonics,
2nd, 3rd, 4th, 5th, etc. Why would a second harmonic be more
damaging to a speaker/transformer than the fundamental? It just
doesn't make sense. The second harmonic of 400 Hz is 800Hz. Both
capable of being generated by a guitar, why would 800Hz be more
damaging?
I do agree with your advice though:
> Final rule of thumb: You can go one step up or down, but
> no further. If you're going down (speaker is lower than
> what the amp is rated for), then always observe the tubes
> for at least 5 minutes after warming up. And re-check them
> from time to time, they may not get stressed out until
> later in their lives.
>
> in re: low level signals......
>
> The same rule of thumb holds true here, except that we don't
> have to worry about blowing tubes or speakers. In fact,
> what we now worry about is fidelity. You can always go
> from a low impedance source to a high impedance load, and
> it's very likely that you will suffer little or no signal
> degradation. The reverse is -not- true, you will definitely
> suffer degradation if you do more than halve the load's
> input impedance compared to the source's output impedance.
>
> And that's it in a nutshell. Without any math, even!
> The rest of the lesson is best taught by example, and that's
> where you come in. Ask your questions, and we'll see what
> would happen (good or bad) if we do this or that or the
> other thing.
>
>
> unlunf
>
