John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
JN

I'll assume your reading skills are off a bit today and not your math ones, as Sqrt(3 x (.57 * .57)) < 1 (I think you missed the meaning of the word "Signal".)

As Scott has approached the maximum current from an amplifier does not occur at maximum voltage into a loudspeaker. George's post #78872 shows this clearly.

As noted by many when an amplifier runs out of voltage "Softly" it really doesn't sound so bad. However when an amplifier runs out of current some sound quite horrible.

That is why I suggest using three non-phase locked oscillators to make the measurement on a 3 way loudspeaker. Using a computer generated test signal will not give the same results unless phase is randomized.
 
As Scott has approached the maximum current from an amplifier does not occur at maximum voltage into a loudspeaker. George's post #78872 shows this clearly.

In that case the load is complex and not a resistance and the frequencies (steady state response) will not be at the peaks or dips in the FR. Assuming minimum phase behavior and that superposition applies the phase wander makes no difference.
 
Last edited:
In that case the load is complex and not a resistance and the frequencies (steady state response) will not be at the peaks or dips in the FR.

Yes Scott real loudspeaker loads are complex. I had one case where the audio power amplifier was clipping at 110 AMPS into a tweeter! As the system used active crossovers the internal one was removed to allow the system to get maximum output.

It is some manufacturer's practice to make loudspeakers that even when actively crossed still have the passive in series for protection. (This has saved me on one occasion and been a pain in far more.)

The phase wander allows for a simple reading of a complex number. As audio power amplifier manufacturers seem to only understand resistors! (Not quite true but there are many amplifiers that will run out of current trying to drive a real loudspeaker as in the above example.) The assumption of minimum phase is probably not a good one, nor even a consistent phase in outdoor applications.
 
Last edited:
If you examine Nelson's load model, you see the reactances. Anytime there are reactances, there will be operation in Q2 and Q4. (unless sines are used at the exact frequencies where phase is zero.)

John

You're right... I was over-simplifying it in my head and assuming they'd match phase-wise.

As to building a FEA model--very doable (although those modules/packages aren't my forte), I simply need to work on more pressing matters. :D Right now DIYAudio is a nice mental break.
 
Yes Scott real loudspeaker loads are complex. I had one case where the audio power amplifier was clipping at 110 AMPS into a tweeter! ..............
Some Members here keep telling me that tweeters do not ever see transients of similar levels as mid and bass speakers.
I always assume that the peak SPL from any loudspeaker is roughly that same irrespective of what frequency range it handles.
 
Some Members here keep telling me that tweeters do not ever see transients of similar levels as mid and bass speakers.
I always assume that the peak SPL from any loudspeaker is roughly that same irrespective of what frequency range it handles.

This was in an American Football stadium. The atmospheric losses are so great at high frequencies that it takes more power for the tweeters than the woofers! The midranges just coast. This system was about 125 KW.
 
JN

I'll assume your reading skills are off a bit today and not your math ones, as Sqrt(3 x (.57 * .57)) < 1 (I think you missed the meaning of the word "Signal".)

No. I was talking about the addition of 3 independent frequency signals.

I did not see what you are talking about.

Edit...ed, TELL me you're not using the 3 phase power relation to describe 3 independent signals on a wire pair.. 120/208 is 57%, that's all I got..
As Scott has approached the maximum current from an amplifier does not occur at maximum voltage into a loudspeaker. George's post #78872 shows this clearly.
Again, what are you talking about? I am talking about the summation of three voltage signals.

snip..., I simply need to work on more pressing matters. :D

Right now DIYAudio is a nice mental break....snip.

Sigh.. How right you are.

John
 
Last edited:
Totally agreed..

The issue is this: The voltages will be a summation as per your statement. However, the currents will also be a summation, as the amp is driving three parallel loads...
Indeed, but since each individual current is 1/3rd the maximum output of the amplifier, the sum will never exceed the maximum current (it will be a worst case, because in real life the vector sum of the three currents will be <1)


There is a problem when one considers the entities as simple real numbers. Operation of a reactive load requires the energy stored in the load be returned.

If one simply uses a real V/I ratio, one calculates negative resistance in quadrants 2 and 4. What is really going on is energy shedding by the load.

Despite any calculations, it is trivial to show that in quadrants 2 and 4, the EMF of the load is HIGHER than that of the amp, so it is trying to return power to the source.
there is certainly reactive energy exchanged between the amplifier and the driver, but the bottom line is:
For a driver in isolation, the equivalent circuit is a series one, and whatever (linear) reactive/energy storage elements you put behind Rdc (except a negative resistance, of course), the magnitude of the resulting impedance will never be smaller than Rdc.
If the driver is preceded by a filter, anything could happen.
For some transient conditions, the instantaneous current might (and will in fact) be higher, but under steady state conditions, the phasors view applies
 
The one and only
Joined 2001
Paid Member
If you examine Nelson's load model, you see the reactances.

And here is the phase plot of that model. If you want wilder, just reduce
the resistance in series with the LC elements.

:cool:
 

Attachments

  • PHASE.gif
    PHASE.gif
    7 KB · Views: 204
Indeed, but since each individual current is 1/3rd the maximum output of the amplifier, the sum will never exceed the maximum current (it will be a worst case, because in real life the vector sum of the three currents will be <1)

Hmm, how best to explain.

If I drive a reactance with a voltage sine, the VI plot will describe an ellipse.
If I add a second hf signal driving it's reactance through the crosover, it too will describe an ellipse. HOWEVER, it's V/I plot will be an ellipse centered on the point of the LF ellipse as it traverses it's V/I plot. There will be times when they have the same direction on the V/I plot, times they are 90 degrees apart, and times where they subtract. Like a spirograph.

There will be times when the LF point is on the zero voltage/max current axis, and the HF ellipse will still extend the current out. Yet, at that point in time, the headroom available to the HF is full rail voltage.

When the two or three signals are correlated in some fashion, say a hard percussive, it's easy to see how various reactances can cause the current envelope to be exceeded despite not hitting the voltage one. Remember, capacitive and inductive reactances rotate opposite in V/I. Even single drivers change from one to the other.

Hey, if it were simple, anybody could do it...

For some transient conditions, the instantaneous current might (and will in fact) be higher, but under steady state conditions, the phasors view applies

In general, I do not listen to music with sines. All my music has time varying signals.

John
 
Last edited:
Root mean squared what? And out of the blue applied to voltage addition?

Adding the voltages is so easy. Why are you converting at all? V clipping and I clamping don't care about rms.

If you're going to apply something entirely out of the context of, what I was speaking of, please preview with some warning..

John

JN

You were quoting me on signals combining. Clipping only occurs when you run out of headroom. This does occur in some systems when playing loudly. In my work we use limiting of the input signal to avoid that issue.

However many audio amplifiers have current protection. This will engage at some value set by an algorithm. The basic bad circuit looks at the output current and drops the signal level at some preset current. This usually sounds quite horrible.

More advanced protection circuits use more of a thermal model of the output devices. So they look at RMS with a peak limiter. So for most of the audio power amplifiers I use actual instantaneous impedance (or in reality RMS impedance) is the important issue.

ES
 
JN

You were quoting me on signals combining.

Signals combining in terms of voltage. You seem to be talking about dynamic compression. Not typical of a home stereo environment.

However many audio amplifiers have current protection. This will engage at some value set by an algorithm. The basic bad circuit looks at the output current and drops the signal level at some preset current. This usually sounds quite horrible.
A typical protection scheme is a combination of limiters. Foldback being one of the "closer to SOA boundaries" options. All work on an instantaneous level, nothing to do with RMS anything.
More advanced protection circuits use more of a thermal model of the output devices. So they look at RMS with a peak limiter. So for most of the audio power amplifiers I use actual instantaneous impedance (or in reality RMS impedance) is the important issue.

ES
Again, what is needed to protect a customer from themselves is not the same as the home environment.

John
 
JN

The Tigers are a bit dated. Just about everything I use today has DSP. But even before that many amplifiers had thermal modeling protection circuits. Also know as an RC network in the current detector loop, or so it seemed.

Many amplifiers just use enough output devices that protection is not needed.

BTY with digital sources such as CD's that have a maximum output level avoiding clipping in a consumer system is fairly easy, don't turn the volume knob to 11.
 
Last edited:
JN

The Tigers are a bit dated.

Now now, play nice!!! ;)

For the day they were so cool. Bad heatsinking, weak supply, poor layout, no fans, one bad resistor choice..overheating zeners. THAT was seat of the pants...soooo coool...

btw, I still have two built into one 3ru chassis, darn thing still works...whoda thunk..

Just about everything I use today has DSP. But even before that many amplifiers had thermal modeling protection circuits. Also know as an RC network in the current detector loop, or so it seemed.

Current limit and foldback is still used. (for others, see pic of the tiger protection.

Many amplifiers just use enough output devices that protection is not needed.
That wouldn't work with me...

don't turn the volume knob to 11.

What's that?? Houston, your breakin up, your breakin up...

pfft. no 11?

yah, riiiiight.. :)


ps blue are the clamps, yellow is the current limit, red is the foldback portion (in conjunction with the 100 ohm yellows...)

Green is a wiring error on the part of SWTPC, remove that or you can lose either Q16, 18, 21, or 23 to current hogging. They've directly paralleled two transistors E and B. Two transistors, pick one.. :eek:

John
 

Attachments

  • tiger 250 protection.jpg
    tiger 250 protection.jpg
    303.1 KB · Views: 228
JN

You can't even color code correctly. Red should be used for the extra transistor sales circuit. :)
What was I thinkin??

The yellow would be more modern with an RC network added.
Why? That can only protect the tweeters... Tweeters must be treated exactly like every other driver, survive, or go away.. I'm sick and tired of people constantly coddling their tweeters.. For me, it's deep end of the pool, learn to swim. Sides...any tweeter that can't handle the clamp transients deserves to die.. Same with the the output die, they deserve to tweet..

Course, power mosfets are an entirely different beast..


But the up to date guys figure once the DSP is in the gizmo why wast any extra parts. DSP is "King" it can do everything.

Sigh.. DSP.. just a crutch I tell ya, just a crutch.. like my smart tv. If it wuz so smart, it'd find me good programs to watch.

Reminds me of the MIG the defector brought over. Tubes for electronics..everybody laughed...

for a while..

John
 
Last edited:
Member
Joined 2014
Paid Member
Sigh.. DSP.. just a crutch I tell ya, just a crutch..

As we say here 'you cannot make a silk purse out of a sow's ear'. Makes a good dog treat tho, or a frankfurter :)
Reminds me of the MIG the defector brought over. Tubes for electronics..everybody laughed...

for a while..

John

F-35 isn't doing too well on SW testing at the moment I hear!

Back in the 90s I had an office where, from the top floor you could watch the aerobatics from the farnborough air show. Vectored thrust was the new thing for war birds and they were showing it off as allowing faster turns. The Russians, with no nanny code on their MIGs worked out that a stall turn could counteract the agility advantage european and American planes had.

Two years later at the next air show oddly all the fly by wire planes had been programmed to allow the pilot to stall them :D. Looked like dolphins at seaworld.

Have to say tho the chicago air and water show is the best spectacle I've seen.
 
Well my view is that the easiest way to explain to my grandmother is to be clear about what I mean, NOT wavering to explain things in terms that are used differently from the use by the rest of the engineering community. Audio has already too much bad rap for sloppy use of terms. What is the definition of PSRR again? :cool:

Jan

Point taken, but I think even you would spec an amplifier like a black box, rather than overcomplicate the datasheet with unimportant information.
 
Status
Not open for further replies.