Speaker Load Simulation

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
typical complex load at audio frequency isn't a linear stability problem - maybe clipping/current limitng and protection circuit activation/recovery could be

if wanted there are spice models of driver and crossover

cable C load ~ O(1 MHz) loop gain untiy intercept can be a amp stability issue - isolating series L, Zobel load are used for decoupling for linear stability with uncertain load Z

usually lumped nF C directly on the amp output is adequate to explore cable loading stability effects
 
Last edited:
Where have you found decent spice models for drivers? I came up with my own, but looking at the models Linkwitz discusses, mine are far short of a realistic electrical model. I just use a resistor with a series inductor. Not good enough.

To the OP. From what I have learned so far, two issues for your tests. Low resistance, i.e what does it do at 3 or 2 Ohms, and capacitance. That's what makes amps go crazy. How much? Don't know. Might punch your amp into SPICE and see what it takes to turn it into an oscillator. I can tell you how to do that with a DH-120!
 
Hi,

You don't need to simulate loudspeaker loads. You just apply rules of thumb.
4R testing should deal with any difficult 8 ohm speaker and 2R testing should
deal with any difficult 4 ohm speaker in reality. Modelling a specific real load
doesn't really help at all as they are all different. Some go for 6R for 8 ohm
and 3R for 4 ohm, as acceptable, probably is, better than AV specifications.

Point is reactive loads are basically equivalent to lower load resistance.

rgds, sreten.
 
Last edited:
In terms of stability (i.e. inductive/capacitive load) I don't think it is necessary to model the speaker accurately as Sreten has noted.

But may be there is a little use (I've never tried such as I think it is also unnecessary) when you design the amp for best performance against varying impedance (with changing frequency).

Speakers typical impedance are well known. And especially when you know your speaker impedance chart/behaviour.

You can model the load with LCR traps and such, which is not a difficult thing to do. You may need it for example to simulate THD20? Nothing to loose imho.
 
Problem is, all speakers are not typical or that well behaved, and some "exotic" wire is very difficult. It may depend on the amps output filters, or lack of them, it's compensation margin etc. I remember some discussion of a load box HK built that was highly regarded by some of the gentlemen here who design amps for a living. No idea what was in it.
 
Thanks for the link, but i couldn't find the most important information: how much does the modelling help in simulation? When they are most needed? And so on (I'm not gonna do the experiment for sure)

Truly speaking I didn't bother for " how much does the modelling help in simulation? When they are most needed? " Linuxguru wanted a spice simulation of speaker and I too am working on spice simulation of speakers. But my purpose is different. and is still on anvil. so I provided the link as a pointer.

Me too is working on the simulation and would like to have a complete spice models for speakers. that would simulate like a real speaker load.
 
Stereophile magazine use this to simulate a speaker:

Real-Life Measurements | Stereophile.com

...the general point that an amplifier should be measured into a load similar to that of a real loudspeaker is a valid one. Loudspeakers can be much more demanding than resistive loads, as evidenced by Eric Benjamin's 1993 AES paper, "Audio Power Amplifiers for Loudspeaker Loads." We had been thinking about how to implement this for some time, when, as a result of a chance conversation, Ken Kantor of NHT and International Jensen sent us a speaker simulator he had been using.

Its circuit diagram is shown in fig.1. Combinations of resistors, inductors, and capacitors produce a load with an impedance magnitude and phase plot (fig.2) intended to represent a typical two-way, sealed-box, 8 ohm loudspeaker. (The small-value resistors shown are the measured series resistances of the coils.) The impedance peak in the bass is the equivalent of the woofer's enclosure resonance; the peak in the low treble is identical to that produced by a crossover filter. The phase angles are also typical; note that the worst-case phase never coincides with the lowest magnitude.
 
Hi,

I can't see the point. You can model a typical basic drivers impedance,
You can model typical midranges and tweeters also. You would then
need to model a typical 2 or 3 way x/o to use with your models.

For what purpose ? A typical speaker model won't tell you
anything useful from testing / simulation that I am aware of.

Some say you should design amplifiers for up to 45 degrees reactance,
some say up to 60 degrees. Basically those loads are more thermally
onerous on the amplifier than pure resistance because the "power"
not dissipated in the load needs to be dissipated in the amplifier.

4 ohms at 60 degrees is a punishing load for an amplifier. The SOA
calculations if your designing output protection are very different
to the 4R case, often not mentioned or covered in textbooks.

But as I said, assuming 4R as the load for an 8ohm amplifier, and
2R as the load for a 4 ohm amplifier usually covers the worst case.

rgds, sreten.
 
Last edited:
A few years back I was testing power amps with four output transistors. I studied the SOA, Safe Operating Area of the transistors with real speakers as the amp load. The transistors were rated 150 watts, 15 amps, 100 volts. The power was derated to 75 watts to allow the heat sinks to run hot. Using an XY scope I monitored the current and voltage of each output device. The protection ckt limited the current to 3 amps when the amp was driven with a short circuit load (zero ohms).
The SOA was +/_ 50 vdc and +/_3 amps. This "Box" was my operating window.
At idle the CRT spot sits at zero amps 50 volts. Dead center of the Box. With program signals and real speaker loads the Box is filled with traces. With sine wave test tones and 8 ohm loads, the traces are just a line. I hope this helps you to understand what is happening.
Don
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.