I also discussed the Tc(jc) with Ian Hegglun and he provided me with a sim model that calculates and integrates this. Next I will try to integrate it in my circuit and see where it leads, so I an estimate when the SOA will be exceeded depending on pulse width and prf.
This could even be fun! 😎
Edit: my app is a 1500V RMS 100mA RMS output stage to directly drive ESLs...
Jan
This could even be fun! 😎
Edit: my app is a 1500V RMS 100mA RMS output stage to directly drive ESLs...
Jan
Nice, gets rid of the xfmr...
I found that by examining the drive signal to the outputs, I could see secondary breakdown as it started. When the circuit started dropping it's drive to the pre's, it was because the output device(s) gain was increasing.
Course, that was bipolars, not fets..
You thinking of any foldback protection?
edit: also, is the output going to provide dc offset as well?
John
I still have to work in the protection, I do have it in the psu. That was the easy part - just a 500v regulator just more volts ;-)
The output stage is basically a BUF634 with floating bootstrapped supplies.
Jan
The output stage is basically a BUF634 with floating bootstrapped supplies.
Jan
How about transferring the energy to a cap thru a diode (so it doesn't go back ) and let a parallel resistor burn it off.
Back emf in an ac motor ( like a 120v synchronous) is what keeps the current, and thus power adjusted for the load. Lots of load, little back emf and most of the 120v is across the windings. Little load, lots of back emf.
V across windings = Vin-Vback emf
Not sure how this relates to drivers but it's probably similar. So more Back emf at resonance where the load is smaller? Or not.
V across windings = Vin-Vback emf
Not sure how this relates to drivers but it's probably similar. So more Back emf at resonance where the load is smaller? Or not.
And if you push on the motor the back emf is greater than the Vin and you have a generator. Which hapens in a driver when the spider pulls the voice coil back after an impulse. This is the energy the amp has to dissipate and what damping factor is about.
Last edited:
No. Almost all of the energy is dissipated in the voice coil resistance. This is what damping factor is all about; a low damping factor means that the amp dissipates more of the energy.cbdb said:
That is not useable with a speaker.
It is similar to drivers.
The spider is not the primary push restoring the driver position at useable frequencies, the air load is.
John
But in the ideal case that the amp Zout = zero, the dissipation happens in the voice coil, not the amp.
Even if the amp Zout = 1 ohms, and the voice coil = 6 ohms, the majority of the dissipation is in he voice coil. The amp sort of facilitates the current loop but doesn't dissipate hat much.
Jan
Even if the amp Zout = 1 ohms, and the voice coil = 6 ohms, the majority of the dissipation is in he voice coil. The amp sort of facilitates the current loop but doesn't dissipate hat much.
Jan
Damping factor is in the best case a misunderstanding. Amps with super-low output impedance becomes even moore sensitive to cable/filter resistanse.
The cure? Thicker cables, from ac to output, trafos included. Makes the amp never run out of breathe
The cure? Thicker cables, from ac to output, trafos included. Makes the amp never run out of breathe
When an amplifier is driving a reactive load like a woofer, it causes the pass transistors to dissipate more energy. This is because an amplifier's output node, even with a zero impedance, does not achieve zero impedance by shunting the current to ground. It does so by using the pass transistors to force the node to zero.
If the amp has to pull 1 ampere to keep the output node at zero volts, it does so by conducting one ampere from the power supply rail.
John
If the amp has to pull 1 ampere to keep the output node at zero volts, it does so by conducting one ampere from the power supply rail.
John
What you've stated, in terms of amplifier design understanding, doesn't mean anything to me.
Can you elaborate with a technical explanation?
John
Damping factor is related to source impedance by a simple and direct mathematical relationship. If you understand one, understanding the other should be at worst, doable and at best easy.
Not a good generality. For one thing, conflating resistance with impedance is at best an approximation, and at worst an error. The load on an amplifier is composed of several elements, the largest of which is generally the loudspeaker. Changing the impedance of the loudspeaker will therefore make the biggest changes on the system.
There may be no disease to cure. It is all about quantification, and the optimal quantities are not difficult to obtain.
Must be some kind of romantic metaphor.
Yes but that energy doesn't come from the EMF so from the speaker perspective no energy is removed, so no damping of that speaker.
I mean, energy is removed by dissipation in the'virtual' Zout of he amp (like 0.1 ohm or whatever) but all the watts dissipated internally in the amp to 'make' that 0.1 ohms are irrelevant for the damping.
Jan
Last edited:
This is cool....
It is also so wrong, because it omits very important influences like voice coil resistance, and does not show the various Monkeys to scale in terms of their actual influence.
Not even wrong.PaleRider said:
True, but this energy comes from the PSU not the speaker. (I know you know that, but others might not.)jneutron said:
I think he must have heard a man in a pub say something like "use thicker cables for mains and speakers; this increases the current driving capabilities of the output stage - we know that base current into the output stage is set by the mains cable".arnyk said:
Jan,
From speaker perspective, it always exist the physical resistance of the wire of the voice coil. Which is roughly 75% of the nominal impedance, 6 ohms for a 8 ohms speaker.
Even if the speaker is electrically shorted, i.e. infinite damping factor, the actual damping cannot be infinite, this is why very high damping factors are useless. Moreover, wiring resistance between amp and speaker can be as high as, say, 0.6 ohm before we can hear audible deterioration of transient response (1 dB) Same things occur with amp output impedance.
The only way to reach further damping is to make negative the output impedance of the amp. This can be done by positive current feedback. I have experimented this 30 years ago. Recently, Nelson Pass has refloated this idea...
From speaker perspective, it always exist the physical resistance of the wire of the voice coil. Which is roughly 75% of the nominal impedance, 6 ohms for a 8 ohms speaker.
Even if the speaker is electrically shorted, i.e. infinite damping factor, the actual damping cannot be infinite, this is why very high damping factors are useless. Moreover, wiring resistance between amp and speaker can be as high as, say, 0.6 ohm before we can hear audible deterioration of transient response (1 dB) Same things occur with amp output impedance.
The only way to reach further damping is to make negative the output impedance of the amp. This can be done by positive current feedback. I have experimented this 30 years ago. Recently, Nelson Pass has refloated this idea...
Last edited:
Fully agree Pierre. If you go the route of neg Zout you probably want to go all the way to say -6 or -7 ohms.
But possibly stability issues can be non-trivial.
Jan
But possibly stability issues can be non-trivial.
Jan