amp and XO gurus: possible oscillation issue?

[Paul Carmody]

The RS28A is a great tweeter, but the drooping top end is a problem that should be dealt with in most situations. The simplest choice would be a contour filter (C4 + R2). And yes, I have used similar filters before. But I'm trying to be more cautious nowadays, and want to ensure I have a safe load for an amp. How do I know how I've gone too far?

Here's what I have so far. Please take a look. Any feedback is appreciated.

[Jeff B.]

Re: amp and XO gurus: possible oscillation issue?

The RS28A is a great tweeter, but the drooping top end is a problem that should be dealt with in most situations. The simplest choice would be a contour filter (C4 + R2). And yes, I have used similar filters before. But I'm trying to be more cautious nowadays, and want to ensure I have a safe load for an amp. How do I know how I've gone too far?

Here's what I have so far. Please take a look. Any feedback is appreciated.

That's similar to the contour I use for the Continuum crossover. It sould offer no low impedance issues for an amplier. In your case the minimum high frequency impedance should still be over 5 ohms, and higher than that realistically, due to the voice coil inductance. Any amp should be OK with that.

[weinstro]

Re: amp and XO gurus: possible oscillation issue?

The RS28A is a great tweeter, but the drooping top end is a problem that should be dealt with in most situations. The simplest choice would be a contour filter (C4 + R2). And yes, I have used similar filters before. But I'm trying to be more cautious nowadays, and want to ensure I have a safe load for an amp. How do I know how I've gone too far?

Here's what I have so far. Please take a look. Any feedback is appreciated.

Basically, electrical phase = 0 is a pure resistive load -- very easy to drive, if the load is not too low.

Phase angles approaching +/- 90 indicate pure inductive or pure capacitive loads. The latter in particular can cause stability problems for amplifiers, especially at higher frequencies. A more complete answer delves into phase margins of the feedback loop and output device safe operating areas.

The system impedance chart attached doesn't show a legend, so I'll assume that the light blue line is electrical system phase. If that's true, your electrical phase angles look OK to me. IIRC, Wolf provides some guidelines for electrical phase limits in one of his blogs. I believe he suggests keeping phase angles <|60| degrees.

Regards,

Rob

[MrkCrwly]

Re: amp and XO gurus: possible oscillation issue?

The RS28A is a great tweeter,

The chart lists the tweeter as RS28F. Did you mean to say RS28F in the above statement?

[Paul Carmody]

Re: amp and XO gurus: possible oscillation issue?

The chart lists the tweeter as RS28F. Did you mean to say RS28F in the above statement?

Oh. I guess the XO diagram is labeled wrong. Don't worry, nothing's in stone here. This XO will probably change before it's finalized. I'll pull these diagrams down then.

[Pete Schumacher]

Re: amp and XO gurus: possible oscillation issue?

Paul, worst case, at high freqs, the RS28 is basically in series with 1.5 Ohm, so you're looking at about 5 Ohms, PLUS the rising inductance of the voice coil.

You're on easy street.

oops, sorry Jeff, shoulda read yours first.

[Jeff B.]

Re: amp and XO gurus: possible oscillation issue?

That's similar to the contour I use for the Continuum crossover. It sould offer no low impedance issues for an amplier. In your case the minimum high frequency impedance should still be over 5 ohms, and higher than that realistically, due to the voice coil inductance. Any amp should be OK with that.

Hey Paul, I took my data files for the RS28a and extrapolated the high pass section's impedance to beyond 50khz. As you go above this frequency the impedance is approaching 6 ohms and will stay above that impedance at higher frequencies and the phase angle is near 0 degrees, with a slight inductive angle as you go higher. I see no cause for concern. Just to be clear, the bypass cap will only cause the circuit's impedance to approach that of the driver itself at very high frequencies (+ the resistance of that first 1.5 ohm resistor), it will not behave like a high frequency short in this postition.

Jeff

[Paul Carmody]

Re: amp and XO gurus: possible oscillation issue?

Hey Paul, I took my data files for the RS28a and extrapolated the high pass section's impedance to beyond 50khz. As you go above this frequency the impedance is approaching 6 ohms and will stay above that impedance at higher frequencies and the phase angle is near 0 degrees, with a slight inductive angle as you go higher. I see no cause for concern. Just to be clear, the bypass cap will only cause the circuit's impedance to approach that of the driver itself at very high frequencies (+ the resistance of that first 1.5 ohm resistor), it will not behave like a high frequency short in this postition.

Jeff

Yeah, it was the "extrapolation" business that had me concerned. My measurements stop at 20 KHz, and so all I see in PCD is the impedance dipping downward at the top end of the spectrum... so my mind was left wondering, "Where's the impedance heading up there?" Also, I had forgotten that the voice coil would have an inductance rise.

[pres589]

Re: amp and XO gurus: possible oscillation issue?

I'm just curious what you're working on here as far as the entire speaker; with that droop in freq. response over 100hz, is this the top of a three-way setup? Am I missing something really obvious like usual? Is it a secret?

[Paul Carmody]

Re: amp and XO gurus: possible oscillation issue?

I'm just curious what you're working on here as far as the entire speaker; with that droop in freq. response over 100hz, is this the top of a three-way setup? Am I missing something really obvious like usual? Is it a secret?

The roll-off below 250 Hz is due to the fact that I used windowed MLS measurements. In reality, these get down to an F3 in the mid 30's. No secrets. Just a 2-way tower in the "Amiga" enclosure.

[fastbike1]

Re: amp and XO gurus: possible oscillation issue?

Hah! I knew it. You get so much use from the Amiga cab that I'm building mine w/ a removable front baffle (usually make the back removable). :D

The roll-off below 250 Hz is due to the fact that I used windowed MLS measurements. In reality, these get down to an F3 in the mid 30's. No secrets. Just a 2-way tower in the "Amiga" enclosure.

[MrkCrwly]

Re: amp and XO gurus: possible oscillation issue?

Rumor has it that Paul is also working on a version with the new RS28F.

[MrkCrwly]

Re: amp and XO gurus: possible oscillation issue?

How many watts should it take to push one of the Amiga/TRS variations? Their efficiency is in the low 80's and I think they reach their max at about 95 db. I am considering a HK3490 stereo 120W, maybe a Pan 1020 7.1 at 110W or DIY ClassDAudio 254/258.

[fastbike1]

Re: amp and XO gurus: possible oscillation issue?

The 3490 should do nicely. I gave one of my sons a 3480 and it has a very strong amp section. W/ low 80s sensitivity you should be able to get 95dB w/ ~32 watts (very roughly).

The ClassDaudio will certainly work as well. Bear in mind the Class DAudio has no controls as part of the kit, so you will need a volume control somewhere in the chain. I would strongly advise against using the digital volume control on the computer.

How many watts should it take to push one of the Amiga/TRS variations? Their efficiency is in the low 80's and I think they reach their max at about 95 db. I am considering a HK3490 stereo 120W, maybe a Pan 1020 7.1 at 110W or DIY ClassDAudio 254/258.