Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
98.216.185.239
In Reply to: RE: Good Experience with FMODs posted by Davey on November 19, 2015 at 13:13:30
Yes, it shows up in the measurements on the ET site and was confirmed to me by Bruce T. It's woofer cone resonance.
I disagree about the balance. The second crossover is 500 hz, with attenuation coming in above that where the woofer has very little real output....except cone resonance.
Recent glowing review of LFTs preferred woofer out of phase. Bruce T says it's a matter of preference.
Follow Ups:
Can you point me to the measurements on the ET site showing this? (I can't find them.)
I see a graph on this page, but it looks like a dip at 1khz, not a peak.
http://www.eminent-tech.com/LFT8specs.html
A 2nd-order (FMOD) filter introduces a group-delay into the woofer path of about 600usec. (Equivalent to about 8 inches.) Any square-wave performance you might have had in that region is degraded when you do that.
Also, the reverse polarity preference noted by REG provides a much better phase match to the midrange driver in the crossover region, but it would degrade the advertised square wave performance too. I'm not sure why that's considered a matter of preference. It's an objective change to the system design.
Anyways, your solution only works for those that are bi-amping.
Fashioning a simple trap into the stock woofer filter could easily address a cone resonance issue. Why has Bruce not implemented that?
Always there are trade-offs with speakers. :)
Dave.
The peak is at ~1.4 Khz.
The real woofer output is well down at > 500Hz, between the 180Hz first order crossover filter and the natural rolloff of the cone. What is being attenuated is the out of band resonance, which obviously contributes nothing to square wave performance.
The sketch of the crossover in the manual shows woofer connected out of phase already.
I don't know why Bruce did not choose to add to the crossover in order to reduce or eliminate this coloration. He does not seem too worried about it and it is a small thing. I do hear it quite clearly on lower key female voices, like Cassandra Wilson's.
"What is being attenuated is the out of band resonance, which obviously contributes nothing to square wave performance."
I didn't say it did. I said implementing a filter and/or reversing the polarity would.
I still can't find a graph on the ET website showing the woofer response alone. Can you point me to it? Is it buried in one of the PDF's in the support pages somewhere?
Dave.
My understanding is that the phase effects of the filter will not be manifested significantly prior to the corner frequency, but perhaps I am not correct.
The manual shows square wave at 500 Hz, but I am pretty certain it is not the woofer. I don't think this cone woofer can do a square wave. I think only the LFT panel can do that due to its magnetic field linearity.
The main response graph shows a clear peak at 1.4 Khz. Bruce agrees that there is woofer cone resonance there. Maybe its not the entire peak. In any case attenuating the woofer above 500 Hz makes this region much less prominent and the nasal coloration on low female voices is gone.
Well, if you're interested I can compute some values for a notch filter. You'd need three components.....which almost doubles the amount of components in the existing crossover. :)
Something with a response like this....as an example.
Cheers,
Dave.
Thank you, but my electronic peak was reached when I opened the woofer cabinet and changed the 4.7uf film bypass capacitor to my preferred Multicap type (RTX).
I still need to listen to LFTs driven by the big Emotivas full range to see if having higher power with these speakers is really beneficial.
Here's a better look at the woofer resonance (blue trace) from the Stereophile measurements of the LFT-VII many years ago.
Dave.
Thank you for finding it.
That's a decent sized peak. I am glad I have a simple way to reduce it.
There is very little information on the FMOD internal construction and sound quality. Their own website has a blurb from AS touting devices' transparency, but I have not used them on the panel itself yet, so very difficult for me to tell.
The same blurb and a fuzzy graph on their web says "fraction of a db loss in band," but my own experience with the woofer section suggests about 2 db loss or so.
I may buy one and break it to see what is inside.
The low-pass FMOD's are just two RC combinations cascaded.
You can somewhat deduce the component values by measuring externally.
A simple resistance measurement from input/output will yield the net of the two series resistors. That net resistance and some arithmetic with your amplifier input resistance will yield the insertion loss.
The second capacitor can (sort of) be measured by probing the output with a capacitance meter.
You can't directly measure the other capacitor and you can't determine exactly the ratio of the two resistors.
Dave.
True, but it wouldn't be as fun as breaking it!
I experienced this firsthand now. :) The resonant peak on the woofer driver is quite horrible. On this set I measured it's at 2230Hz with a secondary peak a bit higher. Clearly audible if you play the woofers on their own with the stock inductor in path.
Dave.
So you went for them! How do you like them overall?
Interesting speakers. They have a couple of issues, but for the most part do everything well.
The woofer is a (modified) Peerless 830667 but, as noted by Dimitry, has a fairly strong cone resonance that's still audible with the first-order electrical filter.
Dave.
Are you planning to add a notch filter?
I've been bothered by a resonance in the IVA's, I think in the midrange. It isn't that loud but it's fairly tizzy and loud enough to be annoying. Not sure yet whether it's a design issue or a consequence of delamination. Either way, I'm looking forward to the arrival of the rest of my Neo 8's -- they guy who shipped them made a mistake so had to send six more. Should bring the quality of the midrange closer to that of the tweeter.
"Are you planning to add a notch filter?"
No.....at least not yet. I think a simpler solution would be to just add an appropriate shunt C.
Dave.
Definitely simpler, though you might lose some driver blending (not sure what the intended woofer level is at that frequency, if it's down at least -20 dB I imagine it wouldn't matter).
If keeping the existing inductor the "appropriate" capacitor value becomes a trade-off. That's why I was hedging. :)
As with all speakers, these need to be bi-amped. (Many issues can be alleviated.)
But I'm going to investigate the conventional path for a while longer.
That said, they're enjoyable to listen to in stock form.
Dave.
I'll be bi amping the IVA's soon, Mini DSP is here and the Yamaha P7000S proved much more, well, normal sounding than the Crown. Which being said, I was definitely hearing some buzzes and rattles from the delamination this afternoon so I'll probably repair that first.
Buzzes, rattles and delamination fixes trump everything else and go to the top of the to-do list. :)
Dave.
Fixing them is less fun than comparing amps, though . . . :-|
So what freq LP FMOD did you get?
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: