Looking good !! ...yes, synergies make for easy driver integration, huh !!
I think your goal/implementation of a 1200Hz CD xover point makes good sense.
With synergies and their tight spacing of mids-to-CD, I don't think the typical advantage of pushing a CD's xover lower is realized.
I think it makes more sense to push a CD lower in conventional designs, to get closer to 1/4 wave spacing, but not so much with synergies using mids where 1/4 is easy.
I know using the 4NDF34's to relieve the low end load on the CD, has certainly added clarity to my syns. With the one caveat the clarity gain could also be coming from not having the low driver try to extend as high as without the mids. ( dunno for certain the relative clarity gain importance)
I'm very interested in where you end up crossing low-to-mid.
Again, super well done....you've built the best speaker i can imagine....a big synergy with rounded secondary flares that needs no sub. 😀
Well, maybe the best thing other than, the same thing even bigger lol
I think your goal/implementation of a 1200Hz CD xover point makes good sense.
With synergies and their tight spacing of mids-to-CD, I don't think the typical advantage of pushing a CD's xover lower is realized.
I think it makes more sense to push a CD lower in conventional designs, to get closer to 1/4 wave spacing, but not so much with synergies using mids where 1/4 is easy.
I know using the 4NDF34's to relieve the low end load on the CD, has certainly added clarity to my syns. With the one caveat the clarity gain could also be coming from not having the low driver try to extend as high as without the mids. ( dunno for certain the relative clarity gain importance)
I'm very interested in where you end up crossing low-to-mid.
Again, super well done....you've built the best speaker i can imagine....a big synergy with rounded secondary flares that needs no sub. 😀
Well, maybe the best thing other than, the same thing even bigger lol
Well it's pretty standard to display a FR with the y axis spanning 50 dB e.g, from lowest line 60 dB to top line 110 dB. Many are used to judge linearity from that scale. If you have 150 dB span, anything will look "nice"...
//
//
Gotcha. I'll re-scale them and post again.
I got a bunch more measurements done yesterday and did some EQ work. Think I'm getting pretty close to getting them dialed in. I didn't get a chance to get any screen grabs though. I'll sit down this evening and get some images together. Should have some good data to share.
Ok, did some re-formatting. Here are the individual drivers' raw response again. This time on a 50dB scale.
And here is where I am at the moment with the full range response.
Starting to look and sound pretty good!
Starting to look and sound pretty good!
A few notes. The mids are SUPER efficient. I have to dial them back something like 17dB in order to level match with the CDs. I was a bit worried that the little APPJ amp I had on them would not keep up, but I don't think that is going to be an issue. I'll replace it before too long, but it is nice to know that with such high efficiency and limited bandwidth requirements, I could use pretty much any low-powered amp I want.
I also wanted to thank ChrisA/Cask05 for his help and inspiration with this one. I learned a lot from him over the years at the Klipsch forum. He helped me understand a lot about how to measure and EQ speakers. His K402 MEH was also obviously a huge inspiration for this project. I wouldn't be listening to hige synergy horns if it weren't for him. I realized I neglected to credit him in my first post so.. thank you Chris!
I also wanted to thank ChrisA/Cask05 for his help and inspiration with this one. I learned a lot from him over the years at the Klipsch forum. He helped me understand a lot about how to measure and EQ speakers. His K402 MEH was also obviously a huge inspiration for this project. I wouldn't be listening to hige synergy horns if it weren't for him. I realized I neglected to credit him in my first post so.. thank you Chris!
Yes, for sure. I have been focused on getting my crossover and EQ to a workable state, but I am getting pretty close there. I'll see if I can get some distortion graphs and a waterfall plot this evening. It has been a while since I have used some of those features in REW, so it might take some tinkering. I have been working with my woodworking tools for too long, getting rusty with the electronic ones, haha.
Just switch view of the measurement you already have above - the last one would be OK. Hit the Distortion tab. In the waterfall view you need to hit Generate button.
//
//
Basically, you're using a small fraction of 1 watt on each channel, even at near-deafening SPL. The reason for this is:
1) like you mentioned, the bandwidth covered by each amplifier channel is notionally 1/3rd of a full range/mono-amping loudspeaker (implying 1/3 the power requirements), and
2) you're direct coupled to the driver with no added resistances or reactances that passive networks must use to mono-amp loudspeakers, in order to level the outputs of the bass, midrange, and tweeter channels.
3) you're completely horn-loaded, which give you a 15+ dB gain, typically.
So, as I might have said earlier, amplification of these type of loudspeakers just isn't an issue from a power headroom standpoint. You're now listening to things like amplifier noise floor and amplifier transfer function (i.e., SPL and phase response) at sub-watt output level of the amplifiers you're using.
Additionally, you're now listening to the loudspeakers using carefully EQed to flat response on-axis (1 metre), so differences with differing amplifier output impedances loading up with passive crossover network and driver impedance responses (particularly high output impedance amplifiers including SETs, etc.) is no longer dominated by small changes in amplifier-to-amplifier SPL and/or phase response (for minimum phase response portions not dominated by all-pass issues), so audible differences in amplifiers will be much less than they were using passive crossovers/mono-amping. This is an extremely good characteristic to have.
Thanks for that. It's a new day in terms of where we can talk about these things (i.e., on this forum instead of the old one). The good news is that I believe more people are beginning to see that DIY loudspeakers using DSP crossovers and freeware/shareware acoustic measurement apps easily beats anything that mainline loudspeaker manufacturers can put into your listening room--and now apparently at 1/10th or greater in cost savings. Perhaps more on that subject in another thread, but I think you're going to find that the loudspeakers you have built will outperform anything you can buy for your home hi-fi system. Not kidding (...and @rickmcinnis can tell you all about that subject...).
The "button bar" just above the REW plot area has a button called Distortion just to the right of the SPL & Phase and All SPL buttons.
I really wouldn't worry too much about harmonic distortion (HD) plot if you're taking SPL measurements below 90-95 dB, because this is a non-linear distortion type that's a strong function of the SPL (and you really can't see what you've got until you crank it up to higher SPL measurements). I think the gentleman that asked for it was going to subsequently ask to up the drive level after you plotted HD. In my experience, unless the HD is weirdly distributed vs. frequency, it doesn't really tell you much other than the quality of the drivers employed and perhaps any issues with horn ports--which isn't something that you're going to tweak using a DSP crossover.
Also, in my experience, Waterfall plots are basically a waste of time. However, the Spectrogram plot tells you a great deal more without having tweak things up, especially if you get your default display settings set up. Here are the Spectrogram plot settings that typically use (below). The "Span after peak (ms)" value of the plot range below can also be set at 10 ms to see sub-millisecond events more clearly:
Chris
Last edited:
I think I have come to this same conclusion. I have heard some very expensive hi-fi systems over the years (Wilson speakers, Audio Research Electronics, etc.), and they can be excellent. However, these are better. Not just better for the price, objectively better absolute performance.
I agree that the non-sensitivity to amplifier performance that you described is an excellent quality, and a huge money saver. I am no longer interested in mega-buck amps. A good SS amp that can drive the bass drivers, and some decent low-powered tubes (my preference) to drive the tops are all that is needed. As you say, once you restrict the bandwidth and eliminate the impedance swings, almost any decent amp will do.
Yes, the spectrogram plot was the one I was thinking of, not the waterfall. I'll see if I can get those this evening.
Here is a spectrogram plot. I am not 100% sure I have my driver delays set correctly. I think they are close, but maybe some of you can interpret this and offer suggestions. It looks like something funky happens at the lower crossover (150Hz)?
Here is a distortion plot. Again, this shows a bit of an anomoly at 150Hz. I seem to be getting a null there. I'm not sure what is happening, as the phase plot is smooth in that area. I meant to save a graph showing phase, but I must not have.
I do all of these measurements on my home PC, which is hooked up to the Hi-Fi system. I then post about this stuff while I am at work, and I often seem to find that I didn't record all of the measurements that I meant to. Oh well, should have a frequency/phase graph to share tomorrow. The phase looks very smooth and flat to me.
I do all of these measurements on my home PC, which is hooked up to the Hi-Fi system. I then post about this stuff while I am at work, and I often seem to find that I didn't record all of the measurements that I meant to. Oh well, should have a frequency/phase graph to share tomorrow. The phase looks very smooth and flat to me.
I should be able to get another distortion plot at much higher volume over the weekend. I do most of my hi-fi tinkering after the kids have gone to bed. I can do frequency sweeps, as my basement is fairly soundproof, but there are limits. The 120dB measurements will have to wait for Saturday!
if you have your drivers perfectly time aligned, you will get a straight vertical line above where the room dominates
the wiggle at your 1200 Hz XO indicates its not quite right there
the big bend circa 150 could be the room or could be woofer time alignment
take measurements really close to lessen room influence to check time alignment. with synergy you don't need to be 2m away, you can put mic right at horn mouth. look at impulse and step responses as well as the spectrogram
the wiggle at your 1200 Hz XO indicates its not quite right there
the big bend circa 150 could be the room or could be woofer time alignment
take measurements really close to lessen room influence to check time alignment. with synergy you don't need to be 2m away, you can put mic right at horn mouth. look at impulse and step responses as well as the spectrogram
I have been taking all of my measurements right at the horn mouth and my room is heavily treated, so I would guess it is not room interaction. I think I need to do some work on time alignment.
Hi, yeah, something is funky with the measurements....smooth flat phase doesn't tie together with the offsets in the spectrograph, or the frequency response null around 150 Hz.
Smooth flat phase ties with a straight vertical line on the spectro.
Looks like the woofer section needs delay...... and a fair bit of it (maybe 4-5ms) ...which is surprising.
What, if any, delays are in place on the mids and CD?
On my builds with similar mouth size and depth, I've found the mic needs to be at least a meter away from the mouth. If the mic is at/in the mouth, relative timings/delays don't hold up real well further away from the speaker. It's kinda like the synergies focus like an eyeball to a specific distance.
I know all speakers do that really. It's just seems to me these big horns exert greater control on the focus distance, ie the measurement distance used for tuning, that normally baffled speakers.