OK, will do a couple more starting above 30hz. I think I need to focus on the crossover before adding any EQ.
Under 30hz your are probably just stressing the amp and drivers is all.
You shouldn’t need eq as such, a speaker should be flat at 1m in an anechoic chamber. Once it’s in room you get a downward slope.
Can you take measurements of the individual drivers with the crossovers applied?
1 Like
I can, from the listening position or 1 meter?
Doesn’t matter really. I was just trying to get a sense of what each driver was up to
As they are full range units, off axis tends to be all over the place. It’s probably worth taking measurements at a fixed distance at various angles.
You might find the rising response is only on axis, which can help compensate for it dropping off off axis due to beaming.
You might find a better toe in angle.
Yeah ‘full range’. I’ll measure at 1 meter and various angles in a bit. Also I’ve chosen a crossover point based on what is probably nonsense I read on the internet. Will post with the other option I was exploring too.
Have you considered just setting fire to them?
Given the fs is 100hz I’d be crossing over at say 150-200hz.
On the main driver I’d probably just start with baffle step compensation and see what it looks like.
2 Likes
I used them as a low cost intro based on a bunch of enthusiastic internet reports. They are serving well for that but I don’t think this is an end state, especially when I see stuff like this
2 Likes
Maybe they could fill in between the bass unit and compression driver? 250-800Hz? Would still need something to deal with the shit from 1.3kHz though.
Your plan has merit!
I’ll set fire to them once I’ve got something else, which I’ll get once I’ve got somewhere with this setup.
Baffle step, would the baffle width be the front part or the wings added on too (191hz vs 93hz)
Probably somewhere between. Just mess about until it measures/sound better 
1 Like
Those were the original measurements, pre soundcard and SPL calibration. Doubt they will be wildly different
only the last picture with the Onken boxes fills me with any form of interest tbh
but getting an Onken design up and running successfully is a whole nother rabbit hole to loose time and money down
1 Like
It’s all horses for courses and the geekery outweighs the hifi really.
I messed about with different crossover points and listened rather than measured, 150hz LR12db was chosen using that method. These are the measurements at 1 meter of 1 speaker. Mic at full range centre.
Both drivers
Just Woofer
Just full range on axis
Just full range approx 30 degrees
Just full range approx 45 degrees
The part I’m sure I’m missing some huge fundamental on is that If I do anything to correct the imbalance above 1khz it sounds way too bass focussed, even when I went through and put notch filters everywhere there was a spike in bass response.
Something still isn’t right. The 15" wont go to 20k so that’s probably noise, which means you are not measuring loud enough or there is another issue.
The full range is 20db up at 2k vs 200hz, this isn’t right, look at the graph I posted, its -+ 5db from 100hz to 10k.
Using white noise rather than pink? Or sweep set wrong? It does look like a 20dB/decade slope.
Just playing both woofers and increasing the playback level of the sweep
So anything above 3khz is noise, that bump at 50hz on the loudest one was the speaker walking off the floor.
I tried taking the mic calibration off, tried pointing it up instead of at the speakers, looked for any settings in REW that might affect it but have come up blank.
Start by just running a tone from REW signal generator and turn up the system gain until you are measuring about 90 dB
Then try a sweep
Why dbc? U should be measuring unweighted
Try 500hz or 1k tone measured at 90 dB at 1m on axis, all filters off