The green trace for the woofer looks right, but not the others
Using REW to measure a different speaker (unfortunately not a known quantity, a sp276/8 in a cardboard box)
I’d say that eliminates the rest of the system and the rising response of the 8" Full range is correct.
Everyone knows gorilla tape sounds bestest on cardboard boxes 
1 Like
It doesn’t explain why the response of both together is lower in the LF than the woofer alone
I’ve moved the speakers, retested and put all the plots together.
apart from the rising full range response which we think is real, the rest looks ok doesn’t it?
It does, kinda. But a listening position measure should slope down rather than up. I’m a bit concerned that your bass is only 10dB above the noise level.
Is that measurement with the crossover in place, or not?
Crossover in place.
I’ve bluetacked the baffles to the floor but there are still some nasty resonance issues, and as before if I flatten it out it sounds massively bass heavy. It also feels that at any sensible spl at that listening position I’m pushing everything a bit too much, especially the bass.
This has partly led to me moving everything around and I’ll put in a listening chair and make that end of the room HiFi orientated.
I’m also contemplating buying 2 more 15s and something else to go between them.
Does the microphone output both channels? Are you measuring using the wrong one? Is phantom power required/on?
With a second order XO at 150Hz the woofer should output nothing at 1kHz, so you have noise at 60dB, but the signal is only 70dB.
Try measuring at 10cm from each driver with the crossover bypassed. That should show us the drivers.
Ant, where abouts are you located? I have some Veravox 5S full-range drivers going spare if you wanted to try something different. And possibly, maybe my Supravox 165-2000 drivers may be spare to borrow at some point… 
Something looks odd on all your plots, I’ve always got decent results with REW.
have you checked the crossover components are wired correctly?
Mic input is left, I’ve selected that correctly, Phantom power is on.
I tried to measure my headphones, using the measurement mic and then a different studio mic. The response from the measurement mic is noticeably rising.
Hythe, Kent. Very appreciated offers, definitely want to get to the point I can trust my measurements before putting good stuff in 
I’m going to play direct into the full range amp shortly and measure at 10cm.
Measuring 10cm from full range, soundcard direct into amp. Then repeating with the mic at the listening position. I have to turn the gain on the mic up quite a bit because the SPL achievable by the speakers is pretty low, mainly because I think I’m going to blow another one up with low frequencies and long excursions.
Well that looks better. You shouldn’t destroy drivers with a short sweep, they don’t have time to get hot.
The measuring mic is the blue one? I’m surprised there is that much of a difference, but it does seem consistent with the story that the mic isn’t recording the lower frequencies correctly. But the 10cm speaker measurement looks ok.
Yeah it’s the blue one, very unscientific test of course. I really need to get a box speaker of some description in here (LVs looking likely) to see what they look like when measured from afar.
Famous last words, another one has gone crunchy, wont be buying another replacement.
Right, what now?
1 Like
I would measure all four drivers without crossover, to design a crossover for them. Measure at 10cm; get phase info as well as frequency response.
That hump in the full range at 100Hz looks a bit concerning, so I would be thinking of crossing them over higher than 150.
Is that a flintlock?
Seems harsh to shoot them