What brought me here to GR Research & Audio was a desire to acquire a pair of floorstanders as they would match the room space better so for £6 I purchased locally a pair of Mission 733's found on eBay that had their polycarbonate cones disintegrating.
I replaced them with a pair of Monacor SPM 165/8 as they are form and fit compatible but unfortunately not functionally equivalent. The result was shite to say the least and I couldn't pass them on so I've lived with them until recently after I discovered that there are some pretty solid techniques and software to measure SPL & Z (REW) and use the measurements to systematically design a crossover to improve their performance (VituixCAD). I've been hooked ever since. The £50 spent on the Monacor could have been considered a waste but it's allowed me to experiment with them and use really cheap components to learn my trade as it were. You can do a lot of retries with very little money this way and possibly go for the ultimate design later with better drivers & components (as I have done by starting a pair of bookshelf monitors).
Before installing REW & VituixCAD these speakers sounded muddy & booming and I experimented with blocking the reflex port up which seemed to improve things a bit but on reflection this was just shooting in the dark.
Seperating the drivers out and driving them with and without the crossover it soon became obvious we were well into woofer cone break up territory and the crossover frequency was far too high. The original had polypropylene cones which could be driven at higher frequency before exciting unwanted modalities.
The other issue was the extremely lacklustre performance of the tweeter which had practically no bandwidth. It was also quite insensitive, preventing the use of notches etc to extend the bandwidth.
It was hard to suppress the cone breakup region and I had to use a 3rd order filter to do this as the SPM165 rang like a bell much above 4KHz and you need to match the underperforming tweeter with as much output as you can close to the tweeters lowest frequency (not much below 3KHz). It was a tight squeeze and I don't like high order filters as the rapid pass band/ stop band phase change associated with high order filters places great emphasis on component tolerances. You can explore this by inverting one of the drivers in the CAD. I wasn't getting the deep null at the crossover frequency I was looking for indicating that it was a bit of a mess.
I recycled as many of the components as possible using REW to measure the inductor values for example and design around them but the speakers were vastly improved by my effort. I took the bungs out of the reflex ports which 'opened them up' - they didn't seem to boom in the same way which must have been an audible illusion (as opposed to an optical illusion). They are now a bit bass heavy and I explored why this was by using the enclosure tuning tool in VituixCAD. This seemed to indicate that the cabinet size should be about 40 litres rather than 22 litres and the reflex port should be about 20cm rather than 12cm so, together with the abysmal performance of the tweeters means that these have pretty much reached the end of the road. It's astonishing how important it is to match cabinet size and port dimensions to the driver if you're doing a reflex enclosure. I now understand why acoustic suspension is so much safer than reflex particularly for small speakers. My 9.1 Litre mini-monitors would need to be at least 20L if I were to use a reflex port.
But it's been a hell of a journey.
PN
