After a bit of hiatus, heres my latest concoction. PTT6.5, 10F and SB26ADC. MiniDSP powered with various amplifiers. Preliminary results are promising. Distortion, what distortion? Port resonance? Not when you dont play that high. Directivity matched? Seems like it. 30hz to 20khz? Check. At 106db? Check (...probably with some distortion lets not get too carried away). Port tuning 35hz confirmed. F3 32hz F10 i think around 25hz. Could it be made a passive? Probably, might be hard work on an amplifier. Downsides? Probably the group delay from ported design. But i havent got the kit to tackle full TL yet. But i will soon. You'll have to excuse the speaker terminals until my speakon connectors arrive. Next, improve the listening room and tweak crossovers.
Edit - not sure why pic is sideways. Mobile ph maybe.
Edit - not sure why pic is sideways. Mobile ph maybe.
Last edited:
This is often the case with some apple iphones/older IOS.
Edit foto, turn around until it is upwards again (4 times one direction) and you *should* be good [emoji6]
Nice work Mainframe! Great choice for the midrange. The best! These drivers are all well behaved. I bet a passive is not hard to do. If you care to post the FRD’s and ZMA’s maybe we can all have a good time fiddling with Xsim. 
Is that a standard issue Denovo knock down cabinet? That will make it easier for others to duplicate.
A TL could be made of course - you could just adapt the one I have.
Is that a standard issue Denovo knock down cabinet? That will make it easier for others to duplicate.
A TL could be made of course - you could just adapt the one I have.
Thanks X. I enjoyed your WAW thread very much. 10f is a nice little versatile driver. No knock down cabinet used but could perhaps be adapted. Fitting the 63mm port and mid enclosure in these dims and 18L (15L nett) was a challenge. Its still early days. I haven't "draft" played with any mfr. data sims in VituixCAD to see what might be required to get the purifi to cross at 3-400hz, but would it not be an inductor the size of the sun?
Good call - with that more sensitive woofer, and perhaps a 3/4" tweeter, you would have a contender for SOA wide dispersion passive speaker. However I was one of the early purchasers of the purifi woofers and I already had it and I use active mostly and I wont go shelling out for more anytime soon. It's fun trying to design around its shortcomings, and my goodness the bass that comes out of it is addictive.
I would cross at circa 1kHz so a 2mH to 3mH inductor is not that big. You can even go with iron plates core to keep cost and size small and get really low DCR and buy +1.5dB vs air core.
Run the tweeter in super tweeter mode and cross at 5.5khz and let the midrange do its flat phase thing from 1kHz to 5.5kHz.
That's sort of my thinking anyhow - there are many ways to do this of course, but setting the frequency, sets the type of filter and I prefer either 1st order or Harsch for nice transients.
Ack! He said "iron", don't say "iron" anywhere near the Purifi staff.
Couple of my favorite quotes from the Erin's Audio Corner YouTube interview with Purifi co-founder Lars Risbo go something like (quoting from memory):
"Iron likes to remember what happened in the past" which I think is the best non-mathematical non-graphical explanation of hysterious ever.
"At Purifi we have grown quite allergic to the grey granular sound of iron in the audio circuit"..
It may be true for tweeters. I have used copper air core and copper foil air core and copper iron core. Not for the Purifi, but for some woofers, they work fine (actually sounds great as a 5.5kHz low pass on my 5in 2-way) and give you low DCR for more sensitivity since the run of copper is so short. Just trying to help MF solve his inductor “the size of the sun” issue. But a 2mH air core is really not that big and ensures no hysteresis.
Hi mainframe99 and xrk971,
My highlighting Purifi’s stance in post #628 on iron or steel in audio circuits is in no way meant to disrespect mainframe99’s 3-way design or xrk971’s correct suggestion that if such a system were built using a passive crossover selecting tradeoffs may result in using an iron core inductor.
Quality is a continuously sliding scale and one of the perennial questions in audio engineering is when have we polished the design with qualities past what is audible? For example on the amplifier side of the Purifi shop they measured that many commonly used binding posts containing ferrous parts add enough distortion to prevent their amp modules from achieving the extremely low THD ratings they are capable of. Of course the flip side of the argument is a good percentage of audiophiles love their tube amps with all kinds of iron in the output transformers and orders of magnitude more distortion than 1ET400A! ‘Ya pays your money and takes your choice.
This side discussion also loops back to favoring mainframe99’s use of DSP crossovers rather than passive speaker side implementations. Use Moore’s Law to get the iron out.
My highlighting Purifi’s stance in post #628 on iron or steel in audio circuits is in no way meant to disrespect mainframe99’s 3-way design or xrk971’s correct suggestion that if such a system were built using a passive crossover selecting tradeoffs may result in using an iron core inductor.
Quality is a continuously sliding scale and one of the perennial questions in audio engineering is when have we polished the design with qualities past what is audible? For example on the amplifier side of the Purifi shop they measured that many commonly used binding posts containing ferrous parts add enough distortion to prevent their amp modules from achieving the extremely low THD ratings they are capable of. Of course the flip side of the argument is a good percentage of audiophiles love their tube amps with all kinds of iron in the output transformers and orders of magnitude more distortion than 1ET400A! ‘Ya pays your money and takes your choice.
This side discussion also loops back to favoring mainframe99’s use of DSP crossovers rather than passive speaker side implementations. Use Moore’s Law to get the iron out.
A very quick play around with the published data for the Purifi PTT6.5W08 (higher sensitivity, lower excursion version), 10f8424 and Hiquphon OW1 in Vituixcad and confirmed in Boxsim2 seems to work really well with almost identical sensitivities. Electrical 2nd order (l/c, c/l) on crossovers at around 600hz and 4.5khz results in an almost flat response on axis and very good phase tracking and a well behaved impedance curve which is above 4ohms at all points. Inductors are under 3.5mH and no really big capacitors (around 35uF max), no resistors used, 10f inverted. Promising.
The magnet used in purifi mid/bass drivers are iron/ferrite though.
The center T might be neo, but the magnet is ferrite by the looks.