New speaker build
Got going this week with the latest project.
I've been impressed with the performance of the compression driver in the Tannoy 295/8. I wanted to progress that favourable impression into something meaningful.
My current favourite speaker manufacture, Monacor, currently make a range of Tannoy-style dual concentric drivers. The SP-3CX series. 8 inch, 10 inch and 12 inch models.
I bought a pair of the SP-310CX 10 inch model for a good price (£120 each).
I spent a long time calculating the cabinet.
Based on the driver's specs it's been designed to work in a ported cabinet. Previously I have always gone for drivers that I know will work well in a sealed cabinet. So for this one, as well as calculating cabinet volume I had to learn a bit about ports and how to calculate them (and their displacement effects on cabinet volume). As long it's within my brain's limited capacity I love doing things like that, so that was fun.
I used 25mm MDF, screwed and glued. Pretty straight forward once dimensions had been calculated.
But the crossover for this was as much art as science.
The main driver for my trying this specific driver was because I like the sound of the compression driver in the Tannoy. It can operated down to 1000Hz. And to my ears it provides dynamics and clarity at that frequency better than any of the midrange cones I've tried. I wanted the Monacor dual concentric to do the same. The frequency curves for the driver show that this is possible.
Problem #1: Monacor recommend a crossover frequency of 1800Hz. However, I suspect that this may in part be for power handling reasons. The driver is rated up to 400 Watts; or rather the woofer part is rated 400 watts. The compression driver tweeter is rated to 100 watts. I imagine that 400 Watts overall would be fine with the crossover set at 1800Hz. Obviously reducing the crossover frequency means more of the power is routed to the compression driver. Possibly putting it at risk with the speaker fed with a total of 400 watts. This isn't a problem for me. My SECA is rated 20 watts. So theoretically the compression driver should be able to handle, from a power POV, the whole frequency range from that.
Problem #2: The frequency curve of the woofer is a little shaky 1000-2000Hz. I know from experience that graphs aren't everything, but I wanted to avoid the woofer in this range, and ignore Monacor's suggested crossover frequency.
Problem #3: The frequency curve of the compression driver is still rising at 1000Hz. It only has about 4dB to climb, but this was troubling, given my plan.
I considered building a Butterworth crossover, as this results in a 3dB lift at the crossover frequency. But in the end decided to go for a Linkwitz-Riley, which is flat, as this is the type I've built previously.
I should have been more adventurous.
With the build's first incarnation completed yesterday I was able to listen.
First thing to tweak is always the attenuation on the tweeter. Once the physical crossover is in place I can do this digitally using a digital high-shelf in JRiver, set at the crossover frequency. -15.1dB sounded good.
I had been most worried about the bass. This is the first ported type I've constructed. So far all good there. In fact I've been genuinely and pleasantly surprised. From the limited listening to date bass is impactful and tuneful. A good amount of air moving through the port. Really pleased with that. The port was a design complication that would have been easy for me to get wrong.
But, as my worries predicted, there is definitely an issue with a slight lack of output around 1000Hz. I can correct it digitally, and the speakers sound great. But obviously I want the speakers to sound great in their own right.
So it's back to the drawing board with the crossover. I've ordered the capacitors and inductors for a Butterworth crossover to give me that 3dB lift at the crossover frequency where the problem happens to be. The availble ratings for capacitors and inductors mean the crossover will actually be around 930Hz. Which I think, based on late night calculations and listening, will be OK. If not, then Plan C will be to raise the crossover closer to 2000Hz.
Once I get things as good as I can, I'll finish them with some sort of covering.
Anyway. I'm enjoying this one. :-)
I've been impressed with the performance of the compression driver in the Tannoy 295/8. I wanted to progress that favourable impression into something meaningful.
My current favourite speaker manufacture, Monacor, currently make a range of Tannoy-style dual concentric drivers. The SP-3CX series. 8 inch, 10 inch and 12 inch models.
I bought a pair of the SP-310CX 10 inch model for a good price (£120 each).
I spent a long time calculating the cabinet.
Based on the driver's specs it's been designed to work in a ported cabinet. Previously I have always gone for drivers that I know will work well in a sealed cabinet. So for this one, as well as calculating cabinet volume I had to learn a bit about ports and how to calculate them (and their displacement effects on cabinet volume). As long it's within my brain's limited capacity I love doing things like that, so that was fun.
I used 25mm MDF, screwed and glued. Pretty straight forward once dimensions had been calculated.
But the crossover for this was as much art as science.
The main driver for my trying this specific driver was because I like the sound of the compression driver in the Tannoy. It can operated down to 1000Hz. And to my ears it provides dynamics and clarity at that frequency better than any of the midrange cones I've tried. I wanted the Monacor dual concentric to do the same. The frequency curves for the driver show that this is possible.
Problem #1: Monacor recommend a crossover frequency of 1800Hz. However, I suspect that this may in part be for power handling reasons. The driver is rated up to 400 Watts; or rather the woofer part is rated 400 watts. The compression driver tweeter is rated to 100 watts. I imagine that 400 Watts overall would be fine with the crossover set at 1800Hz. Obviously reducing the crossover frequency means more of the power is routed to the compression driver. Possibly putting it at risk with the speaker fed with a total of 400 watts. This isn't a problem for me. My SECA is rated 20 watts. So theoretically the compression driver should be able to handle, from a power POV, the whole frequency range from that.
Problem #2: The frequency curve of the woofer is a little shaky 1000-2000Hz. I know from experience that graphs aren't everything, but I wanted to avoid the woofer in this range, and ignore Monacor's suggested crossover frequency.
Problem #3: The frequency curve of the compression driver is still rising at 1000Hz. It only has about 4dB to climb, but this was troubling, given my plan.
I considered building a Butterworth crossover, as this results in a 3dB lift at the crossover frequency. But in the end decided to go for a Linkwitz-Riley, which is flat, as this is the type I've built previously.
I should have been more adventurous.
With the build's first incarnation completed yesterday I was able to listen.
First thing to tweak is always the attenuation on the tweeter. Once the physical crossover is in place I can do this digitally using a digital high-shelf in JRiver, set at the crossover frequency. -15.1dB sounded good.
I had been most worried about the bass. This is the first ported type I've constructed. So far all good there. In fact I've been genuinely and pleasantly surprised. From the limited listening to date bass is impactful and tuneful. A good amount of air moving through the port. Really pleased with that. The port was a design complication that would have been easy for me to get wrong.
But, as my worries predicted, there is definitely an issue with a slight lack of output around 1000Hz. I can correct it digitally, and the speakers sound great. But obviously I want the speakers to sound great in their own right.
So it's back to the drawing board with the crossover. I've ordered the capacitors and inductors for a Butterworth crossover to give me that 3dB lift at the crossover frequency where the problem happens to be. The availble ratings for capacitors and inductors mean the crossover will actually be around 930Hz. Which I think, based on late night calculations and listening, will be OK. If not, then Plan C will be to raise the crossover closer to 2000Hz.
Once I get things as good as I can, I'll finish them with some sort of covering.
Anyway. I'm enjoying this one. :-)
Comments
It's one I've used before and seems to do the job.
https://www.diyaudioandvideo.com/Calculator/SpeakerCrossover/
Interesting to read you're thinking along similar lines re. compression drivers!
Linkwitz Riley produces a 6dB loss for both curves which sums to flat...
But you'll need to build a larger home to accommodate them. :-D
I look forward to seeing any developments.
In now, and definitely worth it!
Much better all round.
Next stage is to do loads of listening as things bed in over the coming days in order to decide what (digitally induced) attenuation is best for the compression driver tweeter. It;s currently sitting at -14.8dB, but that may change a bit as my ears acclimatise. Once I've made a final decision I'll order the resistors required for the L-Pad that will achieve that level of attenuation.
But not having a crossover is one of the primary reasons for going single-driver, surely?
Do you think I'd better ask first?
I'm going to have to plan a trip up to Glawster to hear what you're up to, Ben.
When I've got these ones sorted, we'll sort something! :-)
As stated above, the change to Butterworth curves did make me feel happier.
But I still wasn't satisfied...
So Take 3.
I decided to drop the crossover frequency on the compression driver tweeter to 700Hz, whilst keeping the same 930Hz on the woofer. This creates a more substantial overlap in the region I want to boost.
A 2nd order 700Hz Butterworth requires 20uF cap and 2.5mH inductor.
So I was able to recycle the a 2.5mH inductor and both 10uF caps (2x 10uF in parallel = 20uF) from the Take 1 Linkwitz-Riley crossover...
Resulting in this Frankenstein's Monster...
Things seem to have improved.
BUT I'm currently staying with my parents for a week, so my setup is makeshift in a room with acoustics that aren't my own.
So I can't really make any definitive judgements.
Will do when I get home.
We continue!
Listening suggested -14dB on the tweeters were required.
As luck would have it, I already had the necessary resistors in my bits and bobs box to make L-pads for that.
Sounded just a tiny bit peaky.
Pushing them back about 15cm towards the rear wall has improved that.
Pretty happy with what's going on.
Bass is great. Far better than expected.
That Wonfor magic is shining through!
I wonder, have you considered using all digital crossovers in one of many DSP programs out there? You'd need to multi amp but you'd have so much flexibility, and scope for learning, even if you choose to implement your final choices with passive components. There's hardware options like MiniDSP or Beringer items, but software options are much more powerful (Dirac, Audiolense).
Not making any claims here, but as tools for the journey they may give deeper insights, more flexibility and opportunity to try multiple variations in quick order. Either way, please keep posting as I love reading it!
I think I'm going to use such a program as I mention to EQ my loudspeakers, a single channel at a time, and seek driver time alignment (as well as room correction). I believe that could replicate the significant advantages of multi amping and bespoke crossovers.
This is an excellent article that's giving me food for thought: https://audiophilestyle.com/ca/bits-and-bytes/what-is-accurate-sound-r923/
In principle, yes I have had some very preliminary thoughts about it. Rationally it makes sense to me. And I do have 4 Wonfor amps that I could use to give it a go.
I did think about building these speakers with the drivers directly wired up to 2 pairs of speaker terminals so that I could use them with either an external conventional crossover, or try with a digital one at some point in future. In the end I didn't, my brain couldn't quite cope with that additional complication. I suppose I can always retro fit those in future.
In my head I think I'm still trying to progress my builds along a conventional route to an end point that I'm happy with.
Once satisfied, could be digital will naturally follow as the next grand project.
Loving these from that article!
Hadn't heard of Tekton before.
Which I think means that I have no idea what's going on anymore.
From what I'm reading everyone who's anyone owns a pair of their Double Impact speakers. Which, in fairness, look pretty good value given the number of drivers and size, and reported quality.
It was a little trickier than normal as the amount of sonic overlap of woofer and tweeter between 700Hz and 930Hz made it less easy to predict how digitally reducing the overall level from 700Hz would translate into L-pad values on the tweeter alone.
Anyway. Very happy.
Final steps:
1. Get the speakers covered. Will try a self-adhesive vinyl.
2. Sell the Tannoy 295/8 drivers, and sell the Goodmans 18P drivers. Hopefully get about a grand from those sales. (I have run out of money.)
3. Plan next project!
OMG! What will it be?
Embarrassing, isn't it?
Painted the insides of the cabinets with bitumen paint.
What a lovely substance to work with that is.
The process was made worse when I knocked one of the still-attached-to-the-crossover drivers off the work bench and it ripped out most of the crossover components. And unravelled one of the inductors. Ugh.
What a clumsy oaf I am.
Fortunately I had a spare inductor.
And the time to reconstruct the crossover.
3 coats of bitumen in both cabs (about a third of a litre per speaker).
Sounds dead good. Any cabinet-induced sound reduced a few notches. Perhaps most noticeable in mid-range. Things are more detailed and smoother.
If anyone wants to try this, I would strongly recommend doing so when constructing the cabinet. Doing it retrospectively was horrific.
Also, I'm aware that each of the cabinets is now about 1% smaller in volume. Probably not hugely significant, but in future I would allow for 2mm of bitumen paint on all inside surfaces.
Dr Sticky Mess, himself!
Or are at least worthwhile.
I'm not really happy with the final look. For future projects I need to do more research and work on coverings and edgings.
But sound-wise, I am very happy. Happiest I've been with my system.
Time and effort spent fine-tuning things is a good investment.
Main learning points for me from this project...
Bitumen can be a good internal damping material.
In my room, fixing speakers to concrete slabs is helpful.
These particular drivers sound better with the crossover frequencies overlapped (700Hz - 1000Hz).
And Dave, I was thinking about your comment re. using "upside-down" speakers to avoid having bass drivers close to the floor... I suppose using dual concentrics is an even better option! :-D
Back of fag packet costings (including stands) is £540...
2x Monacor SP-310CX Drivers £250
25mm MDF (cut by merchant) £25
Screws and fixings £10
Bitumen £10
Acoustic foam £20
Crossover components £80
Cabling and binding posts £25
Coverings £40
Monacor badges £15
Stands (MDF and hard wood dowel (cut my merchant)) for £80
At the start of the project my long-term plan was to have these as two-thirds of a big speaker stack, sitting atop of a bass cab handling <75Hz. But... I am sufficiently pleased with the bass repsonse from the Monacor SP310CX drivers that I currently have insufficient motivation to proceed with that. May be a project for next year / next lockdown.
Reminds me of a MEG Monitor.
The RL940C you mean...?
(Absolutely no Googling was necessary for me to just this minute familiarise myself with MEG speakers.
Honest.)
What would have put you off?
Them is some weird looking drive units!