I have bought a pair of Scanspeak D2905/9700 + 8545-00 and a finished enclosure (different), for the purpose of making a pair of Troels Gravesen "Amish" speakers.
It was only later that I realized how much impact the enclosure and baffle design has on the sound and crossover design.. Bummer!
Anyway. The enclosure for the amish design is 190x280x960mm where my enclosure is 198x240x853mm.. Both are ported enclosures.
I would, of course be able to get about the same driver placement as the Amish has since my baffle is wider than the Amish (198 vs 190mm).
Would I be wasting my time completely if I try using the Amish crossover for my enclosure/driver combo, or could it be expected to work alright? Is there anything in particular I could do to optimize the crossover for my enclosure? I know I should probably measure the drivers in my specific enclosure and design a new crossover from there, but I have neither the skills or the tools to do so..
It was only later that I realized how much impact the enclosure and baffle design has on the sound and crossover design.. Bummer!
Anyway. The enclosure for the amish design is 190x280x960mm where my enclosure is 198x240x853mm.. Both are ported enclosures.
I would, of course be able to get about the same driver placement as the Amish has since my baffle is wider than the Amish (198 vs 190mm).
Would I be wasting my time completely if I try using the Amish crossover for my enclosure/driver combo, or could it be expected to work alright? Is there anything in particular I could do to optimize the crossover for my enclosure? I know I should probably measure the drivers in my specific enclosure and design a new crossover from there, but I have neither the skills or the tools to do so..
His cabinet is heavily braced where mine isnt, so the loss in cabinet volume isnt nearly as bad as it seems, unless i end up having to brace it.
While i agree with the point of just giving it a try, im looking at about 200 dollars worth of components, since the crossover uses 5 coils, 4 caps and 5 resistors for each speaker. If someone would say it wouldn't work at all, id rather save the money
While i agree with the point of just giving it a try, im looking at about 200 dollars worth of components, since the crossover uses 5 coils, 4 caps and 5 resistors for each speaker. If someone would say it wouldn't work at all, id rather save the money
His cabinet is heavily braced where mine isnt, so the loss in cabinet volume isnt nearly as bad as it seems, unless i end up having to brace it.
While i agree with the point of just giving it a try, im looking at about 200 dollars worth of components, since the crossover uses 5 coils, 4 caps and 5 resistors for each speaker. If someone would say it wouldn't work at all, id rather save the money
I really don't think that the Xover is an issue, it is probably at about 2 to 3k, and the box is relatively unrelated to anything at those frequencies.
The box will only affect the low end and maybe reflected mid range from the back of the woofer, and this applies in both cases.
I also think the exact dimension of the speaker are not that important (they have an influence, but not that big) as long as the internal volume is the same.
The crossover will work with the same drivers, but with other (even with similar specs) it's not necesairly the case. It can be close if you're lucky, but it will never be tuned exact. So it's better to make a new. But you can try it (i did in the past, and i did build a new because it did not fit).
The crossover will work with the same drivers, but with other (even with similar specs) it's not necesairly the case. It can be close if you're lucky, but it will never be tuned exact. So it's better to make a new. But you can try it (i did in the past, and i did build a new because it did not fit).
While i agree with the point of just giving it a try, im looking at about 200 dollars worth of components, since the crossover uses 5 coils, 4 caps and 5 resistors for each speaker. If someone would say it wouldn't work at all, id rather save the money
That was me and I may have misunderstood you...
My comment was based on the understanding you already had the crossovers in hand.
If you are going to have to build a crossover, no matter what, then going with a fresh design specific to your choices makes a whole lot more sense.
The general rule of thumb is that you can play with dimensions if you stay within about 10% of the original.
Attached below shows why this is true in your case. Here are the comparisons of the baffle diffraction with Troel's original dimensions and then with your new dimensions. What's important is to keep the drivers's placement on the new baffle about the same as the old baffle in terms of distances from the cabinet top and sides. You can see that the differences in response are absolutely minimal and therefore, the xo is going to be perfectly fine in the cabinet you have selected.
More important as has been mentioned is getting the port right for the new net volume of your cabinet. Also, bracing is important, more so when you are using good drivers - you don't want the cabinet resonances messing up the quality you are paying for in the drivers. So keep the braces and redesign the port. Get help with that if you need it. And/or use a sub for the LF's.
The other problem not mentioned so far is the over 100mm difference in height. Ideally, the tweeters should be at ear height in your listening position. You will lose a little bit of HF response by being vertically off-axis. So if this is the case, lower your chair, raise up the cabinets a little or perhaps reduce the resistance on the tweeter by an ohm or 2.
Attached below shows why this is true in your case. Here are the comparisons of the baffle diffraction with Troel's original dimensions and then with your new dimensions. What's important is to keep the drivers's placement on the new baffle about the same as the old baffle in terms of distances from the cabinet top and sides. You can see that the differences in response are absolutely minimal and therefore, the xo is going to be perfectly fine in the cabinet you have selected.
More important as has been mentioned is getting the port right for the new net volume of your cabinet. Also, bracing is important, more so when you are using good drivers - you don't want the cabinet resonances messing up the quality you are paying for in the drivers. So keep the braces and redesign the port. Get help with that if you need it. And/or use a sub for the LF's.
The other problem not mentioned so far is the over 100mm difference in height. Ideally, the tweeters should be at ear height in your listening position. You will lose a little bit of HF response by being vertically off-axis. So if this is the case, lower your chair, raise up the cabinets a little or perhaps reduce the resistance on the tweeter by an ohm or 2.
Attachments
Thanks for clarifying. I am currently learning about crossovers for the purpose of designing my own - at some point. For now I'm pretty sure I will be better off with Troels' XO than anything I would be able to design myself ;-)
That is indeed good news. Thank you very much - both for the conclusions and for your explanations of what to keep in mind.
I have actually spent most of my evening trying to figure out about the reflex port. My hope was to somehow calculate how the port should be modified in order to get the "same" result as Troels' ones. However, I can't really figure out how to do that.
I have gathered various different software to calculate port sizes, but I always hit the wall when I'm supposed to make a decision about which tuning frequency I want. WinISD gave me some suggestions, but I ended up with a tube much larger than the enclosure itself ;-)
The current port in the enclosures I am (re)using is about 68x140mm. Easiest thing would obviously be to just extend or shorten this if thats a possibility, but I'm up for making a new port or adding an additional one.
Any suggestions would be greatly appreciated - thanks already!!
Personally, I like Unibox better than WinISD but it needs Excel.
Unfortunately, vented box F3 is going to be a function of both box size and port length. So you really can't get the same response in a smaller box just by using a longer port. But you can still get a good response with a smaller box albeit with a higher F3. Just tune F3 a little bit higher by increments until you get the best response possible. Most likely that will need a longer port too than the original.
Here's a thought that just came to me. I wonder if you might be able to fashion a new base for each cabinet such that it will raise the tweeter up closer to its original height and by cutting out the bottom of the existing cabinets, also add some extra liters to the net volume.
Unfortunately, vented box F3 is going to be a function of both box size and port length. So you really can't get the same response in a smaller box just by using a longer port. But you can still get a good response with a smaller box albeit with a higher F3. Just tune F3 a little bit higher by increments until you get the best response possible. Most likely that will need a longer port too than the original.
Here's a thought that just came to me. I wonder if you might be able to fashion a new base for each cabinet such that it will raise the tweeter up closer to its original height and by cutting out the bottom of the existing cabinets, also add some extra liters to the net volume.
Thanks. I have been thinking about ways to rise the tweeters as well - I will at least add some spikes to add a little more height.
Currently looking at UniBox, which doesnt make me any less confused. I have managed to load the speaker values end enter the volume, but I'm kinda lost from there. Haha Will give it another go later.
Do you have any idea what kind of port measurements I should end up with for the design?
Currently looking at UniBox, which doesnt make me any less confused. I have managed to load the speaker values end enter the volume, but I'm kinda lost from there. Haha
Will give it another go later.Do you have any idea what kind of port measurements I should end up with for the design?
Last edited by a moderator:
I think you may be agonising over this way too much. It is a fairly narrow baffle and about 50L, but whatever.
I looked at Troels' design as much as he publishes now:
Amish revisted
Interestingly, our Goran did something 25L quite similar:
www.audioexcite.com >> Classic Two – M5
OK, a slightly updated bass 01, but his bass filter is near as dammit the same.
I am no great shakes on reflex, but I think tuning is overrated. There are many different bass alignments. IIRC, smaller boxes get smaller ports You can line a port with a bit of carpet to make it narrower. Or put drinking straws in for a bit of damping.
Wouldn't do much harm to try Goran's simpler tweeter 9500 filter too. I usually just fiddle about with resistors to get level right. All this deep null stuff is overly fussy IMO. Tiny differences. As people have said, 10% change is nothing.
I looked at Troels' design as much as he publishes now:
Amish revisted
Interestingly, our Goran did something 25L quite similar:
www.audioexcite.com >> Classic Two – M5
OK, a slightly updated bass 01, but his bass filter is near as dammit the same.
I am no great shakes on reflex, but I think tuning is overrated. There are many different bass alignments. IIRC, smaller boxes get smaller ports You can line a port with a bit of carpet to make it narrower. Or put drinking straws in for a bit of damping.
Wouldn't do much harm to try Goran's simpler tweeter 9500 filter too. I usually just fiddle about with resistors to get level right. All this deep null stuff is overly fussy IMO. Tiny differences. As people have said, 10% change is nothing.
Troels has done some measurements, and the 9500 is 2dB louder than the 9700:
Why ScanSpeak tweeters 9500 and 9700 aren'
So you'd take out an ohm or two of resistance in the filter.
If the 9700 is a bit louder at the very top, you'd fix that with a little 7.5R plus 1uF Zobel shunt.
All fixable,IMO. I'm not quite clear what the volume of the tower cabinet is internally. But about 20-25L is what most people do.
Why ScanSpeak tweeters 9500 and 9700 aren'
So you'd take out an ohm or two of resistance in the filter.
If the 9700 is a bit louder at the very top, you'd fix that with a little 7.5R plus 1uF Zobel shunt.
All fixable,IMO. I'm not quite clear what the volume of the tower cabinet is internally. But about 20-25L is what most people do.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.