I am looking to build a reasonable 'hi-fi' system and would be most interested in any informed comments on my proposed design, in particular in terms of my choice of drivers, crossover points, number of drivers etc.
The design parameters I am aiming for are:
High dynamic range and good efficiency over the full audio frequency range (15hz-20khz or so), mainly for music purposes, but also occasional HT. Reasonably flat response and reasonable cost (perhaps ~$2000AU). Relatively low distortion. Simplicity so as to reduce the design effort and need for tweaking, which rules out horns! (although they provide *excellent* efficiency I am confident that the result would not have a flat response).
I would like to use active crossovers as this approach appears to simplify the design of multi way systems and give many advantages, such as high dynamic range and power, drivers do not interact negatively with the crossover or other drivers, no endless tweaking of passive crossovers required!!!
Limited disadvantages (such as increased noise introduced by extra op-amps).
My initial thoughts are a 4 way design with 3 way sattelites and an infinite baffle sub for strong and extended sub bass.
Apart from the Audax, I havn't really decided on drivers yet but after looking around these seem to be suitable for my purposes:
The Fostex FT17H tweeter has a fairly flat response and high efficiency of 98.5 dbw/m.
The Audax PR170M0 appears to be an excellent midrange with flattish although rising response curve and especially high efficiency (100dbw/m)
For midbass perhaps something like the Eminence Kappa-12A. High efficiency of 99.3dbm/w and response down to 80hz or so before it starts to roll off. The plan is to cross it over at 80hz and let the subwoofer handle the rest. This is probably the component I am least comfortable with, and given the narrow frequency range it is being asked to handle (80-300hz) I am wondering if I can get away with a lower efficiency driver although I probably wouldn't want to go under 94dbm/w and 12inch is probably as big a diameter as I want for the satellites too (suggestions???)
The crossover points I am thinking of are probably about 300hz for the midbass/mid and about 5-6 khz for the mid/tweeter.
I have heard and it seems reasonable to me that it is best to avoid crossover points within the midrange region of 300-5000 hz because this is where the negative effects introduced by crossovers are most noticeable.
Closed boxes would be used for the midbass and mid drivers.
The slopes on all the crossovers would be 24db/octave.
Havn't really looked at the IB design, but it will probably be comprised of 6x15inch drivers with 500watt or so of amplification.
For the rest, I'll probably just build a 50w*6 power amp based on the SC-480 kit.
DB
The design parameters I am aiming for are:
High dynamic range and good efficiency over the full audio frequency range (15hz-20khz or so), mainly for music purposes, but also occasional HT. Reasonably flat response and reasonable cost (perhaps ~$2000AU). Relatively low distortion. Simplicity so as to reduce the design effort and need for tweaking, which rules out horns! (although they provide *excellent* efficiency I am confident that the result would not have a flat response).
I would like to use active crossovers as this approach appears to simplify the design of multi way systems and give many advantages, such as high dynamic range and power, drivers do not interact negatively with the crossover or other drivers, no endless tweaking of passive crossovers required!!!
Limited disadvantages (such as increased noise introduced by extra op-amps).
My initial thoughts are a 4 way design with 3 way sattelites and an infinite baffle sub for strong and extended sub bass.
Apart from the Audax, I havn't really decided on drivers yet but after looking around these seem to be suitable for my purposes:
The Fostex FT17H tweeter has a fairly flat response and high efficiency of 98.5 dbw/m.
The Audax PR170M0 appears to be an excellent midrange with flattish although rising response curve and especially high efficiency (100dbw/m)
For midbass perhaps something like the Eminence Kappa-12A. High efficiency of 99.3dbm/w and response down to 80hz or so before it starts to roll off. The plan is to cross it over at 80hz and let the subwoofer handle the rest. This is probably the component I am least comfortable with, and given the narrow frequency range it is being asked to handle (80-300hz) I am wondering if I can get away with a lower efficiency driver although I probably wouldn't want to go under 94dbm/w and 12inch is probably as big a diameter as I want for the satellites too (suggestions???)
The crossover points I am thinking of are probably about 300hz for the midbass/mid and about 5-6 khz for the mid/tweeter.
I have heard and it seems reasonable to me that it is best to avoid crossover points within the midrange region of 300-5000 hz because this is where the negative effects introduced by crossovers are most noticeable.
Closed boxes would be used for the midbass and mid drivers.
The slopes on all the crossovers would be 24db/octave.
Havn't really looked at the IB design, but it will probably be comprised of 6x15inch drivers with 500watt or so of amplification.
For the rest, I'll probably just build a 50w*6 power amp based on the SC-480 kit.
DB
HE is a very good idea.
The IB bass is a good idea.
The Audax is a good unit.
The Fostex, I'm not too keen on, preferring real compression drivers and flares to these (I have some FT17's still NIB; used were sold).
Midbass, the Eminence isn't that good. Look for some secondhand JBL 2226.
I will soon be listening to a HE system comprising per channel;
LF: 2 x 18LX60 Beyma's
MB: JBL 2220
Mid: JBL 2123
HF: BMS 4552 or Beyma CP380M or 18Sound ND1090 on 18Sound XT1086 flares. Testing will decide which I'll use and which will go into the PA's.
Initially, all will be based around LR24 / LM3886 xovers and amps, but may dust off the tubes and use them later (space challenged for them).
The IB bass is a good idea.
The Audax is a good unit.
The Fostex, I'm not too keen on, preferring real compression drivers and flares to these (I have some FT17's still NIB; used were sold).
Midbass, the Eminence isn't that good. Look for some secondhand JBL 2226.
I will soon be listening to a HE system comprising per channel;
LF: 2 x 18LX60 Beyma's
MB: JBL 2220
Mid: JBL 2123
HF: BMS 4552 or Beyma CP380M or 18Sound ND1090 on 18Sound XT1086 flares. Testing will decide which I'll use and which will go into the PA's.
Initially, all will be based around LR24 / LM3886 xovers and amps, but may dust off the tubes and use them later (space challenged for them).
I think might have a hard time finding a driver to cover 300HZ to 5k. Figure you want to match the directivity at crossover as best you can so what ever you use can't be too large or it will beaming quite a bit at 5k. Even an 8" driver will be problematic up that high. You might want to drop one crossover point in the middle. You are going to be running active so it's easy to make changes on the fly. Just see what works for you.
I have a couple of active systems up right now and both crossover at 300Hz and 1.5K and they sound fine. The 1.5K crossover points allow me to get a better match up to the drivers directivity through the crossover point. Works for me.
One is JBL E-145's and 2123's with a 2435 compresssion drivers on a PTH1010 waveguides. I have subs under this system that cut in about 50-60Hz. The other is an MTM I am messing with with sealed JBL 121A subs with a pair of 2118J's with an aquaplassed 2426 compression driver with a 2344 horn.
Rob🙂
I have a couple of active systems up right now and both crossover at 300Hz and 1.5K and they sound fine. The 1.5K crossover points allow me to get a better match up to the drivers directivity through the crossover point. Works for me.
One is JBL E-145's and 2123's with a 2435 compresssion drivers on a PTH1010 waveguides. I have subs under this system that cut in about 50-60Hz. The other is an MTM I am messing with with sealed JBL 121A subs with a pair of 2118J's with an aquaplassed 2426 compression driver with a 2344 horn.
Rob🙂
The JBL 2266 looks like an awesome driver ;-) ... something similar such as a JBL 2220 or 2205 would by ok too right? Also I read that the 2266 has some cone breakup distortion at 150hz. I was planning on crossing over a bit higher than that, bit is this only at high volumes that I wouldn't need to worry about? What other drivers might be suitable, and I have recently discovered that the WAF of a satellite speaker with 15" driver is actually not very high...
Am I likely to run into Xmax issues on either the Audax or JBL 2220/2205 ? the Audax only has 0.5mm of Xmax but I figured that wouldn't be a problem over 300hz? I don't want to play at 110db but I would like to be able to reproduce punchy distortion free drums and bass guitar.
I sounds like both you guys are using compression drivers. I know they are efficient but are these drivers capable of 'hi-fi' performance -i.e. smooth response and dispersion?
One final question I have is how do I work out the power requirements for each amp in an active Xover setup?
many thanks for your help...
DB
Am I likely to run into Xmax issues on either the Audax or JBL 2220/2205 ? the Audax only has 0.5mm of Xmax but I figured that wouldn't be a problem over 300hz? I don't want to play at 110db but I would like to be able to reproduce punchy distortion free drums and bass guitar.
I sounds like both you guys are using compression drivers. I know they are efficient but are these drivers capable of 'hi-fi' performance -i.e. smooth response and dispersion?
One final question I have is how do I work out the power requirements for each amp in an active Xover setup?
many thanks for your help...
DB
If you are concerned with the 2226 you might want to look at E-145's. They are 15" musical instrument speaker for keyboards and bass. X max is not an issue for them and they use underhung coils so they are very linear. They will not do 20-40hz octave put with subs you won't have a problem
For power I just figured what my least sensitive driver was and used my listeners distance to determine how much power I needed to get to my target 115db in room. With E-145 that 98db so 100 watts gets me to my target for a pair. The power and SPL figure is consevative. I used the standard distance for outside and did not account for amplifier headroom.
Rob🙂
For power I just figured what my least sensitive driver was and used my listeners distance to determine how much power I needed to get to my target 115db in room. With E-145 that 98db so 100 watts gets me to my target for a pair. The power and SPL figure is consevative. I used the standard distance for outside and did not account for amplifier headroom.
Rob🙂
I'm not aware of any breakup in the 2226 at 150Hz in my experience with it. It's well behaved until quite high, certainly at domestic playback levels.broughd said:
WAF has never been an issue for me, but with careful design, the bulk could be aesthetically minimised I suppose.
No Xmax issues with the JBL's It's 5mm on those mentioned and in most enclosures at any sort of sane domestic level will be fine. The Audax I know mainly from reputation, so others may wish to give definitive comments. I believe it'll not be an issue in practice.broughd said:
Good CD's are better for distortion or dispersion and detail resolution than any dome I've ever heard. Dispersion is very flare / waveguide dependant. With the correct one, directivity will be much better controlled than domes ime.broughd said:
I look at the sensitivities and the SPLs I want to achieve and scale from there.broughd said:
If you go with any of the older Alnico baskets like a 2205 and you have them reconed make sure you have the flux in the gap checked. They can get beat up bad in Pro appications so make sure the baskets have not been demagged at all by hard use over the years.
Rob🙂
Rob🙂
(beginner question...)
Thanks for the suggestion. You say these are linear, but being a musical instrument speker, are these suitable for hifi though, arn't musical instrument speakers designed to introduce distortion or are these an exception? Also, how do they compare in terms of price?
Ok - that sounds reasonable, but I guess I was referring to the power requirements for specific frequency bands that each driver will handle, i.e. I need to work out the power requirements for 10-80hz, 80-300hz, 300-3000hz and 3000-20khz - if they are sufficiently different I might need different rated amps. I believe there is a distribution of power by frequency for normal music, but I havn't been able to find a reliable definition...?
For a midbass driver the JBL 2226 sounds ideal for my purposes, however it also seems to be relatively expensive, I recently saw 4 go on Ebay for $500AU (~$450US). It was listed 'used, as new condition' - is that a normal price - or should I be able to get some cheaper than that? What is the new price?
How much better is the 2226 than other similar spec but cheaper drivers?
e.g. http://cgi.ebay.com.au/NEW-etone-15...ryZ23794QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I'll take it from you as I probably havn't heard good CD's!!! What is a good choice for the highs? I don't need to produce any more than 120db peaks or so, and probably won't be playing at any more than 85db (I value my hearing) so don't need a high power unit - it is smooth response and dispersion that I mainly care about. Oh, I don't want to spend to much $$$ either (willing to pay up for quality to a point though) Do you perhaps have a suggestion as to what would suit me? I'd be interested to know which HF driver you settle on for your system too. The Audax PR170M0 seems like a good compromise to me for a HE mid at an affordable price, the alternative is probably a CD but that is likely to cost a fair bit more and not really give me much advantage right?
Thanks - sounds like good advice - I think I'll just avoid them then because it's a bit hard to check the flux density when buying off ebay!!!
Thanks for your expert advice, hopefully I'll be able to put something decent togerther first try without screwing something up and having to buy more components...!
🙂 DB
A good high frequency compression driver mounted in a good "constant directivity" horn outperforms any dome tweeter in the perceived "definition" because you are getting rid of a lof of reflected ambient sound with the horn, which is the main source of sound "contamination" when listen in any environment prone to ceiling/floor/wall reflections.
A bad HF compression driver won't have a flat frequency response and will need EQ. It may also have notches above 10k. Most CD horns also need very simple Eq to compensate rolloff (very smooth!!) above 5Khz.
A typical bad (medium sized) HF horn will probably have 3dB peaks and dips near the lower cutoff (if it was shrunk too much). It may also suffer from 6dB peaks and notches above 5Khz due to internal reflections. Non CD-horns will be progressively directive above 5Khz and progressively wide-dispersing below 5Khz. When such a non-CD horn is employed in a room (or any environment capable reflecting considerable amounts of off-axis sound back to you), the stuff radiated off-axis below 5Khz (particularly at 1hz-3Khz) will produce a quite honky unpleasant and fatiguing "ambient" effect (particularly at high volumes). Non-CD horns are fine outdoors, though.
In my opinion, good sounding horns for a room are 60x40 CD with nearly square mouths and as much directivity down to the crossover frequency as possible.
Consider building your own midrange and bass horns following the same directivity criteria (big!). For sub-bass you can consider tapped horns for the sake of efficient compactness. Horns may easily cut the THD of your drivers by 10dB.
A bad HF compression driver won't have a flat frequency response and will need EQ. It may also have notches above 10k. Most CD horns also need very simple Eq to compensate rolloff (very smooth!!) above 5Khz.
A typical bad (medium sized) HF horn will probably have 3dB peaks and dips near the lower cutoff (if it was shrunk too much). It may also suffer from 6dB peaks and notches above 5Khz due to internal reflections. Non CD-horns will be progressively directive above 5Khz and progressively wide-dispersing below 5Khz. When such a non-CD horn is employed in a room (or any environment capable reflecting considerable amounts of off-axis sound back to you), the stuff radiated off-axis below 5Khz (particularly at 1hz-3Khz) will produce a quite honky unpleasant and fatiguing "ambient" effect (particularly at high volumes). Non-CD horns are fine outdoors, though.
In my opinion, good sounding horns for a room are 60x40 CD with nearly square mouths and as much directivity down to the crossover frequency as possible.
Consider building your own midrange and bass horns following the same directivity criteria (big!). For sub-bass you can consider tapped horns for the sake of efficient compactness. Horns may easily cut the THD of your drivers by 10dB.
What is an example of such a CD/Horn for HF?
At this stage my intention is to use the Audax PR170M0 midrange. (Would it be worth putting that is a horn by the way???) It starts to get quite directional after 2k, is that a suitable crossover point for the horn?
Are you saying that directivity at higher frequencies is actually desirable to reduce reflections? In that case would I be ok to cross the midrange over higher? I suppose the downside would be that the dispersion would probably not be matched with the horn at the crossover frequency?
Ok, thanks I'll keep that in mind - is that for a HF or mid?
I like the idea of that, but an a little uneasy because I have a sense that it would be very difficult to get right!!!
🙂 DB
It sounds like you want something like a JBL 4430 or a Gedlee summa. Altec model 19 is older and similar principle.broughd said:
Avoiding PA type horns that cost 10$ is a good idea. Youl note that the latest oblate spheroid types have a better sound.
The pattern will slowly narrow from the bass to about 90degrees at ~1khz for the 15" where the horn takes over - the patterns are very similar at the crossover point on purpose.
Up higher past 8khz it might well narrow slightly,that depends on the initial parts of the horn flare.
A constant directivity horn will require some equalisation in the top octaves. This is a small price to pay IMO.
B & C has a good reputation. Youl be paying top dollar for JBL most the time,look out for alternatives.
Perhaps an 18sound HF horn?
If you want to avoid the critical midrange,a conical midrange with a 6" could be possible
http://www.audioheritage.org/vbulletin/showthread.php?t=12126
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1981-05%20AES%20Preprint)%20-%20Monitor%20Loudspeaker%20Systems.pdf
http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1984-4430-35.htm
http://nfp.cba.utulsa.edu/bajaja/FreeAudioReviews/Reviews/AudioMeets/GPAF2005/ReviewGPAF2005.htm
http://nfp.cba.utulsa.edu/bajaja/FreeAudioReviews/Reviews/AudioMeets/GPAF2005/Images/EarlAndDuke.JPG
I think the audax sounds best crossed higher - 4-500 cycles. An MTM in that range with the little 18 sound elliptical horns and the little B&C mylar (i think it's called HE10) one inch compression driver crossed in at 2.5 to 3 kilo-cycles would be sweet.
Bass shouldn't be to difficult to 400 cycles. Maybe build a WWMTM with a good pair of 15's a side to bring up the efficiency
Bass shouldn't be to difficult to 400 cycles. Maybe build a WWMTM with a good pair of 15's a side to bring up the efficiency
Brett
"HF: BMS 4552 or Beyma CP380M or 18Sound ND1090 on 18Sound XT1086 flares. Testing will decide which I'll use and which will go into the PA's"
So how do you rate those three?
Cheers
"HF: BMS 4552 or Beyma CP380M or 18Sound ND1090 on 18Sound XT1086 flares. Testing will decide which I'll use and which will go into the PA's"
So how do you rate those three?
Cheers
I prefer the 380 so far over the ND1090, but that could simply be my bias against Ti diaphragms. The 4552's are on loan.otto88 said:
I have a copy of Soundeasy on the way and will do some more objective testing of these three, plus the Emilar EA175, Altec 807 and a couple of others.
I have 8 of them, and a friend just ordered 6. Great drivers.thadman said:
I got mine before the dollar went screwy and am happy to order more even at 50% increase (because of $A/$US exchange rates) for some projects further down the line. I also have a heap of Pro drivers from JBL etc to compare them with.
Edit: shipping certainly increases the price, but still makes them competitive against many new Pro drivers here, even at trade prices.
- Status
- Not open for further replies.