I’m making my first attempt at building speakers. Here’s what I have in mind:
Small 2 way, Raven R1 ribbons and Scan Speak 15W/8530K01 mid-woofs.
Active crossover, Linkwitz-Riley 24dB to protect the ribbon.
Win-ISD graphs a 9.1 liter sealed for F3 around 60hz
Any problems so far?
Raven's already purchased and I'm just about to pull the trigger on the Scan Speaks....
Small 2 way, Raven R1 ribbons and Scan Speak 15W/8530K01 mid-woofs.
Active crossover, Linkwitz-Riley 24dB to protect the ribbon.
Win-ISD graphs a 9.1 liter sealed for F3 around 60hz
Any problems so far?
Raven's already purchased and I'm just about to pull the trigger on the Scan Speaks....
Promising project!
Hi herm,
Both of the two drivers You intend to use are truly excellent by it`s own IMO.
I don`t have own experience with that particular Scan-Speak model but for a 2-way system and combining with a Raven R1 this appears to me being one of the few promising choices (maybe also one of the Jordan`s full range drivers could be an alternative).
This really might become a very interesting project.
Only one thing I`d like to mention and You should keep in mind IMO:
Altough the spec. sheet says the Raven R1 can be used starting from upwards 2kHz (and the frequency response is reasonable well extented down into that range) my experience with ribbons of this size is to cross them as "high as possible" because of distortion issues. Unfortunately distortion graphs (at different drive levels) are hardly found in ribbons spec. sheets (maybe because they would look not so nice).
Although basicly a good idea, even a high slope x-over (as You want to use) might not completely prevent the ribbon to produce higher distortion (compared to a good dome tweeter) in his lower operating range when actually crossed at around 2kHz.
My first guess therefore would be not to cross below let`s say around 3,5kHz - 4kHz (or even somewhat higher if the Scan-Speak permits that). Probably this is near the upper edge of what the Scan-Speak is able but IMO You should make use of it to as much extend as possible.
As You intend to drive active, anyway experimentation with the x-over point (and slope - don`t limit Yourself on a predetermined filter type - experimentation & measurements combined with extended listening of different filter set-ups is announced - it might become a long journey) is quickly done without spending much money on many big & expensive passive x-over parts (that`s only one of the aspects why I love the active way so much).
Good luck & happy building!!
Hi herm,
Both of the two drivers You intend to use are truly excellent by it`s own IMO.
I don`t have own experience with that particular Scan-Speak model but for a 2-way system and combining with a Raven R1 this appears to me being one of the few promising choices (maybe also one of the Jordan`s full range drivers could be an alternative).
This really might become a very interesting project.
Only one thing I`d like to mention and You should keep in mind IMO:
Altough the spec. sheet says the Raven R1 can be used starting from upwards 2kHz (and the frequency response is reasonable well extented down into that range) my experience with ribbons of this size is to cross them as "high as possible" because of distortion issues. Unfortunately distortion graphs (at different drive levels) are hardly found in ribbons spec. sheets (maybe because they would look not so nice).
Although basicly a good idea, even a high slope x-over (as You want to use) might not completely prevent the ribbon to produce higher distortion (compared to a good dome tweeter) in his lower operating range when actually crossed at around 2kHz.
My first guess therefore would be not to cross below let`s say around 3,5kHz - 4kHz (or even somewhat higher if the Scan-Speak permits that). Probably this is near the upper edge of what the Scan-Speak is able but IMO You should make use of it to as much extend as possible.
As You intend to drive active, anyway experimentation with the x-over point (and slope - don`t limit Yourself on a predetermined filter type - experimentation & measurements combined with extended listening of different filter set-ups is announced - it might become a long journey) is quickly done without spending much money on many big & expensive passive x-over parts (that`s only one of the aspects why I love the active way so much).
Good luck & happy building!!
Thank you for your responses, Per, Joe & Cocolino.
Per – Thanks for the advice on matching. I’ll let you know how it turns out.
Joe – the first cabinet will be an ugly, square box. Once I have experimented with it, I will get a nice cabinet made like this:
http://mywebpages.comcast.net/kidder/Audio/TransLam/translam_alpha.htm
I was told that Bob at http://quicksitebuilder.cnet.com/wrnch2/ can provide this CNC work.
Cocolino – where do you live? I was stationed in Bayreuth (actually Bindlach) for 2.5 years. I would sure like to drink some Maisel’s hefewiezen again!
The reason I picked the Scan Speak is because it can go high enough to cross at 3.5 kHz or maybe even 4 kHz.
I will start designing the active cross soon.
Per – Thanks for the advice on matching. I’ll let you know how it turns out.
Joe – the first cabinet will be an ugly, square box. Once I have experimented with it, I will get a nice cabinet made like this:
http://mywebpages.comcast.net/kidder/Audio/TransLam/translam_alpha.htm
I was told that Bob at http://quicksitebuilder.cnet.com/wrnch2/ can provide this CNC work.
Cocolino – where do you live? I was stationed in Bayreuth (actually Bindlach) for 2.5 years. I would sure like to drink some Maisel’s hefewiezen again!
The reason I picked the Scan Speak is because it can go high enough to cross at 3.5 kHz or maybe even 4 kHz.
I will start designing the active cross soon.
OK, here's my first attempt at building an active cross, and
posting an image.
I am trying to build a block diagram first. Is there anything
I should add to this XO?
How about something to tame the rising response of the scan speak over 1 khz?
An RF filter like Linkwitz uses?
http://www.linkwitzlab.com/xo_eq.htm
Is the level adjust going to be enough for the difference in driver efficiencies?
Should I just put one level adjust (like +5dB) on the mid-woof?
Any comments appreciated.
posting an image.
An externally hosted image should be here but it was not working when we last tested it.
I am trying to build a block diagram first. Is there anything
I should add to this XO?
How about something to tame the rising response of the scan speak over 1 khz?
An RF filter like Linkwitz uses?
http://www.linkwitzlab.com/xo_eq.htm
Is the level adjust going to be enough for the difference in driver efficiencies?
Should I just put one level adjust (like +5dB) on the mid-woof?
Any comments appreciated.
Hi Herm,
I thought the RF filter was to do with Radio Frequency interference - I think you'll need the shelving low pass filter(also on the Linkwitz site) to tame the rise. (I'll be using this on my 15W/8530K00) mids - nearly the same speaker, but with the revelator tweeter. (x/0 at 3000)12Db/oct. ESP boards.
What are you doing about time alignment of the drivers? Linkwitz also has an op-amp based delay on his site..
If it's a 2-way are you going to be using a sub? - It may sound a little light otherwise..
I also considered the Raven, so I'll be very interested as to what you think of your speaks, once they're done. Keep us updated!
Oh, and good luck aswell!
Rob
I thought the RF filter was to do with Radio Frequency interference - I think you'll need the shelving low pass filter(also on the Linkwitz site) to tame the rise. (I'll be using this on my 15W/8530K00) mids - nearly the same speaker, but with the revelator tweeter. (x/0 at 3000)12Db/oct. ESP boards.
What are you doing about time alignment of the drivers? Linkwitz also has an op-amp based delay on his site..
If it's a 2-way are you going to be using a sub? - It may sound a little light otherwise..
I also considered the Raven, so I'll be very interested as to what you think of your speaks, once they're done. Keep us updated!
Oh, and good luck aswell!
Rob
Rob;
Do you think it is worth the trouble to add an RF filter? Maybe I will try without and add later.
I think the shelving lowpass is needed. When we get around to calculating values, let's compare notes.
Speaking of calculating, how much tweeter delay do I need for time alignment? I don't know how much is necessary (or if it's the mid that needs delay, for that matter!)
I will use a sub for now, but my long term goal is to add a woofer.
-herm
Do you think it is worth the trouble to add an RF filter? Maybe I will try without and add later.
I think the shelving lowpass is needed. When we get around to calculating values, let's compare notes.
Speaking of calculating, how much tweeter delay do I need for time alignment? I don't know how much is necessary (or if it's the mid that needs delay, for that matter!)
I will use a sub for now, but my long term goal is to add a woofer.
-herm
Hi Herm,
I'm not going to use an RF filter, but I can't advise as to whether you should. Is RF a problem in the US compared to the UK? I don't know... I'll bet someone who reads this does though!
""Speaking of calculating, how much tweeter delay do I need for time alignment? I don't know how much is necessary (or if it's the mid that needs delay, for that matter!)""
It should be the tweeter, but I don't know much about ribbons..
Hehehee
This subject is being discussed here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=12140
and here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=742&highlight=time+alignment
and here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6030&highlight=acoustic+centre
I'm waiting for the Joe D'Appolito book to turn up before I start properly, as I'm still a week or 2 from finishing my bass units..
The loudspeaker cookbook mentions a little about this, but not enough to be truly helpful..
Also, I'll be measuring my drivers after breaking them in for a bit (I'll use my bass cabs as speaker stands with my b+w 602's on top, and the x/o from my HT amp for now), and then calculating the x/o from these measurements.. Most people here don't have much faith in the published specs
Sorry I can't be more specific. Hopefully I'll have some measurements in the next 6 weeks or so (I'm about to move house so can't do too much at the moment.)
Cheers
Rob
I'm not going to use an RF filter, but I can't advise as to whether you should. Is RF a problem in the US compared to the UK? I don't know... I'll bet someone who reads this does though!
""Speaking of calculating, how much tweeter delay do I need for time alignment? I don't know how much is necessary (or if it's the mid that needs delay, for that matter!)""
It should be the tweeter, but I don't know much about ribbons..
Hehehee
This subject is being discussed here:http://www.diyaudio.com/forums/showthread.php?s=&threadid=12140
and here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=742&highlight=time+alignment
and here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6030&highlight=acoustic+centre
I'm waiting for the Joe D'Appolito book to turn up before I start properly, as I'm still a week or 2 from finishing my bass units..
The loudspeaker cookbook mentions a little about this, but not enough to be truly helpful..
Also, I'll be measuring my drivers after breaking them in for a bit (I'll use my bass cabs as speaker stands with my b+w 602's on top, and the x/o from my HT amp for now), and then calculating the x/o from these measurements.. Most people here don't have much faith in the published specs
Sorry I can't be more specific. Hopefully I'll have some measurements in the next 6 weeks or so (I'm about to move house so can't do too much at the moment.)
Cheers
Rob
Hi again herm,
I´m located about 40km south of Munich - a little far from Bayreuth. Once I have been there for a few days - a beautiful city ("the city of beers" - how they call it).
And of course Hefeweizen....aahhh
...always my favorite.Anyway, here we have Hefeweizen also - probably even better ones as Maisel`s
About Your questions for the x-over:
What You certainly will need is a shelving circuit to compensate for the baffle step drop towards lower frequencies. That`s the circuit specified EQ: half/full space transition on the Linkwitz site its link You provided. You can find a suitable schematic on the Linkwitz site elsewhere.
Taming the rising response of the Scan-Speak also might be a good idea (I`m not sure if is necessary - it all depends on the final values for the filters) as well as RF-Filter at the x-over input (simple RC network).
If I were You I would first breadboard the hole thing. When You use Op-Amps this would be not too much effort. Then while measuring or listening You can easily make changes and also see what else may be still necessary.
After You are sure that it should stay as it is (what may take a while) You can build it up nicer ( on PCB?).
Hmmm... +5dB means a gain circuit. Due to the baffle step loss (respectively the shelfing circuit which is compensating for this) and the big difference in sensivity of the two drivers - my rough guess is that You have to attenuate the Raven by about 10dB if not even somewhat more
so You`ll still have to attenuate the tweeter anyway.What about the following?:
+5dB for the Scans and additionally an adjustable +/- 2,5 db attenuator (which could also be replaced later with fixed resistors if desired).
-5dB fixed resistor attenuator for the Raven
If it turn out that the level is not enough to drive Your amps You could add another (moderate) gain circuit ahead of the x-over.
Depending on the input resistance of Your amps (or better impedance - Your amps might have such a RC RF-filter on its input too) , the cable length (capacitance) between x-over out and amps input and the values (resistance) of Your level adjust circuit(s) at the x-over outputs, an additional buffer at the output maybe is required.
Re: Promising project!
Herm & i had some private conversation wrt the Jordan JX150 for this speaker. In the end he felt more comfortable with the Scan-Speak (me, i'm sitting on a pr of JX150s)
Actually, that Bob is the guy from Creative Sound in Abbotsfor, who i was going to suggest until i followed the link and realized you were talking about the same guy
dave
cocolino said:
Herm & i had some private conversation wrt the Jordan JX150 for this speaker. In the end he felt more comfortable with the Scan-Speak (me, i'm sitting on a pr of JX150s
)herm said:
Actually, that Bob is the guy from Creative Sound in Abbotsfor, who i was going to suggest until i followed the link and realized you were talking about the same guy
dave
cocolino said:
Exactly the kind of task John Pomann's Active Crossover for DIY was developed for.
dave
OK, so here's round two of active filter building blocks.
Any comments? Am I leaving out any major component?
I'll do a bunch of reading this week and start to figure
out how to calculate some values.
Newbie question:
If the sensitivity of the speakers differs by 7 dB (just a guess)
should the level adjust in the XO be 7 dB?
Thanks for the links!!! great info....
Any comments? Am I leaving out any major component?
I'll do a bunch of reading this week and start to figure
out how to calculate some values.
An externally hosted image should be here but it was not working when we last tested it.
Newbie question:
If the sensitivity of the speakers differs by 7 dB (just a guess)
should the level adjust in the XO be 7 dB?
Thanks for the links!!! great info....
If I was building an active crossover
I would make a generic one that
works with any speaker system.
I would insert a level (potentiometer)
between the crossover output circuit
for each type of output (high, low, etc)
and output buffer to allow attenuation
down to zero from max output.
Other exotic ideas;
1. Input gain adjustment to match
the active crossover input to the
source music output. Perhaps something
like this.
reference left op-amp, ignore other
circuits.
http://dreadlordpk.50megs.com/cgi-bin/i/ElectronicStuff/ZapcoPX/ZapcoPX-3.jpg
2. Clipping indicators, one will sense
the input buffer output which will
allow you to adjust input gain
via LED indicator. One will sense
output buffer, etc. Need to
calibrate for specific op-amp.
3. Reed relay on output buffer.
Create a delay circuit that turns
on the relay after a few seconds
after you turn on the power so
any glitches are not sent to the
amplifiers (turn on pops, etc.),
the relay contacts "connect" the
output buffer to the output connectors. Also, place a resistor
network on the output of the reed
relay ... just in case ...
Examples;
reed relay w/ output resistors.
http://dreadlordpk.50megs.com/cgi-bin/i/ElectronicStuff/ZapcoPX/ZapcoPX-6.jpg
turn on delay
http://dreadlordpk.50megs.com/cgi-bin/i/ElectronicStuff/ZapcoPX/ZapcoPX-7.jpg
misc link;
http://sound.westhost.com/projects-3.htm
I would make a generic one that
works with any speaker system.
I would insert a level (potentiometer)
between the crossover output circuit
for each type of output (high, low, etc)
and output buffer to allow attenuation
down to zero from max output.
Other exotic ideas;
1. Input gain adjustment to match
the active crossover input to the
source music output. Perhaps something
like this.
reference left op-amp, ignore other
circuits.
http://dreadlordpk.50megs.com/cgi-bin/i/ElectronicStuff/ZapcoPX/ZapcoPX-3.jpg
2. Clipping indicators, one will sense
the input buffer output which will
allow you to adjust input gain
via LED indicator. One will sense
output buffer, etc. Need to
calibrate for specific op-amp.
3. Reed relay on output buffer.
Create a delay circuit that turns
on the relay after a few seconds
after you turn on the power so
any glitches are not sent to the
amplifiers (turn on pops, etc.),
the relay contacts "connect" the
output buffer to the output connectors. Also, place a resistor
network on the output of the reed
relay ... just in case ...
Examples;
reed relay w/ output resistors.
http://dreadlordpk.50megs.com/cgi-bin/i/ElectronicStuff/ZapcoPX/ZapcoPX-6.jpg
turn on delay
http://dreadlordpk.50megs.com/cgi-bin/i/ElectronicStuff/ZapcoPX/ZapcoPX-7.jpg
misc link;
http://sound.westhost.com/projects-3.htm
I could well be wrong, but I think the EQ circuit mentioned above on the Linkwitz site is to counter the differences between subs on the floor (Phoenix woofer) and mids in the air (Phoenix main panel).
It is not a baffle compensation circuit as it actually boosts the mids and not the bass.
Baffle step is not a problem with dipoles, hence no circuit on the site to compensate.
Steve
It is not a baffle compensation circuit as it actually boosts the mids and not the bass.
Baffle step is not a problem with dipoles, hence no circuit on the site to compensate.
Steve
That`s exactly how I thought it should be.
With this setup You should be prepared to get the thing running okay and to get out the full capabiltiy of this very fine drivers
The delay circuit Rob sugguested is a good idea but depending on Your listening position in relation to the location of the drivers it might turn out that You don`t need it finally - experimentation required.
You can find almost everything needed on the Linkwitz site.
I still believe the difference in sensivity will be rather towards about 10dB.
The fixed voltage divider (L-pad) in the tweeter branch attenutes 5dB.
The gain block in the mid/low branch amplifies by 5dB.
Now with this You`ve compensated for the (my) estimated 10 dB difference already. The +/- 2,5dB (or whatever) is only there to have some headroom playing around.
By the way: I overlooked something while suggesting +5 gain and +/- 2,5dB attenuation for the Scan-Speak:
+2,5dB means to increase the level (of course
) but an resistive voltage divider L-pad (or a pot, whatever You prefer) can only attenuate not amplify. Therefore I think better to calculate the Scans gain circuit for +7,5dB while the following +/- 2,5db level adjust has actually to be a 0dB to -5db attenuator (pot) when in the "middle" position (= -2,5dB) now exaxtly gives what`s desired IMO and this is what I`ve meant originally. Namely overall gain for the Scan of +5dB (when pot/divider in "middle" position) & from this +5dB now true +/-2,5dB adjustment range (overall adjustment range for itself = +2,5dB to +7,5dB) and overall relative (to the Raven as this is attenuated -5dB) adjustment range of +7,5dB to +12,5dB (+10db = pot/divider in the "middle" position) .
(Instead of a pot or a fixed resitor divider You could also use several DIP-switched resistor dividers - just an idea).
here again hopefully to better visualize what I mean:
Scan-Speak: +7,5dB (gain)
+ 0dB to -5dB (attenuation)
= +7,5dB to +2,5dB (gain) >>>
Raven: -5dB (attenuation) >>>
>>>>>>>>>>>>>>>>>
= 7,5dB to 12,5dB in total adjustable difference >>>
= +/- 2,5dB relative adjustable difference
Though this is very simple it`s a little hard for me to find the right wording to express myself
Hope it makes sense to You and does not sound too confusing.
So it seems we are on similar "wavelength" here
Although not down from zero (not really required IMO) to max. out, the variable adjustment is there already (+/-2,5dB). No need for another adjustment in a two-way system - if one branch is adjustable it should be enough. Moreover, as due the big difference in sensivity of the two drivers we have quite some attenuation already (for the Raven) we wouldn`t really want to introduce even more attenuation (and to have to add a gain circuit to compensate for this in turn....).
I mentioned in my second post that indeed a moderate gain circuit ahead of the x-over-input might be required. This depends wether or not "herm" has a preamp already which he want to use and from the input sensivity of his power-amps.
My guess is that although the Scan-Speak has comparatively low sensivity - as there is that 7,5dB gain circuit (5dB in reality with the +/-2,5dB adjustment in "middle" position) the max. avaiable voltage should be still sufficient to drive the amps (unless maybe some "exotics" which need more than 1Veff. or so). In the "worst" case an additional x2 (6dB) gain circuit should be enough.
I wouldn`t use another gain circuit if not necessary.
Yes this makes a good point IMO but depends where the x-over is located finally. When it is near the power-amps and is switched on together with them and when the power-amps. have such a delay circuit already (which is likely) then it would not be necessary.
Only when the amps and x-over are seperated and switched on individually then the relays are definitely a very good idea.
Hi Steve,
I have to admit I haven`t read (still not) the whole article. I just took a look at this block diagram so I don`t know for what particular task in the Linkwitz project this EQ-circuit was intended .
Anyway, a half/full space transition EQ-circuit is exactly what will be required here. You can name it also Baffle-Step EQ-circuit if You prefer but it`s the same thing finally.
Wow, my brain is spinning with all the possibilities!
Let's tackle these one at a time ...
What I was using the shelving low pass for was to
compensate for the scan speak's rise above 1 kHz
(see image)
If I am going to cross at about 3 kHz anyway, this leaves
just a little hump centered about 2 kHz. Should I use
just a notch filter -3db or so, centered at that freq?
I am very tempted to do nothing at first, and see how it
works out.
Let's tackle these one at a time ...
What I was using the shelving low pass for was to
compensate for the scan speak's rise above 1 kHz
(see image)
If I am going to cross at about 3 kHz anyway, this leaves
just a little hump centered about 2 kHz. Should I use
just a notch filter -3db or so, centered at that freq?
I am very tempted to do nothing at first, and see how it
works out.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.