Jennice
You've pretty well got the effect but part of that is also boundary reinforcement. Think of the frequencies below baffle step as being ants. With it out from the wall some of these ants can creep behind the box and get lost (these ants are stupid) but if it's against the wall the ants can't get behind the box and have to go back to the front (as ants do when there is something in their path).
The only formula I use is:
Baffle Step F3=115/Baffle Width in metres
The only tools are my ears.
When I first started mucking around with BSC using a 0.5 driver after reading planet10's info, I used a very crude test. Got 1 speaker and placed it about 1.2m into the room and listened. Then I put the 2nd speaker behind the first (drivers facing the rear), disconnected the tweeter and listened to the bass and the lower mids. The difference was enough for me to build a test box which was taped onto the back of the speaker and trialled with different inductor values starting with a 1st order rolling off at the baffle step F3. This worked well, so I then altered the speaker by adding the 0.5 woofer, increasing the volume and retuning the port due to 2 woofers. More inductor tests until I was happy.
I also tried the other method by adding BSC to the crossover and couldn't get it right or to sound right (used a series xo and also a parallel xo).
The 0.5 woofer behind the box just worked better and gave a higher SPL on the speakers.
The 0.5 woofer worked for me but others have great success doing it in the crossover......... whatever does the job.
You've pretty well got the effect but part of that is also boundary reinforcement. Think of the frequencies below baffle step as being ants. With it out from the wall some of these ants can creep behind the box and get lost (these ants are stupid) but if it's against the wall the ants can't get behind the box and have to go back to the front (as ants do when there is something in their path).
The only formula I use is:
Baffle Step F3=115/Baffle Width in metres
The only tools are my ears.
When I first started mucking around with BSC using a 0.5 driver after reading planet10's info, I used a very crude test. Got 1 speaker and placed it about 1.2m into the room and listened. Then I put the 2nd speaker behind the first (drivers facing the rear), disconnected the tweeter and listened to the bass and the lower mids. The difference was enough for me to build a test box which was taped onto the back of the speaker and trialled with different inductor values starting with a 1st order rolling off at the baffle step F3. This worked well, so I then altered the speaker by adding the 0.5 woofer, increasing the volume and retuning the port due to 2 woofers. More inductor tests until I was happy.
I also tried the other method by adding BSC to the crossover and couldn't get it right or to sound right (used a series xo and also a parallel xo).
The 0.5 woofer behind the box just worked better and gave a higher SPL on the speakers.
The 0.5 woofer worked for me but others have great success doing it in the crossover......... whatever does the job.
rabbitz,
When using the 0.5 woofer for BSC, don't you get a nasty load on your amplifier?
You mention that it is placed at the rear of the speaker. Why at the rear, and what phase (compared to the ones in the front)?
I am considering a speaker with some tweeter (don't know which one, yet), and Peerless SDS164 or HDS 164.
As they'll be playing in a fairly large room (30m2 and rather loud at times), I am thinking of paralleling the two bass drivers, and get a 4 ohm load.
If I added yet another 0.5 driver, it would get nasty
...or would you prefer to have the second driver as the 0.5 unit to compensate for the baffle step?
Any good ideas?
Jennice
When using the 0.5 woofer for BSC, don't you get a nasty load on your amplifier?
You mention that it is placed at the rear of the speaker. Why at the rear, and what phase (compared to the ones in the front)?
I am considering a speaker with some tweeter (don't know which one, yet), and Peerless SDS164 or HDS 164.
As they'll be playing in a fairly large room (30m2 and rather loud at times), I am thinking of paralleling the two bass drivers, and get a 4 ohm load.
If I added yet another 0.5 driver, it would get nasty
...or would you prefer to have the second driver as the 0.5 unit to compensate for the baffle step?
Any good ideas?
Jennice
Jennice said:When using the 0.5 woofer for BSC, don't you get a nasty load on your amplifier?
You mention that it is placed at the rear of the speaker. Why at the rear, and what phase (compared to the ones in the front)?
I am considering a speaker with some tweeter (don't know which one, yet), and Peerless SDS164 or HDS 164.
As they'll be playing in a fairly large room (30m2 and rather loud at times), I am thinking of paralleling the two bass drivers, and get a 4 ohm load.
If I added yet another 0.5 driver, it would get nasty
...or would you prefer to have the second driver as the 0.5 unit to compensate for the baffle step?
Any good ideas?
Jennice
Hi Jennice
With my 0.5 woofer I get a nominal 4 ohm load as I only use the 2 woofers in a 2.5 way. It's positioned at the rear to do the BSC (see the above post of mine on the ants) and if it was on the front it would not be effective for this purpose.
As far as the phase goes, the phase shift is hidden by the box..... do a search as planet10 has a good explanation. Also see this from the Wiki
http://www.diyaudio.com/wiki/index.php?page=Baffle-step+Physics
There are those, myself included, prefer to use the second driver for BSC at the back of the box. Others tend to think it's a waste of a driver. It's up to the builder to experiment to see what suits the application.
What I have found however, is that the 0.5 woofer does not give you the theoretical boost in sensitivity when you parallel 2 drivers. You are supposed to get a 6dB increase but that assumes they are both covering the same frequencies but with the 0.5 only covering the BSC you only get the increase in the frequencies up to the 0.5 woofer roll off (575Hz in my case). So the sensitivity of my speaker remained the same as a 2 way and that was also confirmed by the tweeter padding remaining the same as a 2 way.
If you need SPL, you might have to run the 2 woofers in parallel at the front covering the same frequencies to increase the sensitivity.
I have a question about BSC that has just been looking for the right thread. I'm looking for a good first project. My technical education is pretty rudimentary. I want to start with a well documented project with high quality drivers. It's come down to a choice, I think, between Tony Gees DD8 MKII and John Krutkes upcoming design using the same two drivers. Aside from my interest in a comparison between the series and parallel approaches to the crossovers (anyone care to comment?) I notice that Tony G suggests that the L18 can be used with a range of cabinet volume from 12lt to 24lt.
As far as BSC is concerned, if I'd like to build the cabinets more to my aesthetic taste, what dimensions can I play with without altering the original xo? Can I get away with narrowing the baffle width even a little? Height? Depth?
As far as BSC is concerned, if I'd like to build the cabinets more to my aesthetic taste, what dimensions can I play with without altering the original xo? Can I get away with narrowing the baffle width even a little? Height? Depth?
BillFitzmaurice said:
Bill, you always have such good explanations, thanks. I think I understand this now. If my baffle step starts to rise at (say) 500 Hz, the first inductor in series with the speaker is chosen to start rolling off at this frequency, rather than the chosen x-over, correct? So the tranform function slowly reduces (at 6db/oct) from 500~2500 (x-over frequency), then 12, or 18, or whatever above 2500 Hz. But does the baffle response continue to rise or does it rise to a peak and level off or come back down? If it continues to rise, than this solution is ok, but if the baffle response levels out or comes down, than it would overcompensate. Next, do you need to do anything to the tweeter x-over? Does it need to be padded an additional amount over just getting the sensitivities to match?
Now, some questions on the 0.5 woofer solution. Is it necessary that the front and back speaker be exactly the same? If I am using a fairly small woofer (6.5"), I don't have much bass extension anyway, so could I use a smaller woofer on back, and only boost the frequencies from (say) 100 Hz to the baffle frequency? (just being cheap here, I don't want to buy two woofers if possible!). I am sure the same speaker is the best, but maybe a cheaper woofer on back would work as well. Say I use a Seas L18xx on front, and a Seas P18xxx or something on back? Doable? What happens if I decide to later move the speaker next to a wall?
Thanks, this is a great thread, a lot of good advice so far...
There's an interesting review in issue 34 of Hifi+ magazine . It's of the B&W 704 and 705 speakers. The 705 is a 2 way, the 704 is 2.5 way using the same mid and tweeter as the 705. The 0.5 way driver is front mounted.
The conclusion was that the 0.5 extra way added bass warmth and weight, but reduced overall coherence.
Mark
The conclusion was that the 0.5 extra way added bass warmth and weight, but reduced overall coherence.
Mark
Back 0.5 pros & cons
Back 0.5 woofer must be the same model. We need exact behaviour to accurately track BSC. Other model, or smaller work, but oddly.
Less second driver cost idea gets offset by complexity of different separate compartment since you cant align optimally, differing woofers in common box.
Pros: No loss of nominal single woofer spec sensitivity, transparency and cost gains due to no big coil use on front, exact compensation of step shape, less floor path cancelation due to 2 points relevant to floor path length and room dimensions, simple and accuarately correct solution to ill measuring gear equipped DIYer.
Cons: Second driver added cost, tougher impedance (it can be turned to benefit if wired in series provided the nominal sensitiviry can be high enough for the system), need spacing to back wall (a standmount needs anyway due to correct spatial reproduction considerations).
So here you are, some pros and a couple of not so cons.
Back 0.5 woofer must be the same model. We need exact behaviour to accurately track BSC. Other model, or smaller work, but oddly.
Less second driver cost idea gets offset by complexity of different separate compartment since you cant align optimally, differing woofers in common box.
Pros: No loss of nominal single woofer spec sensitivity, transparency and cost gains due to no big coil use on front, exact compensation of step shape, less floor path cancelation due to 2 points relevant to floor path length and room dimensions, simple and accuarately correct solution to ill measuring gear equipped DIYer.
Cons: Second driver added cost, tougher impedance (it can be turned to benefit if wired in series provided the nominal sensitiviry can be high enough for the system), need spacing to back wall (a standmount needs anyway due to correct spatial reproduction considerations).
So here you are, some pros and a couple of not so cons.
To go back to the very first question - understanding that the results will only be close-ish and not near what you can get if you measure, tweak, etc...
Another tool available on the FRD Consortium site will help make the connection. Actually, two tools.
Export the baffle response from the BDS tool for each driver.
The final tool is the Frequency Response Combiner - however, to use this you need FRD files with the exact same number of data points, which SPLView will do for you (open FRD, normalize over a fixed range [I do 20-20k, 540 data points], export).
To get baffle response equalized into this range, you'll need to add two data points at 20hz and 20000hz (which are 20 0 0 and 20000 6 0) since they're not covered in the export.
Import the response into the first column (raw response) in the FRC tool. Import the baffle step simulation into "baffle target" column, and set the next row in that column to -6 (dB). Set all other values to "none". Combine response.
You'll now see a graph of the driver response including baffle step. The top end should line up, the bottom should drop by ~6db.
Further down there's a "calculate minimum phase" - do this. Make sure it's calculating based on col7 (combined response).
Export the result and use in the PCD tool. You may need to click the "Save" button once before executing the export.
If you want to get an idea of different approaches and possible solutions to baffle step, this will get you there. It'll even get you in the ballpark for tweaking quite often.
C
Another tool available on the FRD Consortium site will help make the connection. Actually, two tools.
Export the baffle response from the BDS tool for each driver.
The final tool is the Frequency Response Combiner - however, to use this you need FRD files with the exact same number of data points, which SPLView will do for you (open FRD, normalize over a fixed range [I do 20-20k, 540 data points], export).
To get baffle response equalized into this range, you'll need to add two data points at 20hz and 20000hz (which are 20 0 0 and 20000 6 0) since they're not covered in the export.
Import the response into the first column (raw response) in the FRC tool. Import the baffle step simulation into "baffle target" column, and set the next row in that column to -6 (dB). Set all other values to "none". Combine response.
You'll now see a graph of the driver response including baffle step. The top end should line up, the bottom should drop by ~6db.
Further down there's a "calculate minimum phase" - do this. Make sure it's calculating based on col7 (combined response).
Export the result and use in the PCD tool. You may need to click the "Save" button once before executing the export.
If you want to get an idea of different approaches and possible solutions to baffle step, this will get you there. It'll even get you in the ballpark for tweaking quite often.
C
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.