Thanks for the explanation. I'm still at the learning stage with this, so if I see anything I don't know how to explain, I have to ask!
I had a go at modelling the 5" Tangband W1685 (http://lautsprechershop.de/pdf/tangband/w5_1685.pdf). It seems to have a huge motor on it, but it has dips at 400, 600 and 800 and 1100 Hz that don't look so bad on a 0-100 dB scale but the final results were not as smooth as the Seas and Peerless models I looked at, so I'll leave it. I also tried the SB17NRXC30-8, as a larger 6" driver, but this had dip at 700 Hz and a peak at 900 Hz - 1 KHz that was made worse when I tried to get any baffle step compensation and cross at over 1200 Hz, so I gave up on that one too. I looked at the larger TB 1721, but it seems to need 18 litres which is too big for rear speakers.
I'm still quite interested in the Peerless 830860 option, as I have seen these drivers used in some well documented builds (Peerless HDS PPB 830860, Elsinore Speakers DIY) and the price seems good.
The tweeters I have tried have not seemed to present any problems to integrate, so I have not explored the many options here. I might try to find out a bit more about the different possibilities for tweeter choice, but it seems that as they are so much more efficient, you're always able to smooth the output down to fairly flat in a way which is easier than for dealing with woofers, where you have to be efficient.
Me too.
Again, be careful with simulations. Adding series resistance to an overly efficient tweeter will exacerbate the FR at Fs. It typically creates a peak at Fs just like it would in a woofer. If you don't have to pad it too much it's no big deal and you can deal with it in crossover but if you have to pad it 10 db it becomes an issue real fast, especially if you have baffle diffraction adding to the peak at FS. Make sure you model and add the baffle and box effects to your simulation or you'll find yourself having to add components to the crossover to pull down a diffraction peak making it difficult to align phase. The goal is to use a second order electric to get BT2 on woof and LR4 on tweet but you might need to go third electric on tweet to make up for Z axis. Then relax the tweet or increase the woof slope to hit your crossover target and align phase. Also, from personal experience, make sure you add enough inductance on the woof to account for BSC. It should have a flat overall level between 200hz and 3K or if you want full BSC.
And again from experience keep in mind that the data from the MFG is not reliable without massaging. The FR response curve may be close but the enclosure and baffle will change it immensely. Most MFG measure on a large infinite baffle so LF does not match your enclosure. Look at my HT as an example. After my second baffle disaster I've decided to cut the front off the boxes again and move the woofer up and tweeter to the side an inch or so. I'm getting a huge peak at 1.7Khz that can be traced directly to a combination of the baffle width, tweeter fs and distance to woofer. I'll keep the same distance to the woofer but change the ratio of left to right placement and move the woofer to reduce the peak and make crossover much easier to design. I also discovered my measurements had the wrong phase data so the final crossover had serious phase issues. I've been using Omnimic for measurement but the phase data can't be used reliably without extracting "minimum phase" and for that I need a gated measurement. Omnimic's "blended" mode is not minimum phase although it's apparently close but it's not really appropriate for crossover design. I've decided to work out the issues I have with ARTA and my Dayton EMM6 mic and preamp and use it instead. I just bought a mic calibrator so I can properly calibrate SPL levels and I just found the cable I was looking for so I can measure the mic gain. ARTA calculates minimum phase and Z axis with the push of a button making the design process much easier. I also like speaker workshop for crossover design. It's a bit tricky placing components but it models very, very nicely.
My Mistakes, learn from them if you can:
The first incarnation below with the tweeter mounted on the top of the cab had very smooth FR. It's an unusual design but it works very well. The only real issue was the small cavity where the tweeter overlapped the woofer. I probably should have left it alone, but that's what experience is all about.
- Good points are nice close tweet/woof placement.
- Good tweeter placement for smooth FR
- Bad points, overlapping tweet/woof frame
An externally hosted image should be here but it was not working when we last tested it.
You can look at FR from the earlier baffle HERE
The second incarnation with tweeter mounted near the center of the baffle is intolerable. FR is so rough it's not worth trying to fix it. I'll cut the front off the current enclosure, cut a new pair of baffles again, flush mount the woofer and tweeter this time and make a usable speaker. Now that I know what's wrong it will be much easier to correct the issues. Offsetting the tweeter will make them a "stereo" pair and help improve off axis response and stereo image between them. I'm convinced this woofer is a decent woofer if I can get it paired with the tweeter in a usable way. It does need impedance correction to work properly but that's as simple as a 5.6cap and a 3 Ohm resistor. I'm also thinking of making another pair of these using the Vifa neo ring radiator. It's said to be very, very transparent. If it's better than this Vifa BC25 it will be an excellent match for a second pair of these. I'll finish these and give them to my Mother for her bedroom.
An externally hosted image should be here but it was not working when we last tested it.
Anyway, what I'm getting at is you'll be much more satisfied if you either purchase a pre-designed kit or use one of the existing designs that's had all the bugs worked out. If you really do want to continue on your own see if you can get someone nearby to measure the drivers on the box once it's finished. It will make all the difference in the world when you get to the crossover.
BTW,. where did you get phase information for the drivers?
Last edited:
There's an excellent inexpensive build on Tek Talk using the Dayton ND-105-4 and the Vifa BC25SC06 tweeter HERE, it's just about exactly what you're looking for and it's finished. It's called the "Lithium" project. That dude has some CLAMPS!
If you haven't found it yet check out this page, you'll love it!
https://sites.google.com/site/undefinition/simulated-measurements
https://sites.google.com/site/undefinition/simulated-measurements
There's an excellent inexpensive build on Tek Talk using the Dayton ND-105-4 and the Vifa BC25SC06 tweeter HERE, it's just about exactly what you're looking for and it's finished. It's called the "Lithium" project. That dude has some CLAMPS!
Thanks, that looks great! Love the clamps too.
I have been using Jeff Bagby's Response Modeller spreadsheet to account for the baffle effects (but not for the Seas, which as I have learned includes baffe step already!) and the baffle diffraction, and finally to extract the phase information from the frd file, since Tuxedocivic pointed out that I had not done it for the things I put up in the first post.
Well, after all this time I think I'm nearly finished. I had some equipment malfunctions and failures and a few other issues but it's very nearly complete now. I changed to the XT25SC90 ring tweeter in place of the BC25SC06. It's a better tweet for this project and doesn't have an "itch" to it on movie soundtracks like the BC25SC06. I'm saving the BC for another project. The semi finished enclosure looks like this:
Enclosure can be purchased flat pack from Parts Express as can all other parts. The cost is almost exactly the same as the original. Crossover will be slightly more due to a notch filter that had to be added to the woof to clean up a breakup mode. Both drivers are flush mounted in this version. This tweet is very clean and transparent. It's also extremely flat. I haven't heard a tweet this flat in a long time. Combined with the now cleaner mid range this is an excellent little HT satellite. I'll update with more as I listen more.
PDC models are below. Transfer function could use a little work, it's a tid bit ragged but impedance is OK, FR is relatively flat and I get a very deep V when I inverse tweeter. Phase tracks very nicely on horizontal axis beyond 20 deg and almost as good vertically to 20 deg.
Inverted Tweeter Sim:
You can load the files in PCD and play with it yourself if you want. Download all files zipped up H_E_R_E
Tom
An externally hosted image should be here but it was not working when we last tested it.
Enclosure can be purchased flat pack from Parts Express as can all other parts. The cost is almost exactly the same as the original. Crossover will be slightly more due to a notch filter that had to be added to the woof to clean up a breakup mode. Both drivers are flush mounted in this version. This tweet is very clean and transparent. It's also extremely flat. I haven't heard a tweet this flat in a long time. Combined with the now cleaner mid range this is an excellent little HT satellite. I'll update with more as I listen more.
PDC models are below. Transfer function could use a little work, it's a tid bit ragged but impedance is OK, FR is relatively flat and I get a very deep V when I inverse tweeter. Phase tracks very nicely on horizontal axis beyond 20 deg and almost as good vertically to 20 deg.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Inverted Tweeter Sim:
An externally hosted image should be here but it was not working when we last tested it.
You can load the files in PCD and play with it yourself if you want. Download all files zipped up H_E_R_E
Tom
That looks great! Thanks for posting. I'm going to be delayed on doing anything for a while as my old TV died (10 year old 26" LCD, so it wasn't premature) and I've shelled out on a Panasonic plasma, so the AV budget is wiped out out to Easter and beyond!
I had already ordered some of these ND91-8 - Dayton 3.5 inch aluminum cone full range driver - Europe Audio and one of these 2x15W @ 4 Ohm TA2024 Class-D Audio Amplifier Board Only 320-330 to make some little active full range speakers for the kids to use with their Nintendos, which I put mp3s on for them to play. I'm going to make them rounded and decorate them to look like R2D2 and R4 (the red one that isn't R2D2) with the speakers on the front. When these arrive I'll have something to play with at least!
You know, I had a similar idea. I bought 3 of these little 18 watt 2.1 amplifier boards on eBay. It requires either two separate 12v AC power supplies or a separate 24v-40v external AC supply with a center tap (called a "split" supply). It uses two TDA2030A amps and a separate TDA2030A for the sub. It has a separate bass filter NE5532 dual op amp chip for boost and cut. You can set the sub XO point from 35-150hz. It would make a nifty amp for an enterprising person with a spare couple full range drivers and maybe an ND-105 4" neo woofer. In 6.4 liters with a 1" x 4" port from PE lined with a little bit of denim insulation the box is tuned perfectly and gets down to below 50hz.
I'm waiting for capacitors to arrive to finish my HT project. I need a 47 mfd cap for the XO and I don't have one. I don't want to use electrolytic so I bought a few plastic caps. I haven't found a local source for crossover caps yet and I ordered the parts to make a crossover prototyping board so I'll just wait a few more days It's a line of dual spring connectors for each driver. Basically I mount +/- spring clips like these
An externally hosted image should be here but it was not working when we last tested it.
in two rows of 5 each. The last used clip on each row is output. One row for woof, one for tweet. I then bridge all the ground connectors on each leg together on the bottom of the board so I can simply clip in the series components I want, add the parallel values and snap in a jumper to the speaker. I can insert up to 4 series components for each driver and as many parallel as I can fit in the clips. This thing will make prototyping crossovers much easier. Now I need to settle on either a connector cup or binding posts for my speaker designs. I can get a 2" cup with dual connectors that will take the place of the finished cup for prototyping. Then I'll finally be set up for easy and quick prototyping. I have several projects in the works, enough to keep me busy for a while. I may even be able to make some money doing this if I can get my **** together and make some good speakers. In the end I think the VV51R is a better speaker than I thought it would be. It doesn't go down very low but it has as good, clear midrange as can be expected from a paper cone and should still sound quite good for vocals and effects. Once I get the sub moved to it's new home and properly EQ'd it will be even better A home theater, done right, should be BETTER than a $20 theater ticket. Now I need to get that 55" LCD up on my wall 
Sorry to hear about your TV. I bought an off-brand Chinese LCD panel about a year ago and I love it. $360 for 42" of 10,000:1 contract 1080P gore and glamor. The brand is SIGMAC. They have a 55" LED panel I'm looking at for under $700 that also does 1080P and has VGA inputs too. The VGA input looks better because the video card drives the monitor H/V so it can synch the monitor to nearly any resolution. It also runs DOT-TO-DOT mode so there is no interpolation and it doesn't run through any additional signal processing hardware so it's a "cleaner" signal. Best of all it has no "overscan". I notice contrast looks much better in VGA mode than in HDMI.
Cheers...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.