need help first time speaker desgin  

Page 1 / 2
  RSS

Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
27/07/2019 11:29 am  

hey guys this is my first time designing a crossover. i probably didnt pick the easiest drivers to design a crossover for but i like to learn things the hard way. the drivers im using are the dayton audio amt2-4 and dayton audio dsa-175. ive got the frequency graph looking the best that i can do and im not sure if it is actually good or not. if someone could help me out that would be great. i also dont know where to place the drivers in the box or if that matters and i dont know much about time and phase.

baker schematic
baker impedence
baker frequency response
baker box

thanks chad


Quote
Douglas Blake
(@douglas-blake)
Estimable Member Registered
Joined: 4 months ago
Posts: 247
28/07/2019 10:35 am  

I'm not the resident speaker expert, I was holding back, hoping one of them would have chimed in by now.

The computer graphs look good.  But do keep in mind they are just a starting point. It is likely the final crossover, after acoustic testing, will be something quite different. 

Also keep in mind that standard values of parts will change it quite a bit. For example it's very unlikely you will find an 8.5 uf capacitor... you'll probably have a choice between 8.2 and 10 which will affect the result.

Here is a list of standard values ... 

For the box, look at the way most speaker cabinets are set up. There is a reason for that... it works.

I've always favoured the sealed box design, to my ear (and admittedly limited testing) this has always given the most pleasing result although it can be a tad bass shy in some cases.  


ReplyQuote

123Toid
(@123toid)
Illustrious Member Admin
Joined: 2 years ago
Posts: 462
28/07/2019 10:38 am  

@toad

First of all, pretty nice try!  We need a little more information before we can say if it is good or not.  First, are these your measurements or the ones off PE's website?  Second, can you post a picture of the crossover region?  You have one of the final response, but you can't see what the individual responses or slopes are.


ReplyQuote
Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
28/07/2019 8:40 pm  

@douglas-blake

  • Thanks for the link. I forgot about checking the values. 

ReplyQuote

Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
28/07/2019 8:51 pm  

@123toid

No its just PE's measurements. I haven't bought anything yet cuz I was to scared that I wouldn't be able to design a crossover waste money on buying everything. What do you mean by crossover region. Im gonna be working out of town this week so I won't be able to post a pic until next weekend. I'll probably be sitting in my hotel room rewatching your youtube videos lol. Thanks for the help


123Toid liked
ReplyQuote
123Toid
(@123toid)
Illustrious Member Admin
Joined: 2 years ago
Posts: 462
28/07/2019 10:04 pm  

@toad

By crossover region I mean to add the curves for S1 and S2 along with the phase.  Instructions in the pictures below.

Step1
Step2
Step3

If these are PE's files then you are a little off.  First it doesn't look like there is any compensation for baffle step loss.  So your high end will be about 6db higher than your low end.  I am guessing you do not have measurement equipment?  If that is the case, we can't ever guarantee perfect or good results, but there are some things we can do to help us get the best we can without those tools.  The first is determine where your baffle step loss approximately will be.  Assuming about an 8" wide cabinet the F3 should be about 570hz and slowly tapering off from there.  So everything really below 500hz (approximately) should be 6db higher on your graph.

We will also want to make an approximate guess on Z-offset.  One we have those, we can start working on the crossover.  But you need those first.  

Of course, the best thing to do is have measurement equipment.  Or find someone that lives close to you, were you can use some for some measurements.  If you live close to central PA, I would be willing to help you out.  If not, maybe ask in the lounge.  Here are my favorite measurement equipment:

 


ReplyQuote

123Toid
(@123toid)
Illustrious Member Admin
Joined: 2 years ago
Posts: 462
29/07/2019 11:19 am  

@toad

After looking a little more in depth at those drivers chosen, I am not sure they are the best fit.  Take a look at the 6.5" response.  THe off axis really starts to separate around 2.5khz, but the response is already on a downward slope.  Seeing as the Tweeter shouldn't be crossed over any lower than 2.5khz, this may become problematic.  First from a frequency response and second from off-axis performance around the crossover region.  These drivers together might pose more of an issue than you would think.  It might be worth switching to a little smaller woofer.  Maybe a 5.5"

 

response

 


ReplyQuote
Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
01/08/2019 8:05 am  

@123toid

Thanks for taking the time to look into this. Ill take a look at some drivers this weekend. I can get the same woofer in 5". I haven't been able to watch those videos you sent me yet due to shitty hotel internet. 


ReplyQuote

Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
07/08/2019 8:27 am  

@123toid

hey toid

i was looking at different drivers and most drivers looked very similar to  me but i think the dayton audio ds135-8 5" is looking better than my original choice or the dayton audio tcp115-4, which is a 4" driver so i would probably want to use 2 of them since i have a large room. im kinda sticking with dayton audio just cuz its easier for me to get the graphs for them. if you can let me know what you think of them that would be great. thanks.


ReplyQuote
123Toid
(@123toid)
Illustrious Member Admin
Joined: 2 years ago
Posts: 462
07/08/2019 9:04 pm  

@toad

I have used the TCp115-4 in multiple build and will definitely give you a nice low end if that is what you are looking for.  Otherwise, you could also check out the RS125-8.  That is a really good driver, that I recently designed some rear speakers with Impulse Audio.  The mid is really nice. But it doesn't get real low.  However if you plan to cross it over to a sub, it is pretty perfect for that.  


ReplyQuote

bjaurelio
(@bjaurelio)
Eminent Member Registered
Joined: 1 month ago
Posts: 49
09/08/2019 2:00 pm  

I've read through the thread, and @123Toid is correct that you really need measurements of your drivers to start designing a crossover. Right now you're in the initial phase of picking drivers and deciding on the configuration. The first thing you want is to ensure there' sufficient overlap in usable frequency from each driver so both are smooth through the crossover region. With a ribbon on top, you're going to have a higher crossover. That one in particular has an FS of 2,300 Hz. The usual rule of thumb is a crossover point should be at least 2x the FS of the tweeter. Your crossover will likely be in the 4.5-5k region. As Toid also mentioned, you want a woofer that matches the directivity of the tweeter around the crossover, which means you need a smaller woofer with matched directivity along with cone breakups above that level where they will be lower in level. If you're set on the AMT, a 4" is probably the largest you want to go on the woofer for directivity. Given the cone breakup concerns, you probably want a full range such as one of the Dayton RS100 variants or a Tang Band W4.

If the tweeter doesn't have to be a ribbon, you can go with a 5" or 6" woofer of choice matched to a standard 1" dome much easier.

Second step after you have your drivers selected is to model the enclosure. You can start with published T/S specs and WinISD or similar software. That certainly gets you really close to the box size and response you can expect from the woofer. Once you buy the drivers, you will want to measure yours. Two options are to build a jig and use ARTA or REW (cheaper but has sound card requirements of stereo input and requires soldering the jig) or DATS. Depending on the manufacturer, some brands can vary quite a bit in the measured vs published T/S parameters. You want to make sure your box design is right before building. Also, if you go ported, you want to be able to run impedance sweeps in box to check port tuning.

After building the first box and mounting your tweeter and woofer, you're ready to measure. Jeff Bagby has a great PDF on measurements. Basically, you want to take gated measurements (to eliminate room reflection from the measurement) on the tweeter axis about 30" away. Then measure the woofer followed by woofer and tweeter. When you import your tweeter and woofer measurements into the crossover design software, you will enter any x-y offsets based on your baffle and then change the z offset until it looks like your measurement of both drivers together. I'm leaving out some extra details about the woofer measurement and splicing in a nearfield measurement since gated measurements will only be useful so low in frequency. 

Just like everything, there's a lot of tedious prep work before you can start plugging in component values to simulate a crossover design. However, if you do the prep work, you're going to have very good correlation from the simulation to actual crossover build. Once complete, it's a very rewarding process knowing all the work that went into making it sound right.


ReplyQuote
Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
18/08/2019 12:50 pm  

I was wanting to stick with an amt tweeter because i have speakers with them and i really like them but i think having a bigger woofer is more important. its been a long process looking for different drivers but i thinking of going with the dayton audio rst28f-4 tweeter. i just noticed that the @123toid and impulse audio used them in a build recently. as for the woofer i was looking at the dayton audio ds175-8. im not necessarily set on that driver. i would like to stick with that tweet since i dont plan on make grills and i dont want my 2 year old kid touching the dome tweeter. i quickly made something up in xsim and this is what i came up with. i understand that im gonna need to get some equipment to measure the devices. i feel like im getting a better understanding of designing a crossover from talking with you guys.

b schematic
b impedance
b frequency

 

 


ReplyQuote

123Toid
(@123toid)
Illustrious Member Admin
Joined: 2 years ago
Posts: 462
19/08/2019 2:30 am  

@toad

I completely get that.  I have no issues with anyone using the FRD and ZMA files to try their hand at crossover design.  It really helps, especially before you get measurement equipment and or decide to get measurement equipment.  We just want to make sure we take the proper precautions, such as trying to figure out what frequency your baffle step will start at depending on box size.  Use this to help determine the frequency in which you want to be about 6db down.

Making an educated guess at Z offset

It won't be perfect.  We are assuming the manufacturers measurements are correct, which, well is often not the case.  But as long as you are aware of that, definitely venture forward.  

Now if you don't mind a critique with your current design, you are way out of phase.  You can tell that by your crossover point and final frequency response.  If you notice, the crossover point and final frequency response are basically all around 85db.  Your crossover point should be about 6db down from your final response (at the crossover point).  So your crossover point in this illustration should be closer to 79db.  This means you are out of phase. If you switch the polarity of the tweeter you will probably be in phase, but will have a massive jump in your frequency response. 

Before you do anymore crossover simulations, I would first guesstimate your baffle-step and z-offset and let us know what you think they could be.  I'll let you know if it seems way off.  Then we can determine phase.  But since the Z offset directly affect phase, it'll be hard to do that without a guesstimate at z-offset. 


ReplyQuote
Toad
 Toad
(@toad)
Active Member Registered
Joined: 2 months ago
Posts: 10
19/08/2019 9:56 pm  

@123toid

i guess i miss understood your video about the baffle step. i thought it was just for full range drivers. yeah ill try and figure the baffle step and the phase. 


ReplyQuote

123Toid
(@123toid)
Illustrious Member Admin
Joined: 2 years ago
Posts: 462
19/08/2019 11:30 pm  

@toad

No worries.  That is why I started this forum. That way everyone could learn at their own pace.  It is really rewarding when you do it, versus someone just telling you how to do it.  You don't really learn that way.  It looks like the DS175-8 is about 87.7db sensitive, which is what your graph is showing.  WIth that in mind, you should be about 6db at the baffle step frequency.  That means everything from the baffle-step on should hover around 82db on your graph. 


ReplyQuote
Page 1 / 2

Share: