My first go at designing a cross over

Everyone knows that idle hands are the playground for the devil, so, I went and did a thing. I am a complete and utter noob to any type of home audio. I have time on my hands (aka boredom) from time to time and have always wanted to do something a little different. In my early 20's, I was head over heels in car audio. Nothing hi-fi, just wanted to see if I could literally shake apart my vehicle and annoy other motorists. Well, some years have passed and the audio bug has reared its ugly head again. Although, this time, I am having a hankering for something more civilized and, from what I'm learning, in a whole other league all it's own. Home audio leaning towards hi-fi. I'm not striving for audiophile levels, getting too far off in the weeds like that would probably lead me to sleepless nights with ideas popping out of no where and it would bug me relentlessly till I either made a note of it, satisfied with that, or testing it on some simulator or program.

So, here we go. This is a list of the components I've chosen for the cabinets.

(1) Dayton Audio DC28F-8 1-1/8" Silk Dome Tweeter​

(2) Dayton Audio DC130B-4 5-1/4" Classic Woofer Speaker​

(2) Dayton Audio SD270A-88 10" DVC Subwoofer​

This list is just for a reference for the cross-over mumbo jumbo that has come together thus far. I know that there are better components out there, these are my chosen ones for this project.

Now for the brain blender. I have used the Xsim program to clank out a crossover for this Frankenstein. I am attaching pics of what I have brewed up thus far and for all to critique and tell me what I am doing wrong (aka suggestions, ideas, or I'm on the right path). Again, I'm not going for audiophile, this is a hobby not an addiction for me.

Thanks in advance for all the kinds words I know are coming my way.
crossoverschematic.jpg

crossoverschematic.jpg
totalfrequencyresponse.jpg
driverfrequencyresponse.jpg
 
diyAudio Moderator
Joined 2008
Paid Member
It's a shame you felt the need to use resistors feeding the woofers, it is more common to use them as the reference for the overall level. If you think they are too loud, I'd ask whether you have included baffle effects?
 
  • Like
Reactions: 1 user
It's a shame you felt the need to use resistors feeding the woofers, it is more common to use them as the reference for the overall level. If you think they are too loud, I'd ask whether you have included baffle effects?
I would always refrain from driving woofers with resistors. You stand the likelihood of uncontrolled/sloppy bass, and even 'boominess'.
Also the component values of C1 and L1 will give you a significant dip in your impedance.
 
Member
Joined 2017
Paid Member
If you would have asked in the beginning, I'm sure most people would have tried to steer you away from a 3-way to start out. But, I actually think you're driver choices are pretty good for a budget 3-way.

Adding to what others have said or at least hinted at....

You're woofer is 87.8 db sensitivity. You need to model the baffle step loss for the woofer and midrange. You should end up with a system sensitivity of about 82 db 1m/2.83v (6 db loss). You will NOT need to pad (have resistors) on the woofer.

The DC130B-4 has a sensitivity of 92.1 db. This will have some baffle step loss that needs to be modeled, but this will be padded down to the woofer level of about 82db. You could get the 8 ohm version and it would be fine.

The DC28F has sensitivity of 89db and will also be padded down to around 82db also.
 
Member
Joined 2017
Paid Member
@ridnroundlost

If you are trying to save money Meniscus is closing down, pretty much everything 30% off. They have 2 of the tweeters and midrange left as of today. They don't have the SD270 but do have two DC300-8 Dayton Classic 10" woofer. It would need a bigger box than the SD270 but has higher sensitivity and it's 8 ohm (if that matters).
 
My understanding is that resistors in series with the woofer are a fire hazard, so you need to avoid that in any XO. I'm pretty sure that there are 3 way designs out there using at least your tweeter and mid which might give you some clues. Some Google searching my prove useful.

Do you have measuring equipment?

Geoff
 
You're woofer is 87.8 db sensitivity. You need to model the baffle step loss for the woofer and midrange. You should end up with a system sensitivity of about 82 db 1m/2.83v (6 db loss). You will NOT need to pad (have resistors) on the woofer.

The DC130B-4 has a sensitivity of 92.1 db. This will have some baffle step loss that needs to be modeled, but this will be padded down to the woofer level of about 82db. You could get the 8 ohm version and it would be fine.
I would like to understand how you arrived at such high 'baffle step' losses, arriving at such a low sensitivity system quote.
This doesn't seem 'real world' to me:confused:
 
Let's see, you have the woofers in series. This means no gain. Subtract -6dB from their net sensitivity, and you are likely at about 82dB. Due to baffle step losses, there is a small peak at the baffle width wavelength, and then a loss of -6dB below that range due to the frequencies wrapping around the baffle and bouncing off the front wall and canceling. This is basic physics and acoustics and is normal for designs placed at 2-3' from boundaries. Closer spacing to walls reduces the loss to 3-4dB, and on or in wall virtually eliminates this loss.
 
Definitely avoid the resistors on woofers if possible. You can do it, providing

1/ The resistors are rated to handle the maximum power input, & preferably more, and
2/ You design the enclosure / baffle alignment from the outset using the artificially increased effective Q as the basis.

Assuming both of those, and you pay attention to heat-dissipation, you should be fine as far as that goes. But it shouldn't really be necessary either; as pointed out above, the LF units are rated as about 87.5dB, although some berk at Dayton has managed to rate them under 1m/w conditions despite them being a 4ohm driver; under normalised 1m/2.83v they're nominally 91.8dB. Wire a pair in series, & that's what you've got under IB conditions as you gain 3dB from the doubled cone area & lose the same amount from the halved current draw. On a real-world baffle, & in most real-world conditions, you'll lose 4dB - 6dB from step loss, so with the generous hat on, you'll have an LF sensitivity of about 87.8dB 1m/2.83v, maximum, and possibly 85.5dB 1m/2.83v if you've got the cabinets away from boundaries & the room is large, so you're not picking up much help from that quarter. So padding should not be required on the woofers, as the midrange units & tweeter are substantially higher.
 
Isn't it generally true that it is nonsensical to create a crossover like this because the cabinet and the baffle have a very large influence on the frequency response.
What conclusion can be drawn from a crossover recorded on an infinite baffle?
Could you see, for example, there is an increase or dip in the frequency response and there I could use the Bafflesteplos or e.g. a standing back wave of the housing or so?
So what conclusions would you draw from that?
 
I wouldn't call it nonsensical -providing you take it purely as, say, a learning exercise to understand the general effects of filters, realise that your levels are off (2 drivers are on 2.83v conditions, one on 1w), then something valuable comes out of it. If you in the process learn about baffle effects & the need to at least simulate those & relative driver positions, offsets etc. (preferably measure, but that's a separate skillset), then I'd say that's a number of positives to take away. What you do with those next...
 
Member
Joined 2017
Paid Member
But it shouldn't really be necessary either; as pointed out above, the LF units are rated as about 87.5dB, although some berk at Dayton has managed to rate them under 1m/w conditions despite them being a 4ohm driver
Where did you see this? The spec page at Parts Express shows them at 87.5db 2.83v/1m with the dual voice coils wired in parallel for a 4 ohm impedance.

Is there something in the TS parameters that makes you think that that is incorrect and really at 1w/1,m?
 
Where did you see this? The spec page at Parts Express shows them at 87.5db 2.83v/1m with the dual voice coils wired in parallel for a 4 ohm impedance.
https://www.parts-express.com/pedocs/specs/295-486-dayton-audio-sd270a-88-specifications-47085.pdf

Back-calculating from the spec. sheet values, its listed SPL is indeed 1m/w, with 2.83v coming in at 91.8dB.

Unusually, Dayton / PE don't seem to be doing themselves too many favours in general with this drive unit, since if you go to the SD270a-88 page on the titular Dayton Audio site & open the spec. sheet link, it in fact comes up with the pdf for the SD215A-88 (!) so as Baldrick would say, they seem to be 'at home to Mr Cockup'. ;)
 
Last edited:
Looks as though I have some homework ahead of me.

First off
It's a shame you felt the need to use resistors feeding the woofers, it is more common to use them as the reference for the overall level. If you think they are too loud, I'd ask whether you have included baffle effects?
The mids are 4 ohm wired in series so the crossover sees 8 ohms and needed the resistors to bring them down to the tweeter. The woofers are dual 8 ohm voicecoils wired in parallel/series so the crossover see 8 ohms also, and again, resistors to bring them down to the tweeter.

My understanding is that resistors in series with the woofer are a fire hazard, so you need to avoid that in any XO. I'm pretty sure that there are 3 way designs out there using at least your tweeter and mid which might give you some clues. Some Google searching my prove useful.
This statement lead me down another road. I took a look at the wattage dissappation and was grateful someone brought this to light. Because here is a chart of that. Thank you very much GeoffMillar
watts dissapated.jpg

So, each resistor in the mids would have seen over 80 watts, each!!! Don't think my homeowners insurance would have appreciated that, but, the fire department would have an eventful time of it.

So, my homework now is to dive in to understanding baffle step and probably choose another tweeter.

Notes: I have not built the cabinets for this project as of yet. Nor do I possess measuring equipment at this point in time, but, it is on my list of thing to get.

I really want to think everyone for your insights, suggestions, and just basically affirming that there is help here. THANK YOU.