Hi everyone!
I'm building a pair of floorstanders with 2x Peerless 830875 and Seas H1189. I would like to use a simple 2 way series filter for them. But the problem is that I can not find information about these filters. I would like to be able to calculate them.
I have everything to measure the impedance curve and frequency response.
Can somebody help me with this?
I'm building a pair of floorstanders with 2x Peerless 830875 and Seas H1189. I would like to use a simple 2 way series filter for them. But the problem is that I can not find information about these filters. I would like to be able to calculate them.
I have everything to measure the impedance curve and frequency response.
Can somebody help me with this?
If you are trying to use a calculator, that won't be optimal. I recommend using an optimizer suite. Jeff Bagby's "Passive Crossover Designer" will do them. If you want some practical opinion and such, Tony Gee's site and Andy G's site will help you there:
Humble Homemade Hifi
ARGOS loudspeakers
Later,
Wolf
jbagby
Humble Homemade Hifi
ARGOS loudspeakers
Later,
Wolf
jbagby
Or you could ask Dave Rosgaard (daver) right here on DiyAudio. He's one of the best in the world.
Best Regards,
TerryO
I don't see any posts from someone with that user name talking about crossovers...but he does go by that on PE it seems.
Last edited:
Yes, but I seem to remember that the overall cost of the parts was less.
It's certainly true that a design intended for use with tube amplifiers can (should?) be designed differently, but then the initial choice of drivers may very well be influenced as well.
Best Regards,
TerryO
Last edited:
Without making any claims as to the relevance of any of my crossover designs (or my abilities), I will admit that I tend to use series almost exclusively. I suppose it started because of the challenge (it is a hobby for me, after all). But when the right combination is made, the resulting sound is something that I've developed a taste for. But then, those are my preferences.
The mathematics involved is not very nice for deriving “simple” formulas for series crossovers, so you probably won’t find anything other than what might get you in the ballpark. Here is about as good of a collection as any I’ve found for starting with Series crossovers. Andy G has a lot of valuable info.
ARGOS loudspeakers
ppfred has also pointed out some other good sources.
A simulation package along with good data acquisition equipment (impedance and frequency response) is probably the only sure way to develop a good, refined series crossover. Unless you are fortunate enough to make good guesses (educated or otherwise). Even with a simulation, more “tweaking” is typically required.
Under the right circumstances (compatible drivers, smooth responses, and proper alignment of the stars), a series crossover can be made with a nearly flat (resistive) impedance. As noted, this can be very desirable for tube amps.
Some drivers are not easy to work with. And especially for those, a series crossover may not be the best choice. In comparison, a parallel crossover might be much simpler to implement, or more economical for some of those drivers.
One of the difficulties in implementing a series crossover is that changing one component value can affect the response of more than just one driver. And certain elements in the circuit are more sensitive to variation than others. This is an issue with even a simple 2-way. I can vouch for the fact that a 3-way can be even more sensitive to component value changes.
Series crossovers are not always the best choice. They are not generally simple, but are often attainable with a good helping of perseverance. And as many have noted, the resulting sound can be very good. I think that's what most of us are after anyway.
The mathematics involved is not very nice for deriving “simple” formulas for series crossovers, so you probably won’t find anything other than what might get you in the ballpark. Here is about as good of a collection as any I’ve found for starting with Series crossovers. Andy G has a lot of valuable info.
ARGOS loudspeakers
ppfred has also pointed out some other good sources.
A simulation package along with good data acquisition equipment (impedance and frequency response) is probably the only sure way to develop a good, refined series crossover. Unless you are fortunate enough to make good guesses (educated or otherwise). Even with a simulation, more “tweaking” is typically required.
Under the right circumstances (compatible drivers, smooth responses, and proper alignment of the stars), a series crossover can be made with a nearly flat (resistive) impedance. As noted, this can be very desirable for tube amps.
Some drivers are not easy to work with. And especially for those, a series crossover may not be the best choice. In comparison, a parallel crossover might be much simpler to implement, or more economical for some of those drivers.
One of the difficulties in implementing a series crossover is that changing one component value can affect the response of more than just one driver. And certain elements in the circuit are more sensitive to variation than others. This is an issue with even a simple 2-way. I can vouch for the fact that a 3-way can be even more sensitive to component value changes.
Series crossovers are not always the best choice. They are not generally simple, but are often attainable with a good helping of perseverance. And as many have noted, the resulting sound can be very good. I think that's what most of us are after anyway.
For my Scanspeak drivers, I found the AR sxo work better than a simple series XO.
The Acoustic Reality Series Crossover, AR-SXO
The tweeter needs to be able to go low and it seems H1189 fits the bill.
Also better to get woofer with smooth high roll off.
For my experience, this also applies for the series XO.
You might want to give it a try if u don't push the speaker very hard and have spare parts around
The Acoustic Reality Series Crossover, AR-SXO
The tweeter needs to be able to go low and it seems H1189 fits the bill.
Also better to get woofer with smooth high roll off.
For my experience, this also applies for the series XO.
You might want to give it a try if u don't push the speaker very hard and have spare parts around
What exactly defines a "simple series crossover" as opposed to a complex series crossover, or for that matter, the Acoustic Reality Series Crossover?
Someone previously mentioned that Dave used "superfluous" items in his crossover. That seems to indicate that perhaps there is, or must be, an optimum (or fixed) number of essential elements and the results will be somewhat superior?
Just interested in what the experts have to say.
Best Regards,
TerryO
Last edited:
I think 'Simple Series' is derived from a first-order series xover, and is not really simple or really effective at times. It can work in a rare case if the drivers are appropriate. I've normally had to go steeper than that.
The AR-SXO is defined by keeping it a circuit 'that amplifiers can enjoy', and mainly placing the padding resistor next to the main cap in series. It's normally a 1st order on both drivers, but the woofer is advised to add a zobel and effectively make it a damped 12dB on the woofer. Mainly, the characterization is the placement of the padding resistor next to the cap.
And to Dave's modi:
Not saying that the results would be superior without some of the notches he employed in a simulation I observed, but I modeled it with a lot fewer parts and less cost and achieved a decent model with the same drivers for the NorCal development project over on PE. The main difference was he had a flatter impedance curve, and I had better acoustic phase relationships.
You should know as a lot of other builders, Terry, that there are many ways to skin a cat on xovers, and designers have their own modi. Why I said what I did was that it seems I got about the same results on paper with a lot fewer and less costly parts count. "Minimalism with regard to necessity" is how I design. You know something has to be there to improve some designs, and if that adds parts, so be it. Let's not just throw a ton of parts at a design and get the same results as something much less complex, as it isn't a very efficient way of doing it.
Just my 2c, and thanks for the compliment, winslow!
Later,
Wolf
The AR-SXO is defined by keeping it a circuit 'that amplifiers can enjoy', and mainly placing the padding resistor next to the main cap in series. It's normally a 1st order on both drivers, but the woofer is advised to add a zobel and effectively make it a damped 12dB on the woofer. Mainly, the characterization is the placement of the padding resistor next to the cap.
And to Dave's modi:
Not saying that the results would be superior without some of the notches he employed in a simulation I observed, but I modeled it with a lot fewer parts and less cost and achieved a decent model with the same drivers for the NorCal development project over on PE. The main difference was he had a flatter impedance curve, and I had better acoustic phase relationships.
You should know as a lot of other builders, Terry, that there are many ways to skin a cat on xovers, and designers have their own modi. Why I said what I did was that it seems I got about the same results on paper with a lot fewer and less costly parts count. "Minimalism with regard to necessity" is how I design. You know something has to be there to improve some designs, and if that adds parts, so be it. Let's not just throw a ton of parts at a design and get the same results as something much less complex, as it isn't a very efficient way of doing it.
Just my 2c, and thanks for the compliment, winslow!
Later,
Wolf
Just so you know- the SXO designer has been included in a later revision of PCD, and is combined in the same spreadsheet here:
jbagby
Later,
Wolf
http://www.diyaudio.com/forums/mark...woofer-1-silk-tweet-combo-16.html#post2337734 and the post 5 after
Thia currently has a simple series XO. 30uF + 3mH. Driver choice, driver placement, XO point, all chosen to make it easy for me.
So far i've just tuned the XO in by ear using parts i had handy, measuring will likely wait till the local monsoon season is over (any day now)
They do sound very musical.
dave
Hi everyone!
I'm building a pair of floorstanders with 2x Peerless 830875 and Seas H1189. I would like to use a simple 2 way series filter for them. But the problem is that I can not find information about these filters. I would like to be able to calculate them.
I have everything to measure the impedance curve and frequency response.
Can somebody help me with this?
Getting back to your driver choices. Overall, nice smooth responses. Those drivers have good reviews.
Peerless 830875
https://www.madisound.com/store/manuals/peerless/830875.pdf
Seas H1189
https://www.madisound.com/store/manuals/seas/H1189.pdf
A couple of things that you'll encounter.
How are you planning on arranging the (2) Peerless 830875 woofers? The reason I ask is that the Re = 6.4 ohms. If you put them in parallel, you'll have a 3.2 ohm load there (and a boost in SPL), which might be a concern depending on the amplifier. If you put them in series, their combined load will be 12.8 ohms. Would a single woofer be a consideration? The SPL looks like an excellent match (the tweeter can easily be padded down to the woofer level)
You can still use a series crossover either way, but the load of the 2 woofers vs. the 4.8 ohm tweeter needs to be accounted for. There are other things that can be done with the circuit as well.
The other thing I notice on the woofer is the breakup peak at 4k~5k. A 1st order series might require an additional notch. A higher order would roll the woofer off steeper and help alleviate that.
Just some things to think about.
Tweeter protection is definitely a concern. Andy G (Agos Speakers) comments (IIRC) that it's really not a big issue, as he's never had a problem with that on the 1st order series crossovers he's done. It seems to make sense, as the inductor (which bypasses the tweeter) is the path of least resistance for the (higher energy) lowpass signal.
What I try to do is make sure that the the tweeter is getting substantially less signal at Fs than in its operating region above that. Typically, I look for 10% - 25% of the current. (Maybe that's too conservative, but it is my habit.) Depending on where the crossover frequency is, and the lower rolloff slope, a higher order might be necessary.
I routinely make the tweeter branch 2nd order for that reason. That is accomplished by placing a cap in line with the tweeter. I guess that's improved protection too. Unfortunately it makes things more challenging for calculations, unless you have a mindless computer running hundreds of calculations per second.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.