Crossover help needed

[Richard10909]

Hi

I'm looking for some help or advice on buying crossover or alternatively building my own for a d.i.y speaker project. It needs to be 2-way and of high quality. The mid bass driver and tweeter I'm using is made by seas. I've included a pic of the crossover I was looking at for 38 pounds each and has a frequency crossover of 2.2 khz, however I've been told the ideal frequency for the build would be 1.8 khz. Would there be much difference between 1.8 and 2.2? The one in the picture is made by Dynavox and not sure if these are good quality.19187

Cheers for any help or discussion on this

[Richard10909]

This crossover is from a pmc speaker which has 1.8khz and would be ideal. Not sure on difference in quality between this and the Dynavox

[walpurgis]

Do you know the model numbers and specs of the proposed drive units?

What are the impedance and sensitivity for instance.

[Arkless Electronics]

I'm afraid there is rather more to it than that Richard. A crossover needs to be designed for a specific speaker system. To do it properly you will need access to a measuring microphone and associated electronics plus suitable software. One built with the cheapest components around but correctly designed for the drive units and cabinet will vastly outperform any generic crossover no matter what the parts quality is.

[fatmarley]

As Jez said, you would need to take frequency response and impedance measurements of the drivers in the enclosure and then import those measurements into crossover design software to do the job properly.

Another option would be to trace the frequency response / impedance using the manufacturers data sheets (I think SPL copy or SPL trace Is the name of the software) Once you've done that you would need to import the measurement Into software that would replicate the baffle step losses.

It may be better to search for an established design that uses those particular drivers or just sell them. If you're the kind of person that wins the lottery on a regular basis, then a generic crossover may work for you. If not, I wouldn't waste your money.

[Reffc]

I'm afraid there is rather more to it than that Richard. A crossover needs to be designed for a specific speaker system. To do it properly you will need access to a measuring microphone and associated electronics plus suitable software. One built with the cheapest components around but correctly designed for the drive units and cabinet will vastly outperform any generic crossover no matter what the parts quality is.


This.

Unfortunately you cannot just buy a crossover and expect it to work with any two way system. Each woofer has it's own unique (to that model) acoustic response with frequency for a given applied voltage and current. It also has it's own unique impedance profile with frequency.

Introducing a capacitor or inductor will have a different effect, or at least magnitude of effect, in the circuit depending on the impedance and acoustic profile of the drive unit, so each louspeaker needs to have it's own bespoke crossover designed to work such that

1) woofer and tweeter are matched for sensitivity;

2) that they are crossed over at the most appropriate point and ;

3) that they are phase correct or coherent at crossover.

None of these generic crossovers sold on ebay or wherever will achieve that.

You can approximate a crossover by using manufacturer's acoustic/impedance data and T&S paraneters, but even that wont be correct for your specific cabinet as those figures are usually derived from measurements taken of a sample placed within a wide baffle measured anechoically.

As Jez says, to do it properly yourself (anything else is guesswork, which is why all professionally competent speaker builders use measurement), you need access to a calibrated measurement mic, and software to measure both impedance profile of the raw drivers, acoustic response of the raw drivers, and to use pulse or chirp short burst signals (fast sweeps or pink noise) gated to exclude room effects. The data is all then imported into a simulator of some sort (freeware by Jeff Bagby is excellent if not as sophisticated as LspCAD but is complex so you do have to know what you're doing to use it), and you design your crossover in that, build a prototype, re-measure and fine tune. Not as straightforward as some think.

Many people do "do it by ear" but it is, for all intents and purposes, shooting in the dark, unless you have the experience to know with precision how the electrical parameters interact with the known speaker measurements (these are still needed). I do know of one major loudspeaker manufacturer (everyone on here knows them and many will have owned a pair of their speakers) whose chief engineer does exactly this, then leaves some clever junior techs to fine tune in software. However, he's been doing this over 35 years and is an electronics engineer who has vast experience of designing loudspeakers.

[struth]

Hi

I'm looking for some help or advice on buying crossover or alternatively building my own for a d.i.y speaker project. It needs to be 2-way and of high quality. The mid bass driver and tweeter I'm using is made by seas. I've included a pic of the crossover I was looking at for 38 pounds each and has a frequency crossover of 2.2 khz, however I've been told the ideal frequency for the build would be 1.8 khz. Would there be much difference between 1.8 and 2.2? The one in the picture is made by Dynavox and not sure if these are good quality.19187

Cheers for any help or discussion on this

Think it will be a bit of a lottery, but either may well work ok. How well, and whether you like result, is only going to be found by trying a few. May well be cheaper in long run, getting the drivers and box looked at by a pro to give you an idea as to whats needed.

[Richard10909]

I'm afraid there is rather more to it than that Richard. A crossover needs to be designed for a specific speaker system. To do it properly you will need access to a measuring microphone and associated electronics plus suitable software. One built with the cheapest components around but correctly designed for the drive units and cabinet will vastly outperform any generic crossover no matter what the parts quality is.




OK thanks everyone. I have the details here. Mid bass driver is seas Prestige H1217-08 CA18RLY. It is 8 ohms and 35-3000Hz range. Sensitivity is 90db. Power handling short term 250w.

Tweeter is 6 ohms impedance and 90.5db sensitivity. With 750Hz nominal free air resonance. Nominal rdc is 4.8 ohms and power rated handling of 90watts.

Cabinet size for the speakers will approximately be 42cm in height and 20cm width and a depth of 38cm.


I'm prepared to build independent crossovers on pcb and solder the components myself if need be.

[Richard10909]

Do you know the model numbers and specs of the proposed drive units?

What are the impedance and sensitivity for instance.


Listed the driver above and tweeter is SEAS 27TDNC/GW H1451-06 there is only 0.5db difference in sensitivity. If this is no good I will look at other drivers

[Arkless Electronics]

No amount of info on the drivers can make up for actually measuring things as so many variables come into play, as Paul outlined. Even the cabinet design can effect the crossover!

[fatmarley]

You could try these - LINK (http://www.troelsgravesen.dk/SEAS_CA18RLY-COMPACT.htm) - but you'd need to buy some new tweeters.

[Richard10909]

Ok well I have some level of electronics and engineering but would not be able to do as Paul suggested. Far to complex for me. Maybe I should just buy the speakers. I actually intended to replicate the speakers with the identical or very close to parts at only a cost of £400. The speakers to buy are well over £2000 so I thought not only would it save a lot of money it would be a cool project. I didn't realise the crossover would make such a huge difference and kind of defeats the whole idea. Alternatively if I make the box from same materials and to exact dimensions I could buy the crossover from the company themselves rather than build or buy the eBay one. The drivers are made by seas not from ebay

[Reffc]

There is a relatively economical option Richard:

Get the boxes sized properly for the woofers, assemble everything, and send or take one cabinet with drivers to a specialist for measurement and crossover design.

The set-up, measurement, file conversion, importing into a simulator and tweaking is around half a day's work at most, sometimes just a few hours depending on what's needed.

Get the specialist to design the crossover for you, and provide you with a circuit diagram and parts list and you can still build them yourself. It'll cost you between £100 and £200 realistically to have that done. Add the cost of your parts, and you're still within a £600 to £700 budget which is way cheaper than £2000!

[struth]

That was my suggestion Paul. Dunno how close he is to anyone doing this so it might need a courier. Be worth it to be right first time.

[walpurgis]

If you haven't bought the drive units, why not take a look through the offerings from IPL Acoustics and Falcon Acoustics? They may offer a DIY package that meets your needs.

http://www.iplacoustics.co.uk/

http://www.falconacoustics.co.uk/

[Richard10909]

Ok thanks for the help. I haven't brought the drivers yet, and it's great to see other people's ideas. I think it would be a good idea to do as suggested and start building the boxes. Fit the drivers and binding posts etc, leaving the crossover part to last which will be the most expensive bit no doubt. This way I can enjoy the practical side of things. Will do a little more research first before or go ahead with this. 😂

[Richard10909]

Also I'm looking to build standmount as I think floor standing is overkill for my living area. I will use these with the small dali 's I have with can go in the rear corners. Also we can spend so much money in hifi gear and room acoustics may alter everything

[Qwin]

Unfortunately everything Jez and Paul have stated is true, the chances of an off the shelf crossover working with your exact drivers and cabinet are less than slim.

Perhaps another approach would be to look at the many kits available, where they have done the design work for you.

This can get good results at a fraction of the cost of a shop bought product, as a major percentage of a speakers cost is the cabinet build.

A word of warning, not all kits are equal, so do your research and buy from a reputable company.
Seas and Scanspeak sell kits.

[system7]

What's the expression? Fools rush in where angels fear to tread? :D

There's actually a lot of 6" SEAS kit designs which have a good track record. Matt pointed out one.
http://www.troelsgravesen.dk/SEAS_CA18RLY-COMPACT.htm
http://www.seas.no/index.php?option=com_content&view=article&id=381:mimir-seas-diy-kits&catid=66:seas-diy-kits&Itemid=365
http://www.seas.no/index.php?option=com_content&view=article&id=380:idunn&catid=66:seas-diy-kits&Itemid=365

I wouldn't be too intimidated by crossover building. Wilmslow and Falcon have most of the bits, including circuit boards, Maplin have some cheapish 40W soldering irons.