soundbyte Posted March 3, 2016 Posted March 3, 2016 Thanks again henry, So i had a play with the Unibox spreadsheet. First time using it, i found a tutorial which showed how to input spes etc, said to design the speaker for 1/2 max rms. So the graphs are based on that. Im a little bit unsure how to optomize box size. The port seems to get quite long, depending on the box size. so i seem to have to go a little bigger with the box to get the port length time. Does it look any good? Also had a play in winisd and attached a diagram of that too. Any feedback on those would also be much appreciated! Thanks again If your driver is this one http://www.sbacoustics.com/index.php/products/midwoofers/nrx/6-sb17nrxc35-4/ then you can use the data in the specs to model your box, without measuring (not recommended) but you will be close with your simulated box in Unibox. Remember to factor in the space taken up with the rear of both the drivers (especially if the woofer magnet is large like this one has) and add that to your box size. Boxes can quite easily be made using MDF (USE a MASK when cutting) or ply (preferred but still use a mask). Use as suggested by @@Sub Sonic the standard pipe sizes available when doing the calculations, measure them as they are not necessarily the diameter that is spoken about eg 90mm drainage pipe is 90OD and 86ID. Crossovers are easy to try from online calculators, see here http://www.erseaudio.com/CrossoverCalculators that should get you close but there are lots of choices it will be up to you to choose what you think is best once built and tested. Crossovers parts are occasionally on SNA just keep a look out and try the Wanted to Buy section. It is a very rewarding task and I hope that you have fun, not worrying too much about making mistakes, but learn from them and I suggest that you keep notes of what you are doing so you know where you are and what you need to do. Best of luck, Russell.
Neo__04 Posted March 3, 2016 Author Posted March 3, 2016 Thanks for the info. Dats v2 has been purchased and on its way. I'll get some measurements as soon as it turns up and post some graphs to make sure im still on track. 1
Sub Sonic Posted March 3, 2016 Posted March 3, 2016 Good decision :-) I used to use an analogue meter to do impedance tests years ago, but the DATS V2 is light years ahead. Easy to use, and can do impedance, resistance, capacitance and inductance and also driver measurements. Very handy. You will definitely need to follow the setup instructions to configure your PC (not hard to do) otherwise it will not read or show correctly.
Neo__04 Posted March 4, 2016 Author Posted March 4, 2016 Awesome, a good little investment i have made then Another question while im going. Im going to buy/make an amp to suit the speakers i will be building. The woofers are 60w rms, and the tweeters are 100w rms. Should i be looking to buy a 50w rms per channel amp, so i can never run it hard enough to cause a problem? Or should i get something bigger like 100w per channel to allow for future upgrades and just keep the volume on the appropriate level. Something like this ok? http://store3.sure-electronics.com/2-x-50-watt-class-d-audio-amplifier-tda7492 http://store3.sure-electronics.com/aa-as32186-1706 Thanks again!
Sub Sonic Posted March 4, 2016 Posted March 4, 2016 I've never used those modules, but others have. Bigger is nearly always better when it comes to amplifier power, speakers are much more likely to be damaged by a too-small amp being over driven. Cheers, SS
davewantsmoore Posted March 7, 2016 Posted March 7, 2016 DATS V2 is light years ahead. I digress but, one thing which DATS can't do, is to measure impedance at multiple drive levels. IME it's quite shocking to most people when the see the impedance curve change at different input powers. 1
davewantsmoore Posted March 7, 2016 Posted March 7, 2016 The woofers are 60w rms, and the tweetersare 100w rms. If you look up the standard they used to calculate these powers, it's with 6db crest factor, and constant power for 100 hours (without damage). The drivers will be able to take more short term power than this. Your tweeter will have a resistor in the passive XO to match the sensitivities of the drivers (as it is nearly twice as sensitive as the woofer, before any baffle loss compensation). A 100w amp, or even more would be ok for these ..... but really the limit for how much power you should apply is going to come from the woofer excursion. If you design a vented enclosure, or if you use EQ to set the closed box response to give you X amount of bass ..... then at some power level, the woofer will reach it's maximum excursion (this will happen before the thermal power limits are reached).
Sub Sonic Posted March 7, 2016 Posted March 7, 2016 I digress but, one thing which DATS can't do, is to measure impedance at multiple drive levels. IME it's quite shocking to most people when the see the impedance curve change at different input powers. I'm interested now! Do you have easy access to any examples? Regards, SS
Neo__04 Posted March 7, 2016 Author Posted March 7, 2016 (edited) Ok, well i've measure all the speakers. Run 5 sets of test of each speaker to average the results. And saved graphs of the impedence and phase curves. Now in regards to your comments Davewantsmoore with the impedance curves, is there a better way that should be done? Without crazy expensive test gear? Whats my next step from here? Edit: Data copy and pasted very badly. Ill attach it in a jpeg. Graphs below -------------- Edited March 7, 2016 by Neo__04
davewantsmoore Posted March 7, 2016 Posted March 7, 2016 I'm interested now! Do you have easy access to any examples? Of the impedance of a driver changing with different drive levels?.... not easy access to my own data showing it (but I could take some to show a few examples, if you wanted - although am a bit pressed for time - I've just become a Dad) Here's one I pinched from the web. It's a vented box, and you can see as the drive level increases, it starts to turn into a sealed box. (Ignore the red curve) ... but remember that as you're listening to music, the drive level is constantly/rapidly changing (it's the music), and so the impedance is fluttering about between all of these curves. 2
davewantsmoore Posted March 7, 2016 Posted March 7, 2016 Now in regards to your comments Davewantsmoore with the impedance curves, is there a better way that should be done? At this stage, don't worry about it at all. There's a million other things to concentrate on. That's why I said "I digress".
Sub Sonic Posted March 7, 2016 Posted March 7, 2016 @@davewantsmoore thanks for the info, you learn something new every day :-)
Neo__04 Posted March 7, 2016 Author Posted March 7, 2016 Ok, So I've had a play in unibox. Redesigned a ported enclosure based on the the new speker driver measurements from Dats. I worked on the speaker specs with the higher fs. Specs & graphs attached. Can you advise if this looks ok? And again, Thanks everyone for your guidance. Learning a lot!
Neo__04 Posted March 10, 2016 Author Posted March 10, 2016 Any thoughts on those specs? About to start the box build and wanna make sure the specs look good Thanks heaps
henry218 Posted March 10, 2016 Posted March 10, 2016 if the purpose is to know more about speaker building, i would suggest to build as big as you can based on your simulation and adjust accordingly to know if you have reached your target as it is a lot easier to make the box smaller (but putting extra panel or fill it with solid object such as bricks... ) than to make it bigger at later time. 1
Sub Sonic Posted March 10, 2016 Posted March 10, 2016 The model up above suggests a 20 litre vented enclosure, how does it model in a 20 litre sealed enclosure?
Sub Sonic Posted March 10, 2016 Posted March 10, 2016 I'd build the slightly larger box at 25.7 litres for sealed, and then you will have the best of both worlds, a decent closed box response, but able to reduce internal volume and add ports as desired.
thoglette Posted March 10, 2016 Posted March 10, 2016 (edited) IMHO you;ll probably be happier (in the long run) with the sound of a sealed box. My brain is not working well enough to do the maths in my head v-v 25l but a little bigger is usually better than too small. The really fun part is now getting the two drivers to work together. And has anyone mentioned baffle step or edge diffraction yet? (I'll through a complete spanner in your works by suggesting you read the tales of a DIY system from 1993) Edited March 10, 2016 by thoglette 1
Neo__04 Posted March 10, 2016 Author Posted March 10, 2016 Well i have sort of been thinking about sealed, Origianlly i was just going to go ported so the bass would be a little nicer. But i think i'll throw a sub in with the speakers, so i might go along the sealed line. Havent heard of baffle step or edge diffraction, but i'll start doing my homework now. Another quick question. Say i run a minidsp for the crossover to start with and play with that till im happy with it, then i can learn to design the equivelant passive crossover. If you make a sealed box, then do your testing with that, how do you then clean up the job and mount the crossover inside? Hope its not a stupid question but do people just screw the back panel on, till they are happy with their tuning, build the crossover, mount it inside, then glue the back panel on? Thanks agian guys, so much help here. Its greatly appreciated
Sub Sonic Posted March 10, 2016 Posted March 10, 2016 Depends on the crossover, but sometimes it can be mounted on the terminal plate. A better option would be to mount it on a piece of wood and carefully screw or glue it to the inside of the back panel. You can put it in through the woofer cutout before the woofer is installed. RE the MiniDSP. I'd use it to have a play around with crossover points and levels until you have matched the drivers fairly well (and hopefully they do match), and then try to replicate that with a passive crossover - this is where it gets tricky. Also, if using more than one inductor, make sure they are cross mounted i.e. at 90 degrees to each other to reduce interaction. Having 3 axes, you can have up to three inductors before you need to start really spacing them out. An easy way to think of this is that if you look straight through the centre of any inductor, you should not be able to see any other inductor. Having a play with your DATS and two inductors in series, in different relative positions, as this will show you what interaction is.
Neo__04 Posted March 10, 2016 Author Posted March 10, 2016 Awesome, thanks for the info., Yep, silly me didnt even think of putting the crossover through the speaker cutout. lol. Its early. I'll get my box built and have a play. I'll put some updates up as soon as i have some
Sub Sonic Posted March 10, 2016 Posted March 10, 2016 Cool. This is probably obvious but use biwire terminals (two pairs of terminals on the same plate) as it allows you to drive each section with a different amp channel during the design phase, and also allows you to play with the passive XO MUCH more easily. Another thought, if using MiniDSP, it's still a good idea to put a passive "protection" cap between the tweeter and amp, to protect the tweeter in case of any amplifier pop/thump etc, or any potential mixup with the MiniDSP program, which is easy to do.
Cloth Ears Posted March 10, 2016 Posted March 10, 2016 Well i have sort of been thinking about sealed, Origianlly i was just going to go ported so the bass would be a little nicer. But i think i'll throw a sub in with the speakers, so i might go along the sealed line. Havent heard of baffle step or edge diffraction, but i'll start doing my homework now. Another quick question. Say i run a minidsp for the crossover to start with and play with that till im happy with it, then i can learn to design the equivelant passive crossover. If you make a sealed box, then do your testing with that, how do you then clean up the job and mount the crossover inside? Hope its not a stupid question but do people just screw the back panel on, till they are happy with their tuning, build the crossover, mount it inside, then glue the back panel on? Thanks agian guys, so much help here. Its greatly appreciated Hey Neo, When using Unibox, the suggested box type is generally fairly good if you're going to be using the box as a mid-bass. So for this particular one, vented might be a really good option. BUT. The simulation you did, with the 20 litre box, shows really awful port characteristics and I think you might get some fairly "interesting" sounds that are not really musically related. When I did the sim for vented, I found a 50 litre box, with a 30cm port (10cm diameter, or 4-inch pipe) with no fill and no leaks gives a fairly good place to start. If you try that out and see how it measures I think you might find it goes well. For sealed, a 30 litre box looks pretty good, and you'd find that it would also sound really quite excellent. But you'd be giving the vented box a whole octave head start. The driver in the sealed box would benefit from a crossover to a subby around 60-70Hz to avoid excursion problems when playing loudly. The vented one could probably be crossed to a subwoofer at about 35Hz. Note that the peak simulated excursion at 50w for both is 34mm for the vented box and 16mm for the sealed (both at 20Hz). For passive xover simulation, I still use a couple of crossover simulators that Murray Hauschild wrote more than a decade ago. He's disappeared from the internet - so there's no way to donate to him. But these are worth it (so if anyone finds him, let me know!). The doco is fairly self explanatory. You can upload measurement files if you want to, as the current simulations are for fixed response, fixed impedance 'drivers' (86.8dB & 3.4ohms for the LP and 90dB & 7.8ohms for the HP). There's probably a couple of old bits of simulation in there - remember that it ignores the cells it isn't using for the particular network.... XO.doc newSXO.zip Note that both the spreadsheets in the ZIP file contain macros (safe, I'm sure - I've been using them since 2005).
Recommended Posts