| Forum Home | ||||
| Press F1 | ||||
| Thread ID: 36024 | 2003-07-29 01:51:00 | Low Pass Filter | Ron Bakker (356) | Press F1 |
| Post ID | Timestamp | Content | User | ||
| 163577 | 2003-07-30 00:11:00 | You got it StandardWireGauge :) | Pheonix (280) | ||
| 163578 | 2003-07-30 01:38:00 | The attenuation in db per octave is a measure of how steeply the unwanted frequencies reduce in amplitude as the frequency is increased. You cant have or dont want a suuden cut off, like in a square wave, but on the other hand you dont want higher frequencies getting through either. Just a parallel capacitor or a single inductor will have a very gradual roll off of 3db per octave ( octave as in when the frequency is doubled), a 2 stage will have 12 db per octave, and the more stages in the network the steeper the cut-off of higher frequencies. Dont worry too much about exact values, it is very approximate, as a speaker is not a constant impedance, 4, 8, or 15 ohms are only nominal values anyway, probably defined as the impedance to a pure sine wave at 1000Hz when mounted in a specific enclosure, but I cant remember now. Most cross-overs excepting sub woofers will be in the frequency range 250-500Hz, much too high for your application Ron. Iron/iron dust etc cores are usually used in commercial equipment, because it is cheaper (less copper) and the coils are smaller, but they may be a bit non linear. Im very rusty on modern audio developments though it may have changed a lot 'since my day'. If you didnt find Bazzas link to the inductor design page it is here: www.colomar.com Ideally for minimum wire length the coil cross section should be about square, but not too important. |
Terry Porritt (14) | ||
| 163579 | 2003-08-01 09:07:00 | So how would I measure inductance. Oh & thanks for your help so far, it has given me a good starting point.. |
Ron Bakker (356) | ||
| 163580 | 2003-08-01 10:03:00 | If you dont have an inductance bridge Ron, it would be ok just to wind the inductor according to the dimensions and number of turns given on the site referenced by Bazza. Even though that site quotes values to 2 decimal places from memory, if you are within about 10% that would be more than accurate enough. Capacitors are not that accurate anyway, and if you cant get hold of a bipolar 200uF (which may be very difficult) then the nearest standard capacity would be two 470uF connected back to back ie. negative to negative, giving 235uF nominal. Or you could use 2x 330uF back to back, which would be a bit closer to the theoretical value quoted on that site. |
Terry Porritt (14) | ||
| 163581 | 2003-08-01 10:08:00 | Cool Thanks Terry |
Ron Bakker (356) | ||
| 163582 | 2003-08-01 10:47:00 | I'll do a few sums based on the data on that site to see how much wire would be needed and whether a Dick Smith reel size would be ok. | Terry Porritt (14) | ||
| 163583 | 2003-08-01 11:32:00 | I had overlooked the fact that your speaker is 4 ohm impedance, not 8, so that the capacity is doubled and the inductance halved. So for a crossover frquency of 80Hz from this page; www.colomar.com you need a capacitor of 351.24uF and an inductor of 11.26 mH. The nearest standard electrolytics are 2x 680uF back to back giving 340uF which is pretty close. From the inductor page: www.colomar.com putting 11.26mH in the box gives you a whole table of former sizes and wire diameters to choose from. He doesnt say what the wire gauge specification is, but being American it will likely be B&S (Brown & Sharpe). Using 22 gauge B&S, the table says the length required is 82.4 metres. Dick Smith Electronics have 100gm reels of 22 gauge, 35metres in length at $10.50 each, so 3 reels would be needed. Catalogue number W3122. Copper is not cheap, you can see why commercial units use iron cores. If you were to use an iron or iron dust type core, then an inductance bridge would definitely be needed to carry out tests to determine the number of turns required, and verify the final winding. |
Terry Porritt (14) | ||
| 163584 | 2003-08-01 12:16:00 | Terry One more silly question I actually have some old transformers lying aroung in the garage that I could maybe steal the wire from. But when I go out there to measure the thickness of the wire I will need to figure how gauge is relative to millimetres. I tried looking up conversions on the net & found heaps of stuff but not gauge to mill conversion. Cheers Ron |
Ron Bakker (356) | ||
| 163585 | 2003-08-01 12:24:00 | OOps I just now see your last post Cheers |
Ron Bakker (356) | ||
| 163586 | 2003-08-01 20:59:00 | I see I made an error in the inductor posting, should be like the same previous one: www.colomar.com You know Ron. for a X-over of 80Hz that is a lot of inductance, and lots of wire. The DC resistance is getting quite high, and 22 gauge wire is the thinnest one could use really, and that is about 4 ohms, so that will put a significant loss into the circuit. This is another reason iron type cores are used. I wonder if it would be more economic to see if a X-over can be purchased? |
Terry Porritt (14) | ||
| 1 2 3 | |||||