Resistor questions............

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi

Question time:

I'm using a paralleled coil (3.5mh, dcr=1r0)and resistor (6R8, 10watt) as baffle step correction for my woofers (2 x Vifa TC18, paralleled, nominally 4ohms).

The resistor is getting hot. By hot I mean almost too hot to touch after two White Stripes songs.

So, if my 10watt resistor is getting hot after 10mins of guitar rock at moderate volume, how many of these things am I going to have to serial/parallel to keep it cool? And, how hot can a resistor get before it's not good.

I'm thinking get 4 x 6.8s, serial two then parallel them to give me 6.8 again but with 40watt power handling.

Any downside to this? How much indictance in a typical wire wound resistor?

Thanks.

Mos
 
The inductance of the four resistors is negligible in comparison to the 3.5mH inductor in the circuit, so I can't see a problem with your idea. It would pose a bigger problem in a low signal level application where it would change the performance of the circuit, however at the power levels you are applying it to, it will be fine.
 
I'm surprised it's getting that hot. I don't use any higher than 5W and they don't even get warm. Maybe something else happening here but then again I've not used BSC in the crossover.... use a 0.5 woofer for that.

You can get non-inductive wire wound resistors from WES or Soundlabs in Oz.

There's some at RS Online with heatsinks (part 160-310) but don't know if they are any good for audio.
 
there is another (easiest) way to deal with the baffle step. A two pole low pass filter, first pole (inductior) at 350Hz for example (depends on your baffle's width) and the second pole (capacitor) to your x frequency. If you ever decide to design a new crossover, you could try it this way, which doesnt spend power for nothing on a resistor ...

Could you post your crossover's diagram ?
 
Bobhayes,

the 3.5mH is to pass the low frequencies to be compensated, while the resistor is to pad down the rest of the higher frequencies. So I think Mos worries it's effect to the high frequencies (not to the frequencies passing the inductor). I heard many people worry about resistor's inductance but I don't know if that can be justified. In series with a tweeter, I could once notice a degradation in the treble using inductive wirewound, but I'm not sure if it is because of the wirewound or just because different resistor.

Rabbitz,

BSC resistor do consume power, that's why some people do it in line level.

Mos,

People usually use about 25% of the output power for the resistor wattage. If you go microscopic, I think hot temperature creates distortion, but I don't know to what extent (and I don't think it is serious). And I would prefer parallelling all the resistors (of high values) than having series. I 'think' I can hear the degradation of series resistors (compare 10 x 56R in parallel with 10 x 0.68R in series!) ;)

Jomor,

I'm not sure I know your suggested network but I guess that will affect the crossover quality?
 
Thanks for the feedback guys.

Below is the xo schematic.

It's thrid order with the first inductor and resister used as the BSC. I'd have preferred 2nd order but the Vifa TC18s have a hump in response above 5khz which is affecting the level of the top end, which is a Morel MDT20.

There's a zobel in there as well. Although taking the zobel out doesn't seem to affect the response much at all.

Rabz, this project has frustrated me greatly. My original concept was a 2.5 way using the bottom woofer as .5 to boost the bottom end for BSC. But, for the life of me, I just couldn't get it to sound good. There was an annoying glare or forwardness in the vocal range, especially with female vocals. That's in the 100-300hz range. My response measurements showed a hump in that region that I couldn't null out. When I went to both woofers paralleled with the 3.5mh//R combo for BSC, it sounded really good. But, using that resistor to burn power is not a very elegant solution. It's bugging me something silly.

Jomor, are you talking about an inductor in series then a cap in series or parallel? One will give me a bandpass and the other is just a 2nd order. I can't quite see what you're getting at. Can you post a quick diagram or a more explicit description. Thanks.

Jay, so if I paralleld four resistors to get my 6ohm nominal value I'd need 4 x 24ohm. That's a big bunch of big resistors :)

I'd appreciate any feedback. Something might click eventually.

Mos
 

Attachments

  • spl_nova.jpg
    spl_nova.jpg
    21.7 KB · Views: 195
Mos

I use the 0.5 woofer at the back of the cabinet so it only adds what is lost on the front woofer. Worked extremely well in my setup and maintains the same SPL as the 2-way without it.... just lifts the BS area.

I'm not familiar with the TC18, but have tried the TC series 4" and 5" (PAE which is almost the same) which are very good value but can be a bit bumpy in the response and they tend to have a nasty peak (PAE 5") or dip (TC4") somewhere..... not the easiest to work with and normally I have to play around with the crossover point and do some funny overlaps or underlaps.

I'm sure you'll sort it out.
 
Mos,

Your paralleled L/R will also be affecting the Qes of your woofers, and therefore the overall Q of the bass system. Are you sure you couldn't get LspCAD to compensate for baffle step with just the filter components?

FWIW, I'm now designing (just for kicks) a 2-way system featuring the Scan-speak 18W8535-00 and 2905/9300. The measured response (including baffle step losses) of the 8535, which had a 10dB rising response between 100Hz and 5kHz, was completely flattened with nothing more than a 3rd order electrical filter to achieve a perfect LR4 cross at 2kHz.

BTW, if you want to revisit the 2.5-way approach, you really need the woofers to be isobarically isolated from each other. Drop me a line at my usual address and we can talk if you like.

James
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.