Announcement

Collapse
No announcement yet.

Crossover components choice

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Dave Bullet
    replied
    Originally posted by AEIOU View Post

    It's a serious mistake thinking that only series and not parallel components matter, if that were the case, then how would you obtain second and third and forth order crossover function? If a parallel component produces distortion, it still affects the entire circuit, because it is in the circuit. Duh!
    I didn't say they were irrelevant. If course they matter to provide all many of shaping. What I said is a generalisation on what people tend to focus more money on. I stand by that. People usually pony up for fancy caps in the tweeter circuit, before they buy that 220uF silver oil or whatever mundorf parallel cap in the woofer parallel leg.

    I did not comment on distortion. Let's assume one parallel component produces more distortion than another and this component loops back to the amplifier. Now... does that carry over to audible distortion through the driver? I dunno. I'd be interested in some expert advice.

    Then there is the matter of whether we are talking fractions of dB difference between the various caps... vs. value tolerances (bigger impact?) and inherent driver distortion (much much bigger impact)

    Leave a comment:


  • djg
    replied
    Looking at the Tarkus cab drawings. Add the top depth measurement to the bottom depth measurement. Divide by 2. That depth measurement gives you the exact volume of the sloped cab in a square cab. That's as far as I would go. I was good at geometry but too lazy for algebra. Jeez was that 10th or 11th grade?

    Leave a comment:


  • Steve Lee
    replied
    Hey!

    Quit being so vowelent!

    Leave a comment:


  • AEIOU
    replied
    Originally posted by Dave Bullet View Post
    What I've gathered over the years:
    1. Crossover components in series with the drivers seem to be the focus for upgrades than those in parallel
    2. Lower frequency sections tend to opt for cheaper components. Either this is price driven (larger values) or we are less sensitive to variations in bass "quality"
    3. If you want to upgrade - series caps in the tweeter.... and less so but possibly midrange circuit are what people do
    4. for all parallel legs - higher resistance inductors (remember to subtract / allow for inductor DCR) and non-polar / bi-polar electrolytic caps are fine. Ideally your XO simulation software will allow you to measure power dissipation across the circuit so you can see if you have a load issue needing to parallel resistors etc...
    It's a serious mistake thinking that only series and not parallel components matter, if that were the case, then how would you obtain second and third and forth order crossover function? If a parallel component produces distortion, it still affects the entire circuit, because it is in the circuit. Duh!

    Leave a comment:


  • Threadbare
    replied
    Thankyou Chris. I will check again until I get the figures to work as they should.

    Leave a comment:


  • Chris Roemer
    replied
    Earlier you wrote:
    "Paul states the top box volume is 13 litres and the bottom one is 62 litres. I have calculated using his internal measurements they are actually 17.75 and 65.75 litres."
    Uh... I DO get 65.7L for the lower box, but (off his orig. drawings) I get 13.4L for the top. He must've felt the magnet & port (in the bottom) consumed about 4L (230 cu.in.).

    Working off his description of a rectangular bottom box, I get 65.7L (the EXACT same I got for his angled box). Again, this is gross internal volume.
    A 4"id x 13" long port (from 4" PVC) uses up about 210cu.in. - leaving about 20ci for the Peerless motor/cone. Seems reasonable (enough) to me.

    Leave a comment:


  • Threadbare
    replied
    Originally posted by PWR RYD View Post
    I think his stated volumes are net volumes. You are calculating the gross volumes. The net volume is what is left of the gross volume after you subtract the space consumed by the port, the driver's magnet, internal bracing, etc.
    After months of delays, bad health issues and pandemic lockdown I'm about to start cutting the mdf and went over the calcs for dimensions once more to make sure. Using Paul's published dimensions for the squared off version of the lower box using 3/4" material brings the gross volume to a whopping 72 litres! Are you sure this is correct? His comments about the build on his site mentions 62 litres. That's a lot of room for the port tube and single driver.

    Leave a comment:


  • Dave Bullet
    replied
    What I've gathered over the years:
    1. Crossover components in series with the drivers seem to be the focus for upgrades than those in parallel
    2. Lower frequency sections tend to opt for cheaper components. Either this is price driven (larger values) or we are less sensitive to variations in bass "quality"
    3. If you want to upgrade - series caps in the tweeter.... and less so but possibly midrange circuit are what people do
    4. for all parallel legs - higher resistance inductors (remember to subtract / allow for inductor DCR) and non-polar / bi-polar electrolytic caps are fine. Ideally your XO simulation software will allow you to measure power dissipation across the circuit so you can see if you have a load issue needing to parallel resistors etc...

    Leave a comment:


  • Threadbare
    replied
    I've already ordered the parts but thanks for the comments. I am still just finding my feet in the "audio" side of electronics. I have built things and repaired things over the years but never had the opportunity to learn about design so any input I get here is welcomed and appreciated. I just steer clear of any topics that might involve snake oil and subjective nightmares about which colour is best for cable ties.

    Leave a comment:


  • PWR RYD
    replied
    I just took a look at Paul's crosover for the Tarkus. C5 is 125 uF. I would definitely just use a 125 uF, 100 V, non-ploarized electrolytic cap there. A 125 uF film cap would be huge, expensive, and might not make any audible improvement in that part of the circuit. C2, the 65 uF cap, is where you might want to splurge a little and pay for a film cap. All of the other caps are small enough I would definitely use film caps. I have had very good luck using the Audyn Q4 caps that PE sells. They are reasonably priced, always measure within spec, and have nice sturdy leads. I think they sound good.

    Leave a comment:


  • Threadbare
    replied
    Thanks for that, Craig. It makes sense now.

    Leave a comment:


  • PWR RYD
    replied
    I would calculate his gross volumes, like you already have, then build your rectangular boxes with the same gross volumes. What is more important is that you keep your baffle widths and relative driver positions on the baffles the same as Paul did. A box volume +/- 5% will be inaudible.

    Leave a comment:


  • PWR RYD
    replied
    I think his stated volumes are net volumes. You are calculating the gross volumes. The net volume is what is left of the gross volume after you subtract the space consumed by the port, the driver's magnet, internal bracing, etc.

    Leave a comment:


  • Threadbare
    replied
    I have ordered the crossover parts and now just firming up on the enclosure dimensions. Paul states the top box volume is 13 litres and the bottom one is 62 litres. I have calculated using his internal measurements they are actually 17.75 and 65.75 litres. Has anyone made the calcs and come up with anything different. The reason I ask is because I want to do simple rectangular shapes instead of sloping. Very easy to do if I know what volumes to aim for. His stated volumes don't really tally with the angled dimensions given. Or do they?
    I have a feeling by the time I work this all out one or more of the drivers will be discontinued.

    Leave a comment:


  • Threadbare
    replied
    That is the most likely reason. He used modestly priced drivers in the design as well. For that I am grateful.

    Leave a comment:

Working...
X