The Hornshoppe Forum
August 19, 2017, 02:08:17 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
News: SMF - Just Installed!
 
   Home   Help Search Login Register  
Pages: 1 [2] 3
  Print  
Author Topic: Home-made "V2" crossover needs a review  (Read 9108 times)
Capt. Z
Sr. Member
****
Posts: 411



View Profile Email
« Reply #15 on: January 04, 2014, 09:35:54 AM »

It's all about what you like/prefer. For me the 3.9 uf was the best solution.
Logged
Tre
Newbie
*
Posts: 24


View Profile Email
« Reply #16 on: January 09, 2014, 09:42:03 AM »

Sorry for jumping in here but what's the target crossover frequency?

From what I am reading here it seems you are all using the cap value as a attenuator?

The Heil AMT is very efficient above about 2kHz.

I don't use a series crossover. I mate my Heil to a JBL 2118J in a ported box that's about 97db

I pad down the Heil by about 5.2db using a 3 ohms series resistor to get the amplitude of the Heil to match the JBL.

This means that, to my ears, the un-attenuated Heil is about 102.2 db.

3.7 ohms (the impedance of the Heil) in series with a 3 ohm resistor gives me a 6.7 ohm driver that is crossed with a

9.1uf cap giving me a 6db per octave crossover at 2600Hz.

I don't know what the efficiency of the horn loaded 126 is but I would guess it is less efficient than the Heil.

BTW a 4.7uf cap will cross the Heil at about 9kHz and the  0.91mH coil will cross the 126 at about 1400Hz (but it won't really without a Zobel in place because of the rising impedance of the 126 caused by the self inductance)

Tre'
« Last Edit: January 09, 2014, 11:30:05 AM by Tre » Logged
Henry
Hero Member
*****
Posts: 802


View Profile WWW
« Reply #17 on: January 09, 2014, 02:17:45 PM »

Hey Tre, here's a schematic of Ed's v2 crossover in case you haven't seen it.  Also, I'm reposting it to make it easier for me to find.  Wink

Logged
Magidrakee
Full Member
***
Posts: 125


View Profile
« Reply #18 on: January 09, 2014, 02:41:16 PM »

Here is my pseudo-scientific attempt at showing differences between cap values using this series crossover on a graph Wink

My microphone isn't very well suited for this, but still I think it gives a fair idea of what I'm talking about:

2.7uf vs 4.7uf  @ 1 meter from middle of heil and horn. Feel free to comment.



Logged
Ed Schilling
Administrator
Hero Member
*****
Posts: 1307


View Profile Email
« Reply #19 on: January 09, 2014, 09:43:16 PM »

Tre',
Quote
From what I am reading here it seems you are all using the cap value as a attenuator?
Jesus, man, don't give it all away! And just maybe the coil value compliments that?

The Horns were measured to be 94 db in free space by John Atkinson for the Stereophile review. Add at least 3 db for corner loading. The series xover as I'm using it is "asymetrical". You'll not find those values with "math". Your assumptions on the sensitivity are 10 db higher than the manufacturer claims. They are at most 93 db from 1k past 20K and almost ruler flat in my room I think CaptZ was here one day when we measured them. My daughter could hear 20K from them at a measured 90db in room while she was at the other end of the house.

Your statement about the rising inductance of the 126 is exactly why textbook formulas and math seldom give the desired results. I really don't care about such things....you see in 40 years of doing this I've not yet heard a "zoebel" actually "fix" anything except in theory and I've always thought the sound suffered every time they were used in my speakers. I understand the theory completely and have nothing to offer as to why they simply have "stole the life" from every speaker I tried them on over the years.

When I first added the Heils I did use a 12db parallel xover with a xover freq. of about 1k and the values were about textbook. Circumstances got me to try the series. The values were derived from what worked and I will admit they did not end up where I thought they would.

Hey Charles, the cap value should not have affected the very top, I don't think. I also think the roll off above 10k is either your software or microphone. They absolutely have flat response from 1-2K past 20K. That rolloff in your graphs is not real unless your Heils are defective! And that is just another example of measurements' ability to lead us astray.

There is nothing about the series xover values I'm using that makes any sense. I see no need to worry about it, I'm pretty darn happy with the sound with any amp and all sources!
Ed


Logged
Magidrakee
Full Member
***
Posts: 125


View Profile
« Reply #20 on: January 09, 2014, 09:51:53 PM »

Hey Ed,

The roll-off you see is my microphone's frequency response (which I included in the graph). My only goal was to compare both crossovers in a relative manner, nothing is meant to be flat here. Sorry if it was not clear.

Oh and by the way I'm not trying to say the original cap value is not good or anything. I'm just trying to understand the impact of the value change on the whole sound, and why I prefer lower value.

Charles
« Last Edit: January 09, 2014, 09:58:57 PM by Magidrakee » Logged
Ed Schilling
Administrator
Hero Member
*****
Posts: 1307


View Profile Email
« Reply #21 on: January 09, 2014, 10:38:55 PM »

Hey Charles, yea, I figured but some reading may not and think that is an actual freq. response. I appreciate the input and please keep it up. I HATE taking measurements! I would encourage any and everyone to experiment with both coil and cap values if they wish.

One thing that bothers me and I do not understand is that when calculating the component values for either a series or parallel xover the values are the same. When I built a series using the calculated values that worked fine with the parallel xover it sucked.
Ed
Logged
Tre
Newbie
*
Posts: 24


View Profile Email
« Reply #22 on: January 10, 2014, 01:51:39 PM »

"They are at most 93 db from 1k past 20K and almost ruler flat ..."

ED, if they are only 93db then why do I have to pad them down 5.2db to get them to match up with my 97db JBL 2118J?

Life is full of mysteries.

BTW With a 4.7 cap the Heil will be about 9db down at 4500Hz. If it's only 93db to start with and the horn is 94db, how does that work?

More mysteries!!!

Best regards,

Tre'

Logged
Ed Schilling
Administrator
Hero Member
*****
Posts: 1307


View Profile Email
« Reply #23 on: January 10, 2014, 05:02:16 PM »

Tre', you are going to love this. I spoke with Rico today the CEO of ESS while he was at CES. He said the official number is 93 but guys have "measured" them to be from 93 to 106 db efficient. So, pick your number Smiley 93 is correct.  Your JBL's are not "97 db" most likely Smiley  And that number means nothing in reality, do you mean "efficient" or "sensitive" ? Efficiency and sensitivity are different things. The total  moving mass of the Horn/Heil combo is under 4 grams. It is both sensitive and efficient. Your JBL's will not be both. The mms  is a "problem". You will not have a speaker that is both sensitive and efficient if it has a very high moving mass regardless of the motor. This is why you have to pad the Heil. You have a "sensitive" driver mated to it that is not "efficient". The exact same problem the original ESS speakers had, IMO.

From the Stereophile review of my speaker, and it applies the the GHH.


Quote
Briefly, sensitivity describes how loudly a speaker will play given a certain voltage; efficiency describes how well the loudspeaker converts electrical power into acoustical power. In other words, efficiency is how easy it is or isn't for an amp to develop and maintain that voltage across the speaker, as a load. Sensitivity is measured in decibels relative to a given voltage and measuring distance (typically 2.83V and 1m, respectively), and efficiency can be assessed by looking, not only at sensitivity but also at a speaker's nominal impedance, in ohms. Efficiency can therefore be gleaned from a graph that plots a speaker's impedance, in ohms, vs frequency, in hertz.

My guess is that you are padding them down to get them to sound right to you. No big deal.

The cap value and how does it work? I have no idea.
Ed
Logged
Steve F
Sr. Member
****
Posts: 405


View Profile Email
« Reply #24 on: January 11, 2014, 08:32:33 AM »

J series JBL drivers are 16 Ohm devices. The Heil is a 4 Ohm. JBL measures their units in half space; the Heils, who knows.
When I ran GHH's, I padded the Heils down just a tiny bit. I suspect that the 126en and the Heil match up very closely in efficiency.
The moving mass of the JBL is 17 gr which is typical of well designed pro mid woofers that size.
I have used several brands of mid woofers, and played them in my room with the Heils. They were less efficient (guessing 1-3 Db), even though they were rated 94-97 Db. Of course they were able to play at volumes that would melt the GHH's, but your ears would melt too.
When I was building those other designs, I found several tweeters, dome, planar, ribbon that were efficiency matches for the Heils. They were literally plug and play, and all rated about 93 Db.
Just my experience.
Logged
Tre
Newbie
*
Posts: 24


View Profile Email
« Reply #25 on: January 19, 2014, 03:55:45 PM »

"J series JBL drivers are 16 Ohm devices. The Heil is a 4 Ohm."

With a 3 ohms series resistor the Heil presents 6.7 ohms to the amp. My Heils measure 3.7 ohms.

The JBL is 13 ohms so some of this is attributable to power distribution.


"When I was building those other designs, I found several tweeters, dome, planar, ribbon that were efficiency matches for the Heils. They were literally plug and play, and all rated about 93 Db. "

The Morel is rated at 92.5db and is not even close to the Heil.

If you are using a 4.7uf cap on the Heil everything below 9kHz is attenuated.

Try crossing them closer to the frequency the 126 is crossed at.

I'm not trying to bust anyone's balls here but I ran the Morel with the JBL and it doesn't keep up.

With 2 Morels in parallel (now we're talking about some power distribution being part of the picture but that ends up being in favor of my position, the Morels are 7 ohms each so 3.5 ohms vs 13 ohms) they almost keep up.

The Heil more than keeps up and had to be attenuated 5.2db to sit where I want it, that is, a little hotter than the 2 morels were.

To me this speaks of the numbers being pretty close unless both JBL and Morel are off by the same amount, in the same direction. JBL 96-97db depending on which data sheet to look at and 92.5db for one Morel, 95.5 for 2.

Please, cross the Heil at the same frequency that the 126 is crossed at. I'm not saying that would sound good but if you (or Ed or whoever) does that, I think you/they will see that the Heil is louder than the 126. Could someone please try this for me? It wouldn't be very hard.

Again, with a 4.7uf crossover cap for the Heil, the Heil is -3db at 9100Hz, 9db down at 4550Hz and -15db at 2275Hz so if that's a good match (with the 126) to you're ear then doesn't that prove my point?

Please don't take any of this as some kind of attack. I'm really just trying to help.

To me, the Heil crossed at the same frequency as the 126 and padded it to match the 126 in output SPL just might sound better than crossing the Heil at 9100Hz and not, otherwise, padding it. I could be wrong but I think it's worth a try.

With all the best intentions, Tre'
« Last Edit: January 19, 2014, 09:58:07 PM by Tre » Logged
Steve F
Sr. Member
****
Posts: 405


View Profile Email
« Reply #26 on: January 19, 2014, 11:09:34 PM »

Tre,
When I ran the Heils, I found them to be a good matchup for the Horns. My Heils were first series, not current production. I don't know what the current ones are like and have lost interest in them. In my rather live room, I padded the Heils down a bit, but not a whole lot. I started playing around with other drivers out of curiosity. I have a few too many hf drivers around, and I wanted to learn more about directivity.

I have heard your JBL driver because the factory dumped a bunch of them in a "tent sale" a few years ago. A buddy of mine got a couple pair. He used them in a humongous system design he found on the internet, and if memory serves me correctly, also published in The Audio Critic. The design might still be available. At least it would be a good read.


I doubt anyone is going to take offense at your comments. You are just reporting your experiences. We all learn new things from each other.

Edit: I ran my tweeter experiments with one of my crossovers, not one of Ed's.
« Last Edit: January 19, 2014, 11:17:39 PM by Steve F » Logged
Ed Schilling
Administrator
Hero Member
*****
Posts: 1307


View Profile Email
« Reply #27 on: January 20, 2014, 10:33:40 AM »

Tre', do not worry at all about your posts/opinions! You will not upset anyone, especially me. That is my job, apparently Smiley. This is a discussion forum, after all and that is what we're doing! I appreciate your posts, you set me straight on the Heil bi/dipole thing and the actual operation of the series xover! Understanding it still did not make the values change Smiley

I don't have much time now but will finish this later and explain the choices that I made.

The parallel xover values were "conventional". The values should be the same for both types according to "theory"......too bad that did not work! More later!
Ed
Logged
sonic
Sr. Member
****
Posts: 394



View Profile
« Reply #28 on: January 20, 2014, 10:38:11 PM »

Yeah, Tre, say what you want.  It's still a free country.  I mean look at what I post and I'm still here.  Cheesy
Logged
Ed Schilling
Administrator
Hero Member
*****
Posts: 1307


View Profile Email
« Reply #29 on: January 20, 2014, 11:02:55 PM »

Damn you Sonic........you made me spit out beer when I read your post Smiley Very funny. Hell yea, you are still here and we hope for a long time. I think it's obvious this forum is about as "wide open" as any and that all opinions and comments are always welcome.

I had a crash on my Gas Gas this evening while looking around the clear cut. I knew better than ride through it but did anyway. My "plan" of trying to "skim over it" almost worked. Almost. What looks like a "road" is not really one, it's just where they have been dragging the trees and is very treacherous. You can not see all the little stumps....it looks smooth but it's not!

Video (oh yea!) and pics in the other section (guns and motorcycles)  tomorrow and hopefully more info on the xover in this thread if I get a few spare minutes Smiley
Ed
« Last Edit: January 21, 2014, 12:16:58 AM by Ed Schilling » Logged
Pages: 1 [2] 3
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.20 | SMF © 2013, Simple Machines Valid XHTML 1.0! Valid CSS!