I found a mistake in the medium circuit, right here:
I think I fix it because now the medium are less in my face and there is no more confusion in the music.
This is the crossover for the bass and tweeter
But also now I think there's something wrong with the voices. Often there's too much sibilance. And especially on the trumpet ( kill bill vol II track 9 if you have it). A lot of cd( not all) sound with too much high.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I think I fix it because now the medium are less in my face and there is no more confusion in the music.
This is the crossover for the bass and tweeter
An externally hosted image should be here but it was not working when we last tested it.
But also now I think there's something wrong with the voices. Often there's too much sibilance. And especially on the trumpet ( kill bill vol II track 9 if you have it). A lot of cd( not all) sound with too much high.
You could try a small RC on tweeter to smoothen things a bit...you should be able to do that without upsetting xo
Play with midrange notch resistor
I seem to miss some RC
There is no paralel inductor(with resistor) on tweeter...tweeter may be stressed a bit
To paralel midrange series inductor with a resistor is "old school" and not the proper way...it seem to work, but it doesnt really
Being a 2.5way design I dont understand why use such a big series inductor...I prefer small inductor with RC and notch
Well, thats my personal experience only
btw...I reckon that Joe use tubeamps, which may make a difference
Play with midrange notch resistor
I seem to miss some RC
There is no paralel inductor(with resistor) on tweeter...tweeter may be stressed a bit
To paralel midrange series inductor with a resistor is "old school" and not the proper way...it seem to work, but it doesnt really
Being a 2.5way design I dont understand why use such a big series inductor...I prefer small inductor with RC and notch
Well, thats my personal experience only
btw...I reckon that Joe use tubeamps, which may make a difference
I didn't add the optionl conjugate filter right because I don't use tube amplifier. I'm using a mosfet power amp and a tube preamp.
I think that tube amplifier doesn't change a lot in the FR if it's linear. I tried a tube amp (not with this speaker) but it didn't shutdown the highs, but more details.
The tweeter in one speaker is not centered by 2-3mm, could be the cause?
For the diffracion control I used 4 pieces of felt 4-5mm thick. 2 pieces are the "INNER FELT" and the other 2 are "OUTER FELT", is it right?
On my cabinets there are still the angle, i didn't cut them at all. Should I do it?
I think that tube amplifier doesn't change a lot in the FR if it's linear. I tried a tube amp (not with this speaker) but it didn't shutdown the highs, but more details.
The tweeter in one speaker is not centered by 2-3mm, could be the cause?
An externally hosted image should be here but it was not working when we last tested it.
For the diffracion control I used 4 pieces of felt 4-5mm thick. 2 pieces are the "INNER FELT" and the other 2 are "OUTER FELT", is it right?
On my cabinets there are still the angle, i didn't cut them at all. Should I do it?
Isn't the tweeter supposed to be in a sort of waveguide? That looks like it's in recessed and flat baffle?
I would recheck the crossover again. And try to do it like you are looking at unknown crossover for the first time, I've found I miss wrong connections simply because my brain expects that what i see matches the schematic. If you follow that...
If all that is correct then it may be you need to tweak the tweeter to your taste. The 810921 is a lively tweeter, with more energy then the response would suggest. I can see some people thinking it to be "too much". I'm still messiing with a design using it and have settled on two versions- and both have a slight downward tilt to the response. The crossover region (1k-3khz) is pretty critical too. I don't what options Joe has in this regard. I will say though that it's worht the effort to tame, it's a really excellent sounding tweeter.
I would recheck the crossover again. And try to do it like you are looking at unknown crossover for the first time, I've found I miss wrong connections simply because my brain expects that what i see matches the schematic. If you follow that...
If all that is correct then it may be you need to tweak the tweeter to your taste. The 810921 is a lively tweeter, with more energy then the response would suggest. I can see some people thinking it to be "too much". I'm still messiing with a design using it and have settled on two versions- and both have a slight downward tilt to the response. The crossover region (1k-3khz) is pretty critical too. I don't what options Joe has in this regard. I will say though that it's worht the effort to tame, it's a really excellent sounding tweeter.
tinitus said:
tinitus, Joes' crossover is carefully designed. I don't think it's a good idea to criticize it upon brief observation. The tweeter has enough protection with a small enough series cap plus a bottomless notch at its resonance. Also, the idea of using parallel RL is not "old school" but should be his design choice that gave a better woofer rolloff than your suggested small L plus RC topology. There are many different network topologies that give similar effects, but there exist almost always subtle differences among them. So, it's up to the designer to choose whatever he likes.
-jAy
augerpro said:
I agree with auger's suggestions. First check the crossover carefully again to see if it's properly assembled.
I've also worked on the 810921. It's an outstanding tweeter except that it may need top octave attenuation depending on a personal taste. I'm sure Joe will have a better suggestion, but as tinitus suggested, a small RC can be a simple solution. Here's my quick simulation of Joe's tweeter network with series RC connected in parallel with the tweeter:
An externally hosted image should be here but it was not working when we last tested it.
The RC is connected AFter R3 in parallel with the tweeter.
yes as said above before doing anything I want to receck all again and be sure I'm listening to the speaker that Joe designed.
On some cd I hear instruments crackling. A guitar in the track 12 of Metallica black album, a trumpet crackling in Paolo Conte. Is strange, insn't it?
Could be that ther's interaction beetween coils or they influenced by the backs of the speakers. The crossover is on the rear panel.
On some cd I hear instruments crackling. A guitar in the track 12 of Metallica black album, a trumpet crackling in Paolo Conte. Is strange, insn't it?
Could be that ther's interaction beetween coils or they influenced by the backs of the speakers. The crossover is on the rear panel.
Jay_WJ said:
The graph generated is fairly representative of what it should be like. As can be seen, the tweeter is well protected and yet has good linear phase/1st order characteristics until the 0.82mH/<0.5R/68uF trap creates a multi-order effect. But get this wrong and the tweeter will distort badly. Is this what is happening?
The Tweeter is reasonably flat when listening (and measuring) just a bit of axis. That is why it is recommended that the speakers are set up so that the insides of the boxes are seen by the listener in the central position. The optimum is around 15 degrees. The top octave will be flatter than your on axis example.
The brain tends to adapt to what we want to perceive, so if a mistake has been made to the crossover or the wiring or its attendant phases, it can be deceptively difficult to spot. I think we've all been there, like where did I leave those keys? You ask the wife in an exasperated tone and she walks in and picks it up just before your face, gives that look and leaves you deflated.
So PeterN, find somebody who can check what you did, and outsider's eye is more likely to spot the problem. Know somebody?
Get them to check the following:
1. The component values.
2. The crossover's wiring.
3. The wiring to and from the drivers, make sure all the phases is correct. Even I have gotten that wrong. Use a 1.5V battery intelligently - it is a useful tool, wouldn't do without it. Also note, this is one of those few speaker systems that make extensive use of series wiring.
Joe R.
tinitus said:
This sounds like good advice and I prefer solid core as well.
Re measurements, if just a simple system impedance measurement could be done and posted here, I would spot a crossover fault quicker than Zorro could draw a sword (and so could others here). This is one of the more simple, if not the simplest, measurement that can be done, with basic tools.
If the impedance looks as expected, then at least we know the crossover should be OK, BUT then we would then look closely at the wiring to the drivers and check the phases are correct. Let's say we get a significant suck-out between the mids and the tweeter, keep in mind that wiring faults create 180 degree phase error, this would make the tweeter sound thin and exposed - even if the crossover is working properly.
But did somebody mention crackling from the tweeter? This would possibly indicate that the lower end of the tweeter's bandwidth is being excessively exposed. THIS would show up in the impedance measurement.
Joe R.
Peter N, for what is worth, swap the phase on both tweeters and see if that helps. That was one big advantage of the external crossovers for me, it makes it real easy to change things to track down and sort out the problems.
Joe, I still believe I have my tweeters wired up incorrectly, but my ears tell me otherwise.
Joe, I still believe I have my tweeters wired up incorrectly, but my ears tell me otherwise.
rob323 said:
No, it means you do have them wired correctly. Trust your ears. It's just that you haven't figured out how you did it. Please note that the phase reversal is wired into the crossover box, that may be what has confused you. I can tell you that it sounds significantly wrong if you get that phase wrong, but then again, you have already experienced that.
Joe R.
Hello, now I think I did it right. With the help of a friend of a friend who has done professional school so he knows something at lest about elettronics.
Anyway here's the mistakes:
1)0.82mH/<0.5R/68uF in the tweeter circuit. Now it doesn't crack anymore.
2) The R3 value in my resistence is wrong. The gave me 3 0,82ohm resistor instead of 1ohm. So now I have 2,46ohm instead of 3ohm
3)We've redone some soldering that wasn't the good enough.
Then I connected all. And the sound was a lot better. But still too bright for any speaker. Than I reversed the polarity of the tweeter and the sound now is right. But I'm sure the tweeter is wired incorrectly because it isn't reversed now, it should be normal polarity.
Before swapping the polarity: 1)the music was faster 2) the cymbal were strange 3) the voices sounded strange but I can't tell you why. I found these things only by playing the cd that my band recorded( poor quality records, not like chesky records).
Anyway to be sure that phase is right I have to check with battery. But I have to attacch the wire from the battery to the terminal separately for woofer, mid, and high or all together?
The sound now it's good, maybe a little bright or sibilant with some bright cds, but it's normal.
So now I'll buy the right value for R3 and if it's ok no RC circuit.
thank you to all
Anyway here's the mistakes:
1)0.82mH/<0.5R/68uF in the tweeter circuit. Now it doesn't crack anymore.
2) The R3 value in my resistence is wrong. The gave me 3 0,82ohm resistor instead of 1ohm. So now I have 2,46ohm instead of 3ohm
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
3)We've redone some soldering that wasn't the good enough.
Then I connected all. And the sound was a lot better. But still too bright for any speaker. Than I reversed the polarity of the tweeter and the sound now is right. But I'm sure the tweeter is wired incorrectly because it isn't reversed now, it should be normal polarity.
Before swapping the polarity: 1)the music was faster 2) the cymbal were strange 3) the voices sounded strange but I can't tell you why. I found these things only by playing the cd that my band recorded( poor quality records, not like chesky records).
Anyway to be sure that phase is right I have to check with battery. But I have to attacch the wire from the battery to the terminal separately for woofer, mid, and high or all together?
The sound now it's good, maybe a little bright or sibilant with some bright cds, but it's normal.
So now I'll buy the right value for R3 and if it's ok no RC circuit.
thank you to all
Elsinore builders in USA
What are you using for the 25mm MDF Joe specified for much of the enclosure? I did some Googling for a source but came up empty so far.
Are there any of your Elsinore folks near Chicago? I would love to listen before I embark on the building project.
Thanks,
Francois
What are you using for the 25mm MDF Joe specified for much of the enclosure? I did some Googling for a source but came up empty so far.
Are there any of your Elsinore folks near Chicago? I would love to listen before I embark on the building project.
Thanks,
Francois
adding the optional conjugate filter makes you loose 1-2db of sensivity?
http://www.customanalogue.com/elsinore/elsinore_mk2.htm
Thank you
http://www.customanalogue.com/elsinore/elsinore_mk2.htm
Thank you
Peter N said:
Should not affect sensitivity at all. The above graph simply indicates the difference the source impedance changes from a theoretical zero Ohm and adding a 3 Ohm resistor simulating a 3 Ohm output impedance. This approximates many single-end triode (SET) or zero feedback tube amps. The drop is caused by the resistor and not the sensitivity of the speaker. Both curves have the conjugate filter.
What to look for is that the difference between the two curves stays the same at all frequencies. This indicates that the 3R is not modulating the frequency response. This is the purpose of the conjugate filter - without it the variation between the the curves is much greater and less flat with 3R.
So the purpose of the conjugate filter is 1) keep the response flat even with higher impedance amps and 2) keep the electrical phase flatter overall. It makes the Elsinores more amplifier friendly, especially tube amps.
Joe R.
Joe R.