seas thor crossover redesign help

[jimangie1973]

A lot more detail. Air.
Now all the small details in the upper region comes through.
This is exactly what I was looking for.

Now I just have to check that it isn't too sharp...

Great. Glad it has worked for you. If it's too much, replace the 12 Ohm with a 10 Ohm. :T

[KarlXII]

Yeah. It may be a little on the sharp side.
But it has a whole new impression.
It now DEMANDS the foot to stomp!

However, I will try out a ten for sure.
I have a couple of 2 ohm resistors...if I add it one of them in series with the old 8 I'd get 10, right?

What configuration are you running at home now?

[jimangie1973]

Yeah. It may be a little on the sharp side.
But it has a whole new impression.
It now DEMANDS the foot to stomp!

However, I will try out a ten for sure.
I have a couple of 2 ohm resistors...if I add it one of them in series with the old 8 I'd get 10, right?

What configuration are you running at home now?

Correct, you can put the 8 and 2 in series to get 10.

I running with the 1.33 Ohm series resistor, and the 8 Ohm in parallel with the tweeter. I sometimes think I may want a hair more top end, which is why I was playing with the simulation. I will try the 1.2 Ohm and 10 Ohm combination tonight.

[KarlXII]

Sounds good.
I'll stay on this track a little longer.
Let's compare notes later.

[KarlXII]

Yep. 10 is better, that´s for sure. Sharpness is gone but not the high definition.
If I'm tweaking again it will be for 9.
I'm really surprised what large differences there is between 12, 10 and 8 ohms.
I hope it is not just all in my mind. :lol:

[jimangie1973]

Yep. 10 is better, that´s for sure. Sharpness is gone but not the high definition.
If I'm tweaking again it will be for 9.
I'm really surprised what large differences there is between 12, 10 and 8 ohms.
I hope it is not just all in my mind. :lol:

I tried the 10 Ohm, 1.2 Ohm combo and it was too bright for my ears. I went back to 1.33 Ohm, still too bright with the 10. I'm back to 8.33 (paralleled a 50 with the 10) and it sounds right. I will try 9 Ohm tonight.

I think I went through this process before when I settled on the 8. :roll:

I was surprised too how such a small change can be easily audible.

[KarlXII]

I've been listening to the 10 all day, and I am pretty pleased.

Any thoughts on the nine yet, Jim?

[jimangie1973]

I've been listening to the 10 all day, and I am [i]pretty]/i] pleased.

Any thoughts on the nine yet, Jim?

Does the 10 sound a little too bright to you? If so, the 9 should be perfect. I thought the 10 was too bright, but I have no idea how close your system matches mine (driver spec tolerances, different boxes, room, etc).

[KarlXII]

Does the 10 sound a little too bright to you? If so, the 9 should be perfect. I thought the 10 was too bright, but I have no idea how close your system matches mine (driver spec tolerances, different boxes, room, etc).

It might be. I haven't decided yet. :??
I'll just have to go ahead and try.
Actually it is not too bright but perhaps too sharp at times.

How about you - are you going back to 8?

[jimangie1973]

It might be. I haven't decided yet. :??
I'll just have to go ahead and try.
Actually it is not too bright but perhaps too sharp at times.

How about you - are you going back to 8?

Using 9 now. Haven't listened long enough to know for sure if I like it better than 8. I know what you mean. :??

Could the 'sharp' you are mentioning be lower treble frequencies? If so, you can raise the 1.2 Ohm to 1.33 Ohm. If you have any 1Ohm resistors around, you can series that with the 3 you have in parallel with the 2. Confused yet?

[Renron]

Hi Karl ! Thanks for the link to this thread. If I may ask Jim a question....?
Thanks for optimizing the Xovers for the Thor. I have built a set of wonderful small Thors and love'em. Always trying to get more out of the drivers I am curious about this new Xover incarnation and would like to build it to compare with a Premium version of the original Xover.
Question: How does the ga. effect the workings of an Inductor?
I am using 12ga. Goertz Alpha Core foil Inductors and SE Clarity Caps with Mills wire wound resistors. I think they (the original Xovers) sound pretty good, but would like to improve the sound if possible.
Anyone with the new Xover want to exchange (temporarily) and test out an Upgraded original Xover? I live near Sacramento CA.
Thanks for all your time.
Ron

[jimangie1973]

Hi Karl ! Thanks for the link to this thread. If I may ask Jim a question....?
Thanks for optimizing the Xovers for the Thor. I have built a set of wonderful small Thors and love'em. Always trying to get more out of the drivers I am curious about this new Xover incarnation and would like to build it to compare with a Premium version of the original Xover.
Question: How does the ga. effect the workings of an Inductor?
I am using 12ga. Goertz Alpha Core foil Inductors and SE Clarity Caps with Mills wire wound resistors. I think they (the original Xovers) sound pretty good, but would like to improve the sound if possible.
Anyone with the new Xover want to exchange (temporarily) and test out an Upgraded original Xover? I live near Sacramento CA.
Thanks for all your time.
Ron

Hi Ron,

I would say that if you love your small Thor's the way they are, stick with them. I only heard a varient of the Odin with newer drivers (rubber boot) which sounded poor with the original crossover.

The inductor guage effects the overall sensitivity of the woofers, and the frequency response (system Qts is raised with higher series resistance). With my crossover, the higher series resistance of the 16AWG inducters reduces the baffle step by about 1 dB relative to the 14 AWG, which works better in most listening rooms. The downside is slightly more amplifier power is needed to get the same volume level.

The effect of the higher series resistance on the frequency range up to 100-200 Hz depends on the box. If someone with the Thors or small Thors could post data from an impedance sweep, I could show exactly how the series resistance effects the system response for the TL design.

My intention was not to bash the original design, but provide an alternative for those unhappy with it.

Good luck,
Jim

[Renron]

Jim,
Thank you for your quick response.
I'm happy with my curved Thors but am always tweaking something on my system......I just can't seem to leave well enough alone. :lol:
Thank you for the explanation of inductor ga. . I think I'll buy the parts for your Xovers and give'em a listen. Here's a photo of my build.
Edit; my drivers also have the "rubber boot".

Ron

[KarlXII]

Using 9 now. Haven't listened long enough to know for sure if I like it better than 8. I know what you mean. :??

Could the 'sharp' you are mentioning be lower treble frequencies? If so, you can raise the 1.2 Ohm to 1.33 Ohm. If you have any 1Ohm resistors around, you can series that with the 3 you have in parallel with the 2. Confused yet?

:lol:

No, actually I'm getting a hang of it. :W
Problem is, I haven't got any 10 ohm around.
But, I do have a few 2 ohm.
So, in order to get 9, I will parallell a 2 with the 2 in series with the 8.
If I'm not happy with that I may take off that 2 and parallell that with another 2 and switch that combo for the 3.
Confused yet? 8)

[KarlXII]

OK, I'm on 9 now.
Sweet. Less sharpness definitively, however now that some time has passed (and some filter versions) I can no longer remember how it sounds with the 8. :lol:
But there was less airiness (Is that an actual english word?) with 9 than 10 so I might go ahead and test the 10 with 1,33. Not that I really miss it, but hey - do we want to optimize or what?
Actually I find 9 to be the most balanced version so far.

[Renron]

:lol:

No, actually I'm getting a hang of it. :W
Problem is, I haven't got any 10 ohm around.
But, I do have a few 2 ohm.
So, in order to get 9, I will parallell a 2 with the 2 in series with the 8.
If I'm not happy with that I may take off that 2 and parallell that with another 2 and switch that combo for the 3.
Confused yet? 8)

ROTFL...... :rofl:

Aren't // calculators wonderful!

Jim, after re-reading your posts I now understand that changing the coils from 14ga to 16ga raises the resistance and thusly would require more power (watts) for the same dBs.
If I am following along correctly, changing the resistors from 8-->9-->10 , reduces the shunt value and increases the current to the tweeter? or is it the voltage to the tweeter? both?
Learning as I go.
Ron

[jimangie1973]

Ron,
For a given source voltage, raising the resistance parallel to the tweeter raises the voltage (and current as a result) across the tweeter. In reality it's not that simple though because we're not dealing with purely resistive components. That's why a simulation is needed to determine the effects of changing values.

Karl,
I've currently got 8.5 Ohm / 1.25 Ohm going. At this point, i'm not sure if the perceived changes are plecebo.

Jim

[KarlXII]

Karl,
I've currently got 8.5 Ohm / 1.25 Ohm going. At this point, i'm not sure if the perceived changes are plecebo.

Jim

I'm gonna do one last tryout with 10 / 1.2 just for sports.
I know it is possible to detect changes on a half dB, but that is just on a single note. I suspect we are more sensitive on the frequency scale, where we pick up sound on frequencys relative to other frequencys. I suspect that is why we can tell difference from one speaker to another although they have a fairly even response. The character is in the harmonics.
So when we are tweaking the curve this way I believe we can detect very small differences.



...but I may be wrong, too. :B

[Renron]

Ron,
For a given source voltage, raising the resistance parallel to the tweeter raises the voltage (and current as a result) across the tweeter. In reality it's not that simple though because we're not dealing with purely resistive components. That's why a simulation is needed to determine the effects of changing values.

Karl,
I've currently got 8.5 Ohm / 1.25 Ohm going. At this point, i'm not sure if the perceived changes are plecebo.

Jim

Yep, Impedance vs. Resistance.
thanks for confirming what I thought I knew. Thank you for your time in upgrading the Xovers
Ron

[kappa546]

you guys be sure to let us know which sounds best after you're done tinkering. if i'm a metal dome fan (think 27TBFC/G), would the 9ohm resistor be the one you think? i got lost in the process and don't know what the 1.2ohm part was referring to.

[jimangie1973]

you guys be sure to let us know which sounds best after you're done tinkering. if i'm a metal dome fan (think 27TBFC/G), would the 9ohm resistor be the one you think? i got lost in the process and don't know what the 1.2ohm part was referring to.

Are you saying you want to replace the Millenium with the TBFC/G? If so, the whole tweeter circuit will most likely have to change to get optimal performance. I don't have measurements for that tweeter in the same baffle so I couldn't even design it correctly.

The 1.2 Ohm resistor we're referring to is the first series resistor in the tweeter circuit.

[KarlXII]

I think he wants us to recommend the setup that will give the millennium tweeter that metal dome character the most.

If so - yes, I think kappa546 will be pleased to go with the 9 ohm setup.
Treble is very pronounced, but still not too dominant.

I'm still on 9, will do some changes tonight.

[kappa546]

I think he wants us to recommend the setup that will give the millennium tweeter that metal dome character the most.

If so - yes, I think kappa546 will be pleased to go with the 9 ohm setup.
Treble is very pronounced, but still not too dominant.

I'm still on 9, will do some changes tonight.

yea that's what i meant.

[KarlXII]

Using 9 now. Haven't listened long enough to know for sure if I like it better than 8. I know what you mean. :??

Could the 'sharp' you are mentioning be lower treble frequencies? If so, you can raise the 1.2 Ohm to 1.33 Ohm. If you have any 1Ohm resistors around, you can series that with the 3 you have in parallel with the 2. Confused yet?

I switched the 3 ohm for a 4 ohm, so now I'm on 1,2 / 9 ohm and I think I'm going to leave it that way. Best so far IMHO.
At least until I get it measured.
How are you doing, Jim?

[jimangie1973]

I switched the 3 ohm for a 4 ohm, so now I'm on 1,2 / 9 ohm and I think I'm going to leave it that way. Best so far IMHO.
At least until I get it measured.
How are you doing, Jim?

I'm back to 1.33 Ohm / 8.33 Ohm. Anything greater than 8.33 starts sounding too bright.

I'm wondering about the consistency of the millenium tweeter over different batches. Bill Fuss is also running the crossover and needs more BBC dip to reduce what he says is a very forward harsh sound. He raised the 4 uF tweeter shunt cap 8.7 uF. I haven't tried this myself as mine are not forward or harsh sounding at all, and my measurements verify it with a flat power response to at least 30 degrees.

I'm really looking forward to your measurements to see how they compare with mine.

Jim

[babbadoo]

Yes, the crossover is complete, no more tweaks in the future. The change I made to get 1 dB less BSC was to use Madisound 16 AWG sidewinder coils instead of the Erse 14 AWG. Values are the same, 2 mH and 0.7 mH. I am using the Madisound coils now in my system and prefer it. This is the configuration Brian listened to the second time he stopped by. ... and another benefit, it saves about $50.

See the schematic above done by planet10, but substitute in the Madisound coils.

Since Im a newbee in this area this question might be a bit simple/stupid/odd?
In the first specification you calculated the xo with the 14g Erse coils with the values:
- 14g 2.0 mH 0.31 ohm
- 14g 0.7 mH 0.18 ohm
then you have the alterative with:
- 16g 2.0 mH 0.52 ohm
- 16g 0.7 mH 0.28 ohm

So different diameter and different resistans will these still sound the same or what will be the change in sound?

If I put in a couple of 15g, since they are eaiser to buy for mee, with the vaulues:
- 15g 2.0 mH ?ohm
- 15g 0.7 mH ?ohm

Should I get close to the 14g or the 16g resistance and will I hear the difference?

[KarlXII]

Perhaps that bright stuff was addictive?
Anyways, I´m back in 3/9 Ohms now and really enjoying it.
The speakers really sing. Clarity and transparancy galore.

I can't get hold of the guy that was supposed to help me with measurements so I'll do them myself. I ordered a Behringer ECM8000 and a M-Audio MobilPre USB plus a mikrophone cable today. I will get them February 1st.
It may come in handy in the future. I'm looking to build a pair of surround speakers. Tripoles or bipoles...

[Brian Bunge]

The smaller gauge coils have a higher DCR value than the thicker ones. This reduces the baffle step compensation a little. I remember Jim stating that he felt the bass response was a little heavy in his room so he used the smaller coils to reduce that a bit. I don't know the exact amount, but I believe it only dropped the bass output by a dB or so.

[KarlXII]

Perhaps that bright stuff was addictive?
Anyways, I´m back in 3/9 Ohms now and really enjoying it.
The speakers really sing. Clarity and transparancy galore.

I can't get hold of the guy that was supposed to help me with measurements so I'll do them myself. I ordered a Behringer ECM8000 and a M-Audio MobilPre USB plus a mikrophone cable today. I will get them February 1st.
It may come in handy in the future. I'm looking to build a pair of surround speakers. Tripoles or bipoles...

Still on the nine, hopefully doing some measurements tomorrow.
Got the equipment today.

[KarlXII]

Got some measurements done today. Finally! :T

Since these were my first-ever, they may or may not be correct. :E

I set it up outdoors to minimize influence from the room.


The setup was a PC with Svante´s Tombstone and the stuff in the post above.

This is from 3 feet:






Comments?
Suggestions?


Edit:
After some discussions on another forum, and comparison with other plots I think the above is closer to the truth so I pull the seond one.

However, I don´t think it´s accurate in the bass portion; it does not reflect the bass coming out of the port. When listening to the speaker, one does not get the impression that it is so weak in the low bass.

I'm adding the port graph in stead.

[ThomasW]

You couldn't ask for a better plots than those... :T

[KarlXII]

You couldn't ask for a better plots than those... :T

Thanks, Thomas.
I'm really happy with the performance of this speaker, but there's always room for improvements.

[Jed]

Thanks, Thomas.
I'm really happy with the performance of this speaker, but there's always room for improvements.

Yeah, but that is about as good as I've seen anywhere. If it sounds as good as that response looks then I'd say set it and forget it and enjoy the music.

[KarlXII]

Just did a new measurement. The author of the test software hinted that I could measure the port and the elements combined if i put the microphone inside the speaker. That and a 12db/oct tilt would make it fairly accurate to appx 100 hz. Said and done.

This is how the SmallThor rolls off in the low bass.

[jimangie1973]

Looks great Karl!! That bass response is impressive. It looks like the consistency of the drivers is pretty good. That response is almost exactly what I got.

What resistor values did you end up with in the tweeter circuit? I ended up with 1.33 Ohm first series, 2.2 Ohm second series, and 8.33 Ohm shunt.

I have kept the shunt cap with the tweeter at 4 uF. Bill Fuss raised this to get more BBC dip.

Jim

[KarlXII]

Thanks Jim!

I'm running 1,2, 2,2 and 9 ohms and I'm staying with that.
I also have the 4 uF shunt cap.
I think it's great!

Can you show your plots? Should be interesting.

I guess tweaking of the speaker is complete now.
Now it's time to start work on the room.

And new surround speakers. :P

[jimangie1973]

Thanks Jim!

I'm running 1,2, 2,2 and 9 ohms and I'm staying with that.
I also have the 4 uF shunt cap.
I think it's great!

Can you show your plots? Should be interesting.

I guess tweaking of the speaker is complete now.
Now it's time to start work on the room.

And new surround speakers. :P

Sure, I will do it tonight. I'll have to gate it to about 3-4 msec so I won't have an accurate measurement of the low end.

[jimangie1973]

Here's my on-axis measurements at about 1 meter with no mic calibration used on my ECM8000.

Tweakable tweeter values used:
First series resistor: 1.33 Ohm
Series Lpad resistor:2.2 Ohm
Shunt Lpad resistor: 8.33 Ohm
Shunt cap: 4 uF

[Brian Bunge]

Jim,

Those look great. I've really got to get my wife's Microbe cabinets finished so we can take some measurements of the drivers. The red dye on birdseye maple is looking pretty sharp! Just have to paint the baffles and backs now...

[KarlXII]

Hmmmm.

I'm not that used to measurements.

What's the difference between the graphs?
What's the dip in the last one?
What software did you use?

[KarlXII]

I just happened to do a distortion measurement.
Unfortunately I am not aware of any standard reference level i db, I just measured it playing..well, loud.

[Dennis H]

I'm not Jim but I'll give it a shot. Remember he's measuring indoors so he has to deal with room reflections.

1. Very short gated measurement. Just measuring the direct signal before any reflections arrive. Accurate at high frequencies but not accurate below several hundred Hz.

2. Ungated measurement. Accurate at all frequencies but it shows the full effects of room reflections so it looks more ragged.

3. Measured with the wires reversed on the tweeter. Deep null at the XO frequency shows good phase alignment.

[KarlXII]

I'm not Jim but I'll give it a shot. Remember he's measuring indoors so he has to deal with room reflections.

1. Very short gated measurement. Just measuring the direct signal before any reflections arrive. Accurate at high frequencies but not accurate below several hundred Hz.

2. Ungated measurement. Accurate at all frequencies but it shows the full effects of room reflections so it looks more ragged.

3. Measured with the wires reversed on the tweeter. Deep null at the XO frequency shows good phase alignment.

After reading up a bit, I'm sure you are right.

[jimangie1973]

Hmmmm.

I'm not that used to measurements.

What's the difference between the graphs?
What's the dip in the last one?
What software did you use?

Dennis answered the first two questions perfectly. As for the third, I'm using ARTA for frequency response, and LIMP for impedance. I export the ARTA/LIMP data in text format to Matlab where I do the plots. I also do the crossover simulations in Matlab. I use Matlab mainly because it's what I'm good with. I have used it for years at work. I also know exactly what the simulation is doing because I have written it from the ground up.

Jim

[HYPERTUNE]

Just wodering if the W18E and W18EX are interchangeable in this application?

TIA...

EDIT:

Spoke to Elliot from Zalytron and he gave me the rundown. Seems the W18E is the one to get. I'll definately be putting an order in with him. Price is good too