Series crossover problem

[F1 FAN]

Hi all ,I recently built this crossover for use with my VIFA M17SG and SEAS H1212 tweeter. But when listening at hi levels the 6.8 ohm resistor in series with the 67uf cap gets extremely hot. I removed these two components assuming they were just for impedance eqilization. Then I noticed that the 6.8 ohm resistor conected to the tweeters pos terminal was now getting hot. Is it normal for a series xo to disipate this much power through resistors or is this crossover design flawed?It actually sounds better than the parallel network I previously used but the excessive heat conserns me.Any input on this would be much appreciated .Thanks Fred http://audioclone.free.fr/New BABY2000.gif

 

[Jennice]

I'm by no means a pro (rather the contrary), but that looks like a rather creative cross-over.

Where did you get hold of such a schematic?

I can't see the point in the RC part across the terminals, either

Jennice

 

[Andy G]

1. what power rating resistors are you using?
I suspect you would need at least 50w for the 6.8ohms in the RC, because its across the woofer as well and therefore equalising the impdance over the whole speaker !!

2. contact the designer !!!!

 

[jmateus]

series crossover

I'm a great fan of the series crossovers that I've implemented over the years on all of my speakers. Especially the great sound
stage that gives to the speakers is the real advantage of
this type of crossover.

From your design I can see the over care put on this particular
one, see I'm used to simplify these crossovers just where the
big gurus recommend.

I never had resistors overheating in this type of crossover so
I'm a little puzzelled as to why either one of the resistors gets
so hot. May be the main reason is the complexity of the design
I don't know, but again there might be a reason.

I suggest you a google on the subject there are good sites you
can get ideas about this excellent topology.

One I must say is a must, a calculator and designer.

http://www.pvconsultants.com/audio/crossover/scd.htm

 

[Notax]

1. what power rating resistors are you using?
I suspect you would need at least 50w for the 6.8ohms in the RC, because its across the woofer as well and therefore equalising the impdance over the whole speaker !!

2. contact the designer !!!!

Its equalising of impendance at certain frequency, or range (800-2500 Hz). But frequency is high enough, so it is not needed resistor which can hold more than 10-15 W.
One more thing. 68 uF for equalisation is very much. It puts down impendance from 8-9 Ohm to 4-5 Ohm. About two times. Because of that, maybe is needed more powerfull resistor. I dont like that x-over. Whole.

 

[F1 FAN]

Thanks everyone for your responses. The resistors I am using are 15 watt. Hi Jennice ,I found this crossover on a french site called Audio clone. This design is said to be a copy of the expensive ASA Baby Monitor speaker. Hi Jmateus, thanks for the link to the series design program it looks real good.

 

[jmateus]

series crossover

Hi Fi Fan

Just for what's worth Rod Elliott has a nice discussion of differences between series and parallel crossovers and reasons
why you prefer one to the other. Interesting stuff

http://sound.westhost.com/

 

[F1 FAN]

Hi jmateus, I will definitely look that up.I have previously read about Active xo's on Rod's site. His site has to be one of the very best on the net just full of excellent info.

 

[F1 FAN]

Another question I have for anyone knowledgeable about series xo's.What is the purpose of the .33mh coil across the tweeter? I have never seen a coil in this location on any series xo schematic.

 

[Andy G]

Another question I have for anyone knowledgeable about series xo's.What is the purpose of the .33mh coil across the tweeter? I have never seen a coil in this location on any series xo schematic.

yep, got me rather puzzled too..

 

[tinitus]

I guess only - 0.33mH in paralel with tweeter, it probably will lower impedance on tweeter at low frequency resulting in a "short" in the end

But 0.33mH together with tweeter (resistance) could also form some kind of simple "notch" filter or "tilt"(eQ) on bass

But offcourse the obvious is that it is part of filter in paralel with tweeters 5.6uf (12db)

Thats why I dont like series filter - hard to see what happens as components may have several funktions

 

[jcake5]

Gents,

You now have a third order filter network with the inductor in parallel to the tweeter.

Cheers,
Eric

 

[lndm]

This really looks like a series/parallel combo.

 

[jcake5]

Gents,

Refer to the following for more information.

http://members.aol.com/klehma/radioseries.html

The given equations are a starting point for your modeling efforts.

Real example follows:

http://www.humblehomemadehifi.com/Crossover.html

Cheers,
Eric

 

[F1 FAN]

Hi Eric Thanks for the links. Acording to first one my xo looks like a combined 3rd order hi pass 2nd order lowpass.I have read the Humble article before and it is very informative but none of samples show a coil across the tweeter just LCR networks.But I think he was just covering the design of 1st and 2nd order. 3rd order series must be a rare Beast.

 

[Cloth Ears]

Looks to me like it's trying to be a 3rd order (tweeter) plus 2nd order (woofer) with a Zobel for the woofer and RC impedance on the lot. But the extra 6.8r in series with the tweeter has me a little worried.

When I try to sim it, it gives me an interesting picture of a roller-coaster ride.

 

[F1 FAN]

Thanks Cloth for taking the time to sim it. Now you have me worried that it is a bad design. Unfortunately I don't have the tools to do proper frequency response measurements.But just listening to them nothing would suggest a roller coaster type response.They actually sound very smooth but a little layed back. Also the sound is very consistant even at larger off axis angles vertically and horizontally .The reduction in vertical lobing is a big improvment over the previously used parallel network. But this may be the result of a lower xo frequency Im guessing this one is about 2k the parallel one was at 3k.

 

[lndm]

If a design is too complex to comprehend (and this includes the driver to crossover interaction) then you will probably have to take it for granted. Some of my crossovers can look a little strange, too, but every component is there for a reason and small changes can upset the balance. There is not really that much difference between a great crossover and a good one (if you take my drift).

That said, tweaking (as in a form of reverse engineering, if you like) may be your key to understanding the workings of this crossover. Otherwise, you may be up to ditching it for a fresh one.

 

[PeteMcK]

my first impression is that the .33mH should be connected to the other end of the 6R8 resistor...???

 

[Cloth Ears]

Thanks Cloth for taking the time to sim it. Now you have me worried that it is a bad design.

F1 Fan (I hope you are,

Don't treat a sim as a replacement for your ears. From the sounds (heh heh) of your first note, you're not worried about the sound, only about the heat generated over your resistors.

If the sound is good, then think about replacing the 6.8r (in series with the 67ohm cap) with a high power resistor (say a 50W one with a heat-sink attached). This should alleviate any worries you have about the heat being generated. As it's not directly in the signal path, it should affect the sound too badly - especially if you can match them in the shop (take your meter along).

If the sound is not too good, then we have another problem.

I use SXO.xls (and PXO for parallels), a spreadsheet I got from the web, written by Murray Hauschild - but I can't seem to find it now. The bad sim is probably due to bad data for the 2 drivers (I don't have measurements for them.