Repairing old stereo crossover

[victor]

Me and a friend was repairing an old stereo he have, and the speakers have a a passive crossover inside each box. All tweeters failed, so my first guess was that the crossover was faulty and took the tweeters with then.

So I kinda R.E. the circuit, but it doesn't seems to make much sense to me why they built it that way. It is not any typical filter configuration I have seem.

The stereo is very very old, the Caps are HUGE, I also never seem that kind of color-coded looking caps, I found googling that they are of polyester type.

The stereo is a Polyvox 5000 like this one



[![xover by victornpb ca5f5437754d4c88 - Upverter]
(http://upverter.com/victornpb/ca5f5437754d4c88/xover/#/)

I have no clue about the inductor value, the tweeters are 2 16ohm in parallel.
2 of the caps of of the high-pass part are cracked on the sides.

[oldway]

The only failure of the cross over which could damage the tweeters would be if one of the 2 condensators C17 or C18 is shorten.
Are you sure your amplifier is not oscillating at high frequency?

[David_AVD]

I agree that HF oscillation was the likely cause of the tweeter failure.

[Yago]

Couldn't the amp being overdriven into distortion blow the tweeters?
With it being in a rack with a mixer on top indicates it had been used in a working environment and might have been pushed too hard.

[David_AVD]

Also a possibility of course.  Passive crossovers rarely go faulty apart from physically dislodged parts and cracked joints.

[victor]

checking the data sheet of the amp it says it goes from 0Hz DC to 80KHz.
is it possible that it is amplifying some HF stuff like high order harmonics, intermodulation or something like that?
I wish I had a Spectrum Analyzer

I know C17 and C18 makes a 54uF High-pass filter, but is C9 a low pass filter?
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/filcap2.html

Notice the two caps on the left they have cracks on the side, is it OK or a sign that they gone bad? (C19 and C18)


This equipment is very old, probably has twice my age or more.
Like these Electrolytic caps are 33uF but they are HUGE like 2 inches taller, I guess they couldn't make a higher capacitance in a compact package.

[G7PSK]

The cracks in the plastic sleeves is due to shrinkage, likely due to heating.

[SeanB]

Tweeters normally die because you are playing things like Rock or other heavily compressed music into them at high volume. Old tweeters were designed on the assumption that most would play classical music, with a very low level of high frequency content and mostly at low level with the odd peak of high frequency transient like a cymbal or other instrument. Modern music has a lot more energy at high frequency, and this tends to overload the older designs burning out the tweeters. You probably had 10W or more of power going into them, while they are rated for 3W actual power ( rating for the midrange and woofer would be around 100W) so they eventually burn out.

Had that with some speakers, and the replacement tweeters were rather expensive. If you can use higher power rated units or piezo units, as these will be better suited to high level listening.

[Mr Smiley]

Has your 5-32 ohm shunt resistor gone o/c  between wiper and ground

 

[victor]

someone could help me determine, what would be the cutting frequency of the High-pass filter, and what order it would be?
none of the calculators meet the design of a standard filter.

http://sim.okawa-denshi.jp/en/CRhikeisan.htm

I want to know because, if I add a simple 6db CR HPF direct on each tweeter, how effect the actual crossover would have on a second one?

[Anks]

Going by the capacitors that are cracked I would have to say that this is most likely due to being over driven and the amp supplying to the crossover what is essentially a square wave.

The best way to protect the tweeters would be to use a light bulb of the correct value to limit the signal. What happens with this arrangement is that when the current is excessive the light bulb lights and burns off the excess. The bulb also has a slow off and on time that will round off the edges of the square wave.

If you don't like the idea of a light bulb being in series with the signal another method is to use a resettable fuse (yet again of the right value) in parallel with the bulb so that the bulb is only in the circuit when it is needed.

The other way to prevent this happening again is not to drive the amp into clipping.

Here is a schematic of a Community SLS920 professional loud speaker. It shows the method I have just described and is pretty much the industry standard for high end passive xovers in the sound reinforcement industry.

[Shredhead]

Hey man, nice stereo!  Gotta love the old stuff.  You said that your amp claims to go down to 0Hz and it has a transformer on the crossover which makes me think that it might be DC coupled.  That would mean it has no protection from DC offset going to your speakers if things go wrong in your amp or pre amp.

With no load on the amp, I would turn it and the preamp on and see if it is sending out any DC voltage.  That would account for dead speakers.  On older amps, preamps and receivers like that you usually have to go through it all and check for broken solder joints, replace a bunch of electro caps and re-check the bias.  I would check the values of all of the caps on the crossover boards too.  My guess is that the values have drifted quite a bit over the years and it might be cheaper to start over with new crossovers than to replace all of the bad components.

[David_AVD]

Has your 5-32 ohm shunt resistor gone o/c  between wiper and ground

Traditionally, the variable resistor (on front panel) was to adjust the HF response to the listener's taste.  Since the user is free to turn this all the way up, the wiper going open will not cause a fault condition.

Hey man, nice stereo!  Gotta love the old stuff.  You said that your amp claims to go down to 0Hz and it has a transformer on the crossover which makes me think that it might be DC coupled.  That would mean it has no protection from DC offset going to your speakers if things go wrong in your amp or pre amp.

The "transformer" on the crossover is an inductor for the bass driver.

Most amplifiers have DC protection with a relay on the speaker output line.  DC on the output would not blow the tweeters.

[Mr Smiley]

Has your 5-32 ohm shunt resistor gone o/c  between wiper and ground

Traditionally, the variable resistor (on front panel) was to adjust the HF response to the listener's taste.  Since the user is free to turn this all the way up, the wiper going open will not cause a fault condition.

The adjustable pot is across the tweeters, so some of the input power is dumped to ground whether it's set to either 5 or 32 ohm or between. It's value could be chosen to specifically limit the max power delivered to the tweeters.

If it goes o/c to ground then all the input power is dumped directly into the tweeters.

 

[David_AVD]

Oh, I see what you mean now.  You mean the end connection going open.

I was talking about the wiper going open, which is no problem.

[victor]

The variable resistor is not open, it is actually a bunch of power resistors soldered to a rotary switch of 5 positions.

I think I'm willing to not use the original HPF and built a custom Band pass filter, that cut low frequencies and high frequencies as well, something like 5KHz to 20KHz.

Any design suggestions?! anything simple and effective, must be a passive filter.

[Richard Head]

Those "electrolytic" capacitors are in fact non-polarised capacitors. I hope they aren't blown as they will probably be difficult to find nowadays.
That "old" equipment is generally far better built than most of the stuff available today. Really solid stuff compared to the plastic crap these days.

Dick
 

[victor]

Using this calculator for Narrow band-pass filter http://archive.pac-audio.com/crossover/crossoverbp.asp
This is the values I got.

https://www.dropbox.com/s/nl6d5rkgplm3ioa/Screenshot%202014-03-27%2018.25.10.png


I think I will built the 2 order, because it require a small inductor, what do you guys think?

Inductor-L1: 0.187041904945961
Capacitor-C1: 1.04190614297251
Inductor-L2: 0.133404274391348
Capcitor-C2: 1.46082358039584

A 187uH still a quite large inductor.
How do I calculate what voltage ratting the caps should be?

Some Air core inductors calculators that I used
http://www.colomar.com/Shavano/inductor_info.html
http://www.circuits.dk/calculator_multi_layer_aircore.htm
http://www.diyaudioandvideo.com/Calculator/Inductor/