Seeking Crossover Help

fisherphile

New Member
Reaching out to the gods for ideas. Upon rebuilding these D9s I discovered poor woofer performance with some rubbing, garbled distortion and unusual excursion. I opted to replace the voice coils and spiders. On the bench with direct connect to amp the freeair results sounded good. But installed back in the cabinets I still get the garbled distortion at high volume. Woofer movement is now perfect and smooth and the sound is good but still distorts at high volume. Running gear is plentiful in the power department and performs flawlessly on every other speaker I have built in the last few years. My guess is crossover. I replaced the elctrolytics with PE caps like usual but the only unusal thing I did was parallel gang multiples to achieve the original values. The woofer circuit has 40uF between power and ground after the inductor. I parallelled a 22u,F17uF and 1uF to get 40uF. Im thinking this is the culpret or possibly the inductor. Ive never ganged caps before and coincidentally never had an issue before... Any ideas anyone?
 

Attachments

  • 37609507_10213711269989315_2792261107891306496_o.jpg
    37609507_10213711269989315_2792261107891306496_o.jpg
    94.8 KB · Views: 33
Send your post to AK member Gordonw. He’s a busy guy, but he is also a crossover guru. If he responds, he should be able to give you an idea of what is going on. He also builds amazing tube amps. I heard one, and to me it hung with the VAC Statment amps I heard at the same time. The guy knows his stuff.
 
Parallel smaller capacitors have better performance than a single larger unit. This is the basis for bypassing. As long as the wiring is correct and the capacitors are properly functioning, this will not be a problem.

One caveat. Placement of the inductors is crucial to avoid cross-coupling. So best practice is placing eat at a right angle to the other and placing further apart as the field strength varies as r^2. I refer you to the work of Troels Gravesen, the master on the subject:

Yours are far enough apart that coupling should not be happening. I wonder about the capacitor on top of the inductor, as the capacitor windings can function like an inductor, albeit a very small one, and the coupling may affect performance.

I suspect the driver rebuild did not go as well as one might have wanted.
 
Ok.

That suggests you have the crossover miswired. If you have the woofer overlapping the midrange you'll get peculiar effects.

Try running the crossover with the other drivers disconnected, so you only have the woofer. If you have this distortion with the second-order crossover (inductor followed by capacitor) you've narrowed it down to that portion. You can then try a first-order crossover (only inductor) with the woofer to see how it sounds. If you have no problems, move on to the midrange bandpass. If the midrange capacitor (floor of filter) is too low you'll have overlap which can account for the symptoms you report.

Isolate the problem by removing variables and adding them back in, one at a time.
 
That sounds more like it, I can tell you that high powered clean sine waves (IE 40Hz) doesnt produce issues. It comes up when the signal is more "complex" or "pink"
 
PS I looked up the specs on these electrlytics I used and they have up to 10% tolerance and rated only to 200watts rms. they are 100v. These vegas are the highest power speakers I've done with a rating of 350 watts continuous and I am using a CE4000 power amp. Could this simply be an issue of using poorly suited caps?
 
Wait a minute, you used electrolytics? I thought you replaced them with PP film? Oh, sorry, I misread that.

Electrolytics are horrible! Rectification effects, non-linear effects, high ESR, etc. Yuck.

Yeah, that's your problem. Capacitor distortion.

Get some Part's Express Dayton (house brand) PP film. Distortion will vanish like morning dew.
 
Ive gotten away with it in many speakers, but none of them had this kind of power demand. I use films on the top end and electros on the bottom end which is how this particular crossover was actually done from the factory. the original woofer cap was 40uF 75 volts. but then again I never got to hear what it sounded like 35 years ago.. hmmm Ill play around with it. I have some theories. I did eliminate the caps once already but ran out of playing time to really verify a result. I did notice that with the caps out of the equation the woofer plays some vocals through it. My theory is that the low frequencies pass right by the caps while basically grounding out the high frequencies hence its wiring between the positive and ground. Ill keep listening to possible suggestions, Iv'e only been studying crossover witchcraft for a couple weeks so I'll take all the education I can get.
 
Caps were causing 90 percent of distortion... The inductors as well probably because they are iron core. I circumvented both the caps and coils and just gave full signal to the woofer similar to some of the later model vegas ive done like AT15 and such.. changed the sound but for the better after all these are not easy listening speakers. Lesson learned in the future Ill rebuild with more robust components.
 
Yeah, cored inductors have higher distortion because of non-linear behavior. Electrolytics always sound worse in the signal path; their only virtue is delivering high values in a small package at an affordable price. The rectification issue, higher ESR, and short lifespan make them undesirable for crossovers. Except, of course, if one needs a giant value for a midwoofer or to filter DC at the amplifier output. Then there's no other solution because film would have great size and even greater cost.

I always use PP film, building larger values from smaller capacitors, and bypassing. Greatly improves the sound. The Dayton house brand of Parts Express is very nice. PP is a wonderful capacitor, the only slightly better dielectric is PTFE (tradename Teflon) at higher cost. I doubt anyone could hear the difference, as it is difficult to measure.

The second-order crossover has steeper slopes than the first order, so you'll remove more of the midrange by shunting whatever makes it through the inductor. The problem with higher-order crossovers, of course, is additional phase shift and filter ringing.

I suggest you use air-core inductors with low DCR. No need to spend money on ribbon inductors. A good quality, low gauge inductor will be perfect.

Also, the Dayton PP film capacitors, bypassed as per the above.

That will get you where you need to go.
 
The woofer uses a 100uf electrolytic and that's what you should use. But a good one. The rest of the caps are below 16uf. I have a pic on here somewhere of how I do them with Dayton polys.
 
The woofer uses a 100uf electrolytic and that's what you should use. But a good one. The rest of the caps are below 16uf. I have a pic on here somewhere of how I do them with Dayton polys.

Nope. Just because film did not exist in those sizes at a manufacturable price point doesn't mean today's rebuilder must honor that cost-cutting measure as holy writ. Ah rekkin' Ah didn't git that thar memorandum o' foolishness, and Ah fer shit sure ain't readin' that drivel if it does end up on mah desk.

A 100 uF Dayton film is not horrendously expensive. A 25 uF @ 5% is $7.72 apiece. Four is $31, and two would be needed. The bundle is big, but not horrendously so.

So for $62—I am neglecting the cost of the bypass capacitors which would be needed for the electrolytic—one could have distortion free sound.

Amortize that over a year and it is $5 per month, not even the price of takeout lunch. It is the price of two decent dinners.

Electrolytics sound bad. This is not one of those situations where options do not exist.

Build the crossover the right way and you'll have joy every time you hear it. Do it the wrong way and you'll learn what cheap sounds like.
 
That was in response to "Loose ends". I am in agreement that poly are better. Just noting that my particular value is 40 not 100

As was mine.

I don't understand why people think that choices made at a time when options were constrained must be slavishly followed as if the designer intended use of what we would today consider to be inferior components. Nuh-uh.

Electrolytics are bad, bad capacitors and the fewer of them in the amplifier, the better.
 
So here's a silly question: If I wanted to start eliminating electrocaps in my amplifiers what specific type of cap might I use for that application?
 
PP film. Perfect for that. Inexpensive. Work great for coupling and cathode bypass. I suggest removing the ceramic capacitors used for coupling as these have bad properties as well.

Some will suggest PTFE (tradename Teflon) but that's just going to cost you more. The DA and leakage for PP vs. PTFE are nearly identical; it is PIO which is far worse, and electrolytic which is horrendous. PIO adds a lot of second-harmonic distortion which is seen as making the amplifier "warmer" and "more lively"; naah, it's just distortion of a type we find pleasing.

Many AK threads cover this.

I suggest leaving the electrolytics for filtering, as this the intended application and they work great at that. It is a needless expense to replace them. But a caveat: when removing electrolytics from a power supply it is best to ruin the filter's Q by adding a small series resistor of a few Ω to simulate the ESR. The reason has to do with dissipating the ripple current as heat vs. having it oscillate. Having a filter with too low an impedance, i.e. too low an ESR, often causes ringing. This is a big issue when large numbers of decoupling capacitors are parallel connected, as in computer motherboards.
 
As far as the OP's problem, I'm convinced that a bad electrolytic was the cause. I'm 67 so I've done a lot of Vegas in my time. A 100uf Dayton electrolytic and Dayton polys is the way I go. Best bang for the buck. I have a pic on here somewhere of how I do it. Not a lot of room.
 
Back
Top Bottom