Using Batteries to Bias crossovers into Class A, how does it work?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Well some posts in another thread I started got me thinking about the JBL Everest Speaker project, and I took another look at the Crossover Schematic, available on the web site. In the Owners manual, it says that the Mid, High, and Ulta High frequency crossover networks incorporate battery bias so that the capacitors will operate in Class A. I will give the links to both, but I'm wondering how this works. The crossovers are passive high level crossovers, not active ones. I typically think of biasing into Class A as dealing with amplifiers, and so would be done with the opamps, as it effects the switching (Right?). What does biasing the caps into Class A do exactly, whats it mean? How does it work?

If someone who understands what they are doing could present maybe a simplified schematic so I could better understand where these are hooked in, that would be nice. LF2 confuses me most as I don't understand where all the wires are going to. In the manual it refers to this as there proprietary Charge Couple method, and applies a voltage to the twin capacitors used in each position, which creates a voltage difference between the plates, but no DC found at output.

I'm kind of curious to know if this works, it seems simple and cheap to implement if it does work to reduce distortion, but I'm not sure how to incorporate it.

Owners Manual
Technical Document with Schematic
 
diyAudio Chief Moderator
Joined 2002
Paid Member
There was a thread where SY measured electrolytic bipolar capacitors for possible crossover distortion after an Earl Geddes suggested method. He could not find any. So we doubted the bias helping such a hypothesis. I had a suspicion that the capacitor's electrostatic dimensional stability is enhanced instead, but it remains to be checked. There is subjective evidence that biasing bipolar capacitors in crossovers helps sonics.
 
Long ago Walter Jung wrote a landmark paper on capacitors.
I think it was in Audio magazine and may have been titled
Picking Audio Capacitors. He has some measurements of the
distortion reduction biasing 2 back to back caps and I have used
this and it did work for me.
 
Its real easy to make a biased network. This is how you do it. Lets say the original cap was 1uf. To set one up biased for the same value you would wire 2X 2uF caps in series. You then place a 2meg resistor between the caps and tie a + 9Volt battery to the other side of the resistor. The negative on the battery goes to system ground.

What you end up with is a T shaped circuit. With the caps up top and the resistor going down to the + on the battery. It works by the way. All my crossovers are Charge Coupled the same way as Everest. The only issue can be cost as it can get expensive doubling both the capacitor count and the values.

Look at the UHF section instead of the LF-2, it's much easier to understand.


Rob:)

http://www.audioheritage.org/vbulletin/showthread.php?t=3555
 
diyAudio Chief Moderator
Joined 2002
Paid Member
In this JBL paper the electrostatic dimensional stability suspicion of mine seems to be discussed. They talk about crossover type distortion too - that SY could not find - and I think that there must be some loose terminology problem here. They probably discuss dielectric memory. In such speakers as JBL big ones the amplifier driving voltage that they can withstand will rise far above 9V bias so I can't see how their bias would keep the caps ''class A'' if that was the real problem.
 
" In such speakers as JBL big ones the amplifier driving voltage that they can withstand will rise far above 9V bias so I can't see how their bias would keep the caps ''class A'' if that was the real problem."

That's not DC so the signal still goes through 0 volts. Also the improvement is subtle and audible at low levels during quite passages. You get more clarity

Rob:)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Done that. They get cleaner indeed. What do you mean by not being DC? Under 9V or above the driving signal will always be AC. I guess there assumption is that by keeping the cap on a DC threshold it will not reverse charge under that at AC zero cross.
 
Honestly I never thought that the parasitics from an additional capacitor would be significant enough to be audible. This is one of those easy to try out for a DIY guy type changes. You take a simple 2 way speaker with a simple crossover and try it out. If you don't like it your done.

Rob :)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I liked it when big bipolars or MKTs had to be used. On expensive smaller polys it did not add. I don't know if it is exactly the parasitics, but double the caps in series is more stuff to pass the signal through isn't it?
 
Well since they are double the value the resistance will be lower, so wouldn't the parasitic series resistance be about the same? Also extremely low by any definition, most good polys are in the .0005 ohm range, not .1 ohm like some electrolytics.

I finally realized what all those darn wires were for, and what everything was going to. They designed the crossover so that if you biamp the speakers it bypasses the woofers highpass and tweeters lowpass crossover.

Well I think I understood this concept better than I realized, as that is what I thought they were doing. Sounds like the impact is debatable, and that a member attempted to measure the effect, but could not. Salas, I see you mention that he measured electrolytics. Now they biased poly caps in the Everest, any chance there would be some difference in the effect. I mean, I would think that if anything, distortion would be lower in a poly over an electrolytic, but I don't know, might account for Harmens results as compared with SY's. Also, not to doubt SY's ability, but did he have the appropriate equipment to measure low distortion? I mean, with some of this stuff, the distortion is on the magnitute of 100ths and 1000th's of a percent, which most readily available distortion measuring systems can not measure. When I contacted a company about measuring distortion with their software, they told me that computer based systems are limited by the quality of the ADC plus additional computer noise, and that they have never done better than .1% at their lab. When I talked with an Agilent engineer about the 8903A I have, he told me that it would measure much lower distortion than any computer based system not using an external filter, amplifier, etc. Again, he told me there just aren't good enough ADC's out there, and that its been a problem for them in designing better new measurement equipment.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
SY is ace in measuring. If he sees the thread he will inform you the set up and the method I guess.

I doubt the crossover distortion theory. Must be piezoelectric stability and dielectric memory effect getting better with the battery. That is why it subjectively helps plastic caps too. I have enhanced big MKTs that way.
 
DC Bias

This sounds suspect. Often times it seems argumentively convenient to confuse the MACRO-world with the MICRO-world.

If one studies materials science applied to electronic theory, one learns that the dielectric effects occur within the space-charge region around the dipole molecular structure. This region is measured in the MICRO world. I doubt that a 9 Volt MACRO bias has much of any effect on the MICRO dipole kinetics.

John L.
 
Interesting reading in the link and testing as well. Why 25hz for the test frequency?? If you look at the systems that JBL uses this technology in it is in the upper frequency ranges. In many cases the woofers shunting capacitors are not charge coupled and the series midrange capacitors in the bandpass and up is where it gets used. I think you may have been looking to low to see the effect. If you think about it any small shift may not be significant at a lower frequency but can become significant as the frequency rises. Think of it as time/phase shift

In the Everest they use Non Polar Electrolytics in paralel with polypropylene in the bass crossovers and part of the horn bandpass. They change over to all poly above that range.

Rob:)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.