Loud Speaker Crossovers and power amp placement

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
1)Is it better to have the crossover outside the cabinet?
Not just for being able to get at it, but for the curcuit itself etc.

2)Is it best to have a short run of interconnect and a long run of cable? Or a long run of interconnect and a short run of cable?
 
The advantage to having the crossover external to the cabinet would be the additional distance of the inductors from the drivers, which would reduce cross-talk.

Keep the interconnects as short as possible, the speaker cables long. Interconnects are subject to noise pickup, which is then amplified. Any noise picked up on the speaker cables will be less audible and won't be amplified.
 
BillFitzmaurice said:
The advantage to having the crossover external to the cabinet would be the additional distance of the inductors from the drivers, which would reduce cross-talk.

Keep the interconnects as short as possible, the speaker cables long. Interconnects are subject to noise pickup, which is then amplified. Any noise picked up on the speaker cables will be less audible and won't be amplified.
Unless - heaven forbid - you use optical interconnects. :O
 
Balancing act

Just how good are balanced lines?


Does anyone know if the White Noise Audio Balanced line driver/receiver kit is any good?

I would quite like to have my gainclones and a couple of other amps on/in my new cabs. I was thinking of having GC + JLH class A + simple valve + digital amp to swap about and play with.

Having them on or very near the speaker would be useful.
 
The main reason why balanced is used almost exclusively in the professional ranks is that it removes the ground line from the signal chain, and it is the ground line that is responsible for the majority of the noise pickup in unbalanced systems, primarily through ground looping. Careful attention to detail, and a good bit of luck, can keep ground induced noise out of unbalanced systems, while in balanced systems the problem seldom arises, if at all.

The bottom line is that when you have the option to use balanced do it. Even better is optical, as there is neither a ground line nor an analog signal there that can be corrupted by outside noise sources.
 
Bill, maybe you know it but just adding an inverting opamp will only make the interconnect "pseudo-differential" removing much of the benefits of going differential. However, with three opamps (or the readymade DRV134 from BB) you end up with a fairly good differential output stage. (True differential signaling requires a transformer AFAIK).

But to answer JRKO, yes it is simple to do a differential ("balanced") interconnect which IMO is superior to a single-ended ("unbalanced") connection despite the extra circuitry necessary. The best thing of all is that you get rid of those horrible RCA connectors and can stock up on some nice XLR's :D

/Magnus
 
Perhaps you misunderstood the topology I'm alluding to. By taking the signal output and running it through a unity gain inverter you create a negative output (180 degrees out of phase) which when used along with the original signal gives you a fully balanced output; many commercial designs use precisely this configuration. Of course you can also use a single line-driver opamp, but essentially all that does is to derive both the positive and negative output sources from a single chip. In a retrofit situation I'd go with the additional inverting stage only for simplicity, as one dual chip would handle the inverting chore for both channels. I've done this many times converting unbalanced to balanced and it functions perfectly. I'd go with the line driver for a new construction.
 
Bill,

Yes I know the topology you are speaking of very well, and it's what is generally called "pseudo-differential". And yes, it is very common in ("pseudo"!) pro gear because it is much simpler and cheaper than a true differential output.

I will try to briefly explain why this configuration is not true differential, but you would be better off reading some articles about the subject, coz' my explanation might not be the best. ;)

First, I generally use the word differential signaling rather than "balanced" because strictly technically speaking "a balanced circuit is symmetric with perfectly matched elements on either side of an axis of symmetry" ("Analysis and Design of Analog Integrated Circuits 4th Edition", ISBN 0-471-32168-0). This is thus just a theoretical situation which will never be reached in practice.

What is the meaning of truly differential? In a perfect differential audio interconnect you could connect the output source (say the preamp) to the input source (say the amp) with just the ground line and the + line (HOT) and leave the - line (COLD) open and you would then have NO signal transmission. This is because the outputs are "floating" with respect to ground - all signal information is carried between the + and - conductor and leaving one of them open leads to no signal transmission. On a sidenote, a speaker could be said to be nearly perfectly differential. It only acts on the voltage difference between it's terminals and not to the (normally non-existent) grounded case.

This would not be the case with the "pseudo-differential" output stage. In fact, what you do is just to add another single-ended output but with reversed polarity. You could do the same experiment as above, but now you would indeed have signal transmission when leaving the - conductor open (assuming the input stage of the amp is not transformer-coupled, i.e. fully differential, it rarely is). Adding the - conductor would just yield an input signal voltage that is multiplied by 2.

As a technical figure of merit for a differential output stage you state the differential-mode impedance vs the common-mode impedance. A perfectly differential ouput would have zero DM impedance and infinite CM impedance. The "pseudo-differential" output stage would typically have say 75 ohm CM impedance and 150 ohm DM impedance, clearly not very differential at all! The DRV134 from Burr-Brown has typically a DM impedance of 50 ohm and a CM impedance of 1.6 kohm, pretty good in other words.

As I said, my explanation is quite bad and I would suggest some reading on the subject. Have a look at these articles from Jensen:

http://www.jensen-transformers.com/apps_wp.html
Also, more info and some DIY designs from Rod Elliott:
http://sound.westhost.com/balance.htm

As to answer JRKO, the DRV134 from Burr-Brown (now Texas Instruments) is an integrated monolithic differential line driver using the three opamp configuration with on-chip precision-trimmed resistors. All you have to do is pretty much to power the chip with a split power supply (+/- 15 V) and you're all set to go! Go to Texas' web site and do a search for DRV134:
http://www.ti.com
They even offer free samples so you could be up and running for no cost at all in a few days!

Cheers
/Magnus
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.