I once had a cart and I placed it before the horse. I don't understand why, but it never did drive well.
It was because your horse had increased IM distortion.
Dave.
It is far worst than that. If you study amplifier input stages, you will see where this is a disaster for distortion too.
Now, a passive filter isolated by buffers between the preamp and power amp. Excellent idea. We call it an active crossover.
Passive crossovers are used because they can be pretty good, are what 99.9% of the market expects and on a system wide basis, cheap. Active crossovers are superior in all performance aspects but are complicated and due to all the extra amps and cables, not very viable in the market unless as was mentioned, incorporated internally in self powered speakers. If you want to understand the details, The good D. Self just published over 500 pages on the subject. Not the last word, and some points a few might argue with, but a worthwhile read. It contains far more knowledge to be absorbed than this thread. Why am I here? Second cup of coffee before I mow the yard and get back to simulating with my DCX what I intend to build in analog active crossover next.
It is a simple as that. If you want to look for the dime you dropped, go look where you dropped it, not under the light because you can see better. There ain't nutin' there to find!
Can you explain why a low impedance at the input gives distortion. My limited knowledge doesnt understand.
An active filter is not a passive filter isolated by buffers but uses an amp in positieve feedback configuration at the edge of oscillation making them critical, to make poles in the complex domain to simulate coils.
An active filter is not a passive filter isolated by buffers but uses an amp in positieve feedback configuration at the edge of oscillation making them critical, to make poles in the complex domain to simulate coils.
Are you sure? I don't understand.
Please explain in language we/I can understand.
A 2nd or higher order filter needs capacitors and coils. Since coils are expensive, bulky and has a lot of paracitiec problems, coils are avoided in active filters. This can be done with a Gyrator or an amplifier with an amplification factor of exactly 1. The filter elements (capacitance and resistor) is put in the in phase feedback loop. Since the amp only has an amp factor of 1, it will not oscillate despite the in phase feedback. However this is very critical. If the amp factor is slighty higher then 1 the filter has an oscillation tendency that gives noise and distortion.
How do you ensure that the filter looks into the same load before and after the amp?
I don't make replicas, but only own stuff. But I would measure phase and frequency curve.
I just don't see how you can take a passive filter - let's say 2. order linkwitz reiley (Not electrically of course but measured) - and place it before the amp, and still have a 2. order linkwitz reiley measured on the speaker!
The impedance that a passive filter drives can make a huge difference. The driving impedance will also make a difference.
If you want a passive filter before your amp, you'll have to completely redesign it for the amp's input impedance - if you want the same filter function as when driving a speaker.
Unless this is just a bad joke.
This is why I asked a number of posts ago. When I consult webpages about calculating passive line level filters, they end up very differently from speaker filters.
But still, the original poster uses her speaker filter before the amps and says that it works well. Which surprised me. How could that be?
I seriously doubt it worked well - the filter function would be completely different.
For a passive crossover to work well, you need to know the impedance curve of the device it's driving. In the case of a loudspeaker that is a fairly complex curve and low impedance. For an amplifier the input impedance will be much higher and much flatter.
Of course the original question (as I understand it) was: Is it better to bi-amp and use passive filters than other methods? Not really a DIY question, but an interesting one.
For a passive crossover to work well, you need to know the impedance curve of the device it's driving. In the case of a loudspeaker that is a fairly complex curve and low impedance. For an amplifier the input impedance will be much higher and much flatter.
Of course the original question (as I understand it) was: Is it better to bi-amp and use passive filters than other methods? Not really a DIY question, but an interesting one.
You don't (necessarily) need to know the impedance curve of the driven device.
You can make a transfer function (differential) measurement at each of the driver terminals using the power amplifier output as the reference. This yields an exact electrical transfer function of the stock, passive filter sections since it includes the interaction with the driver(s) complex impedance.
Now you have a "target" response that your before-the-amplifiers crossover (whatever type/topology you might select) should achieve in order to maintain the exact characterisitics of the orignal speaker designers objective.
It's a fairly simple mechanical process that's been outlined/performed by a number of members (including me) of this forum through the years.
It's a silly idea to transplant an existing speaker-level crossover to a position in front of the power amps since that environment of lower signal levels and higher impedances allows to use components much better suited to the task than the large capacitors/inductors used in a high-level network. And as mentioned previously, it places a load on most preamps that they're completely incapable of dealing with.
Anyways, the original posting (and thread title) seems to advocate a bi-amp setup where the existing crossover is split and multiple amplifiers are used with their inputs simply paralleled. Nothing wrong with that. It's called a "passive bi-amp" configuration and has been used by some folks for many years.
Unfortunately, we seem to be dealing with a poster here who doesn't have a good understanding of the issues involved with various bi-amping schemes.
Cheers,
Dave.
You can make a transfer function (differential) measurement at each of the driver terminals using the power amplifier output as the reference. This yields an exact electrical transfer function of the stock, passive filter sections since it includes the interaction with the driver(s) complex impedance.
Now you have a "target" response that your before-the-amplifiers crossover (whatever type/topology you might select) should achieve in order to maintain the exact characterisitics of the orignal speaker designers objective.
It's a fairly simple mechanical process that's been outlined/performed by a number of members (including me) of this forum through the years.
It's a silly idea to transplant an existing speaker-level crossover to a position in front of the power amps since that environment of lower signal levels and higher impedances allows to use components much better suited to the task than the large capacitors/inductors used in a high-level network. And as mentioned previously, it places a load on most preamps that they're completely incapable of dealing with.
Anyways, the original posting (and thread title) seems to advocate a bi-amp setup where the existing crossover is split and multiple amplifiers are used with their inputs simply paralleled. Nothing wrong with that. It's called a "passive bi-amp" configuration and has been used by some folks for many years.
Unfortunately, we seem to be dealing with a poster here who doesn't have a good understanding of the issues involved with various bi-amping schemes.
Cheers,
Dave.
LOL. Yeah, it makes a good joke.
I've often used this to get the exact curves and EQ from an active crossover - it works for that too.
Yes, this a great way to proceed. Then you know exactly what you've got.
I've often used this to get the exact curves and EQ from an active crossover - it works for that too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.