Multi-channel amps for active speakers?

[asliarun]

Just thought I would throw in this option (although looks like you have found a better value/money option).

Model 7075 Seven Channel Power Amplifier

Model 7125 Seven Channel Power Amplifier

Both are 7 channel power amps, and deliver true 75W and 125W per channel while still maintaining decent specs at those levels when continuously driven. IMHO, these are priced very well for the performance they deliver.

 

[tcpip]

Model 7075 Seven Channel Power Amplifier

Model 7125 Seven Channel Power Amplifier

Yes, they seem to be pretty good. More than adequate for any 3-way stereo active speaker stack.

I found similar value from the Rotel RMB series of power amps, available at UKP 100 or 150 on eBay UK.

 

[tcpip]

The Minidsp is far more flexible and useful if you have unruly speakers (read metal drivers). but for well behaved speakers the DCX is much simpler to setup) (and allows multiple configurations) (and if you do some of the well documented DCX mods - it is supposed to sound really good too)

Prodded by some recent comments, I was looking at the 2x4 MiniDSP module's specs. It's pretty good, actually! I'd never looked at it very closely because I was focused on the Darbari and I knew I needed 3-way actives, so 2x4 wouldn't do. But here's what I like about the 2x4:

• 24-bit ADC and DAC, which means volume control in the digital domain will work just fine. They've not scrimped on the sample size -- they've just cut the sampling rate to half compared to the 4x10. I can live with that.
• support for a pot based volume control. This means that I can buy a motorised pot remote volume control kit and build a preamp around the 2x4. Have an input selector to select from a few inputs, feed the selected input to the 2x4, control its output volume using the pot, and I have an inexpensive and perfectly usable preamp with 4 channels of signal out for actives
• It has all the flexibility for biquad programming which the 4x10 has, allowing me to do crazy filters to really tame metal-cone drivers (one of my original reasons to look at the 4x10 for the Darbari).

Fantastic. Once the Darbari starts singing, I can almost feel a 2-way coming up, with Dayton RS 150 metal-cone mid-bass drivers from diyaudiocart. Take the output of this xo-preamp, feed four channels into my currently unused Yamaha RX-V800 receiver, and am all set up for a nice small-room system. You can argue about my choice of power amplification of course, but you can't deny that the 2x4 forms the basis for an excellent xo-preamp for two-way actives. Why, one can even make a single-chassis integrated amp with the preamp, xo, and four channels of LM3886 from diyaudiocart at an amazing price point. Just putting heatsinks on the sides of the chassis will be enough to run four of these modules, and everything will easily fit into a 2U or 3U box. Even heatsinks with horizontal fins will work well at these power ratings, thus making it easy to buy heatsinks in India.

The madness continues.

 

[kapvin]

Prodded by some recent comments, I was looking at the 2x4 MiniDSP module's specs. It's pretty good, actually! I'd never looked at it very closely because I was focused on the Darbari and I knew I needed 3-way actives, so 2x4 wouldn't do. But here's what I like about the 2x4:

• 24-bit ADC and DAC, which means volume control in the digital domain will work just fine. They've not scrimped on the sample size -- they've just cut the sampling rate to half compared to the 4x10. I can live with that.
• support for a pot based volume control. This means that I can buy a motorised pot remote volume control kit and build a preamp around the 2x4. Have an input selector to select from a few inputs, feed the selected input to the 2x4, control its output volume using the pot, and I have an inexpensive and perfectly usable preamp with 4 channels of signal out for actives
• It has all the flexibility for biquad programming which the 4x10 has, allowing me to do crazy filters to really tame metal-cone drivers (one of my original reasons to look at the 4x10 for the Darbari).

Fantastic. Once the Darbari starts singing, I can almost feel a 2-way coming up, with Dayton RS 150 metal-cone mid-bass drivers from diyaudiocart. Take the output of this xo-preamp, feed four channels into my currently unused Yamaha RX-V800 receiver, and am all set up for a nice small-room system. You can argue about my choice of power amplification of course, but you can't deny that the 2x4 forms the basis for an excellent xo-preamp for two-way actives. Why, one can even make a single-chassis integrated amp with the preamp, xo, and four channels of LM3886 from diyaudiocart at an amazing price point. Just putting heatsinks on the sides of the chassis will be enough to run four of these modules, and everything will easily fit into a 2U or 3U box. Even heatsinks with horizontal fins will work well at these power ratings, thus making it easy to buy heatsinks in India.

The madness continues.

hi tcpip,

I just saw this post.

Strangely enough, I have a pair of minidsp 2x4 as well. I used it in my car for a 4 way audio

i refer to it at the end of this thread on my car build
Optra Magnum Semi DIY ICE install - Page 6

if you have the time and patience you can read through the whole thread..

essentially, the system way playing from a mid-fi head unit to a hifi amp with this crossover in between.

the passive crossover was barely optimised and used el-cheapo components, but it soundest a bit more "lively" than the active system. I think the ADC is to blame. I have not tried it an all digital system (that was part of the plan which never got executed- digital out from the hU to the dsp and a passive volume control) - currently the car is with my dad in bangalore (did not have the heart to sell it) with a semi-active crossover (and sounding the better for it)

i'd invest in the digital input if I were buying 2x4 for hi fi home use.

I don't remember if I packed up the minidsp 2x4s when I shifted (i think not) and if not, you can borrow it to try out when I next come to India.

best wishes

Kapil

 

[tcpip]

the passive crossover was barely optimised and used el-cheapo components, but it soundest a bit more "lively" than the active system. I think the ADC is to blame.

I guess you know this well, but... a difference in "liveliness" level or "detail" level or "warmth" in speakers or crossovers is almost certainly due to differences in their frequency responses. If two amps sound different in these areas, then the problem could be deeper.

It's almost impossible to build two crossovers with absolutely identical frequency responses, specially in the case of your experiments where you have one built using passive components and the other using active. Therefore, it is possible that the problem will be due to differences in the SPL curves, not the ADC. I have heard even my HTC cellphone's output on my very high-end ER4PT earphones and am amazed at the accurace of modern DAC and ADC. I have read Seigfried Linkwitz being impressed with the DAC output of some setup where his Orions were being driven by a nano-Digi followed by some fairly ordinary FiiO DACs. I will therefore suspect differences in SPL curves first, and ADC or DAC issues later.

Will read your thread. Thanks.

 

[kapvin]

I guess you know this well, but... a difference in "liveliness" level or "detail" level or "warmth" in speakers or crossovers is almost certainly due to differences in their frequency responses. If two amps sound different in these areas, then the problem could be deeper.

It's almost impossible to build two crossovers with absolutely identical frequency responses, specially in the case of your experiments where you have one built using passive components and the other using active. Therefore, it is possible that the problem will be due to differences in the SPL curves, not the ADC. I have heard even my HTC cellphone's output on my very high-end ER4PT earphones and am amazed at the accurace of modern DAC and ADC. I have read Seigfried Linkwitz being impressed with the DAC output of some setup where his Orions were being driven by a nano-Digi followed by some fairly ordinary FiiO DACs. I will therefore suspect differences in SPL curves first, and ADC or DAC issues later.

Will read your thread. Thanks.

Thanks for your post tcpip. Frequency response variations is what I thought at first too.

But do remember that I have reasonably good measurement equipment(by DIY standards) and measuring in the car, while a cramped affair, has a high degree of repeatability, because,ceterus paribus, esp with the listening position locked within a few inches.

And while I used a final target response that was different from the analogue! the issue wasn't resolved by getting a close match to the analogue fr.

I hate using audiophile words as they do not suit me, but would "enjoyable" and "listenable" sound more appropriate.

Again, feel the adc, not the DAC or the processor is the culprit.
Of all ADCs that I have(DEQ, DCX, minidsp4x0 and this) I think the minidsp 2x4 sounds the least involving.

 

[tcpip]

But do remember that I have reasonably good measurement equipment(by DIY standards) and measuring in the car, while a cramped affair, has a high degree of repeatability, because,ceterus paribus, esp with the listening position locked within a few inches.

And while I used a final target response that was different from the analogue! the issue wasn't resolved by getting a close match to the analogue fr.

Okay... now I am worried. :sad: Can't respond off the cuff without more detailed discussion about your setup --- too many variables.

If two setups are being fitted in the same "environment" (i.e. car in your case) and have closely matched FR, and the same drivers, then I can't suggest any pointer to the problem.

When you replicated the xo from passive to active, did you apply the same xo to each driver? I mean, it's not good enough to have the same measured summed SPL and phase, if you want to compare the two xo. You need each driver's xo to be identical in XO 1 and XO 2. For instance, you may be able to build a passive xo with the Fc at 2.5KHz and a second xo with the Fc at 2.0KHz, and both of these may be made to sum and give very similar summed response. But they will sound different, because the different Fc.

Thought I'd ask, since you mentioned something about the final target response "was different".

 

[kapvin]

Okay... now I am worried. :sad: Can't respond off the cuff without more detailed discussion about your setup --- too many variables.

If two setups are being fitted in the same "environment" (i.e. car in your case) and have closely matched FR, and the same drivers, then I can't suggest any pointer to the problem.

When you replicated the xo from passive to active, did you apply the same xo to each driver? I mean, it's not good enough to have the same measured summed SPL and phase, if you want to compare the two xo. You need each driver's xo to be identical in XO 1 and XO 2. For instance, you may be able to build a passive xo with the Fc at 2.5KHz and a second xo with the Fc at 2.0KHz, and both of these may be made to sum and give very similar summed response. But they will sound different, because the different Fc.

Thought I'd ask, since you mentioned something about the final target response "was different".

No problem I explaining.. Though the details are very boring.

Passive crossover was a proof of concept, so not optimised, but was patched together through available components. The .zma of all drivers and the frds, were fed into pcd and the transfer functions optAined. As you know that the transfer function for passive crossover depends on the impedance.

When I switched to active, the first thing I did was to mimic the transfer function digitally. Hope that explains the first. Part.


When I moved to active, the final target response was mapped to a house curve that I had worked out. But what I was comparing was the passive crossover and it's digital copy.

Hope that explains.

 

[tcpip]

No problem I explaining.. Though the details are very boring.

It's these details which make all the difference to someone like me, struggling with speaker design.

When I switched to active, the first thing I did was to mimic the transfer function digitally. Hope that explains the first. Part.

How did you mimic the transfer function? Did you use software to accept the passive XO transfer function as input and spit out a set of biquads to exactly mimic it? Or did you use "the eyeball method" to copy it?

Did you take measurements of the two crossovers after they were built? (These measurements show the impact of component tolerances, etc.) In the final analysis, if the two crossovers differ by even 1 dB in parts of the spectrum, they will sound different. In a home environment, I have heard differences in overall tonality due to 0.5dB changes in tweeter level.

And after all this, if you found no differences measurable in SPL, then I will be curious about the distortion added by the passive xo. But that's a different story. :sad:

 

[kapvin]

How did you mimic the transfer function? Did you usesoftware to accept the passive XO transfer function as input and spit out a set of biquads to exactly mimic it? Or did you use "the eyeball method" to copy it?" - haha. not eye ball.. remember that both transfer functions are mathematical - you really do not need the eyeball at all.

Did you take measurements of the two crossovers after they were built? - I measure each component before putting it in the crossover. since no component is matched 100% for stereo i take the average of the two.(These measurements show the impact of component tolerances, etc.) In the final analysis, if the two crossovers differ by even 1 dB in parts of the spectrum, they will sound differentI know - and you will find that .5db can make a difference in an active design. so the answer is, no the crossovers were not matched to .5 or even 1 db. in the final measurements, they were matched within +-1.5 dB afaik. But since it was an active, I tweaked the values to "straddle the passive response curve. no joy . In a home environment, I have heard differences in overall tonality due to 0.5dB changes in tweeter level.

And after all this, if you found no differences measurable in SPL, then I will be curious about the distortion added by the passive xo. But that's a different story. :sad: -i was as dogged about it as you, but the minidsp always sounded just a bit smeared

this is difficult to explain - but there is a marked difference between an analogue input into the 2x4 and the 4x10 minidsp - like i said, try and you'll see. the issue is not tonal balance, it is just that it is a bit dull

 

[tcpip]

this is difficult to explain - but there is a marked difference between an analogue input into the 2x4 and the 4x10 minidsp - like i said, try and you'll see. the issue is not tonal balance, it is just that it is a bit dull

Okay, I guess I'll have to wait till I build something with the 2x4. I already have the 4x10HD, so I will be able to compare. Thanks for your patience.