Jump to content
Ricci

Amplifier Comparison SpeakerPower SP2-12000 and Powersoft K20-DSP-Aesop

Recommended Posts

Break-in could definitely explain some of that difference, but I'd expect the gain to be limited to the low end.  I also noticed the shift of the impedance peak.  AIUI, suspension compliance can also be very sensitive to temperature, with more compliance seen at higher temperature.  That may be a stronger effect than that of the temperature on the air density and speed of sound.

 

I hadn't noticed the transformation of humps into peaks in the upper end of the response.  That is very interesting.  What do you mean by differing polarity?  I'm pretty sure both amps are running in bridged mode, so the voltage on the output terminals should always be inverted exactly, relative to one another.  Perhaps what we're seeing here is another consequence of temperature and/or different amounts of break-in on the soft parts.  I can see these things causing upper frequency resonances to shift or change.

Share this post


Link to post
Share on other sites

I simply mean that one amp is reversed signal polarity from the other. It shouldn't affect anything.

 

Here are some graphs including the max burst for both.Note that while there are some differences in the 21Ipal FR they aren't anywhere near the RF differences. I've never seen anything like that. It almost looks like there is some kind of oscillation as the two measurements seem to flip flop back and forth. I'm stumped.  

 

 

post-5-0-19946400-1473886555_thumb.png

post-5-0-14143800-1473886580_thumb.png

post-5-0-30121400-1473886615_thumb.png

post-5-0-83045900-1473886620_thumb.png

post-5-0-90792000-1473886703_thumb.jpg

post-5-0-12207000-1473886728_thumb.jpg

 

Share this post


Link to post
Share on other sites

I'm planning to go ahead and do the tests again using the SP amp and the K20 both on the same day to reduce the margin of error. Really the only cabs I have that will somewhat handle those amps are the dual 21Ipal and the RF T3. The RF T3 is a bit sketchy on the 6000 during the long term sweeps. It was taking a LOT of power. Not sure how many of those it can take. I'll need to allow a cool down period in between to get the temps down and allow the sensitivity to settle back into the starting point.

 

I'd like to do more of this type of testing of amps for fun but I need a good 4ohm nominal load and multiple drivers with a lot of excursion. Problem with the RF and Ipal drivers is they are single coil only and the impedance is so low. I really don't want to buy yet another 2 ohm T3 just for a 4ohm load for this type of testing. For example an Inuke 6K isn't even rated for 2 ohms nominal. It would be best to have multiple smaller drivers with dual voice coils to keep the size down and allow more wiring flexibility. Perhaps a whole boatload of low cost 10" or 12" drivers.

Share this post


Link to post
Share on other sites

I'm using two of these: https://www.amazon.com/Noctua-NF-F12-PWM-Cooling-Fan/dp/B00650P2ZC/ref=sr_1_1?ie=UTF8&qid=1473874503&sr=8-1&keywords=noctua+120mm

 

EDIT: WHOOPS! It was actually these ones, MemX. Hope you didn't order just yet.

 

https://www.amazon.com/gp/product/B00BEZKX8Y/ref=oh_aui_detailpage_o08_s00?ie=UTF8&psc=1

 

HUUUUGGGGEEE drop is fan noise.

 

Will obviously not cool as well as the stock fans though.

 

Awesome, cheers!

 

I'm liking them specs noise-wise:

 

http://noctua.at/en/products/fan

http://noctua.at/en/nf-s12a-flx

 

 

Size
120x120x25 mm
Mounting hole spacing
105x105 mm
Connector
3-Pin
Bearing
Blade Geometry
S-Series with Anti-Stall Knobs
Rotational Speed (+/- 10%)
1200 RPM
Rotational Speed with L.N.A. (+/- 10%)
900 RPM
Rotational Speed with U.L.N.A. (+/- 10%)
700 RPM
Acoustical Noise
17,8 dB(A)
Acoustical Noise with L.N.A.
10,7 dB(A)
Acoustical Noise with U.L.N.A.
7,4 dB(A)
Static Pressure
1,19 mm H₂O
Static Pressure with L.N.A.
0,73 mm H₂O
Static Pressure with U.L.N.A.
0,51 mm H₂O
Max. Input Power
1,44 W
Max. Input Current
0,12 A
Voltage
12 V
MTBF
> 150.000 h
Scope of Delivery
  • Low-Noise Adaptor (L.N.A.)
  • Ultra-Low-Noise Adaptor (U.L.N.A.)
  • 3:4-Pin Adaptor
  • 30cm Extension Cable
  • 4 Vibration-Compensators
  • 4 Fan Screws
Warranty
6 years

 

 

<20dB is QUIET - my background noise level is something like 30dB in my place, IIRC.

 

 

 

I can't remember what the flow rate is on the standard fans, I found them somewhere previously but it was amost a googlewhack :lol: 

 

I think they're something like 100cfm??  Which is a LOT but I guess fair enough for an amp that is designed to be backstage at a gig, running 5kw of bass through it for 3 hours solid.

 

 

 

EDIT:  Yep, found the specs - https://www.avforums.com/threads/fan-replacement-thread.1799166/#post-19561094

 

Jamicon JF1225B2UR-R

100.5CFM / 170 cubic metres per hour

45.7dB(A) (!!)

 

 

Given how underworked these amps usually are day-to-day when just at 'normal' volumes on music and TV, I think a change would be good... :D lol

Share this post


Link to post
Share on other sites

Or maybe just buy another amp or a bigger amp. My fans in my Crest arent as good as the Noctua's but I cant say I have ever pushed my amp hard enough to worry. I will in the future be using the Noctua fans for any other amp mods although I plan on just saving my coins and buying three SP amps. Two four channel amps and one sub amp. I cant keep up with the Jones's so eight amplifier channels will be good for any of my crazy ideas.

Share this post


Link to post
Share on other sites

I simply mean that one amp is reversed signal polarity from the other. It shouldn't affect anything.

 

Here are some graphs including the max burst for both.Note that while there are some differences in the 21Ipal FR they aren't anywhere near the RF differences. I've never seen anything like that. It almost looks like there is some kind of oscillation as the two measurements seem to flip flop back and forth. I'm stumped.  

 

 

attachicon.gif21-IPAL DO Seal Amp FR Comparison.png

attachicon.gifRF T3S2-19 Amp FR Comparison.png

attachicon.gif21-IPAL DO Seal Amp max long term Comparison.png

attachicon.gifRF T3S2-19 Amp max sweep Comparison.png

attachicon.gif21-Ipal DOS Max Burst.jpg

attachicon.gifrf t3s2-19 sealed MAX BURST AMP.jpg

OK Josh (and other bass overlords), here's my hypothesis 

 

Underlying assumptions

At such low impedance loads, neither amplifier's power supply can sustain max-burst levels for any significant period of time. The Speakerpower's power supply can produce more current and has greater output capacitance. The Powersoft's power supply produces much higher voltage rails. The drivers compliances have loosened up since testing with the K20. Rigorous testing also broke in the cones/spiders/dust caps, enough to shift breakup modes. This last assumption is a stretch, but it's the only explanation I can think of for the last conclusion.

 

Explanation of max burst results - K20 wins on the high end

At the high end, the max burst performance is voltage limited, not distortion limited. You estimated the K20 to be able to produce ~400v bursts. At system resonance, current draw is minimal, so the inferior current capabilities make no difference. Same at the high end where driver inductance drives up system impedance. 

 

Explanation of max burst results - SP2-12000 wins on the low end

This is likely to be the breaking-in of the driver. Below system resonance, the compliance starts pulling the cone toward the center before the motor finishes pushing it all the way to it's peak positions. We already know that this is the main source of distortion at the extreme low end. The loosened compliance helped reduce distortion further, contributing to the higher achievable SPLs before breaking CEA-2010 thresholds. The lower system resonances also lowered the frequency of the impedance peak, reducing current and extending Xmax at a slightly lower frequency than in previous K20 testing. Better output capacitance could also prevent the rails from sagging, which could introduce some distortion in the K20 as well.

 

Explanation of long term results - K20 wins on the high end

Higher available voltage offers greater achievable SPL on the high end. Tightened compliance during these tests also helps high end performance. Higher impedance in the high end due to the impedance peak and driver inductance minimizes the effect of the lower-current power supply.

 

Explanation of long term results - SP2-12000 wins on the low end

More current capacity from the power supply offers more low-end maximum performance.

 

Explanation of sensitivity measurements - SP2-12000 wins on the low end

I don't see how this could have anything to do with the amplifers themselves. I think this is solely due to loosened compliance.

 

Explanation of sensitivity measurements - peaks and troughs seem switched between tests

Same with this anomaly. I don't see how going from one flat-response amplifier to another could cause any change here. I think the breakup modes have shifted. On both of these drivers, breakup modes are very mild, exhibiting well damped resonances that create soft lumps in the FR instead of sharp peaks. It doesn't look to me like the peaks have become troughs and vice-versa, it looks to me like all but one or two of the resonances have shifted lower in frequency. The one that I can't really explain is the blip around 150Hz. This can be seen in all your measurements of similarly sized boxes, which is sure to be a rear chamber resonance. I don't see how this could have moved if the enclosure is the same size. Maybe the damping material has settled, affecting speed of sound and therefore perceived depth of the rear chamber to the acoustic waves?

  • Like 1

Share this post


Link to post
Share on other sites

 

Explanation of sensitivity measurements - SP2-12000 wins on the low end

I don't see how this could have anything to do with the amplifers themselves. I think this is solely due to loosened compliance.

 

Explanation of sensitivity measurements - peaks and troughs seem switched between tests

Same with this anomaly. I don't see how going from one flat-response amplifier to another could cause any change here. I think the breakup modes have shifted. On both of these drivers, breakup modes are very mild, exhibiting well damped resonances that create soft lumps in the FR instead of sharp peaks. It doesn't look to me like the peaks have become troughs and vice-versa, it looks to me like all but one or two of the resonances have shifted lower in frequency. The one that I can't really explain is the blip around 150Hz. This can be seen in all your measurements of similarly sized boxes, which is sure to be a rear chamber resonance. I don't see how this could have moved if the enclosure is the same size. Maybe the damping material has settled, affecting speed of sound and therefore perceived depth of the rear chamber to the acoustic waves?

 

Changed compliance surely plays a part. The impedance results indicate that the peak has shifted down some. I guess it could be possible that some of the breakup nodes have shifted as well. You can see a baffle step in all of the results as well. Nothing in the enclosures has changed.

 

I'll be revisiting this at some point so I can do a comparison back to back on the same day. Might be interesting to check that same RF driver every year and see how much it changes over the next few years of use. Makes me wonder how far the suspension will soften and what the original brand new driver will look like compared to 3 or 4 years of use later.

Share this post


Link to post
Share on other sites

Changed compliance surely plays a part. The impedance results indicate that the peak has shifted down some. I guess it could be possible that some of the breakup nodes have shifted as well. You can see a baffle step in all of the results as well. Nothing in the enclosures has changed.

 

I'll be revisiting this at some point so I can do a comparison back to back on the same day. Might be interesting to check that same RF driver every year and see how much it changes over the next few years of use. Makes me wonder how far the suspension will soften and what the original brand new driver will look like compared to 3 or 4 years of use later.

 

These changes are very small, what I have found in the past is that temperature difference can make a larger impact.  As an extremely example, a motor chilled below 0 deg. can yield as much as 1.5dB greater burst output with max power which is pretty remarkable.

Share this post


Link to post
Share on other sites

That old freezer is about to become DB's top rated LLT.

 

Any day now.

 

haha

Ya. me wonders about the cheap liquid cooling options for PCs these days and how they might work on motors...

they are probably not nearly big enough but the idea seems interesting :)

Share this post


Link to post
Share on other sites

Would be interesting if it could be made to work as such.

 

I think of the cooling on some of the Sundown and Stereo Integrity drivers with lots of little holes and how those could be coolant pipes or something. I don't know how to get liquid coolant into the motor and around the voicecoil without saturating the open gap with liquid coolant and most likely preventing proper action of the system.

 

There is no way other people haven't thought about this sort of thing. I guess it would require some serious R&D and engineering just to test it.

 

Might be easier to water cool an amplifier.

Share this post


Link to post
Share on other sites

Maybe have the coil rest in some ultra-low viscosity, thermal-conducing fluid, sort of like Ferrofluid tweeters. What we need is Ferrofluid subwoofers! 

Share this post


Link to post
Share on other sites

Would be interesting if it could be made to work as such.

 

I think of the cooling on some of the Sundown and Stereo Integrity drivers with lots of little holes and how those could be coolant pipes or something. I don't know how to get liquid coolant into the motor and around the voicecoil without saturating the open gap with liquid coolant and most likely preventing proper action of the system.

 

There is no way other people haven't thought about this sort of thing. I guess it would require some serious R&D and engineering just to test it.

 

Might be easier to water cool an amplifier.

Air cooling still has a lot of potential before liquid cooling is needed, I have some ideas for some really crazy stuff with unique air cooling system(I do have some ideas for liquid cooling as well), but nothing that's really "practical" either way. That wont stop me though :ph34r:, just might be a while.

Share this post


Link to post
Share on other sites

You'd think there'd be a way to use some type of ceramic material to coat the VC vs. glue.  I think I remember Nick from SI saying even the best VC glues will melt around ~450-500F. 

 

The Kilns used to make ceramic pottery get up around 1,800-2,400F, so I highly doubt any amount of a heat from a VC could damage it.  Ceramic tiles protect the space shuttle from re-entry for crying out loud...talk about heat dissipation!

 

I suppose when the coils heat up and expand it'd just break/crack ceramic material?  Maybe some sort of Artic Silver type of "filler" could be a buffer between the coil and ceramic shell?  Could that be done without creating air pockets?

Share this post


Link to post
Share on other sites

That's true.

 

I feel like I've asked this before but why are we still using voicecoil technology and not doing something like the old ServoDrive ContraBass system? That had a conventional looking cone but instead of a normal voicecoil it was belt driven by an electric motor system.

Share this post


Link to post
Share on other sites

Modern voice-coil subwoofers are currently capable of maxing out the linearity of even the largest FEA optimized surrounds. They're much simpler to make and have far fewer moving parts. There's also no static friction to cause distortion at low levels, or kinetic friction to cause wear at high levels. Cost is much lower and reliability is much better. Plus they can often be useful over much greater bandwith due to less inertia.

 

Danley's original ServoDrive transducers had 12-16mm xmax. Outstanding when 4mm was the norm, but pretty lackluster by today's standards.

 

Surely a servodrive with modern materials and engineering methods would be quite impressive indeed, but the cost would certainly be unreasonable. I don't know how you could manage surround noise, let alone mechanical noise with such excursions.

 

If Danley's patent expired, it's telling that no one has bothered to come up with a copycat product. If he still holds the patent, it's equally revealing that he doesn't bother making speakers with it while the idea is still protected.

  • Like 1

Share this post


Link to post
Share on other sites

That's true.

 

I feel like I've asked this before but why are we still using voicecoil technology and not doing something like the old ServoDrive ContraBass system? That had a conventional looking cone but instead of a normal voicecoil it was belt driven by an electric motor system.

 

 

Modern voice-coil subwoofers are currently capable of maxing out the linearity of even the largest FEA optimized surrounds. They're much simpler to make and have far fewer moving parts. There's also no static friction to cause distortion at low levels, or kinetic friction to cause wear at high levels. Cost is much lower and reliability is much better. Plus they can often be useful over much greater bandwith due to less inertia.

 

Danley's original ServoDrive transducers had 12-16mm xmax. Outstanding when 4mm was the norm, but pretty lackluster by today's standards.

 

Surely a servodrive with modern materials and engineering methods would be quite impressive indeed, but the cost would certainly be unreasonable. I don't know how you could manage surround noise, let alone mechanical noise with such excursions.

 

If Danley's patent expired, it's telling that no one has bothered to come up with a copycat product. If he still holds the patent, it's equally revealing that he doesn't bother making speakers with it while the idea is still protected.

 

The patent is long expired now, but the real issue was always cost, engineering, parts sourcing and testing required to get the mechanics refined and durable enough.  There is much more possible now with modern cone/surround design.  Before Tom left ServoDrive, followed by myself and later Jeff (JTR), we did have a pretty cool 2nd generation prototype in testing.  There was a concept for a preferred 3rd generation that would require more development and work with parts suppliers.  Ultimately you need applications where high motor force and high excursion are needed for the advantages to outweigh the hurdles and cost, and/or you would need to have a custom built and optimized motor made.

 

It does start to get interesting again when comparing to 6" VC's the motor strength and excursion required for 18-32" cones.  If someone has $50-100k they want to gift away to get them to production I'd be happy to make it happen.  :rolleyes:

 

-Mark

  • Like 2

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×