hmmm, henry your recommendations seem quite specific. do you by chance have a design based on the satori + d2608 crossed over at 2.2-2.8khz?
I've used the d2608 with phl1220, at first my xo was 1.9khz, but im not 100% happy with it, so i changed to 2.7khz, thats where the sound became so fluent, detail and yet so easy to listen to. I didnt apply any corrections with its rising response as i designed them for 15-20deg off axis listening window (to smooth out some diffractions)
As for satori, im using raal 70-10d with xover @2.8khz.
So from that experience..
The Satori is probably the best 6" hifi mid/bass ever made, it's hard to wrong with it whatever direction you choose to take it.
As the TPA3116 are low-ish voltage devices you are certainly better off going for the 4 ohm version if you can as this will allow you to deliver more power.
As you are using a DSP, do not overlook the benefits of using a Linkwitz transform circuit. Certainly with only a smaller amount of available power you will not be able to EQ the thing down particularly low, but I'm thinking more along the lines of simply getting you to a realistic target of say an f3 of 60Hz with the alignment that you desire. Even if you don't adjust the f3 of the system you can always use the LT just for setting the Q down to where you desire.
I have used the TPA3118 amplifier in a few systems, one of which is a low SPL nearfield application and they sound very nice. In my application I used them in conjunction with a PCM5142 and used its onboard DSP for system EQ.
TI have recently produced a chip called the TAS5756M. This basically combines a DAC, miniDSP and TPA3118 style amplifier in a single chip. A pair of them would certainly be a nice way to make a simple high quality, digital input, system. That's if you're good with the DIY side of electronics.
As the TPA3116 are low-ish voltage devices you are certainly better off going for the 4 ohm version if you can as this will allow you to deliver more power.
As you are using a DSP, do not overlook the benefits of using a Linkwitz transform circuit. Certainly with only a smaller amount of available power you will not be able to EQ the thing down particularly low, but I'm thinking more along the lines of simply getting you to a realistic target of say an f3 of 60Hz with the alignment that you desire. Even if you don't adjust the f3 of the system you can always use the LT just for setting the Q down to where you desire.
I have used the TPA3118 amplifier in a few systems, one of which is a low SPL nearfield application and they sound very nice. In my application I used them in conjunction with a PCM5142 and used its onboard DSP for system EQ.
TI have recently produced a chip called the TAS5756M. This basically combines a DAC, miniDSP and TPA3118 style amplifier in a single chip. A pair of them would certainly be a nice way to make a simple high quality, digital input, system. That's if you're good with the DIY side of electronics.
What is it that leads people to make that claim, besides the fact that the basket/motor look really cool? The Klippel measurements I've seen were nothing special. Decidedly inferior compared to a Scan Illuminator. True, the Scan is more expensive.
Illuminator is better for bass but Satori is better for midrange. The Satori 4ohm is 91db, its much more appealing for me than others
Klippel might not be stellar, but Zaph's distortion measurements show the most linear motor on a 6" that I've ever seen. Gornir's measurements reinforce this too over at audioexcite. Then there's the extremely well designed cone. The only issue is the small resonance at around 1200Hz, but this only impinges on the 2nd order distortion, so it's not much of an issue.
The only issue I remember seeing of the Satori was that it didn't have particularly good Klippel data out at large excursions. If the driver losing a bit of linearity beyond ~3mm of excursion is important in your application, then yeah, using one of the illuminators would probably be a better choice. But as was said above me, if midrange reproduction is what you're after, Satori for the win.
Besides wasn't the issue with the Satori mainly suspension related? I'd expect this to increase low order distortion more than anything else and be rather innocuous and again only be an issue at larger than average SPLs.
One thing that'd be interesting is if SB made a version with an aluminium cone. That should completely do away with the resonance issue at 1200Hz. Yes it would limit the upper end usability, but for some people that wouldn't be seen as a problem.
SB apparently have a 5" and 7.5" in the works too, although these are not likely to appear particularly quickly.
Well 5th, you explained it a LOT better than my 1 sentence
Yes, i think its inevitable the family will grow
I only use them as midrange, but i can hear a very different presentation of bass than the phl. Both are excellent, although PHL have that certain raw performance appeal to me but most of the time listeners will prefer satori for the transparency and yet its still have that warmth of paper cone like.
thanks 5th element, very helpful!
for now I will try the TPA3116 since they were so cheap, but I might decide to upgrade to hypex 180watt amps in the future. The other chips are very interesting. I have done a little electronics work, but I find it too hard to work with surface-mount chips. Eval boards are no problem for me though.
I will definitely consider a Linkwitz transform circuit.
Maybe I am missing something, but the 8ohm satori models a little better with an F3=65hz while the 4ohm models with an F3=81hz. Even though the 4ohm is more efficient and would allow for more DSP eq gain, wouldn't the 8ohm require less eq to get the right amount of bass correction? I think (based on the unibox) I would only need 4db of low shelf eq to flatten out the 8ohm 15L satori to around a 50hz F3. I think the 4ohm satori would need closer to 6db low shelf to get a 50hz F3. The 8ohm is modeled with 7.7watts and the 4ohm needs only 6.5watts.
But yeah if the 4ohm tests as cleanly (low distortion, clean midrange) as the 8ohm, then maybe it would be better to go for the more efficient 4ohm satori. They model close enough, and the TPA3116 is rated for 50watts @4ohm.
I attached unibox models for both 15L and 17L sealed for both 4ohm and 8ohm satori based on the published driver specs.
I still have not completely decided if I will go for sealed or vented. I would prefer sealed for the bass transients, but I also want to have an F3 in the 50hz range. So far the 15L sealed satori (4ohm or 8ohm) looks I could hit an F3 ~55, F6 ~45, F10 ~35 with a little bass shelving which would be wonderful.
for now I will try the TPA3116 since they were so cheap, but I might decide to upgrade to hypex 180watt amps in the future. The other chips are very interesting. I have done a little electronics work, but I find it too hard to work with surface-mount chips. Eval boards are no problem for me though.
I will definitely consider a Linkwitz transform circuit.
Maybe I am missing something, but the 8ohm satori models a little better with an F3=65hz while the 4ohm models with an F3=81hz. Even though the 4ohm is more efficient and would allow for more DSP eq gain, wouldn't the 8ohm require less eq to get the right amount of bass correction? I think (based on the unibox) I would only need 4db of low shelf eq to flatten out the 8ohm 15L satori to around a 50hz F3. I think the 4ohm satori would need closer to 6db low shelf to get a 50hz F3. The 8ohm is modeled with 7.7watts and the 4ohm needs only 6.5watts.
But yeah if the 4ohm tests as cleanly (low distortion, clean midrange) as the 8ohm, then maybe it would be better to go for the more efficient 4ohm satori. They model close enough, and the TPA3116 is rated for 50watts @4ohm.
I attached unibox models for both 15L and 17L sealed for both 4ohm and 8ohm satori based on the published driver specs.
I still have not completely decided if I will go for sealed or vented. I would prefer sealed for the bass transients, but I also want to have an F3 in the 50hz range. So far the 15L sealed satori (4ohm or 8ohm) looks I could hit an F3 ~55, F6 ~45, F10 ~35 with a little bass shelving which would be wonderful.
Attachments
OK looking at the published frequency responses for the 8ohm and 4ohm satori, it looks like the 4ohm rolls off less and more evenly below 100hz. These are from SB acoustics, but from what I have read, their specs are close to what the independent testers have found (zaph, audioexcite...)
Maybe only 8db down at 50hz for the 4ohm
while the 8ohm looks to be closer to 12db down at 50hz
Is that bump at 30-40hz real?
OK I think I am convinced on the 4ohm Satori.... now just to decide on a tweeter. Still think I will go for the RS28F based on what I have read.
Maybe only 8db down at 50hz for the 4ohm
An externally hosted image should be here but it was not working when we last tested it.
while the 8ohm looks to be closer to 12db down at 50hz
An externally hosted image should be here but it was not working when we last tested it.
Is that bump at 30-40hz real?
OK I think I am convinced on the 4ohm Satori.... now just to decide on a tweeter. Still think I will go for the RS28F based on what I have read.
Last edited:
There are a few things that one needs to pay attention to when considering these things.
For a start the SB measurements are done on an IEC baffle, which is a relatively big baffle and done in free air. There is no cabinet involved. This means that the drivers response will be dictated by it's Qts now being equal to what would be a sealed box Qtc and the fs of the driver now represents the box tuning. As you can see the free air measurement has little in common with a closed box alignment. That is just because one driver has what appears to be superior performance in free air, does not mean it will have when placed within a box.
The second thing to pay attention to with the simulations is the impedance of the driver at the given frequencies. It might be that you require a certain amount of gain to reach the same SPL, but if you look at the impedance, the 4 ohm version might require less of an overall power increase (relative to the flat area with no EQ) to reach the same level of output.
All of that said, in your application, you are most certainly going to be voltage limited, rather than current. What you should be looking at is the required amount of voltage swing necessary, for a given frequency, with both the 8 and 4 ohm, to see which one has the potential to 'go the loudest', with the limited voltage swing that the TPA3116 amplifiers will have available.
For a start the SB measurements are done on an IEC baffle, which is a relatively big baffle and done in free air. There is no cabinet involved. This means that the drivers response will be dictated by it's Qts now being equal to what would be a sealed box Qtc and the fs of the driver now represents the box tuning. As you can see the free air measurement has little in common with a closed box alignment. That is just because one driver has what appears to be superior performance in free air, does not mean it will have when placed within a box.
The second thing to pay attention to with the simulations is the impedance of the driver at the given frequencies. It might be that you require a certain amount of gain to reach the same SPL, but if you look at the impedance, the 4 ohm version might require less of an overall power increase (relative to the flat area with no EQ) to reach the same level of output.
All of that said, in your application, you are most certainly going to be voltage limited, rather than current. What you should be looking at is the required amount of voltage swing necessary, for a given frequency, with both the 8 and 4 ohm, to see which one has the potential to 'go the loudest', with the limited voltage swing that the TPA3116 amplifiers will have available.
thanks 5th, that was very well explained.
I looked up the spec sheet on the TPA3116 http://www.ti.com/lit/ds/slos708c/slos708c.pdf and from what I can tell the max voltage is maybe 14v or 17.5v with max current maybe 7.5amps. I think this just means 50watts 4ohms, 25watts 8ohms, 12watts 16ohms. At lower impedance it becomes current limited.
Looking at the impedance of the 4ohm and 8ohm satori I can see what you are saying now. ohms law I=V/R P=I*V. Even the 4ohm satori at 40hz is 18ohm or about 10watts max from the TPA3116. I will not be getting a lot of bass from these drivers with this amp in a sealed box.
I really want to try an aperiodic enclosure for these speakers since it is supposed to flatten the driver impedance peak. Aperiodic Loudspeaker Enclosure Design
I can definitely see that with my application and these amps that the 4ohm Satori is the much better option.
I looked up the spec sheet on the TPA3116 http://www.ti.com/lit/ds/slos708c/slos708c.pdf and from what I can tell the max voltage is maybe 14v or 17.5v with max current maybe 7.5amps. I think this just means 50watts 4ohms, 25watts 8ohms, 12watts 16ohms. At lower impedance it becomes current limited.
Looking at the impedance of the 4ohm and 8ohm satori I can see what you are saying now. ohms law I=V/R P=I*V. Even the 4ohm satori at 40hz is 18ohm or about 10watts max from the TPA3116. I will not be getting a lot of bass from these drivers with this amp in a sealed box.
I really want to try an aperiodic enclosure for these speakers since it is supposed to flatten the driver impedance peak. Aperiodic Loudspeaker Enclosure Design
I can definitely see that with my application and these amps that the 4ohm Satori is the much better option.
That isn't quite what I was meaning. The bass response that you will get out of the driver will closely mirror what you get in the simulation program. Amplifiers are constant voltage devices and vary the current they put out and thus the power delivered, based off of the impedance they are driving.
What I was trying to get at was that to work out the power delivered to the driver you need to look at the impedance too. It might appear that the 4 ohm version needs 6dB of boost, by a given frequency, to reach 'flat' and the 8 ohm version needs only 4dB of boost. But if the 4 ohm version has reached an impedance maximum, around the box tuning, then the power required to actually give you 6dB might be lower than what the 8 ohm version needs.
If the impedance shoots up and the power therefore goes down, this is a good thing because it means the efficiency around that point has increased.
TI rate their class D amplifiers at 10% THD, which is ridiculously high and you will never want to reach this point during normal listening. If you look at the curves of power vs distortion you will see that it remains roughly flat up to a point, then sky rockets. You will be wanting to use the amplifier up to the point before it sky rockets.
I think I am getting this. I should look at the response and impedance from the models not the IEC open baffle response from the SB graphs when considering power delivery. And I need to consider both response and impedance.
If I put in a sealed box model of 1000L I approach open baffle and the response and impedance of the model does look like the SBA plots. I can also see that changing the fill (min fill, walls covered, heavy fill) changes the impedance quite a bit.
Here are the impedance plots for Satori 4ohm and 8ohm in 15L sealed (min fill). 4ohm satori impedance peaks at FB=58hz and 46ohms, the 8ohm satori peaks at Fb=59hz and 67ohms. At 50hz the 4ohm satori impedance is 20ohms, while the 8ohm satori is 27ohms at 50hz.
I am trying to do some calculations with power requirement to boost at these low frequencies but they don't seem to make sense. I think I am missing something here. I will keep studying up on this.
5th thanks so much for your patience trying to teach me this. I will trust you on your 4ohm Satori recommendation for use with the TPA3116.
Overall the 4ohm does look good, more efficient, and the amp distortion looks better at 4ohm in the higher frequencies... as long as I keep the power / max SPL down.
I think I can also see also why you recommend the linkwitz transform network. ESP - The Linkwitz Transform Circuit
The miniDSP can do this with a custom biquad
Linkwitz Transform | MiniDSP
If I put in a sealed box model of 1000L I approach open baffle and the response and impedance of the model does look like the SBA plots. I can also see that changing the fill (min fill, walls covered, heavy fill) changes the impedance quite a bit.
Here are the impedance plots for Satori 4ohm and 8ohm in 15L sealed (min fill). 4ohm satori impedance peaks at FB=58hz and 46ohms, the 8ohm satori peaks at Fb=59hz and 67ohms. At 50hz the 4ohm satori impedance is 20ohms, while the 8ohm satori is 27ohms at 50hz.
I am trying to do some calculations with power requirement to boost at these low frequencies but they don't seem to make sense. I think I am missing something here. I will keep studying up on this.
5th thanks so much for your patience trying to teach me this. I will trust you on your 4ohm Satori recommendation for use with the TPA3116.
Overall the 4ohm does look good, more efficient, and the amp distortion looks better at 4ohm in the higher frequencies... as long as I keep the power / max SPL down.
I think I can also see also why you recommend the linkwitz transform network. ESP - The Linkwitz Transform Circuit
The miniDSP can do this with a custom biquad
Linkwitz Transform | MiniDSP
Attachments
since I am using a miniDSP it should be ok to have multiple profiles
one profile with linkwitz transform when I go stand alone without sub (normal day mode) to give me flat to ~50hz
another profile with bass roll off to integrate with the subwoofer on the weekends or when I really need to go full-range and need to check the sub-bass frequencies on my mixes.
this is one of the reasons I really wanted to go DSP/active, to have this kind of flexibility. but you are right that I don't want LT when I plan to integrate the sub.
one profile with linkwitz transform when I go stand alone without sub (normal day mode) to give me flat to ~50hz
another profile with bass roll off to integrate with the subwoofer on the weekends or when I really need to go full-range and need to check the sub-bass frequencies on my mixes.
this is one of the reasons I really wanted to go DSP/active, to have this kind of flexibility. but you are right that I don't want LT when I plan to integrate the sub.
jeshi, did you take a look at the Scan Speak 15W/8530K01? It is like the 15W/8530K00 but better suited for a sealed box due to the higher QTS/QES, so it should be able to go lower in frequency than the K00 in a sealed box: http://www.scan-speak.dk/datasheet/pdf/15w-8530k01.pdf
I simulated it on WinISD and I get an F3 around 55Hz with a 12.2L box.
I simulated it on WinISD and I get an F3 around 55Hz with a 12.2L box.
Last edited:
I've been working on exactly the same question. What is the "ideal" size for the 8ohm Satori? Your modeling above really helps. If I could get an F6 at 40hz, I think I would be happy.
My project, much like yours, will be active with two bass modules. I already have the HDS tweeter, so that's what I'm going with. i may save up for the Raal tweeter. The DSPs real strength is the flexibility.
My project, much like yours, will be active with two bass modules. I already have the HDS tweeter, so that's what I'm going with. i may save up for the Raal tweeter. The DSPs real strength is the flexibility.
I`ve had the K00 Scan and the SB 15NRXC ( paper ) and 15MFC ( poly ). If i had to choose today, I would go for the SB paper version, its 5 times cheaper than the Scan and comes very close to it, was really hard to say which one is better but the Scan is less sensitive. The paper SB has a very soft rubber suspension.
Based on specs, I would definitely pick up the Satori, looks like an exceptional midwoofer.
Based on specs, I would definitely pick up the Satori, looks like an exceptional midwoofer.
The 4 ohm isn't more efficient than the 8 ohm, it just has a higher voltage sensitivity.
As an example if you have 8 volts across 8 ohms 1 amp will flow and 8 watts will be dissipated.
If you now replace that 8 ohm resistor with a 4ohm one but keep the voltage the same, 2 amps will flow. Now 16 watts are dissipated. If you want to dissipate 8 watts, like above, you need to reduce the voltage drive down to 5.66 volts.
It is easier to get more watts if the resistance is lower. It is the same with the Satori. If you are voltage limited and cannot turn the volume up any higher then the only way to get more output is to lower the impedance of the driver.
For a given SPL both drivers will require the same amount of power, it's just that the 4 ohm version will get there with less voltage than the 8 ohm version will.
As you can see, from this point of view it is much better for you to use the 4 ohm version as it matches better with the TPA3116.
What I was getting at before was that if the 8 ohm version, at 1 watt, produces 85dB where it's impedance is flat. But down at 50Hz it's 27 ohms, the power will drop by a factor of ~3.375.
The 4 ohm version will also produce 85dB for 1 watt and down at 50Hz it is now 20 ohms. Power here will drop by a factor of ~5.
The 4 ohm might require a bit more boost in terms of dB, at 50Hz, to reach flat, but the power could either be identical to the 8 ohm version, or less.
Obviously you're still better going for the 4 ohm version as it will have more headroom with the TPA.
The LT can have a number of useful applications and lends itself well to both situations.
If you've put the Satori in a slightly too small box, such that the Qtc might be 0.9, it is quite useful to use the LT just to reduce the system Q down to 0.5, even if you're not going to lower the tuning too. This actually has the effect of lowering the required power around the old system resonance at the expensive of increasing it a little lower in frequency.
In an unassisted (sub less) environment it also gives you the option to lower the response, which is something you need to take care with, as it is easy to over do it, but it is something I wholeheartedly recommend. I mean if you're only going to listen at low SPLs, then by extending the response down to 50Hz, you're just making better use of the resources you've got. Does the driver have the excursion? Yes. Does the amplifier have the power? Yes. Then go for it.
In the assisted situation the LT also makes a lot of sense. Say you want to cross over the sub to the mains at 80Hz. In this case you set the LTs target response to 80Hz with a Q of 0.7. In addition to this you apply a high pass filter with a target of 80Hz and also with a Q of 0.7. Like this the LT and the high pass will sum to give you a 4th order roll off with a Q of 0.5, or a 4th order LW Riley. This is perfect for mating to a sub.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.