Mini Scoops - Reality check |
Post Reply | Page <1 1112131415> |
Author | |||
fatfreddiescat
Young Croc Joined: 15 October 2010 Location: N.E.Wales Status: Offline Points: 1081 |
Post Options
Thanks(0)
|
||
Hi Diaz, any chance of posting an impedance plot?
|
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
It doesn't strictly need to be, it would just help. It entirely depends on how loud a signal you are testing with etc etc. The graphs shown makes perfect sense even if there is some error in them. But yeah you do want it ~1m away for a far field measurement of both driver and port. Sticking it right in the port is fine if you just want to measure what the port is contributing. In this case though, putting the mic into the port and also right in front of the driver for a near field measurement is what's been suggested throughout this thread and it is certainly interesting. It looks like you have the port tuned to just under 40Hz when I zoom in on the graph due to the symmetrical boost in output around that point and the rapid falloff of the driver output when you go lower as it becomes unloaded. It is very interesting to see the dip in output ~100Hz which is presumably the rear wavefront from the port being 180 degrees out of phase with the driver output. But I would expect the dip in response to come from a far field measurement as opposed to mic in front of box or in port. The way the port is showing 10dB of gain at this frequency is a bit odd, if it was a true port it shouldn't be doing much here. Looking at the graph it is happening at about 97Hz which at 340m/s for sound is a wavelength of 3.5m, 1.75m half wavelength, and 88cm quarter length. Do any of the main dimensions of the plan correspond to those numbers? I am thinking standing wave resonance perhaps. Some wadding might help to reduce it if so. To change the tuning you can add wadding, or reduce the rear chamber volume. Is there any chance you can get an impedance plot? Then you can see easily what any changes to the design do to move the tuning very quickly and easily. What sort of volume did you make the graph measurements at? It'd have to be fairly quiet to avoid one of them affecting the other even at near field. |
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
Also could you take a frequency vs SPL reading of the port with microphone at its centre of area (avoid wall effects), and halfway down the length of it? I'd really like to know exactly what it is doing without the effect of the front driver wave involved. Make the measurement quite quietly too. Doesn't matter what absolute dB value it works out too as long as the signal to noise is ok.
|
|||
Diaz
Registered User Joined: 09 November 2011 Location: SBG | Austria Status: Offline Points: 36 |
Post Options
Thanks(0)
|
||
First, we did the measurement at very quiet level to avoid room effects and disturbance of the horn/speaker.
There is this gain around 78Hz we detected as standing wave of the width of the cabinet. The Proto is poorly braced and i think that's why this peak appears. The 88cm Peak could be refereed to the hight of the cabinet (90) or could be the second peak of the horn as it also shows in the Hog Plot around 80Hz? In my understanding the two measurements add and to get rid of the heavy 55Hz excursion i have to damp the back chamber and perhaps give it a little more volume, right? Also more "resistance" /"damping" at the coupling to the horn would be useful so i have to make the horn mouth a little smaller/tighter, right? So then the green line (measurement in front of speaker) in our measurement would skip a few hz down i think. Is it right that this could be achieved through erase Panel "T" and make panel "B" longer a few cm? http://hornplans.free.fr/miniscoopms18mkii600581.jpg Impedance Plot follows in a few days - i hope... it disappeared somehow/or we messed it??? |
|||
SUBSISTENCE Soundsystem https://t.me/subsistencesoundsystem
Mini Scoop Soundsystems Group: https://www.facebook.com/groups/749452582177469/ |
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
Ah right, it is actually a horn. My bad, I thought you had made a
miniscoop with constant cross sectional port area. Well that makes a lot
more sense then. I wondered why there was so much SPL content between
what looked like port resonance and the first higher frequency
resonance. Put it down to a non-ideal measurement but horn loading is
also a good reason!
Panel bracing issues should show up as quite small spikes in the impedance plot, so that is a good way to look for them. And if it is acting as a horn you will see larger spikes in the impedance according to that as well which should correspond with the spl peaks from the nearfield. Tuning a rear loaded horn is a bit of an odd thing. I always assumed that if you could get rid of the rear chamber in a scoop entirely, it might be a design bonus. In a front loaded horn the rear chamber sizing provides some damping and then affects the high frequency roll off point which we don't really care about <90Hz. Maybe in a rear loaded there is very little loading on the air side of the driver so rear compression chamber volume does become more important. I wonder if anyone has ever made a proper scoop without compression chamber, just driver mounted straight to horn throat. I keep saying it but one day I am going to get pen and paper out and model a rear loaded horn enclosure from first principles to really get an idea of what is going on. Anyways... The first must do task if you want to tune it is to take an impedance measurement. Since you only have one of these built (I assume) then it is probably not working as an ideal horn. I.e. the mouth is not large enough in area compared to the length/expansion to get ideal radiation down to cut off and some of it is being reflected back to excite standing waves. There is not a lot you can do about this except build more cabs. Maybe Staiper will comment on his design thoughts if you are lucky. Once you've made an impedance measurement you can have a play at tuning. I don't think wadding in the rear chamber is going to do much for tuning at all but it might reduce sound bouncing off the wall behind the driver and so sound a bit better. Try it and see. Same goes for wadding in the horn. Putting some in the first 1/3 of the expansion might help a bit, or it might not. People do not generally put wadding in horns but giving it a go costs nothing. It may help to damp some standing waves. You can try reducing the chamber volume with blocks of wood or similar but I don't think that is going to do anything significant to your low end output. If the compression chamber is important it might do something. An impedance sweep will tell you instantly if it has done or not. If the peaks move in frequency, resonance has changed. If they shrink in height they are more damped. I don't think you can do much about taming excursion at 55Hz except building more cabs and putting them next to each other, which should load the driver more effectively as you approach the low cut off. I'd be surprised if excursion at 55Hz was more then 40 or 30Hz, have you tried testing the cones with a test signal of constant amplitude as opposed to music. Maybe you did mention that already, apologies if so. Measure the impedance, and then we can see what is going on. I expect it will be behaving like a reflex/transmission line at the lowest frequencies and horn loaded somewhere above that with increasing efficiency Edit - Made some things a bit clearer, removed some waffle :) Edited by odc04r - 26 January 2017 at 6:43pm |
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
The comments regarding the Oberton vs PD sound are interesting, that is the sort of thing that you cannot predict using simple models with idead pistons acting as ideal resonant systems. Drivers do real things, they are not perfect and they do flex under load, bigger cones probably even more so than smaller sometimes. They do have different magnet structures, and different manufacturers are going to have different production tolerances.
Having said all that on paper the biggest reason I would pick for the difference between the 2 is Fs. The PD is significantly higher than the oberton and much more in the region of interest. The PD also has a less lossy suspension (Qms) which I wonder may have something to do with it sounding a little more musical. Electrical damping always dominates when braking a driver, but mechanical damping plays its part when it is being driven too. If you have the kit and knowhow Diaz, it'd be interesting to see an impulse response from both drivers. But you have to be quite careful how that is measured to avoid room reflections, Look into time gated measurements. |
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
3rd thought in a row (really must do some work), it is possible perhaps that if you size the rear chamber of the horn correctly its properties might be able to contribute to reluctance annulling of the non-ideal horn path. That's just a thought from left field, I know it can be done with a front loaded horn but that's a sealed chamber. Not sure the principle is still valid once you put a horn on the other side of that chamber and not the other side of a driver. Becomes series instead of parallel. Once again, something for the pen and paper one day. Could be another reason why some drivers prefer different chamber sizes anyway apart from Q and Vas/Cms.
|
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
Quote is not edit...
Edited by odc04r - 26 January 2017 at 6:43pm |
|||
Diaz
Registered User Joined: 09 November 2011 Location: SBG | Austria Status: Offline Points: 36 |
Post Options
Thanks(0)
|
||
Thank you odc04r for this usefull input and sorry for the late reply!
First, we didn't try it with a sweep but only with music. My friend took the chance this weekend and made a measurement again. It's a goundfloor measurement like last time at 2m distance at a concrete floor. It had -10°C so the Speaker was cold too but it's an outdoor measurement this time. So here we go: Oberton 18NXB1600 in an Staiper Mini Scoop MS18-MK2: It seems we have been wrong with your estimated peak around 55, its obviously around 65. Tuning looks good for sub duties where 35Hz is just right. Underneath it falls like a typical bass reflex... What also makes me wonder are the broad impedance peaks or to say it in other words the small/narrow valleys ;) I'm not sure how to proceed? How to get the 65Hz under control and perhaps the tuning a little deeper... 35Hz are quite ok obviously. Should i got for a smaller rear chamber and a smaller horn throat? Could the MK3 Version be an option? Btw. PD is gone... |
|||
SUBSISTENCE Soundsystem https://t.me/subsistencesoundsystem
Mini Scoop Soundsystems Group: https://www.facebook.com/groups/749452582177469/ |
|||
fatfreddiescat
Young Croc Joined: 15 October 2010 Location: N.E.Wales Status: Offline Points: 1081 |
Post Options
Thanks(0)
|
||
for getting the measurements done, @ -10, makes me feel like a complete woose as have been putting off a measurement day at the unit as to cold - some 15C warmer than where you are.
Impedance plot looks smoothed? Not that it matters. Dip at 35 ish looks like it is a full on helmholtz (reflex) damping effect as impedance drops, to what I would guess to be roughly the DCR of the driver, but is asymmetrical, possibly suggesting a quaterwave effect from the horn, upper half wave effect which is stronger is at approx 95Hz so quaterwave would be half that, this would also push the apparent 65Hz peak up from around 55 ish if it was just acting as a reflex box. There also appears to be a shallow dip in the frequency response above the tuning frequency- possible interaction between the quaterwave mode and the resonant portion of reflex? De-tuning the reflex part or shortening the horn may separate the tuning of the horn and reflex effect far enough for them not to interact? I would at this stage try de-tuning by extending the board behind the driver to restrict the throat area and see if that would work. Thanks again for posting the measurement |
|||
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
||
You're not wrong about the tuning, clear impedance minima at 35Hz like you say. And I agree that at this point the cabinet is acting as a reflex enclosure. Could also be acting 1/4 wave to some extent, but I'm going to have to go with reflex as a 1/4 wave pipe at 35Hz would be 2.5m long and I don't think the horn path is that. Tuning is a touch peaky but I wouldn't worry about that personally. To be fair I said tuning of the port looked like it was about 40Hz from the nearfield SPL plot so I'm pretty happy with that guess! The 30Hz peak is your driver, 65Hz peak is the sealed box resonance where the mass of air in the port makes it look effectively closed. Peaks after that are horn or transmission line related. There doesn't seem to be any real region of horn gain, apart from perhaps 80Hz up. Which as you'd not typically use a box like this much higher does lend weight to the argument of it is effectively a reflex box with some extra port gain. Which makes sense as if it was a true horn I'd expect to see a more peaky lower response anyway due to low mouth area. You can see that the second horn impedance peak is already very damped due to the sharp folding of the cabinet at about 150Hz. And around this region you have the classic 1/2 wave cancellation notch too. Have you got any damping material in the rear chamber like the plan suggests? If not it would both lower the impedance peak at 65Hz, and shift the port tuning slightly lower. I'd definitely try that. Apart from that I don't think the cabinet really needs anything else. Experiment with different amounts of stuffing in the upper chamber until you find a measurement and sound that you like. The 65Hz resonance is interesting, think of it as the box having some ability to store energy when operating at this frequency. The more damped the peak is the quicker it will release that energy when you stop driving it. There is a little peak in output that corresponds to it. I wonder if this is perhaps a good feature when it comes to rolling basslines which often tend to have a lot of content here. You may find that if you put too much stuffing in the chamber and the peak becomes damped that the box sounds a lot more dry maybe? Less fun? Thanks for making the measurements! I've not heard these cabs in the wild before but the SPL measurements suggest they could be fun. I expect with a few together crossed over at 80Hz they could do quite a bit of damage. |
|||
B_Bender
Registered User Joined: 28 October 2016 Location: Surrey Status: Offline Points: 118 |
Post Options
Thanks(0)
|
||
Looks good for 35hz. 40-50hz though, not so much.
|
|||
Post Reply | Page <1 1112131415> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |