Erin's Audio Corner gets a Klippel NFS!

[amirm]

So I understand this. To get that last little bit of precision, you would like to modify the transition from the microphone to the end of the pipe?

No. The test Erin has setup is not the way NFS should be used for actual measurements. It is highly exaggerating the reflections from the fixture due to very short distance to the driver in his test. While I have also removed the cage and built my own boom, the effect you are talking about is not there once you optimize the other aspects of NFS operation. The microphone front has far larger surface for reflections than the edges of that pipe for example.

There are also limitations to boom design which if done wrong, can interfere with the operation of NFS. I am surprised Erin is jumping into modifying the system before making a few measurements and learning those other aspects. Every time I modify the boom it takes me hours and hours to re-align the entire system to make sure this doesn't happen in actual use. The worst sound you can hear is the grinding of the stepper motor as the mic fixture crashes into the rest of the system!

I need to emphasize that you need to have a ton of experience with NFS under your belt before removing the cage. One wrong keystroke can shove the microphone into the speaker with drastic consequences during setup! Without the cage, you have no safety measures against this. There are also horrors you can face when the system is running especially unattended. If not done right, the system can slap the mic into the speaker, wiring to the speaker, rest of NFS system, etc. not just once but over and over again! I have had all of this happen after I removed the cage but thankfully, with no serious consequence as I was there to catch them. And testing was done with my own speakers and not others. After I removed the cage, I also watched the system like a hawk for many runs. I have now optimized the physical parameters of the measurement system which has sharply reduced chance of this happening. But at the start, it was hell.

 

[mdsimon2]

This is a very interesting find Erin!

I’m rather surprised it makes such a big difference!

I wonder how much of a difference a standard mic holder, on a tripod, affects the same top 3 octaves.

I mean, this is for people doing measurements out at 0.5-1M, using the old fashioned (non-robotic) manual system.

As Erin mentioned the more serious DIYers are well aware of this issue. See link below for write up / measurements from Troels Gravesen on the impact of discontinuities close to the microphone.

http://www.troelsgravesen.dk/measurements.htm

If you are making acoustic measurements of drivers / speakers definitely worth taking some time to get this right.

Michael

 

[McFly]

I wouldn't hold a Mic with a metal pipe, I'd use uPVC pipe or similar if possible. Just me

 

[hardisj]

I wouldn't hold a Mic with a metal pipe, I'd use uPVC pipe or similar if possible. Just me

PVC is too flimsy. Been there. I actually modded a piece of PVC to have a strong piece of metal going through it to hold it up but that was in a pinch and kind of silly when using the right piece would have made sense.

 

[hardisj]

No. The test Erin has setup is not the way NFS should be used for actual measurements.

Ironically, the idea was borne from a conversation I had with Christian at Klippel (the guy who is mostly responsible for the NFS). We've had a few skype sessions and this was one topic I asked him about last week. And just in case anyone thinks I'm making this up to cover my own butt, here's some "proof".

Actually, Christian said they use 3 screws at the end to hold the mic in place. Which is what I did originally - with PVC - and then I found the PVC pipe was too large to fit the mic holder. The inner wall thickness prohibits me from using any size of PVC (at least, the SCH 40 sold locally).
Here is a picture of that. But with that said, I was going to do that with the tube I purchased but I would have to get smaller set screws because the ones in the picture are too large and stick out about 1/8" from the pipe.

Honestly, this thing really isn't that complicated to set up. I had it up and running fault-free my first time. I've already measured three speakers as trial runs (different sizes). Maybe if I use a very irregularly shaped speaker it might be a bit more troublesome to make sure the "cylinder" is free of things. As you may have noticed, I also strapped down all the wiring to the support pole. I understand you're trying to help but I'm not a total doofus.

 

[hardisj]

As Erin mentioned the more serious DIYers are well aware of this issue. See link below for write up / measurements from Troels Gravesen on the impact of discontinuities close to the microphone.

http://www.troelsgravesen.dk/measurements.htm

If you are making acoustic measurements of drivers / speakers definitely worth taking some time to get this right.

Michael

Agreed. And it is such an easy "fix" that there is no reason not to.

 

[617]

Ironically, the idea was borne from a conversation I had with Christian at Klippel (the guy who is mostly responsible for the NFS). We've had a few skype sessions and this was one topic I asked him about last week. And just in case anyone thinks I'm making this up to cover my own butt, here's some "proof".

View attachment 112561




Actually, Christian said they use 3 screws at the end to hold the mic in place. Which is what I did originally - with PVC - and then I found the PVC pipe was too large to fit the mic holder. The inner wall thickness prohibits me from using any size of PVC (at least, the SCH 40 sold locally).
Here is a picture of that. But with that said, I was going to do that with the tube I purchased but I would have to get smaller set screws because the ones in the picture are too large and stick out about 1/8" from the pipe.

View attachment 112568






Honestly, this thing really isn't that complicated to set up. I had it up and running fault-free my first time. I've already measured three speakers as trial runs (different sizes). Maybe if I use a very irregularly shaped speaker it might be a bit more troublesome to make sure the "cylinder" is free of things. As you may have noticed, I also strapped down all the wiring to the support pole. I understand you're trying to help but I'm not a total doofus.

You're a mobile audio guy, I'm sure you're used to being underestimated

 

[McFly]

Well done Erin. Very hands on. Johns idea of setting no go zones for the mic makes a lot of sense to me, for people that might not be so mechanically talented.

 

[tktran303]

Look a lot of it is language and personality and also Internet communication.

I always preface things “I wonder if...” to be polite; even though I’m 99% confident of the problem. Because I value humility and I believe in what is now called “growth mindset” and really I don’t want overstep myself.

Some people are bit more forthright or assertive or seem aggressive, depending on how it’s put forward, or received.

I’m sure Amir doesn’t mean to call you a dufus, or even imply that, and yes, sometimes his language can rub people the wrong way.

 

[amirm]

I understand you're trying to help but I'm not a total doofus.

It has nothing to do with that. Or any knowledge of measurements or acoustics. The robotics of the system is nothing like I or you have ever used. It is a proprietary system to be learned and importantly, respected. It has the potential to do serious damage to itself and what you are testing. How it behaves, how it computes the locations it scans, how you adjust the boom, etc. are things have have to be learned as there is not any documentation on it.

My NFS system reminds of my CNC machine. I thought I had it all figured out after practicing on a few pieces of MDF. Then I put a large sheet in it and hit go and after a few minutes, the thing decided to grind the 3 horse power router into the MDF as if to come out the other side of earth! You have not gotten scared until you see a carbide tip come to a grinding halt and the hum of the servo motors still try to jam it in further into MDF! The problem was that the table was not fully level and gradually elevated (it was 12 feet so very hard to align) and as the machine kept going, it started to cut deeper and deeper until it could do no more and horrifically jammed.

It is entirely possible that you are much smarter than me and have envisioned all the ways the system works so you won't face these things. I commented because the conversion was going in the wrong direction, implying major issues with the system that in practice are not very severe. Worrying about the pipe edge on that boom??? Once you go there, might as well not use the machine because there are reflection points elsewhere which you can't deal with. Fortunately these things can be worked around and excellent results can be had.

To be sure, the robotics is where the limits of the system come into play. The math assumes there are no limits in how you measure a speaker where in reality there is. So you could make comments that make sense and criticism that is valid. I am just asking you to use the system at least a bit end to end so you see all the other aspects that come into play. The boom and how it is configured is the heart of the system with heavy, heavy interaction with the rest of the system. As I said at the outset, I don't know how anyone comes into using the system and just intuits all of this. I am looking at your pictures and I can see you have not crossed this bridge yet. Or bridges I should say.

 

[hardisj]

Worrying about the pipe edge on that boom??? Once you go there, might as well not use the machine because there are reflection points elsewhere which you can't deal with. Fortunately these things can be worked around and excellent results can be had.

Oh, for sure, man. I don't disagree that worrying about the pipe edge itself is overkill (on the updated boom). Math tells you the influence from it will be above 20kHz and the measurements already show a solid improvement. But a little piece of foam or adding an additional inner layer of tubing to help shore that up and keep the mic steady in the tube can't hurt.

I am looking at your pictures and I can see you have not crossed this bridge yet. Or bridges I should say.

I would appreciate seeing your setup as is. I know others have asked in this thread and you haven't replied (I assume to not bog this thread down) but I personally would like to see the pictures of your NFS setup with some explanation about what it is you mean in the above quote. If you have posted them somewhere else a link will suffice. If you would rather PM that's fine, too.

 

[Steve Dallas]

Klippel fight!

 

[hardisj]

Mic boom is now complete. I flared out the steel rod. Took some copper pipe (I had on hand for an antenna project) that also happens to fit right over the mic with very little room to move. I filed the copper down on the outside about halfway up and that allows me to slide it in to place of the steel tube and press-fit it without it moving once pushed in to place (requires a bit of force to separate the copper and steel once pushed in place). Holds the microphone well. Looks ugly. But functions perfectly.

 

[jtwrace]

Klippel fight!

No, I sure hope not. We need more NFS' that "agree" rather than work against each-other.

 

[amirm]

I would appreciate seeing your setup as is. I know others have asked in this thread and you haven't replied (I assume to not bog this thread down) but I personally would like to see the pictures of your NFS setup with some explanation about what it is you mean in the above quote.

My Klippel is surrounded with thousand household items in the garage which I prefer to not share. Here is a shot of the boom however:

The boom is a new aluminum tube that is much shorter than stock. This allows you to extend the range of the boom without it hitting something. In my case you can see that shelving part of the roof that it could hit. Using the maximum reach of the system horizontally is key to getting low frequency response right for complex radiators. This is why I mentioned that you would not be operating the NFS with the Mic so close to the speaker. I have squeezed every millimeter of r-axis range for this, necessitating the shorter tube (and having room around NFS).

On mounting the mic, I first put it in the tube but I could not find a way to get the boom to then operate without the adjustment blocks hitting the r-axis, or not being able to reference the mic with the cone on the bottom. After spending a lot of time, I settled on it being next to the tube, not inside it. To then deal with potential reflections, I borrowed batting material from my wife and built what you see there. Inside that cone is fluffy batting material. This was useful when I had the cage. It is basically an acoustic absorber not only for the area around the mic but also past it. Since frequencies are high, a couple of inches is enough to give you high alpha (absorption factor).

Once I extended the measurement range though, the absorption didn't seem to make hardly any difference. The path length of reflection is so long that its level is literally in the noise (we are talking about half a meter). I have left it in there for good measure and for near-field measurements.

The rest of the system looks like yours so not much in showing a picture. The fine points matter there which is to make sure you use the full reach of the r-axis per above without the Z axis hitting anything. You potentially need to slide the the r-axis left or right to get maximum extension. This stuff is not documented and pictures in the installation manual are very hard to read and interpret in this regard.

It is hairy alignment work as I mentioned in my last post in the way mistakes can cause the system to crash into itself, speaker, alignment cones, etc.
Work on getting the lower speaker reference point right. If you do, then you never have to touch it again. Set it wrong and the mic will run into wiring, or will crash into r-axis.

I think this is all you need to get the optimal setup.

 

[amirm]

Oh, this is the replacement aluminum tube I purchased: https://www.ebay.com/itm/222921582426

It is only $13 and lightweight so doesn't put stress on the z axis.

 

[hardisj]

reference the mic with the cone on the bottom

Yea, when I first calibrated the setup I realized at that moment the movable range of the unit was less than I was thinking it was going to be. If you know what I mean. Basically, it took the height I thought I was going to be able to achieve and cut it down by about 10 inches. :/

And, yea, I get what you mean about the other axes. I will be limited to certain size speakers but maybe I can do something about that in the future if I find the need to get the z-axis extension.

Thanks for sharing the other details.

 

[hardisj]

BTW, is that a high-tech rubber band on that towel I see there?

 

[Hiten]

Microphone long foam covers are readily available. Just closed part needs to be cut.

https://www.google.com/search?q=mic...1nwAEB&sclient=img&ei=cHcqYOSgH_fj3LUPrIWa-Ac

Regards

 

[FAST]

Ironically, the idea was borne from a conversation I had with Christian at Klippel (the guy who is mostly responsible for the NFS). We've had a few skype sessions and this was one topic I asked him about last week. And just in case anyone thinks I'm making this up to cover my own butt, here's some "proof".

View attachment 112561




Actually, Christian said they use 3 screws at the end to hold the mic in place. Which is what I did originally - with PVC - and then I found the PVC pipe was too large to fit the mic holder. The inner wall thickness prohibits me from using any size of PVC (at least, the SCH 40 sold locally).
Here is a picture of that. But with that said, I was going to do that with the tube I purchased but I would have to get smaller set screws because the ones in the picture are too large and stick out about 1/8" from the pipe.

View attachment 112568






Honestly, this thing really isn't that complicated to set up. I had it up and running fault-free my first time. I've already measured three speakers as trial runs (different sizes). Maybe if I use a very irregularly shaped speaker it might be a bit more troublesome to make sure the "cylinder" is free of things. As you may have noticed, I also strapped down all the wiring to the support pole. I understand you're trying to help but I'm not a total doofus.

FYI, you may want to try SCH 80 . which you can typically get at Home Depot or lowes, Its grey and has a thicker wall.