Jub-Like Drone Build (Jubilee)

[PrestonTom]

Hey, this looks like fun!

To get you started, let me orient you on the impedance graph that Mike posted (conventional Jubilee without a drone)

First, the impedance as you go lower in freq is headed toward about 3 Ohm. This shows that the two drivers (each about 6 Ohm DCR) are wired in parallel.

Second, There is a big peak at about 40 Hz. This peak shows two features. First is that the Jubilee is using "reactance annulling" to get good low freq response. If the back volume were larger, the peak would shift to the left. If the back volume has open cell foam rubber added, the peak will shift to the left (to a certain extent). If the cabinet has an air leak in the back camber, then the peak would be reduced overall.

Third, peaks at about 100, 200 and 300 Hz. These are caused by relections from the mouth back to the throat and are due to an undersized mouth (or shortened horn length). These peaks will be increased (to a certain extent) if the cabinet is pulled away from the corner (losing boundary or corner gain) or elevated off the floor. If a second cabinet is placed on top, these reflections (and likewise the impedance peaks) will be reduced. That comes from the two cabinets showning "mutual coupling" (sort of like providing a "boundary").

I'll defer to others to explain cabinet venting and impedance graphs and tuning.

[Panacea Engineering]

Tom,

Now that you have completely lost me.........please tell me what your thoughts are on the overlay? Remember, I am a total idiot when it comes to this and i realize that MANY more test will have to be run. This was for me, more of a learning experience using the software/hardware than anything else, at this point. The WT3 program has many options that i will need to explore and with some help, hopefully someone will be able to tell me which of them need to be done.W. C.

[PrestonTom]

WC, My comments were meant to be helpful. Use the impedance plot as a tool to 1) check for air leaks, 2) check to see that the drivers are correctly wired together (in parallel) and 3) check to see if the cabinet has an air leak, 4) be consistent when measuring the cabinet (keep it in the corner), 5) see the effect of adding foam rubber in the back chamber. The most important use will be when tuning the drone. Others need to chime in on this last point.

If your measurment rig is set up correctly, then the "Golden Jubilee" (reference cabinet) should show a measure similar to what Mike posted.

[Guest " "]

When are you going to do an SPL test of the passive unit vs the golden jubilee's in the area where the expected radiator benefit will be (40hz)?

[Coytee]

Why does the plot posted by Mike have a more linear increase on the second half?

[mikebse2a3]

Why does the plot posted by Mike have a more linear increase on the second half?

This is a good lesson in how even though the data might be the same changing the X-Y axis values can falsely make someone think that something looks better or worse than it really is just because it is displayed differently even though it might be exactly the same data. Anyone who wants to see this for themselves just take the same test data and then change the graph's x/y values and watch how things can look better or worse visually.

The X-axis and Y-axis Graph parameters WC used are different than Roy's plot causing the plots to visually look different (due mostly to the Y-axis display parameters in this case) as displayed but if you actually look at the data points they are very close at all points considering the different measurement conditions and equipment. Also note: Roy's Y-axis for Z(ie: ohm) is logarithmic scale versus WC's Y-axis is linear scale for Z and this changes the visual appearance only while the data is still the same regardless of which scale type is chosen for a plot.

I haven't used the program WC is using but there might be options for setting the X/Y scales for either Logarithmic or Linear if he chooses to get different views of the same data. Many times the scale limit parameters and Log/Lin graphing should be chosen to display the data gathered best for the areas of main interest.

I would suggest WC use a setting close to 5-10Hz for his lower limit for his testing.

[mikebse2a3]

Just some thoughts for you WC and hopefully some others with more experience will chime in with their experience.

It's been close to 25 years since I played with ported loudspeakers so I'm no expert by any means but after looking at the Speaker A / Golden Jubilee overlay plot it looks to me like the K-12 passive (used as is with no additional weighting) is in the ball park and maybe toward the high side of the tuning curve range that you might want to experiment with. I believe you should run a Frequency Response Test of both systems to see what it looks like compared to the Golden.

WC I would be very surprised if this passive doesn't run into excursion limitations at higher SPL considering it has to deal with the air volume capabilities of the (2) K31 woofers. So I'm thinking you might also want to run the Frequency Responce Test at several drive levels to see if it exposes this limitation (which I believe will show up as a loss of low frequency extension as the SPL is raised).

mike tn

[Guest " "]

(which I believe will show up as a loss of low frequency extension as the SPL is raised).

I am curios about this as well......

[djk]

"
I am curios about this as well......"

The PR may need to displace as much as 3x what the two 12s can displace, that's why I suggested the TC Sounds PR.

[mikebse2a3]

Hey WC

I was thinking.....do you have the clear hatch on the Drone cabinet and if so can you see the movement of the drone's cone?

I'm not sure what your capabilities are as far as test signals but if you have a frequency generator(or possibly a computer program that can generate single frequency tones) that you could sweep/step through the 20Hz-50Hz range and watch for were the maximum drone excursion falls. This also might give you an idea of the maximum SPL possible before the drone runs out of linear excursion.

mike tn

[Guest " "]

if you have a frequency generator(or possibly a computer program that can generate single frequency tones) that you could sweep/step through the 20Hz-50Hz range

yep...and if you have an iphone, there are free apps that can generate tones.

[DrWho]

Hey WC, you might find this helpful:
http://polaraudio.blogspot.com/2012/01/calibration.html

I've not read through it, but at the very least it would be a common point of discusstion as you dive into the REW world. For what you're doing, calibrating the absolute SPL and calibrating the soundcard itself won't be necessary, but I think it is important to understand the concepts as to why you'd want to calibrate them.

Btw, looking at your impedance plot, I would try to understand that ripple at ~70Hz. A real good way to do this is to play a constant test tone (which you can do in REW) and then slowly change the frequency until your hear something weird. You'll be looking for air rushing sounds, or buzzing (panel resonances).

It also looks like you're tuned to the 30Hz to 35Hz region, which is about where you'd want to be. If possible, it would be a lot more informative to get an 1/8 space (corner) measurement. That's usually pretty hard to achieve outdoors though.

[Bigdnfay1]

Try the audio test bench at www.higherfi.com/software.htm Big D

[Panacea Engineering]

Just an update to let you guys know that I haven't passed away...............

Everything is ready, I have studied till I am crosseyed and I am looking for a couple of un-interupted days to start work on the testing. Looks like everytime I think I have some time, something comes up. Either Doctors appointments, wheather (Huge Factor), having to do some stupid research and filing of paper work for my retirement, dealing with my damn Lawyer, or building construction issues. As I am the General Contractor, I have to deal directly with Code Enforcement (calling in, setting up and meeting with the various Inspectors), numerous sub-Contractor issues, paying the bills, etc....You guys get the idea, I am sure. I just wanted to post a picture of where the building is and why getting to the testing is taking so long. This project will run well into the Spring (The Speaker testing and the Building Construction) so please bear with me. I want all of the testing to mean something so I have to do it the only way I know how, by learning as I go. All of you guys have been GREAT and I am just asking you to be patient as I think this is going to be a very worthwhile project (Jub-Like Drones).

Here is the New Building (pay NO attentuion to the date, as I had to buy a new battery and forgot to set the date.

[seti]

Now that is a shop! You may need some KP600 stacks as well as jubilees

[Panacea Engineering]

I wanted to add a couple of pictures of Speaker B (the one loaded with K-31's and the TV Sounds Adjustable Mass Passive). I had a few finishing details to complete and took some pictures back when I was loading the drivers. I am starting out with a weight of 398 grams ( I think djk reccomended 400), close as I could egt to 400. The total adjustment that can be added to this passiev is a whopping five pounds of washers (you can see them in previous pictures). Again, I have installed a 3/4" piece of Plexiglass on the Top Hatch to be able to observe the Passive in action. I do not really think this is of any real value other than being Neat as hell....! It will allow me to observe (if we add a lot more weight what is going on in the Dog House. I will be looking more for physical damage than anything else.

Here it is installed in the middle position wht two 3/4" spacers. The 1-1/2" spacing is needed because the cone travels at least an inch and a quater, while it is in operation

[Panacea Engineering]

Now that is a shop! You may need some KP600 stacks as well as jubilees

This is something that I have been wanting to do for years....! I will finally be able to build a dedicated sound room large enough to do justice to the larger Pro and Heritage speakers. Since I have accumilated all of this rest gear, it will be nice to have a place to finally put it to good use. I feel sure that I will be asking for advice on building the New Sound Room in a future thread. The possibilities are endless and I am REALLY looking forward to warmer wheather.

One other thing that I failed to mention above; It is getting time for me to start working at the Flying Field in SC, for the upcoming event in May. One good thing (now thast I am retired) is I will be able to spend time during the week, rather than having to work around weekends....!

W. C.

[WMcD]

W.C.

Thank you for your email and pm inviting me to aid in the testing project.

The use of the passive is terra incognito and I've don't have any special knowledge. But I'll give you my comments.

In my view, the whole reactance annulling theory might be good for horns which are large or theoretically perfect. Yet I see that finite and short bass horns present reactances at the throat which are going plus and minus j. So the back chamber can help or hurt.

It seems to me that PWK was most concerned with getting the system to resonant as low as possible. This is to get the woofer / back chamber give a boost as low as possible.

I rans some simulations with a big P-Spice file regarding throat impedance below cutoff. PWK had an article examining the same. He showed there was some horn loading effect. Generally mine showed the same with some additional information. The acoustic resistance is better than without a horn, but nowhere near what you get well above Fc. Also, reactive effects have died away as we move below Fc. (I better check this.)

In looking at impedance plots, the lowest (in freq) peak is indeed where the active woofer(s) are reaching system resonance.

It seems to me that the first order of business is to get the passive radiator (plus the box and active drivers) tuned to some freq which can be identified in an impedance plot. Right now we have the mass of washers but no real idea of whether there are too few or too many or where we are going..

It would be nice if we could find an accelerometer to put on the passive and see at which freq it is moving. A tall order. A clear hatch has been suggested. As an alternative:

Gosh, I know W.C. doesn't want to hear this, because of the effort involved. But. I'd suggest building a box of the same volume as the back chamber as the bass unit and mount the active and passive drivers in it. Then fool with the mass washers added to the passive. Note that the passive could be put in the box backwards so that the washers are on the outside so we don't have to take the box apart to fool with the washers..

So: what is the target for the resonance of the system overall? I'm thinking that the resonance of the passive (by adjusting the number of washers) has to be made close to that of what we see as that of the active drivers.. That is pretty much in keeping with port design theory. There are many "alignments" in T-S bass theory.

The one which is most easy to understand is that the woofer mass and compliance spring is exactly equal to the box and mass in the port. So just when the woofer wants to resonant, the box and port are an alter ego. It gets more complicated with a passive radiator which has it's own spring. Meaning I'm not sure either..

In any event. Just to collect thoughts. Conceptually, It seems to me that when we look at performance of a K-Horn below Fc, we are mostly looking at a Heresy type with a little better acoustic resistance because of horn loading. Extending this. W.C.'s design is like a Forte II with a little better acoustic resistance. Thus we have to look box design with a passive. But in this, we have to have inforation on the passive's tuning.

WMcD

.

but not so much in practice when we're looking at relatively short horns.

[Guest " "]

I still wonder what the passive will actually do.

In a box design, depending on the size of the box and the mass of the passive, a relationship can be established which can contribute to the low end response of the box. Adding weight to a passive really just lowers the upper range of the passives response window, and if done correctly with minimal effect on the targeted tune point which is usually within half of the FS of the woofers.

In a horn environment...the loading of the horn is now influenced by a cancellation relationship between the active and passive. Likely to occur in the lower half of the performance band of the passive. So if the desired tune point of the passive is 30hz +- half an octave, too much cancellation between the active and the passive will raise the lower edge of the response band of the passive, while on the other hand...adding weights will lower the upper edge of the response band. End result the half octave a passive normally adds value to a box design, shrinks and for the sake of illustration lets say it's 1/4 octave now. but then....where is the center of the tune point?

One piece of information that would be useful to determine what needs to be tuned is a baseline response curve of the test "not-a-jube" box with out the passive (install a blocking plate), then another baseline test with the passive installed.

[WMcD]

I started reading up on this.

One website has an explanation.

http://www.diysubwoofers.org/prd/

It seems that I was wrong to suggest the passive resonanting at the driver resonance. It must be lower. Mea culpa. I should have remembered more as I typed.

It seems to me that a box port system is tuned to Fs or nearby in the various alignments. The slug of air in the port does move out of phase with the main driver at some point in freq and could cause a notch in overall output. But this doesn't create much interference because the loading on the back of the active driver is very strong and prevents significant movement / output. Exactly how this is loading parameter is quanitified (Q issues?) . . . I'm not clear on But therefore we're getting very little out of phase interference.

It must be that the passive does not load the pressure in the box very much and therefore we do get a notch. The goal is move that notch down below the range we are interested in. The link says the notch is at the resonance of the passive (as if it is free air resonance of the unmouted passive). Mabye it is more correct to say it is at the resonance of the entire system where the the active and passive are moving in tandem -- and this is where one moves out and the other moves in.

Maybe we can accomplish some good results in the project, in theory. But I see problems.

A) There are writings saying that the passive should be larger (or equal) in size compared to the active. Something to noodle / read about. In the Forte they are equal and in the Forte II the passive is larger. But in the project, the passive is half the size of the two actives (combined). So I'm wondering if we need two passives and one active.

I'm still thinking that a good starting point for design is to assume we are just making a box type direct radiator(s) and passive(s) system. We can look at designs on software simulators. But in the project we can't vary box size and the choice of drivers is fairly well fixed, and the only thing we can vary in the passive is the mass. That is the good news and the bad news.

Thoughts out there?

WMcD