[cjxj]

The above makes sense to me, especially the mid range push. I've built a couple of budget CIEM 2 ways with 30120s. Used a common CI/ED 10ohm, .22, .22 uf, orange, green recipe. One was for a friend and was paired with a 2389. He's very pleased with how they sound. The other one was for me and is paired with an ED. I'm still a newbee novice (only 5 sets so far). Having fun and kind of flying blind until I get my test rig going, but subjectively compared to a set I have with HE's, I feel the lows/mids are a bit stronger compared to the highs. For general purpose and live music monitoring use, I actually like them enough not to want to crack them back open right now to fiddle with resistors and dampers, etc. Maybe sometime, but for now, I have a fresh set of shells and a mixture of BA's just waiting for me to find time and figure out what I want to try for my next build. ;>).

 

[rayhe19]

I have a question about phase. I understand that if you reverse the polarity on one BA, it can cause an electrical phase issue where one driver will cancel out audio from another driver. However, I'm reading up freqphase from JH audio and I don't really understand whats going on. According to Jerry Harvey, if the drivers are stacked in parallel lengths of tubing, the wavelengths will arrive at different times and cancel each other out. Isn't the speed of sound generally consistent and mainly dependent on temperature and composition of air? Because of this, wouldn't it actually make more sense to keep everything parallel? According to https://en.wikipedia.org/wiki/Speed_of_sound#Effect_of_frequency_and_gas_composition mainly 10hz-100hz is affected by frequency and its at a rate of about 0.1m/s.

I guess my question is how do you guys solve the issue of acoustic phasing canceling out certain frequencies at the crossover points if there even is one. Also, isn't acoustic phase more of a problem when you're using multiples of the same driver?

 

[Bad Username]

I have a question about phase. I understand that if you reverse the polarity on one BA, it can cause an electrical phase issue where one driver will cancel out audio from another driver. However, I'm reading up freqphase from JH audio and I don't really understand whats going on. According to Jerry Harvey, if the drivers are stacked in parallel lengths of tubing, the wavelengths will arrive at different times and cancel each other out. Isn't the speed of sound generally consistent and mainly dependent on temperature and composition of air? Because of this, wouldn't it actually make more sense to keep everything parallel? According to https://en.wikipedia.org/wiki/Speed_of_sound#Effect_of_frequency_and_gas_composition mainly 10hz-100hz is affected by frequency and its at a rate of about 0.1m/s.

I guess my question is how do you guys solve the issue of acoustic phasing canceling out certain frequencies at the crossover points if there even is one. Also, isn't acoustic phase more of a problem when you're using multiples of the same driver?

You don't. Not perfectly anyhow.

You are right, speed of sound propagation is more or less the same regardless of frequency. The delay we are dealing with is due to the crossover. It's called phase shift.
You cannot build a crossover without a capacitor. Capacitors take time to charge, and therefore have a tiny delay before passing a signal through. The delay is 90 degrees of one cycle per order of crossover. Because of the position of the capacitor in a crossover circuit, the high frequency always leads,and the low always lags. A 2nd order crossover will have a 180 degree phase shift. And therefore be perfectly out of phase, equivalent to wiring one driver in reverse. We can correct this shift by wiring either the woofer or tweeter inverted, and they will be in phase (at the crossover point). A 4th order will be 360 degrees out of phase, an entire cycle, and therefore will still technically perform as perfectly in phase at the crossover point.
The flat graph JH audio is posting must have its scale zoomed out very very far, marketing trick. In reality a flat phase is not possible so long as you have a crossover. Because every single frequency is delayed by a different amount (the lower the frequency the greater the delay). We can only achieve phase coherency at the crossover points. The phase on a graph will still look like a linear upward slope, not flat. Using tube length differences will again only shift phase at the crossover points. It does not result in perfect phase like they show on their graph. What they are claiming is simply impossible as far as I am aware.

 

[ForceMajeure]

45deg per order not 90

 

[Bad Username]

Well, assuming you are applying the crossover for both highpass and low pass. Then the total phase shift is 90 degrees at fc.

High pass will be +90 deg at DC, +45 deg at the crossover frequency, and 0 deg at high frequency.
Low pass is 0 deg at DC, - 45 deg at crossover frequency, and - 90 deg at high frequency. So you are looking at a 90 degree shift to deal with.

But yes, technically you have +/- 45deg per order of crossover at fc

 

[rayhe19]

Do you guys know what the crossover setup is in a knowles GV? I was planning to use one as a driver for my ciem. Also, would you recommend adding a CL with a high ohm damper as a subbass driver? Or will that be too much bass? Thanks!

 

[Firedrops]

it all depends on your design in the end. as you can see the 30120 looks good but once you incorporate it with other drivers the midbass to lower mids can be a bit too much if you go for the damping technique as a lowpass...on the other hand on the 22955 have a faster drop but 2k mids comes back so it could add itself to other drivers and make the 2k bump too high... it's all about finding the right balance.

My thoughts exactly. The 30120 seems to have a lot of mid bass boost, and while the shift to 3kHz is something you'd desire for fullrange driver the question is whether this driver is actually suitable for fullrange applications. It could be good for budget 2-way design probably - something like CI-30120 + WBFK.

Very good points there, I didn't pick up on the mid boost at first. Guess I'll stick to the 22955 for my 3-way then. Thanks for the discussion!

 

[Squirg]

,x,x,x

I like the purple one. Nice color.

 

[tranhieu]

Not sure if this's allowed, I will delete this post upon request. But I am having several 2389, 2356, 1723, 2600, 23e25, 26e25 lyring around in the drawer here. Bought them in bulk for R&D but ended up using only a few (the design was completed much faster than we thought).

Just want to collect back what we paid, or even less than that is fine, any amount is also ok.

 

[Im1fan2nv]

I've looked and looked for a way to splice. I just can't fit 3 tubes so I need to splice and make a Y-splitter so I only have 2 tubes. Any help with photos would be greatly appreciated. I've made some that seem fine until I put them in the shell and they break and leak because I need to bend the tubing to fit my drivers in the shell. I'm using Foto to bond with a coat a laq on top of that.

 

[Xymordos]

I've looked and looked. I just can't fit 3 tubes so I need to splice and make a Y-splitter. Any help with photos would be greatly appreciated. I've made some that seem fine until I put them in the shell and the break apart because I need to bend the tubing to fit my drivers in the shell. I'm using Foto to bond with a coat a laq on top of that.

I sometimes squeeze the end of two small tubes a bit to fit them into a bigger tube. The bigger tube then extends to the end of the bore.

 

[Firedrops]

Just came across this while looking around taobao.

The top label is the most interesting one - it says that the electrical component (referring to capacitors/resistors) can be directly mounted across the dynamic driver's 2 terminals.

What exactly does this do? Is it actually just acting as a low-pass filter (having a cap parallel to the driver)?

 

[Im1fan2nv]

I just wanted to post a photo of my attempt to use the Ballistic Gel for a mold. If it's done correctly, it works great. Clear as can be. My impression was not as smooth as I needed but you get the point as to what the gel can do for you. I ended up making an acrylic impression to make the final mold. Wax is not an option because the melting temp for the gel is about 250F. If anyone is interested in the process I used, just let me know.

 

[Im1fan2nv]

Do you guys know what the crossover setup is in a knowles GV? I was planning to use one as a driver for my ciem. Also, would you recommend adding a CL with a high ohm damper as a subbass driver? Or will that be too much bass? Thanks!

Shilohsjustice posted GV crossover pg. 442

 

[ForceMajeure]

I made a rotating machine.
The cost is about 10 dollars.



http://customiem.blogspot.jp/2017/07/blog-post_22.html

That's a cool solution.
It looks like it might be turning too much (eyeballing it it looks like ~30 turns per minute). Can you make it turn slower?