C-Note MT Bookshelf Speaker Kit Pair & Other kits

[jhollander]

Without the 2 uF to ground (C3) in the Tweeter circuit you would get a rising response like you have shown

[datrumole]

Without the 2 uF to ground (C3) in the Tweeter circuit you would get a rising response like you have shown

thanks man, this is the best i could to illustrate the XOs, kinda need to review em all since it gets flipped over

[datrumole]

well, i tried a few different measurements, and was asked in another thread to measure the tweeter without the crossover

see the rew measurement names to get an idea of color coding

- xover measurements -

the below shows a full sweep assembled

sweep with just the mid bass (mb)

sweep with just the tweeter (tw)


- no xover -
just the tweeter

you'll note the HF rise is being caused by the xover

edit: the tweeter only measurement was run at -24dbFS, the rest were -12dbFS for illustration purposes

now, i managed to get this on 3 completely separate builds, so one would think i would have gotten 1 right, but all 3 wrong...well...not sure what to say on that, ha. not shown is the other two identical sweeps from the other two assembled speakers (building another set for a friend)

[jhollander]

I can't tell from the pics. I'd build a tweeter circuit free form (not with the woofer and not over under on a breadboard) and measure separately.

[Chris Roemer]

I've run several sims now (for this thread), and I'm wondering why your tweeter w/NO XO isn't shown at its rated +10 to +12dB louder than your midbass?
How are you doing your measurements (the mic shouldn't move), and with what software?
Can you run a 20-20k impedance sweep of the entire speaker to show us?

[datrumole]

I've run several sims now (for this thread), and I'm wondering why your tweeter w/NO XO isn't shown at its rated +10 to +12dB louder than your midbass?
How are you doing your measurements (the mic shouldn't move), and with what software?
Can you run a 20-20k impedance sweep of the entire speaker to show us?

thanks chris

the +10-12db question, i was running at -12dbFS for all my tests, ran with the tweeter at -12dbFS and is was REALLY loud, so i re-ran it with a -24dbFS signal as to not push it so hard, and to illustrate a little better as well. which perfectly fits with your +12db attenuation you mentioned. apologies it wasn't noted

measurements are with a umik1 in REW with the calibration file, mic pointed directly at the unit (using teh standard/non-90deg calibration file), at 1ft. i get same response at 3ft as well, however willing to remeasure with the standard 1m if that would help. the mic and speakers are in the middle of a wide open room both on kitchen stools so there are no early reflection points. mic remains stationary and i have relative markers for the speaker so they are put back in the same spot between tests. probably inches of margin of error on that front

i'm not entirely sure i have any idea how to run an impedance sweep tbh, can i do it with what i mentioned above?

[jhollander]

Good point Chris the tweeter should be at a much higher volume level. Looks like a different tweeter.

[Chris Roemer]

So do you think If I trace your driver plots (filterless) and add +12dB to the tweeter, that I can sim your situation using the stock C-Note XO (and it looks like you added a 0.22uF "tank" cap across the woofer inductor)?

I looked at your XO pics BTW, and don't see anything jumpin' out at me.

[datrumole]

So do you think If I trace your driver plots (filterless) and add +12dB to the tweeter, that I can sim your situation using the stock C-Note XO (and it looks like you added a 0.22uF "tank" cap across the woofer inductor)?

I looked at your XO pics BTW, and don't see anything jumpin' out at me.

as of now, the plots provided are the xover full, just tweeter, and just midbass

ill provide the unattenuated (-12dbFS) of just the raw tweeter and midbass, would that help?

and yes, i did the .22 cap mod

edit, also, is there anywhere in your sim that if you removed maybe like C3 you would get my same response? just trying to rule out user error (me)

[datrumole]

ok, here are the raw driver measurements, all with no crossover and -12dbFS from REW

[Chris Roemer]

I can't see a way to generate a +2-3dB "lift" near 12k (using a passive filter) as you show in your measurements, whether I use mfr. FRDs or traced ones off your last post.
Pretty odd.

If that bump bothers you, you could notch it probably. I couldn't hear 2dB @ 12k anymore, although my wife could.

[datrumole]

I can't see a way to generate a +2-3dB "lift" near 12k (using a passive filter) as you show in your measurements, whether I use mfr. FRDs or traced ones off your last post.
Pretty odd.

If that bump bothers you, you could notch it probably. I couldn't hear 2dB @ 12k anymore, although my wife could.

thanks chris

if C3 is not in the sim (perhaps i failed to connect it properly or something) does the response look similar? i'm just tyring to fully rule out anything i did that i can correct to get this to look at the simulated/intended response. heck, even the same response as others have posted

also, to me, it looks like the hump is starting around 7k-ish, no?

i'll do another round of measurements with my denon in direct mode (no processing) as well, i believe the tp22 has poor response in the high end with high resistance, but it still doesnt kill the behavior

as for 'living with it', i'm sure i can, i can eq it, i can do a number of things. but this thread started with me ensuring it was properly constructed. at no point does it look like the intended response, nor even others who have built the kit and posted their measurements. so either something is wrong in construction, part tolerance (again, reproducible on all 3 of the 4 i've measured). if it's not construction based, than i'd say this design needs some rework if for some reason the parts are no longer manufactured the same as when it was designed. if it's construction based, i'm happy and hoping i can address it

[Chris Roemer]

Show us your impedance plots from your sims, please. Did you use mfr. z-files?

You can SEE the effect of not connecting C3 properly, you don't get a "lift" near 12kHz. Curious.

[datrumole]

Show us your impedance plots from your sims, please. Did you use mfr. z-files?

You can SEE the effect of not connecting C3 properly, you don't get a "lift" near 12kHz. Curious.

i apologize if i wasn't clear, i dont have any sim anything. i was curious if on your end you could modify the design in such a way that it could simulate the behavior/resposne i'm getting. meaning if like the resistor was literally taken out in the sim, that would explain the rise, and then i would go look to ensure the resistor was working or connected or something. thats what i was getting at.

and yeah, a lot of people are saying C3 is the 'pull down' so that's just why i focused on that

if you are saying in teh sim, if C3 isnt connected, that it wouldnt cause the rise, then i think we can rule it out as the culprit of the issue. or perhaps it's not a rise, but a failure to pull it down?

anyway, anything else i can do, or are we saying here that the speakers are functioning as intended?

[Chris Roemer]

Orange is your filtered woofer, red is w/C3 connected, yellow is leaving C3 out, and green is how your C-Notes measure.
Like I've said, you can't really get a "boost" @ 12k by changing your component values.
Almost the only way to get a boost in your summation (above the filtered tweeter plot) would be if the woofer (@12k) was loud enough to be significantly "additive", but I THINK (@12k) it's down by -40dB?