Altec Attenuation Circuit Help

Please reread my previous post ( I had edited it > adding + subtracting some text ).

Tweeter = JBL 077 ( UHF for > than 8K ) or equivalent in Earl-speak.

:)
 
A few thoughts;

(A) It's impossible to know how much response your 802s have over 10K unless you measure with a program that accepts a calibration file ( from a calibrated mic ).
- You haven't said which mic? you used with your RTA program.
- Perhaps your mic rolls off at 10K ( leading to false conclusions ).

(B) In Review: Generally the whole idea of HF compensation is to drive ( turn up ) the HF device so that everything above 10K is just a few db's under the 1K level ( as a reference ).
- Then everything 2K-8K is turned down by some attenuation circuit ( such as the 30923 ).
- The bypass cap value determines just how broad ( or narrow ) the spectrum of attenuation is.

(C) Your diaphragm is most likely a 20275 ( not known for extended HF response ).
- It can be replaced by GPA with a more modern ( extended ) diaphragm.
- I would recommend the 8 ohm version ( model # 34647 ) since using a lower impedance will allow you to drive the horn circuit harder ( getting a bit more UHF ) before attenuating its' midband response ( 2K-8K ) > with the 30923.
- FYI, the last version of the 9844(D) used an 8ohm driver ( 902-8B ).
- The 9844 was the pro version of your speaker.

:)
 
Last edited:
A few thoughts;

(A) It's impossible to know how much response your 802s have over 10K unless you measure with a program that accepts a calibration file ( from a calibrated mic ).
- You haven't said which mic? you used with your RTA program.
- Perhaps your mic rolls off at 10K ( leading to false conclusions ).

(B) In Review: Generally the whole idea of HF compensation is to drive ( turn up ) the HF device so that everything above 10K is just a few db's under the 1K level ( as a reference ).
- Then everything 2K-8K is turned down by some attenuation circuit ( such as the 30923 ).
- The bypass cap value determines just how broad ( or narrow ) the spectrum of attenuation is.

(C) Your diaphragm is most likely a 20275 ( not known for extended HF response ).
- It can be replaced by GPA with a more modern ( extended ) diaphragm.
- I would recommend the 8 ohm version ( model # 34647 ) since using a lower impedance will allow you to drive the horn circuit harder ( getting a bit more UHF ) before attenuating its' midband response ( 2K-8K ) > with the 30923.
- FYI, the last version of the 9844(D) used an 8ohm driver ( 902-8B ).
- The 9844 was the pro version of your speaker.

:)
I have a calibrated mic from PE like you linked to earlier. I ran a coupe of RTAs with the UMM-6 microphone into the REV software on my MacBook laptop. Very small change on the graphs but the sound changes much more than it measures. The circuit does seem to pull back the most noticeable frequencies of the pink noise. First is without the circuit:
7:22:18 No attenuation 2 Feet.jpg

Second is with the circuit with 13ohm resistance dialed in via a potentiometer.
7:22:18  With Attenuation.jpg
 
Hi,

The mid-band attenuation is hard to detect because you're using a vertical scale of over 120db within REW.

One needs to use a vertical scale of 50 to 60 db to clearly see what's going on.

A 50db ( vertical ) scale is typically used by professionals when they are looking to make performance/quality assessments.

:)
 
I have something weird happening with the circuits I built into my new X-Over. It really seems not to make much of a difference how it is set.
8:5:18 Full Open Both Circuits.jpg 8:5:18 Full Open HF Full Attenuation Mid-Range..jpg 8:5:18 Middlie Both Circuits.jpgIMG_20180805_154003.jpg

I know I have the HF wired out of phase from the LF . I think that is how it should be since these are second order crossovers. Seem correct?
 
Flipped polarity ( or not ) isn't just determined by the order of the crossover, it's also determined ( just as much ) by the Z-Axis offset.

Whatever gives the best, most solid response at Xover , is the correct polarity to use.

The crossover does look nice!

The following ( older ) pics show the 30923 circuit has previously worked properly ( in your earlier crossover mock-up ).

It's anyone's guess, as to what's happened since then with your new build.

index.php

index.php


:)
 
Last edited:
Crossover Closeup.png

I'd relocate that Ground Terminal further away from the High Frequencies buss Terminal.

A partial short occurring there might be the root of your recent problems//observations.

Apart from that, it's a nice tidy arrangement.

:)
 
I'll look into it tonight. Thanks for the compliment on the build.

You're welcome!

You might want to use your iPad setup to duplicate your first 2 measurements ( #2 of 2 is seen below ) .
- Use the same scaling displayed in the example below.

index.php


They have that handy note telling one where to take a comparative measurement from.

This will give you a reality check to see if your new networks are good-to-go ( or not ).

:)
 
Last edited:
Here are my latest measurements. I do not see a short but that was a good thought.

IMG_20180812_151549.jpg

IMG_20180812_151855.jpg
Not really sure it is the best response. When I wrote the note, I didn't see the notch at 500Hz.

IMG_20180812_151103.jpg
The attenuation from the 16ohm L-Pad is crazy!

IMG_20180812_150746.jpg

IMG_20180812_150412.jpg
This is the configuration I am currently listening to. In general I'm pretty happy with the sound. It seems the notch at 500Hz is problematic at the one foot distance. I have the horns crossed over at 1060Hz I believe.
 
Hi,

Nice pics!

I've been travelling ( so I couldn't comment till now ).

Comments? Overall I think the response is quite good, with the HF being as flat as is possible with that horn/driver combo ( & going as high as expected ).

If those speakers were mine I'd be looking at somewhat reducing that 1K peak.

That 1K peak seems to be coming from the twin 12's.

index.php


You might try adding some capacitance onto the 10.5uF cap found within the woofer circuit ( creating a new total of, as much as 24uF ) to see if you can level out that frequency area.

index.php


:)
 
Last edited:
Back
Top Bottom