UREI Model 829 crossover design

[Steve Mac]

Thanks and very interesting...
Why auto lamps?
I've seen them used in older circuit designs where they wanted the resistance to decrease with increased current flow(I think), what do they do here?

Plus there's what looks like an LRC circuit just before the woofer with L4-R10-C4. Is this impedance correction for a certain frequency?

[donaldpatten]

Steve
Just the opposite ; the more current, the hotter the filament, the higher the resistance... the more protection for the drivers. Also it looks Cool when the lamps flash.
C4, L4 & R10 form a notch filter @ 900 Hz to flatten out the woofer mid freqs. With the pot ( R10 ) at max ( 25 Ohms ), C4 & L4 are 5 ohms, giving the most loss. With the pot at min ( 0 Ohms ) the mid's are not attenuated, as if the three parts were not in the circuit. The range is about 6 dB.
Don

[Steve Mac]

That is interesting, thanks.

So it looks as though UREI built in some analog type compression by introducing the filaments, at least to some degree. So, in reality they are not necessary except for current protection. But that kind of works along with the fuses which are before the filaments.

[donaldpatten]

Steve
Remember these speaker systems were for recording studios and were played at top volume. So a little compression, better than blowing them. I think their ears were operating in compression mode too.
Don

[Earl K]

Hi Steve

- Thanks for the postings of the original 829 network. I haven't seen that before .

- Regarding C8 going to ground . I suspect it's meant to be as you've drawn it. My calculations show a turn over frequency of around 10K .
- As the 25 ohm pot ( R7 ) is turned ( & its' effective impedance is reduced somewhat ) , C8 which "sees" ground will pull down the high frequencies, ( as if it's the 2nd pole of a lowpass filter ). The correct wiring of R7 into the circuit is critical because one needs to make sure that there is always some inline circuit resistance ( or else at HF the amp will see a dead short ) . That would be why the wiper of that pot is wired such that it is after R6 ( the 10 ohm resistor ). I imagine maximum HF attenuation occurs through this routing. I'll guess that R7 is wired so that when it is "off" , high frequencies actually bypass around R6 and route fully through R5 on their way to the driver. R7 , when "off" ( fully counter-clockwise ) needs to be wired in a way that the measured resistance is the maximum available / between the wiper and the terminal that leads to C8 ). This resistance is necessary to keep 10K and beyond from having a permanent downward slope. Also, when the pot is fully clockwise, their needs to be significant measurable resistance between the wiper and the input to that pot ( for all the same reasons about keeping some inline resistance / amp-isolation ).

- I believe a HF trim like this is pretty "old-school". It harkens back to the days when not everyone insisted on a ruler-flat curve to mix by. I've seen variations of this circuit used with older Tannoy duplexes ( in the HPD series actually ). Their implementation was a bit more elegant IMO . All in all, this is quite interesting to look at.

- The DCR of L6 is quite critical . L6 & C9 form a series LCR notch filter , strapped in parallel across the lines. The DCR of L6 will determine the depth of the notch. My calcs. show it to be centered at around 1453 hz . This LCR filter seems to assume the role of the typical 2nd pole in a normal highpass/ but instead of using a coil that gives ever increasing attenuation below it's turnover frequency / the LCR is just attenuating an area of response right in the horns' crossover region . Going this route gives a gentle 6 db per octave slope with added protection in the crossover area. I've read that LCR traps like this don't impose phase changes ( though I've never tried to confirm this ) . Tom Dawson ( over at H.E. , A.A. ) indicates that he uses some circuits such as this ( the LCR part with a single pole filter ) .

Well, you asked for some thoughts , so I rambled off a few about how I see the horn circuit working . I'm certainly open to others' thoughts on my interpretation.

regards <> EarlK

[ckdf]

Thanks for posting this circuit, its indeed very interesting. I did some analysis and simulated it with PSPICE and the results are quite interesting.

Basically the circuitry implements a very steep crossover between the woofer and the high frequency horn
at about 1.6 kHz. Most notch filters that you find in there are used up to "steepen" the slopes of the crossover filters, except for C4/L4 which as someone mentioned before creates a dip at about 900 Hz.
The depth of the dip in response can be adjusted by R10. I assume that the bass driver creates some
peak due to cone breakup in that range that can be tamed by this filter.

The slope created for the HF horn is more than 24db per octave (i.e. like a 4th order filter) and for the woofer it is even much steeper. Obviously they wanted to as much as possible limit the overlap range where both
drivers operate to avoid issues with interference. Already 1.6 kHz is pushing a 15 inch driver a bit beyond its limits and above that things get worse. So you want to "turn it off" as quickly as possible above the crossover frequnecy.

The high pass section is designed to create a rising slope above the crossover range, probably to compensate a HF rolloff of the tweeter horn. As also already stated before this can be reduced with the setting of R7. R7 turned to a position where the tap connects directly to C8 gives a flat response, the more the tap is turned away from this position the more you get a rising HF response.


Regards

Claus K

[donaldpatten]

Steve
Earlk is correct : C9/L6 form a notch controlled by the AWG wire size. About # 28 offsets the HF driver impedance peak below xover. I recall the DCR was 4 or 5 ohms.
Don

[pbalzano1]

I just found this message board when looking up Urei 829s.
I'm not a tech guy, just a singer. My friend gave me a pair of these that came out of CBS studios. He's in a band with my wife and they wanted to do some remixes of some demos at my home. I recently sold off most of my recording gear but I have enough left over - except for good speakers - to help them out. I connected the speakers, they sound ok with all the crossover settings up all the way ( that's how I got them ) but I have no idea how to go about tuning these things for my room. Any generic starting points? I know it's probably a silly question but there's nothing I can find on the web about these. It seems they lack punch at the bottom unless you turn them up. Also, the highs and high mids really wear on your ears. Is it just a personal preference thing about the eq or shouldn't these sound somewhat balanced without going crazy on the eq? It would probably be a nightmare to try and mix anything the way they're set up now. Thanks for your time, Phil

[pbalzano1]

Sorry, I meant 811a with the 829 crossover.

[donaldpatten]

pbalzano1
Trust your ears. They are the best measurement tool you will ever have. You said your wife is a singer,and you know her voice, so start with a recording of her solo. Then listen and adjust the
two controls to get her voice. The studio engineers adjust the pots for max level , not flat response.If your pr are like the others out there, the pots are burnt open on the max end. Just reverse the wires to the pot pins 1 & 3. Now the max is min , at least it works.
Enjoy, Don