 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.105.104.136
In Reply to: RE: Alternative to the EV ST350 horn tweeter on EV Sentry IV-A's posted by Geary Lyons on October 23, 2013 at 20:20:43
Geary
As the pic in Grindstone's post alludes to, the ST350 was meant to be used with a passive EQ circuit following the X36, to compensate for the falling high freq. response of the CD horn. The ST350B EDS shows a 7.5 Ohm resistor in series with an 0.18 mH inductor shunting the plus to minus, followed by a 2.5 to 3 uF cap in series with the ST350 plus terminal. Without this (or something like it) the ST350 will sound very shouty. You don't mention whether your DIY XO incorporates an EQ circuit like this or not. I would try the stock 3500 Hz XO freq. along with this EQ circuit, while adjusting the cap value. Your choice of a 4700 Hz XO point may have somewhat of a similar effect. Just my take on it.
Paul
Follow Ups:
Geary
Note that the reference by Grindstone and me is the Electro-Voice ST350B EDS (which is gone now from the EV discontinued products website!), while the one linked by you above (and what comes up with a google search) is the University Sound ST350B EDS, which has nothing about the EQ circuit which would follow the X36. Products which were superseded by newer EV products ( but were still being made) were frequently re-badged as University by EV, and got a gray paint job. The Sentry lVA EDS indicates a pro-sound application where active EQ would be present anyway, which explains why the passive circuit following the X36 is dropped in the later versions, but there was a fixation on not blowing the tweeter, which is a non-issue nowadays with a low power SET amp.
Paul
Paul this is great! Thanks. The good news is the XO is still on the test bed board so I can try this! The bad news is that I thought I had the XO done and was finished tossing $$$ at it!Would you provide your thoughts on the possible best way to proceed?
1) Keep my current XO point at 4700hz and do the prescribed EQ after the XO. Reading the new EDS, it sounds like the circuit works for any XO point "at or above" 3000hz?
2) Keep the upper band pass at 4700hz, lower the ST350 high pass to 3600hz and use the prescribed EQ circuit. Would this give me a bump in FR at around 4150hz?
3) Change both mid and high to 3600 and use the prescribed EQ circuit.
I in my listening tests I liked the 4700hz configuration better than the stock at 3500. The transition was more seamless and the high frequency smoother. Additionally the higher point potentially offers a bit more tweeter protection as I removed the STR circuit. You are quite right about the tweeter blowing paranoia. Removing the treble attenuation and STR were big improvements.Thank you for taking the time to "bonehead" this for me!
Cheers,
Geary
Edits: 10/24/13 10/24/13
Geary
The 4700 Hz XO point would have a similar effect to the EQ circuit following the X36, that is to say suppressing the midrange hump at the lower end of the ST350's response as seen in the EV EDS. It's no surprise that you preferred the 4700 Hz to a 3500 XO, the ST350 sounds really harsh without the EQ circuit, and more like a T350 with it. I would at least try it with the 3500 XO and the EQ circuit, and you have the option of adjusting the cap value to fine tune it. Lowering the cap value will decrease the steepness of the ramp beginning at about 2500 Hz and tilting downward towards the highs, at the expense of losing sensitivity as you suck more energy out of the lower end. Use only air core inductors (no iron)! The concern with the 4700Hz XO is that there might be a dip in the response in the area where the 1823M and ST350 integrate, but you'd have to measure this and see. The 1823/1824 family don't have phase plugs and just can't go very high, hence the EV tweeters going lower than others like the JBL's, and why you don't typically see JBL tweeters with these EV mids., but EV T350's can work with JBL etc mids.
You would have to measure the response of the EQ circuit with a 4700 Hz XO to see what happens, and how it reacts with the higher XO point, but adjusting the EQ cap value should have similar results.
I would try working with the ST350's first before going to other tweeters. The Fostex 900's that RCA Fan recommended look great from the specs, but at $1800 a pair they may make the ST350's start to seem like a good deal. Anyway, let us know how it goes. Good luck from the Electro-Voice Brotherhood!
Paul
Paul,
I am going to stick to the stock ST350's. The EDS that grindstone posted shows 3 alternative XO/EQ examples. 3kHz, 3.5kHz, (X36) and 6kHz. I could go to a 3.5kHz, Linkwitz-Riley alignment, (that is what the X36 appears to be, then add the EQ, but that is more components in the signal path. And I would have to buy new air core inductors. Also, I am concerned with taking the ST350 that low without the tweeter protection circuit, which adds even more! In your opinion, should I be concerned?This is the 3kHz:
I like the current 4700Hz, but you shared some concerns. Since 4500Hz is 1/2 octave above 3000Hz could I not just multiply these values by the SQRT of 2, (1.4)? That give me a C 3.4uF and the L .448mH. That calculates pretty close to a straight Butterworth filter at about 4100Hz, so I assume the EQ is from the filtering of the 3kHz hump in the raw FR.
The 6kHz example has me confused. My electronic mathematical skills are limited, but if I am reading this correctly are they saying to parallel 3 inductors to have an effective inductance of .12mH? .
Here is my confusion, .12mH in the crossover would not get a filter anywhere around 6kHz. Am I missing something here?
The 3kHz bumped half an octave seems the cleanest alignment, with the fewest components. (And close to good foil inductors that I have to hand!)
Is my logic faulty on the "bumped" component values to get to approximately 4.5kHz AND have the desired EQ?I really appreciate your, and everyone's, input on this Sentry odyssey!
Cheers,
Geary
Edits: 10/28/13
The "Fig 5" circuit is a 12dB/octave Butterworth high-pass with a corner of 5300Hz. It therefor attenuates the energy going to the tweeter, 12dB/octave below 5300Hz. So it gives some pretty good tweeter protection.
In conjunction with the tweeter's falling response, then net total acoustic response is roughly equivalent to a highpass at 3000Hz, according to the manufacturer specs being quoted. But that is NOT what the electrical filter is doing, that's just the total effect of filter plus tweeter response.
Thanks PJ,
That makes sense. I can do the XO math, but understanding the total systemic effects mostly alludes me!!! So without more knowledgeable input from experts/wizards/engineers I wind up with lowest common denominator solutions. Just trying to make an even better match for the Paramours!!
Cheers,
Geary
This has a lot of similarities to what we did with the old "Climax" speaker design. A lot of tweeters have non-flat response (especially horns) that can be compensated in the crossover, but the book values have to go out the window to do it.
Thanks PJ,
I am pretty much a book values beginner on this. I started with the original EV crossover box and listened, measured and tweeked from there to get to my current XO. But, admittedly, I used "book values", i.e. the XO formulas in Excel and/or the online crossover calculators.I think the speakers sounds pretty good, with the exception of the high frequency steep roll off. They sound exponentially better than the factory XO boxes, but that is likely due to the much better components that I used versus the nearly 50 year old components. Hopefully the EQ circuit will improve on the tweeter results.
I am always grateful to those who share their experience and expertise, because that allows me to achieve much better results than my limited expertise would provide.
Cheers,
Geary
Edits: 11/04/13
Geary
The Linkwitz-Riley XO comes from direct radiator land where it's useful to focus the main lobe radiated in a 2 way box speaker. In a rig like the Sentry lV, the mid and tweeter horns will dominate the dispersion in their pass bands, and the natural steep bass roll-offs of the horns will dominate at the lower ends, so a 2nd order Butterworth should be fine.
As I recall you mentioned using a 45 or 2A3 SET amp which would indicate a music listening application for the Sentry lV's, and it's rather unlikely you would blow an ST350 with a sub 5 Watt amp. You would have to push the amp into hard clipping with a test signal long term to do it. I use my ST350A's as rear firing tweeters in a 3 way di-pole horn rig with 6 drivers per side, and I cross them at 2500 Hz with a first order series crossover. The forward firing tweeters are T350's, and I've been running them with 2500 Hz 1st order XO's for well over a decade and I have never blown one. Ditto for the T35's which the ST350's recently replaced. If you are concerned about this, you can use the STR tweeter protection modules, but I had these back in the 80's and never heard one switch out even with a 100 WPC SS amp playing loud rock'n roll, so they went in the parts box. BTW the EQ circuit will in itself offer some protection to the tweeter as it's shunting energy through the resistor. I just don't see tweeter protection here as an issue, but as I don't have control over your rig, I'm not gonna guarantee your voice coils ; )
I don't have a freq. resp. graph for the 1823M, but my 1823 and 1824M's start to roll off quite steeply much above 2500 Hz with my DIY tractrix horns, hence my choice of this XO. Crossing the 1823M and ST350 at 6 K would guarantee a dip in the response between them. Though a 4500 Hz XO may give acceptable results, a dip could be expected there too. Once again, you could straiten this out to a degree with active EQ, which would be there anyway for feedback suppression in a pro sound setting.
I'm not sure exactly what you mean with the "bumped" XO idea. If you parallel three 0.12 mH inductors, they would function as a single 0.04 mH inductor in a circuit. Why EV spec'd it this way is hard to say, perhaps it was cheaper to do it this way than a single 0.04 unit due to parts availability, and they had 0.12's on hand from something else. You can connect inductors in parallel to get a lower value, or in series to get a higher value. Caps have a reverse effect and are connected in parallel for a higher value, and series for a lower value. There are online calculators for this (do a search with "parallel inductor calculator" etc), as well as calculators for XO's and notch filters (which the ST350 EQ circuit is), see link below.
BTW The spec'd 7.5 Ohm resistor in the EQ circuit can be hard to source, so I used two 8 Ohm resistors in series, and paralleled with one 8 Ohm resistor from RS which is close enough. I use a 0.20 mH Jantzen foil inductor, and I settled on a 2.20 uF cap. Larger caps (up to 10 uF) did'nt seem to do much, while smaller caps (down to 0.68 uF) shunted so much energy out of the ST350's that the old T35's could easily drown them out, along with some roughness appearing at the upper end of the ST350's with the 0.68. I think you're just going to have to break down and buy some parts. Don't worry, you'll use them from the parts box sometime down the line. I always do.
Paul
OK! OK! 3500Hz it is!! ;-)!! You are quite right on the 1823M roll off, as well. This is from the 1823M EDS with PA horns but, probably similarly applicable with the ST120's.
My reference to the X36 being Linkwitz-Riley was due to the values of the capacitors and inductors that they used. A L-R alignment with those values is about 3500kHz.
My alignments are Butterworth. I prefer Butterworth, so...will the EQ be affected with approximately C = 4uf and L = .5mH for the XO before the EQ rather than the X36 values?
Just for grins & giggles,I am not sure about what EV engineers meant for inductor value on the 6K. I couldn't find what the referenced part number value is. I got the same .04mH paralleling 3 x .12mh and thought "no way!". So three of those referenced parts yielding .12mH was my conclusion, but still doesn't work out to 6K for me!!!
Off to make a BOM!!! Butterworth or X36 values? What did you use?
Thanks for all of your practical experience!!!!
Cheers,
Geary
Geary
The X36 is a Constant K crossover, the main feature of which is a low insertion loss, and secondary to this is the use of two identical value inductors: 0.727 mH (see the X36 EDS for more, if it's still up on the EV site). Constant K's are seldom seen now except in line level stuff, but EV championed them back in the 70's, which always seemed odd to me due to the fact that the T350 (107 dB) and 1823/1824 (105+ dB) mids had to be padded down to mate with even high efficiency direct radiators in boxes. Anyway, the three parallel inductors in the 6K XO for the ST350 start to make sense in the light of this. The sensitivity of the ST350 is listed in the EDS as 101 dB with the aforementioned EQ circuit used, and the use of 3 paralleled inductors was most likely an attempt to maintain as low a DCR as possible to maximize sensitivity, without having to source a single low DCR (and thus heavier wire) choke of the necessary value.
I did'nt take the 6K XO seriously in fig. 6 due to the fact that you would realistically need active boosting in the area between the 1823M and ST350 to make it work, but it 's a stand alone XO for the 1823M and ST350 none the less, which slipped by me. You can use one of the online calcs I linked to design your own 6K XO for the ST350, and then design a notch filter to pull down the ST350 mid hump centering around 2.5K Hz., and see what you get for comparison. As Paul Joppa's post notes: the XO, the accompanying EQ , and non-flat response of the CD horn ST350 are all a working unit. You would have to try the EQ circuit with a Butterworth and see how it performs, but I would'nt expect any drastic departures from what I've seen.
For the easiest way to go, just combine the X36 with the EQ circuit already discussed. You'd need just 2 caps, 2 coils and a few resistors from Parts Express, and then see how you like it. You can probably even run the ST350 wide open with no L-pad. Note Inmate's post from a few days ago concerning matching the coverage pattern of horn drivers. The ST350 is probably a better match to the ST120 mid in this regard than many others. I love the Fostex alnico horn tweeters, but they have been criticized by some for being a bit beamy. You may run into problems however in a small room (or a rectangular room with the Sentrys on the short wall) with reflections off the side walls from a mid horn with a wide horizontal coverage pattern like that, resulting in smearing of the stereo image, but you don't seem to be complaining about that so far.
I moved on from my X36's (and X8's) some years ago after measuring the response of my 1823's and T350's with no crossover. First order 500/2500 Hz parallel seemed the obvious way to go, and later first order series crossovers. That's perhaps a bigger can of worms than you wanted to open though.
Paul
Paul,
I have a plan....I will go with the 3500hz, but with the Butterworth alignment rather than the X36 alignment. I brought the woofer/mid up to 550Hz, may help a woofer dip from about 250-400Hz. Comparing all of the available EDS FR plots, (lot's of assumption/extrapolation, since none really are the component mix in the Sentry's), I do think I will get the smoothest FR with this alignment. Thank you for pointing out the 1824M drop off above 2500Hz. The 1823M, with the PA horns, looks like that drop starts around 3000Hz to 3400Hz so that really focused the XO point.
I validated using the calculators in the link that you provided. It is more sophisticated, allowing the selection of an initial slope of 3dB or 6dB. (BTW, great link. Much good stuff there!! I saw why I was getting vastly different component values from different formulas and calculators! That's my handicap...good at math, weak on what the results mean acoustically in the real world!) If the EQ gets the ST350 10kHz -15kHz up eliminating the steep drop off, I will be a happy camper....and an EV purist!! ;-)
I will go with this to listen/test/tweak. I have some 15ohm Mills 12W, (overkill) I can parallel for the 7.5ohm and caps from 2-3.3uf to try. I will probably start with a 2.2uf as I have some Russian mylar/paper/oil caps that may be perfect for the tweeter. So I have to hit Parts Express for the .18mH copper foils. Timing is everything...PEx sent me a "Come Back" 5% discount code!
Thanks again for your time and valuable input!!! If you have other thoughts it would be great to hear them.
Cheers,
Geary
Geary
The XO between the mid and bass is an area to look into as the 1823's roll off steeply around 500 Hz at the bottom end, and I see the Sentry lV uses a 400 Hz XO point.
Once you get everything working, you can try getting the 1823's and ST350's in phase. To do this you can build a sled so you can mount the tweeters on top of the cab, and slide it back and forth. Reversing the tweeter connections from their normal hookup, you can set an oscillator at the XO freq., and slide the sled back or forth to find the point where you get the greatest cancellation at the XO freq. with a mic placed at your listening position. The test tones on the Stereophile Test CD will work for this if you don't have an oscillator, and a reasonable quality mic plugged into something with a VU meter will suffice. When you reconnect the tweeter normally, the mids and tweeters will support each other at the XO freq.
Keep us posted on your progress!
Paul
I guess I was thinking about the Fostex T500 which currently costs $1034.40 ea., while the T900A actually costs $535.50 ea. from Madisound.
Paul
I will only add that, frequency curves aside, you should also be looking for a smooth transition from the mid to the tweeter with regard to dispersion. This has a major effect on the reverberant sound in the room.:)
Edits: 10/25/13
Hey Geary,
Paul has a good bead on this. I would only remind you that the x-over isn't just about FR curves, it's also about SPL and diaphragm excursion. The ST350 phenolic diaphragm is fairly "delicate" (which partially accounts for its excellent HF performance). Depending on your SPL desires, you'll want to cross to it at a reasonably high frequency and/or steep slope, so you don't make it go too far one way or the other. While E-V specs (I think) 3000 Hz as a low end, that depends on how loud you're playing the system. I cross to my T350 at about 5500 Hz, at 12 dB/octave, for home use, and I can crank it 'til I'm happy, with no worries.
:)
Agreed! I doubt I listen to these much over 80-85dB. I was listening more for smooth integration and transition from the mid horn to the ST350 than FR. Quite honestly I started at the stock 3500hz and tried XO points every 250hz. Right about 4750hz sounded really nice. I don't think I heard much difference at 5KHz, so stopped. Maybe I should have tried higher!?
The ST350 is 3-4dB/1W/1M less efficient than the 1823M/ST120 mid combination. The tweeter attenuation in the stock crossover was for protection, not level matching or FR.
This discussion has been most helpful and explains what I was hearing and my measurements. Since this EQ is only on the ST350 I won't be mucking up the midrange.
Wish I started this discussion months ago!!!
Cheers,
Geary
Thanks Paul, that sheds some light on the plot and dispersion graph. Here is my current XO schematic. When I build out the final board it will be with bi-wire for the bass bins. I need to remeasure with this xover. I measured with a xover closer to stock, but need to borrow a mike again.
Thanks to Grindstone for the data and polar plots, and thanks to you for the x-over and EQ insight.
All of that jives with my recollection of the ST350 driver.
Ya know, it's funny. Sometimes, we think "heck, I'll never need this spec sheet again" and then toss it out. Then, 20 years later, ya get onto an Internet forum, and someone's asking about that exact product!
Another reason to never throw anything out!
:)
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
