JE Labs Speaker Building Tips

 

Drivers:

After several years of searching for SE amp friendly speakers, I found out that only paper cone driver units are most suitable for this application due to its low mass. Good vintage units had cone edge suspension which moved anywhere from 1/8" to 1/4". This design philosophy was not for ultimate bass extension but excellence in midrange response. Decent 8" low frequency units typically have frequency response from 70-80hz to 5,000 and 10"- 15" units, from 50-60hz up to 1,000-3,000hz. The 8" Altec 755 series go a bit lower and higher than most of its 8" contemporaries and are considered full range.

Accordion edge suspension [increased compliance] started becoming popular from the late 50s to the 60s and bass went down to the 30-40hz region and only a few units maintained decent response up to the midrange like the 12" Altec 414 and the 15" 416. When looking for a woofer, listen to it [I use a pocket radio tuned to a familiar FM station with alligator clip leads plugged into the earphone jack or my SE171A amp] full range [in free air or use a baffle board] and watch out for peaky resonance in the 1,000- 3,000 range. The midrange has to have a natural and lifelike response because we want to take advantage of its gentle roll off in the top end instead of "choking" it with an inductor [choke] before adding a tweeter.

Likewise in searching for a sweet sounding tweeter, I do the "pocket radio test" and listen to the definition of human speech [even if they sound like little Martians]. If there is no excessive sibilance, spit or sizzle, they may be suitable candidates.

Most vintage drivers used Alnico magnets but some later ferrite versions are not bad sounding either.

Crossover:

Simplest is best! 1st order, 6dB/octave slope only for the high frequency driver. Let the smooth sounding low frequency driver roll off mechanically. Based on the "full range test" you can experiment with various 1st order [6dB/octave] crossover points to fill-in the loss in the top end which is easy to determine by using this crossover calculator as a baseline and then tweak [often coupled to a variable L-pad] the final value using your ears. You may also consult vintage brochures for specifications about the driver in question and start experimenting from the manufacturer's recommendations.

If you find old paper caps that are not dried up or shorted, use them because they sound good. Actually the 2uf caps I find in the Jensens and Heathkits remind me of Jupiter Beeswax. Please don't use plastic caps, they either sound damped [expensive Teflon] or spitty and sibilant [mylars and polypropylenes]. Paper [dry or immersed in oil] is best in my opinion. Paper in oil can types can still be found surplus for ~ $1/uf at hamfests and radio shows last time I hung out, so it might be worthwhile checking these events.

I've tried many types of crossover chokes from iron core to low DCR [DC resistance] OFHC copper flat ribbon and they all slowed down midrange response on any woofer I tried, vintage or modern. All it does is suckout the midrange leading one to think [psychoacoustically] that there is more bass and treble an effect we typically hear from modern high-end speakers. Much of the problem that make it difficult for SE amps to drive modern efficient speakers is due to the series/parallel inductors in the circuit. It may only be a tenth of an ohm DCR, but imagine the length of copper the signal has to travel before reaching the cone? Bad idea!

During preliminary testing, I advise using an L-pad to match the efficiency of the tweeter to the woofer. Once you have found the right sonic balance and alignment, you may remove the L-pad and measure the impedance across the 3 terminals and substitute fixed resistors. 8 or 16 ohm L-pads are affordable and available from Parts Express. Many times, I just leave them in circuit and forget about fixed "non-inductive wirewounds" which are expensive and hard to find in values I actually need.

Don't worry about padding down high frequency compression drivers. It is the nature of the design principle of compression drivers to be 10 dB or more efficient than high efficiency direct radiating woofers [at best, 100dB/1W/1M]. This is one of those real world compromise I learned to live with. The purist notion of not padding down is a high-end audiophile misconception in my opinion. Even modern dome tweeters are more efficient than their woofer counterparts. If you open any modern speaker and examine the crossover, there are fixed resistors to properly align the tweets with the woofs. This is the reason why Altec had built in attenuation terminals or an L-pad in their stock crossovers.

Please remember that my recommendations above apply only to those who use SE amps with 8W or less output. 1st order/6dB slope crossovers offer minimal protection to the drivers, especially fragile tweeters like the Altec 3000. Even a Dynaco ST70 can fry these units if overdriven.

Click on images to zoom or print the Jensen cabinet and open baffle plans

Speaker cabinets and open baffles:

Only use plywood! I agree with Walt Bender in his "Ask Walt" series in Audiomart, if MDF or particle board have good acoustical properties, how come no luthier has made a stringed instrument with this material? Even as late as the 1960s, BBC engineers specified solid wood as the material of choice for the classic LS3/5A cabinet however by the time they reached the market the cabinet was plywood.

Refer to vintage cabinet making plans from Jensen, Altec, EV and etc. for ideas. Those designers grew up listening to acoustic instruments in proper performing venues and knew what they were after because their research was based on using primitive tools and didn't have Ivie or Sound Technology spectrum analyzers and computers to misguide them in anechoic chambers. In the final analysis, they only had their ears to trust.

Open baffle sounds most lifelike to my ears and it is amazing how much bass a well designed moderately sized baffle projects. Horn loading [front or back] also don't work well in typical sized listening rooms. I was disappointed when I temporarily replaced my custom 614s with Altec 816 Voice of the Theatre loaded with 416s in my main listening room. Front loaded horn bass bins need a lot of space for the sound to breath or else you just end up with mostly "direct sound" which isn't lifelike. The problem I encountered was mainly due to the front horn loaded 416s acting too fast and the bass reflex port could not keep up at my listening position.

Altec 825/828/816 cabinets are best used if you have a big loft wherein you can actually accomodate a string quartet with about 20-25 people listening, they were designed for theatres. I also have a problem with rear loaded [and firing] bass/midrange horns because of the tunnel like effect on the midrange and no matter what you do, the bass will never synchronize with the upper midrange front firing horn, it will always be behind the beat.

The best compromise for a wide frequency range speaker system I found for a typical sized listening room is bass reflex for the low frequency with an upper midrange horn like the 811 in my Altec system. It may sound a bit boomy at a couple of mid bass frequencies due to cabinet resonance but this can be controlled by voicing the amp/speaker or source component interface and/or judicious damping of the panels [with larger cabinets, light damping may be required as I did with my 614]. When the cabinet is done right the resonance does not do significant harm to interfere with the more important midrange. I used 3/4" birch ply which looks nicer but on hindsight cheaper grade and higher resonance 3/4" may have better controlled midbass resonance. As the cabinet gets smaller like the Duette or Heath SS1, this is even less of a problem.

Speaker placement and tuning:

We go to concert halls with good acoustics because the sound we appreciate from instruments during a performance is a combination of direct and reflected sound. In my opinion, this concept is very applicable when trying to make a system sound right in a listening room, contrary to what some experts may say. For example, can you imagine the recorded sound of a violin 1 foot away from a fine Neumann U47 microphone in a dead [anechoic] studio? The sound will be horrendous even if it's a good player with a Strad. In fact when me and my colleagues are choosing an instrument, we try them in a good hall as a final test. The difference between a finely crafted modern violin and a Strad or del Gesu will not be so apparent in a small room. Sometimes the modern instrument will even sound louder or better nearfield. But in a good concert hall the difference will be significantly in favor of the Strad or del Gesu. This is due to harmonic overtones those revered instruments possess which can only gel from the combination of direct and reflected sound in a proper acoustic space.

I question the logic of a domestic environment tuned to behave like a studio. The point is, a properly designed speaker system should be able to function so that the sound it projects into the listening environment has the proper blend of direct and reflected sound by the time it reaches the listener's ears. This is the reason why I am fond of the open baffle concept. Due to its simplicity, factors that are detrimental to the purity of sound are eliminated. In spite of its limitations if loaded with a good driver, it can project a pure midrange and present depth and airiness with minimal obstruction. I always try and keep this in my sonic memory when tuning any type of speaker system.

Sources and references:

all rights reserved

Back to main page or Shop Talk