Sep 262016

I’m excited by what I’m hearing from my Altec A5 Voice of the Theatre loudspeakers.

a5 votts

I’ve made a few modifications to the crossovers I’m building, and to the wiring from the crossovers to the drivers, that really are working great together.

I still have a few experiments I want to try with the crossovers, but I feel like I’m *almost* there. Actually, even if I stopped right now the performance would impress most listeners mightily I think.

Duke Ellington Indigos

Last night on Facebook, I was telling friends:

“I’m listening to Ellington Indigos on vinyl right now while I’m dialing in the crossovers on my vintage Altec A5 Voice of the Theatre loudspeakers. It sounds spooky real!”

“Ellington at live-like levels sounds naturally live, and with a sense of dynamic ease that is spellbinding. Good stuff!”

My crossover diagram was getting so marked up with changes that I decided I’d better make a new one for you that would be easier to read (below).

I used some Mac compatible software called Electric Design to draw the crossover circuit and it only costs $17 and works great – a bargain! Mac users can download Electric Design from the Mac App Store. 


At the urging of Pete Riggle, I installed a pair of 16 Ohm L-pads (see above) into the high-frequency circuits of the Hiraga-inspired crossovers I built, and it has really helped me to get them dialed in so that high-frequency and low-frequency horns are well-balanced and blending together beautifully.

At Yazaki-san’s suggestion, I also have a pair of Clarostat potentiometers (R3-VR in the circuit show above) installed on the high-frequency notch filters at R3 to give me a little adjustability for the depth and shape of the notch filter.

Pete Riggle was so kind as to run some SPICE simulations of the combination of the L-pad and the variable resistor (VR) in the notch filter for various values, and even drew the frequency response plots and labeled them by hand! I must admit to being awed! Thank you, Pete!

Pete: “Attached please find Bode diagrams (frequency response plots) showing the electrical output of the 16 Ohm Hiraga crossover into 16 ohm resistive loads. The assumption is that the tweeter has an L-pad which allows the output of the tweeter to be adjusted relative to the woofer.”

“Figures 1, 2, and 3 show the results assuming the tweeter/L-pad combination is set to be 12 dB more sensitive on axis than the woofer.”

Figure 1

Figure 1

Pete: “Figure 1 has the 24 Ohm value specified by Hiraga for resistor in the notch filter. This give a very nice result, with a 3 dB bump around 500 Hz and a 1 dB dip around 4 kHz.”

Jeff: With my A5’s, even though the Hiraga 24 Ohm value at R3 for the notch filter measures very nicely, I found it to be overly aggressive in the highs, so I have tried values for R3 between 24 Ohms and 35 Ohms, which makes for much more musical results in my system / room.

Figure 2

Figure 2

Pete: “Figure 2 has a 32 Ohm value for the notch filter resistor. Note that, relative to the results of Figure 1, this change of resistor value depresses the response output by an additional 1 dB at 500 Hz and 2 dB at 4 kHz, relative to the response shown in Figure 1.”

Jeff: Even though Figure 2 shows a ‘bumpier’ plot, my listening tests have revealed that R3 values between 30 Ohms and 35 Ohms sounds better in the musical sense. At 24 Ohms I cringe at the musicality, and between 30 Ohms and 35 Ohms I smile, while cranking up the volume and grooving out to the music! At the moment my total resistance for R3 is 30 Ohms, and it is sounding very nice.

Figure 3

Figure 3

Pete: “Figure 3 has a 16 Ohm value for the notch filter resistor. Comparing the response with that shown in Figure 1, note that with the smaller notch filter resistance value the response still has the 3 dB bump around 500 Hz, that we see in Figure 1, but that the output at 4 kHz has risen 2.5 dB from the response curve of Figure 1. The response curve of Figure 3 has a little misleading because of the expanded vertical scale provided by auto-scaling by Spice.”

Jeff: A 24 Ohm resistor in the notch filter (R3) sounds nasty to me, so I can’t even imagine going to 16 Ohms! That will be the R3 experiment that I don’t try!

Pete: “Figures 4, 5, and 6 show the results with the tweeter shelved 6 dB higher than in Figures 1, 2, and 3. The combination of tweeter and L-pad is set to be 18 dB more sensitive on axis than the woofer.”

Figure 4

Figure 4

Pete: “Figure 4 shows the results assuming the tweeter with its L-pad set to be 18 dB more sensitive on axis than the woofer. Here we have shelved the tweeter 6 dB higher than in the configurations of Figures 1, 2, and 3. With the 24 Ohm notch filter resistor we see a 7 dB bump in response at 800 Hz, slumping down to a trough at 3.5 kHz, the trough bottoming at 4 dB above the woofer baseline output. As with all these calculations, the tweeter output rises after the trough.”

Jeff: The Figure 4 combination would kill me! 😉

Figure 5

Figure 5

Pete: “Figure 5 is for a configuration like that of Figure 4, but with a 20 ohm resistor in the notch filter. The 7 dB bump in response at 800 Hz remains; but the trough at 3.5 kHz bottoms at 6 dB above the woofer baseline output.”

Jeff: Arrrgh!

Figure 6

Figure 6

Pete: “Figure 6 is for a configuration like that of Figures 4 and 5, but with a 16 ohm resistor in the notch filter. The bump has now moved to 1 kHz, and increased to 7.5 dB. The trough is now at 4 kHz, where the output is about 7.5 dB above the woofer baseline output.”

Jeff: Arrrgh times two!

Figure 7

Figure 7

Pete: “Figure 7 gives the result for a case like that of Figure 1, but with the tweeter shelved 6 dB lower. Here the L-pad/tweeter combination has a sensitivity 6 dB above that of the woofer. The notch filter has the 24 ohm resistor. There is a response bump of 1 dB at 400 Hz, and a response trough of 7 dB at 4 kHz. Again, the auto-scaling provided by Spice has given an exaggerated view of the vertical scale.”

“One would think that the best sound would be with the settings as used in the configuration which gave Figure 1.”

“Of course these figures show only the electrical response of the crossover, with adjustment to show the effects of tweeter/L-pad settings. In the real world there would be small deviations attributable to the electrical impedance characteristics of the drivers with their horns and enclosures,  and changes attributable to the on-axis and off axis frequency  response characteristics of the drivers. These considerations aside, the Spice results give a pretty good idea of how the crossover itself behaves.”

“One thing I would point out here is that it is worth studying the nature of the changes attributable to the L-pad versus those attributable to the notch filter resistance.”

“Sometime tomorrow I will email a figure of the Spice model including the metering elements used to get the response curves.”

Figure 8

Figure 8

Pete: “Attached is Figure 8, a copy is the Spice circuit I used for the 16 Ohm Hiraga crossover circuit. To model the 8 ohm circuit, the resistance and inductance values need to be halved, and the capacitance values doubled.”

“Something I forgot to include in yesterday’s discussion of the response curves is the expected nature of the sound corresponding to the nature of the curves. The reader may wish to refer to Figure 1.”

“Here are my thoughts: The bump centered around 400 or 500 Hz might give the mid-bass a little welcome warmth. This bump occurs high enough in frequency that it doesn’t risk masking the deep bass. There is a range of frequencies centered on 3 or 4 kHz in which “hardness” of reproduced sound tends to show up. I believe this is also referred to as the “presence” range. My experience using the DEQX with the Po’ Boy speakers indicates that a trough with a few dB reduction of the response curve in this range tends make a speaker to be more listenable. This is exactly what the Hiraga network does. Also, the bump centered on 400 or 500 Hz may be helpful in making it possible to execute the trough without losing the “presence” of the sound.”

“A few words may be in order to explain the metering circuit. Spice calculates the electrical power delivered to the metering resistor, converts that power to decibels, and creates a frequency response curve (Bode diagram). The power delivered to the metering resistor is an analog to the acoustical power which would be delivered by a perfect powered-speaker (flat on axis response, fixed input resistance) driven by the crossover network. The 10 kOhm used as a meter resistance is high enough not to significantly disturb the behavior of the crossover circuit.”

“I did some careful thinking about how to meter the combined (complex or phasor) output of the woofer and tweeter as indicated by the Spice circuit for the crossover. One might think that because we want to add the acoustical output of the woofer and tweeter, that we should be seeking to meter the phasor sum of crossover output voltages for the woofer and tweeter.  However we intend to wire the tweeter out of phase with the woofer. Therefore we need to meter the phasor difference between the crossover woofer and tweeter outputs. I came to the considered conclusion that connecting the metering resistor between the crossover woofer and tweeter outputs does meter the phasor difference between the outputs.”

“That is it for now.”

Jeff: Wow! Thank you, Pete, for all the thought you’ve put into this and your efforts in creating these diagrams.

Test pressing from Analogue Productions of Chet Baker's 'Chet'.

Test pressing from Analogue Productions of Chet Baker’s ‘Chet’.

What I can tell you, is that between being able to adjust the high-frequency attenuation with the L-pads and the R3 notch filter, my A5 crossovers are really getting dialed in and working extremely well, and it is amazing what these ancient Altec A5 theatre loudspeakers can do!

For example, listening to the Analogue Productions test pressing of Chet Baker’s Chet (above) with L-pads turned down between a third and half-way, and the R3 notch filter  with about 30-35 Ohms total resistance, sounds astonishingly musical and mesmerizing, which is somewhat contradictory to what the SPICE plots of the 24 Ohm R3 calculations of Pete’s suggests. Chalk it up to system / room / personal preference / position of the planets / variables, I guess.

Here’s the bottom line: Use calculations to help understand how the variables affect the response curves, and use that to inform your decisions on what to do, but at the end of the day trust your ears.

For me, balancing the adjustment of the L-pads down, and keeping the notch filter R3 up at higher resistance, gives incredibly musical results.

About wire: I’ve got Duelund DCA16GA as speaker cables going to my Altec’s, and it trounced my Western Electric WE16GA in that position, sounding more musically compelling and natural.

The Duelund DCA16GA as speaker cables is really nice.

The Duelund DCA16GA as speaker cables is really nice.

However, for the wire from the crossovers to the drivers the tables were turned, and the Western Electric WE16GA ended up sounding more musical in that position, being more musically natural and compelling in the way music was presented.

To be fair, the Western Electric has a ton of hours on it, and the Duelund DCA16GA had no hours on it, so I should probably run-in the Duelund on the Cable Cooker for a while before making a final decision.

Ok, that’s all for tonight, and I want to thank Pete Riggle for all the work he put into doing the SPICE calculations and working up the figures for the various L-pad and notch filter combinations, it was very illuminating!

As always, thanks for stopping by, and may the tone be with you!
 Posted by at 7:38 pm

  16 Responses to “The Altec A5 Voice of the Theatre Project – Wow!”

  1. I would like to make 2 comments:

    1) regarding the notch filter… I still don’t understand why you are using two R and the variable resistor. Why not a singe R and the L-Pad? Will you replace all with a fixed value R once you settle on a value?

    2) The HF phase is inverted on your speaker/crossover combo. The + is the – , and the – is the plus, but inverted 180˚ phase as in the original crossover. I need to look but this is either from the non-alignment of the LF and HF or due to the phase shift from the crossover. This is beyond me but you might want to get some feedback from Pete regarding whether the L-pad should of on the HF -. I’m not sure it makes any difference but I think it would be something to do a sanity check on.

    And one last comment / suggestion. The whole L-Pad attenuator could be scraped and replaced with an auto former from Mast Mutter… this would give you a better damping value on the HF side of the circuit and less R noise. I have a pair for my 604 project, and he will now sell them al la carte if you ping him and ask. The price is in there low $200s. Something to think about but it might just come down to whether is makes a big difference, and/or is it worth the not insignificant price.

    Love the Altec magic. Thanks Jeff. -Josh

    • Hi Josh,

      The Hiraga circuit inverts the HF +/- as depicted in the diagram in my post above, and I don’t think any of the R3 changes I’ve made affect that. The correct way to connect the L-pad is 1 connects to in/out negative, 2 connects to out positive, and 3 connects to in positive, as per the diagram. You ask a good cautious question though, and I actually did have Pete check the circuit prior to publishing the post, and he gave it a thumbs up.

      You’ve raised the questions before about why R3 is the way it is, and the use of an autoformer for HF attenuation, but my answers have not really changed.

      R3 uses 2 resistors and a VR because at the time we started planning for the project the 20 Ohm resistors were out of stock so we went with 2 10 Ohm resistors. The circuit above reflects what we did then. Since then they’re back in stock, so I ordered some 20 Ohm resistors and have them on hand for experimentation. The main advantage of using 1 20 Ohm resistor with the VR in R3 is to make the circuit more compact, which is not an unworthy goal, but I’m not sure it really matters from a performance perspective.

      Even after I figure out what level of overall resistance I prefer in R3, I will not be replacing the R3 sequence of resistors with a single resistor, as I want to maintain the adjustability to the notch filter to help optimize my A5’s for the different rooms I will use them in.

      As I have mentioned in earlier posts, I bought a pair of autoformer based crossovers that I had Werner Jagusch in Germany build for me, and I was not impressed with the results on my A5’s (I haven’t tried them on my A7’s yet), so I’m not all that keen about using autoformers in my crossovers.

      While I don’t doubt that Mast Mutter builds very nice autoformers which could be incorporated into my modified Hiraga-style crossovers, my crossovers sound so much better than the Jagusch autoformer crossovers do that there’s simply no comparison, so I’m not keen on changing something that’s working so extremely well on a rather expensive hunch, but I appreciate the suggestion.

      The adjustable notch filter in combination with the adjustable attenuation of the L-pad provides a very powerful way for dialing in the crossovers, and a rather remarkable level of performance. I recommend you give the approach a try before you dismiss it as a possibility.

      And as Pete’s plots suggest, when what theory says should sound best, doesn’t, it’s better to trust your ears and go with what works best for your system, room, and tastes, and it might be different for yours than mine.

      Thanks for stopping by!



      • Yah, the out of phase wiring between the HF and LF driver is as per original Altec 2nd order / 2 order crossover wiring as per their “1960s Loudspeaker Enclosures – Their Design and Use” guide shown here in Fig 19 so the Hiraga crossover schematic is still consistent with this. I guess has to do with the depth alignment of the LF and HF drivers in the z plane to match phase at 500Hz crossover point:

        The L-pad is definitely nice for dialing everything in. I have the original Altec 25 ohm variable R for mine. It’s kinda like an Lpad hah because it still have 3 pins with one path in line to the driver that goes (0-25) ohms and another parallel path that simultaneously goes (25-0) ohms and through an additional 20 ohms R to ground to help maintain a more flat impedance curve. Most people I see replace the variable resistor with a fixed R and keep the parallel R in place but miss that there is additional R from the variable R. Anyway long story short, yah I played with the position for a long, long time before pulling everything and checking values with a Fluke DMM, then going from there played with fixed values before finally settling on 18-20 ohms in my Flamencos. It was also fun to play with different R types, and this could also affect the exact value a bit. Vishay Z needed to be a bit higher in R, compared to the variable R which I guess was due to some inductance not sure. The Allen-Bradleys and Mills MRA-12 seemed similar with regards to value selection. The copper fittings made swapping these out easy and I really appreciate you and Pete from suggesting these in your crossover builds.

        On a final note about the auto former transformers you have… I haven’t had a chance to finish the 604 build or evaluate the autoformers I will just build a modified Altec N-1600 crossover which also does HF/LF phase inversion and I think it is 2nd order on both drivers. I’m not sure about the crossovers you had with the AF but I thought that he used a 1st order on the HF side which always seemed like it would ask to much of the HF driver down low. And then also raises the question about phase because 1st order doesn’t flip phase while 2nd order does.

        It’s nice to do what you are doing by listening… the modeling and theory only get you so close. Having the control is nice, and then in the end you have something the exact way that you prefer which is great!

  2. Thank you Jeff and Pete
    i am gone follow this close 🙂 and just order 16 ohm Lpads
    and tomorrow i get my Hiraga crossovers wired up again
    maybe this will outperform my Jagush crossovers till now i had no luck with the Hiraga crossover it sound a lot worse than the Jagush crossover
    but its alway good to try things
    thanks a lot for share this with us
    and all the clear information


  3. Hi Jeff. Can you tell me where you order for L-Pad please? Is it 16ohm mono?
    Kind regards

  4. Hi All:

    About Tweeter Phase:

    As I recall, for a two way crossover it is conventional to wire the tweeter in phase with the woofer for first order, and out of phase with the woofer for a second order. This is explained well in the “Loudspeaker Design Cookbook.” One has to be careful because a positive electrical pulse applied to the terminal labeled “+” causes a positive pressure pulse for some Altec compression drivers, and a negative pulse for others. Of course, we want to label “+” the terminal that creates a positive pressure pulse when energized by a positive electrical pulse. This may require relabeling the terminals of some drivers. The rationale is that in the early days, Altec labeled “+” the terminal that caused the diaphragm to move toward the mouth of the compression driver. However some phase plug arrangements are such that when the diaphragm moves toward the mouth of the compression driver, a negative pressure pulse is created. Later Altec got smart and standardized the labeling so that the “+” terminal that gives a positive pressure pulse when energized by a positive electrical pulse.

    About Notch Depth and Shelving:

    One would expect that the optimum notch may be deeper or shallower than that created by the circuit as Hiraga specified it. The preferred notch depth will depend on many things, including the driver, the horn, the room, the tweeter shelving level, the listener, and maybe even the preferred listening volume.

    Best Regards,

    • Thanks for the great comment, Pete! I’m looking forward to have you give it a listen of Sunday to see what your take is on an optimum setting. Best, Jeff

  5. Hi Jeff-san,

    I’m happy for your success in this (on-going) Altec journey.. If ever you try the Altecs with SET amps, you may want to experiment with adjusting the density of the (pink) fiberglass batting for the bassbins. The bass response would most likely change with the relatively higher output impedance of zero feedback SET’s giving a somewhat looser bass response compared to the more traditional PP pentode amps with feedback. Also, it is good that you have adjustability for the compression drivers’ XO. SET’s higher output impedance will also change it’s response somewhat.

    BTW, if Jupiter ever makes Tin Foil in Wax-Paper caps, those would be most interesting to try.


    • Hi Rick,

      Thanks for the tips – appreciated! I’m looking forward to trying an SET with the A5’s tomorrow when Pete brings his Yamamoto 45 over, it should be fun!

      I would love to hear some Jupiter tin-foil in wax paper caps, who knows, maybe they’g give them a try. I like the Red Astron’s but they’re tin-foil and mylar. They sound great in some of the applications I’ve tried, but one can over do it with too many of them, I’ve found.

      Kind regards,


  6. ampohm made some tinfoil paper in wax capacitors
    i used them in one of my amps i like the sound but never compare these with something else.

    Have fun Jeff with listening to the 45 its a wonderful tube i like it on full range drivers like lowthers
    on the A5 i like it also but a 2a3 is for me the minimum tube on this speakers .
    but you should hear a trace of the magic that the 45 has
    the Yamamoto is a good fine amp.
    have a good time 🙂

    • Hi Robert,

      I do like the 45 tube. I’ve had two 45 amps in the past that I used on my Avantgarde Duo loudspeakers. The first was a custom set of 45 monaural amplifiers that Gordon Rankin at Wavelength audio made for me, and the second was the Yamamoto 45 that I reviewed for 6Moons some years ago.

      It will be interesting to hear the Yamamoto 45 on the Altec A5 loudspeakers – I’m looking forward to it!

      Kind regards,


  7. Hello Jeff,

    I am looking forward to read how the Yamamoto 45 is doing! I read just some nice information about this fine amp. The looks are in some way funny but at the same times very nice.

    Greetz, Have a nice listen!


    • Hi Siedy!

      I am looking forward to hearing the 45 Yamamoto SET on the Altec A5 Voice of the Theatre loudspeakers.

      The Yamamoto 45 is the same 45 Yamamoto that I reviewed for 6Moons years ago, and has been in friends’ systems since then, first Bill VanWinkle, a beloved sightless piano tuner near me who I was able to share a number of wonderful adventures with, and that purchased the Yammy after I reviewed it, and then willed it to his friend Pete Riggle upon his passing.

      You could say there is a bit of a sentimental attachment for us here at Jeff’s Place for this particular Yammy 45, so it will be a wonderful experience to hear it again, and to reminisce about our friend, Bill.

      Thanks for your good wishes!

      Kind regards,



  8. Hi Jeff,

    I just read you review in 6moons from……2004! I loved reading it and to (re)discovered which paths you followed in your ‘audio journey’. An impressive journey! I recognize a lot of my audio traveling in your journey.

    Now your are learning a lot about the amazing Altec Lansing speakers. And, we as your readers are learning a lot too, now!

    I heard a lot of Altec Lansings in the past. I always liked the sound. Impressive as always was the speed and openess of the sound and also the way they can deliver dynamics. I liked especially the AL 604. This AL 604 sound has a lot of common with the Alnico DC unit of the Tannoy Westminster. Perhaps, in the future, you wil do some experiments with this 604 unit.

    Again, have a nice listening and, indeed, it is a very nice reminisce of your friend Bill.


  9. Hi Jeff,

    Just a short completion on my previous message.

    The AL who I admired a lot was the: AL model 18, in a so called 620 cabinet. The speaker unit was the 604-8G. We call them also Duplex speakers. My good friend Bert has owned them.

    We compared them at time with the Tannoy 15 DMT mk2. He choose to go on with the 15 DMT. Although he knew that the AL 18 was the more dynamic speaker! I am not sure how many years it is that he owned the (10 years?). Sometimes he is thinking to buy the speakerunits and built the cabinets himself. In the past he buit the cabinet of my Tannot GRF. So, I know how good he is in building cabinets!

    Secretly I hope that he wil start building and assemble them! I will be at his place when these speakers will start blowing us away!


 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>



%d bloggers like this: