RETRO-MEASURE: Dynaco A-35 speakers (1972-1976). Technological maturity and audio. Comparison of the qSpin predicted room response vs. in-room measurement.

Audiophilia has a rich history filled with many products and ideas over the decades. Many of the "vintage" products can still be found these days on the used market and many carry with them an interesting story or might have developed "cult followings". I find it interesting examining these older "retro" products to help provide context when we think about the sound quality we have these days.

Some companies like the one we're looking at today - Dynaco - founded by David Hafler and Ed Laurent back in the mid-50's are well known. The company's products were popular through the '60-'70s including a foray into quad-channels. The company eventually got acquired and the brand decommissioned by 1980.

Dynaco was known for their amplifiers and "Dynakit"s they sold in the emerging hi-fi market back in the day to audio enthusiasts. Along the way, by 1969, they got into the speaker market with their "A-Series" line of products including the very popular A-25 (>1M sold worldwide according to Ed Laurent) and the speaker we'll be examining today - the less popular, larger Dynaco A-35 sibling.

Thanks again to linnrd for digging up this pair of old speakers for me to examine from his gear archive. It looks like these have been very well taken care of over almost half a Century. Well, at least there's no need to "break in" these speakers. ;-)

The 2-way design and drivers are the same between the A-25 and A-35. There's a 1.5" SEAS 87H tweeter with metal grille and a 10" SEAS 25TV-EW woofer. Given that this is the early 1970's, there are no fancy materials here, just the 1.5" 'textile' dome tweeter and 10" paper cone woofer. The woofer surround is made of butyl-rubber which is less prone to deterioration than foam, and AlNiCo magnet for the drive unit - the strongest permanent magnet one could get before the development of rare-earth magnets (eg. neodymium).

The exterior looks pretty good but as typically found with '70s era goods, there are some rough edges and machining imperfections here and there.

As you can see, the tweeter is set just above the woofer and off-centered to the left. With the grilles off, IMO, this asymmetrical, non-mirror image look is a bit unattractive so I suspect most will keep the grilles on.

Here's a nice photo of the speaker line-up:

We can clearly see that the A-35 is a larger unit and marketed as being "most accurate" within the line-up. They all have light-colored inset fabric grilles. One nice thing about the A-35 grille is that they're held in place with small Velcro bits for easy removal and cleaning; I believe the A-25's grille is glued in.

Dynaco A-35 fabric grille...

This is what the smaller Dynaco A-25 looks like. Notice the stuffed lower port. (From YouTube.)

With grille off, we see that the A-35 is missing the dampened port of the A-25 at the bottom front. The A-35 enclosure is reasonably sturdy but sounds hollow when tapped, material is said to be wrapped plywood with walnut veneer. These are by no means "small" bookshelves measuring around 32cm W x 57cm H x 25cm D. Each speaker weighs about 10kg. 

The A-35 is described as a "dual chamber aperiodic enclosure" (a bit different from the usual principle behind the aperiodic box with a dampened vent) with no external ports. The internal dampened narrow slot vent allows some flow between the woofer and tweeter compartments. In doing so, the goal isn't to lower bass extension, rather it's to lower the "Q" of the woofer, reducing stored energy, and allowing for better bass control.

Here's a look at the cabinet design. Volume-wise, the compartments appear to be the same size. I'm not sure what kind or how much "stuffing" there might be in those compartments to help optimize performance:

See plans including the A-25 here.

Let's now look at the back of the speaker:

Notice the presence of a 5-position control knob which changes the level of the tweeter. All the way clockwise on the tone knob will result in a "brighter" sound, and all the way counterclockwise (to the left) will dampen the high frequencies. The midpoint as you can see has been labeled "Normal", or perhaps more accurately their most "neutral" setting. Two standard binding posts accept banana plugs, bare wires, or spades. Proudly "Made in Denmark" on the label.

I have seen descriptions of the crossover being slightly different between the A-25 and A-35 (eg. here's a post suggesting use of an 8μF capacitor in the A-35 as opposed to 5μF?). Specs sheets I've seen suggest that the crossover is somewhere between 1-1.2kHz. Otherwise the 5-position switch to change the tweeter level is the same. Here's a peek inside the A-25 (here are more pics):

From https://www.updatemydynaco.com/A25LoudSpeaker.html.

As you can see, the crossover is very simple and for reference here's the circuit for those who want to open up the A-25/35 to do some tweaking or refreshing:

See here for the source.

Back in the day, I believe these Dynaco speakers were popular with the college kids and in dorms. Looking at the price sheet, the A-25 had an MSRP of US$94 / $72 kit each, and the A-35 for US$125 / $100 kit each. If we run that through an inflation calculator, $250 (2 x $125) for a pair of A-35's in 1973 would be around US$1700 today - so while we can perhaps say that these are "value" loudspeakers, I would not call that particularly cheap.

[Interesting inflation fact, the price of an average loaf of bread was quite stable through the 1960's at around $0.20-0.25. With inflationary pressures, by early 1973 it had gone up to US$0.35. During that year there was a big inflation shock to $0.49/loaf by the summer of '73. Alas, prices have never really looked back... These days in 2023, cheap white bread would be around US$1.90-2.00/loaf.]

Objective Results:

As usual, I listened to the speakers in my sound room before measurements. As a write-up, in my opinion, it's the objective stuff that will outlast any subjective opinions I might offer when conveying performance to readers so I'll present this information first.

Let's start with the electrical characteristic of the speakers - impedance and phase. For convenience, I arbitrarily picked one of the speakers as the "left" speaker for measurement. They both were in good shape and performed similarly:

As a 2-way enclosed speaker, we see the resonance peaks for both the woofer (attenuated "semi-aperiodic") and tweeter. The A-35 looks to be relatively easy to drive with minimum impedance of around 6.4Ω at 150Hz (cursor). Phase doesn't swing to extremes and is maintained within +30 to -20°. I've included the calculated EPDR curve for reference as well which dips down to 4Ω.

Notice some wrinkles in the impedance curve particularly around 300-400Hz and up at 600-700Hz which could reflect areas of resonance which we'll keep an eye on with the acoustic testing.

The graph above was done with the tweeter control knob set to "Normal". Let's have a look at the impedance if we turn the knob to the far left and right settings:

As expected, doing this will change the resistor values connected to the tweeter, resulting in increase resistance when turn to the left to attenuate the treble, and decreased turned to the right (thus increasing treble output). We'll look at this again below with the acoustic measurements.

I see that one of the "specs" sheets for the A-35 lists the sensitivity as "98dB". I don't see how this is realistic if measured as dB/2.8V/m given the design whether anechoic or in-room. Here's a more realistic measure:

I'm seeing 86.5dB/2.8V/m average for this speaker using a quasi-anechoic technique averaging output from 500Hz to 7kHz. We're seeing also the on-axis frequency response in that graph.

Speaking of frequency response then, here's the "qSpin" result with a blending of measurements at 1m and in the nearfield:

Overall, we're looking at a speaker with both bass and treble roll-off. There's a mid-bass bump at 125Hz, and a more gentle roll-off below as one would expect with a sealed box speaker. If we use the level at 1kHz, the -3dB point would be around 75Hz and -6dB at 55Hz in the bass. The high frequencies drop off beyond 12kHz which is OK but not great high fidelity these days and hits -6dB (ref. 1kHz) at 15.5kHz; this will reduce the amount of "brilliance" one might experience in the music depending on one's hearing acuity.

Notice there's a bit of irregularity in the frequency response between 500-700Hz which looks like it correlates to an impedance "wrinkle". In the higher frequencies, we've got some irregularities particularly from 2-4kHz. The "Listening Window" curve which consists of an average of frequency responses from straight on-axis to +/-30° looks rougher than something like the vintage Spendor SA1 from the mid-70s, measured back in 2021.

When we look at the directivity index graph in the bottom right, we again see quite a bit of irregularity between 2-4kHz; not as uncontrolled as the Radio Shack Minimus 7, though ;-). Clarity can be compromised as a result in these sensitive frequencies for human hearing.

We can see the irregularities in the horizontal and vertical directivity polar maps:

I was somewhat surprised by the relatively narrow directivity across the frequencies. Note the irregularities in the 2-4kHz zone horizontally and also vertically where there are some anomalies down to 1.5kHz; be mindful of listening height and try not to get too far off tweeter axis.

For the measurements, I directed the microphone at tweeter level pointing midline to the speaker. When the grille is on, we can't see that the tweeter is offset from the center of the speaker. If we look at the horizontal plot, notice that the intensity of the high frequencies is offset a little below 0° from about 1.2kHz which correlates to the crossover region where the tweeter takes over. Below 1.2kHz, the centered woofer is in control and the dispersion pattern looks more symmetrical around 0°.

In the time domain, let's have a look at the step response:

It's a typical pattern where both the tweeter and woofer are oriented in positive acoustic polarity. Overall this looks reasonably smooth.

The cumulative spectral decay "waterfall" taken on tweeter axis isn't clean but still not bad. There's a bit more energy in that 2-4kHz zone than I've seen on better speakers (like say the vintage Spendor SA1). But at least it's not the 4kHz "ringing" I found with the old Radio Shack Minimus 7.

Let's take a peek now at some distortion measurements:

As we have discussed over the years, harmonic distortion usually isn't too objectionable although I think it's always good to aim for  <0.3% (below -50dB) THD based on empirical testing. In the harmonic distortion graph above performed at 85dB SPL at 1m, we see that from 100Hz and above, the Dynaco generally keeps THD below -40dB or 1%. Certainly not one of the lower distortion speakers I've measured but this should not be too objectionable.

The cursor is placed at 1kHz and you can see the individual distortion component values in the legend. In total, at 1kHz, THD was -53.5dB (0.21%) with 3rd > 2nd order harmonics through much of the audible spectrum. Good to see little higher order harmonics beyond the 4th.

To make things more challenging and for some measurement numbers, we can look at intermodulation distortion (IMD) using dual 300 + 1300Hz tones as I have done with other speakers at a range of output levels:

We see IMD consistently around -50dB across various amplitudes; this is alright. I like that it can maintain -51dB down at 60dB SPL indicating low distortion at low levels. However, most speakers I've tested tend to improve when the volume is increased up to 80 or 90dB SPL. The fact that it doesn't with the Dynaco suggests that these won't be able to play as loud without distortion compared to most other speakers I've tested.

We also get a sense that there are fidelity limitations with this speaker on the multitone burst test for output level linearity (tones range from 300Hz to 2.5kHz demanding both woofer and tweeter participation):

Notice that as we push output above 95dB SPL, the measured level doesn't quite keep track and deviates down from the target. While I know these speakers can play loud, watch out for audible distortions creeping in a bit earlier than other speakers.

To finish off, let's check on a few loose ends... First, here's the effect of the original cloth grille:

Looks like the effect of that grille is minimal - "transparent". Just a slight dip around 5-6kHz and at 

11-14kHz. Not all speaker grilles are this innocuous.

I measured the electrical effect of the rear tweeter tonality knob on the impedance above. Now here's the effect acoustically across the 5 knob settings:

As expected, turn the knob to the right and we'll boost the treble, turn it to the left and we'll make the sound a bit "darker" with lower relative "presence" and "brilliance".

Subjective Impressions:

Notice the size of the Dynaco with their 10" woofer compared to my slender, but much deeper, floor standing Paradigm Signature S8 speakers which use multiple 7" bass drivers. Although the stands are a bit high for such large speakers (unstable!), the tweeter level is at the same height as my Paradigms.

As I noted above, my usual procedure is to listen to the product first then put it on the test bench, then put it back in the system for longer listening sessions. As you can seen in the image above, although I placed the Dynaco closer together than my usual speakers, I compensated by sitting about 3 feet closer to better approximate the "triangle" between speakers and listener. Amplifier duties provided by the Hypex nCore NC252MP, and I used the Sabaj A20d 2022 as DAC, streaming from my Roon server to the fanless MeLE Quieter2Q MiniPC (discussed here) running Linux.

I listened to a variety of music through the Dynaco including the new P!nk album TRUSTFALL (DR6, 2023) which is typical modern compressed pop-music fare. Of course, music production didn't sound like this in the early-to-mid '70s with lots of synthetic instruments and high dynamic compression when these speakers were first made!

Considering the low-res nature of albums like this, the Dynaco sounded really quite good and perhaps even made the sound less harsh thanks to their more "mellow" character and roll-off of the extreme high frequencies. P!nk's vocals were accentuated and the soundstage was wide even though precision of the placement of sound elements and clarity isn't as good in a complex mix - like the background vocals in "Never Gonna Not Dance Again" which tended to get a bit muddy. The piano-driven track "Lost Cause" and duet with Chris Stapleton "Just Say I'm Sorry" accompanied by the electric guitar in the back sounded great with these speakers as examples of quieter vocal-oriented tracks.

Falling back to more "audiophile" music, I had a listen to the recent Saturday Night in San Francisco: Live 12.6.80 (DR12, 2022, 24/96 digital release). Audiophiles love acoustic guitars and Di Meola / McLaughlin / De Lucia's "Saturday Night" live concert has been entertaining us and providing endless remasters to buy over the last few decades. ;-)

Overall, the sound was pleasant but I thought the speakers couldn't quite muster the temporal precision I'm used to hearing. The best I can describe is that the attack wasn't as clean and the space between the burst of guitar plucks (like say 6:00 into "Splendido Sundance", or the opening to "Trilogy Suite") just didn't have the usual impact. I double checked this with the track "Moonlight on Spring River" from The Dali CD, Vol. 3 that I've used over the years and likewise the speed of the pipa didn't give me the usual goosebumps. I think the high frequency roll-off had a negative impact on the subjective sense of transient "speed".

Let's now get a bit heavier... AC/DC's Highway to Hell from 1979 is about the right vintage pairing as I'm sure many college students would have listened to this with the Dynaco back in the day! This provided an opportunity to turn the music up and hear how these perform at higher levels. First, it's clear that this speaker is weak in the lower bass and don't expect to hear much if any sub-bass. Pairing with a sub is recommended although I wish given the size they had more oomph down to 70Hz for cleaner integration. Otherwise, the music is still enjoyable and the dynamics from this old recording came through nicely. I suppose with hard rock, one just doesn't need great fidelity to enjoy distorted guitars, loud drums, and the late Bon Scott's harsh vocals. ;-)

Yeah, I think the aperiodic enclosure does provide very tight well-controlled bass, but deeper bass is still very much missing!

Finally, let's get more refined and enjoy some classical music.

I had a listen to the Hamlet Piano Trio's Beethoven Piano Trios (DR14, 2017, SACD rip) from Channel Classics - they always do a good job with the recording and production (including multichannel). This album sounded really good through the Dynaco. With a natural acoustic recording like this, there's no need for very extended bass nor extreme highs. While the A-35 is clearly technically not a high-resolution speaker (in more complex passages the separation between violin and cello in particular seems harder to tease apart), I appreciated the illusion of "space" with a nice sense of depth as if the trio is playing in my room. I can imagine this getting better with more optimal placement.

In summary:

When introduced in 1969, the Dynaco A-25 was reviewed by J Gordon Holt at Stereophile and was given good grades for its sound compared to other similar-priced speakers of that era (Acoustic Research and KLH) though not for it's "transparency". Holt says: "the A-25s had a dry, almost grainy quality and a somewhat dead sound". Although part of the "A-Series" family, the A-35 is a different beast, but I can concur with the idea that transparency is a bit lacking here even if the A-35 was marketed as "the most accurate" model.

I would not use the words "grainy" or "dead" to describe the sound of the A-35 however. While it does lack deep bass and extended sparkle, turning the tweeter control knob one click to the right of "Normal" helps, but this doesn't really make up for the lack of frequency response >12kHz. I thought the speakers sounded smooth with vocals and did a fine job with dynamics at medium playback levels. The description from Holt of a "somewhat dead sound" certainly would not apply to the A-35; instead, at worst, I would use the adjective "boring", perhaps "too mellow" depending on the musical content one is trying to play. So while the Dynaco speakers wouldn't particular rank high on my list of favourite transducers I've heard (for the record, I much preferred the Spendor SA1), I can certainly attest to the polite pleasantness of the sound quality.

On a side note, I think this poster was right that the A-25 could very well have deeper bass than the A-35 due to its vented port design even if bass control might not be as tight.

I would have been a babe in diapers when these speakers were new! The sound I'm hearing with these speakers is what I've always associated as "vintage" hi-fi (other adjectives I could use might include "romantic" or "sentimental") when thinking about growing up in the '70s and '80s. There is a nice midrange here that will complement vocals, atmospheric classic jazz, maybe even small classical ensembles. The tonality is non-fatiguing at the "Normal" tweeter setting. However, it's still obviously an "editorialized" kind of pleasantness which holds us back a bit from full "hi-fi", and hearing all that's on a recording.

I'm sure many audiophiles will still be quite subjectively happy with this sound though. Just don't tell me it's anywhere faithful to the source when playing some psy-trance electronica like say Ghost Rider's "Make Us Stronger" off Part of the Dream (DR4, 2017), okay? ;-)

On a practical note, given the moderately large size of these speakers yet relatively poor bass extension, I would recommend placing them on low stands for stability and angling them up towards listener ear level. Place them closer to a wall or even corners (boundary loading) to boost the low end. Experiment a bit and watch for excessive boominess. You might even find that it sounds best with asymmetrical toe-in due to the tweeter offset.

Thanks again to linnrd for letting me borrow these speakers for awhile. ;-)

--------------------

In trying to be a more well-rounded audiophile, I think it's important to not just read about new products, but also make sure we look back in history and experience what has come before. It gives perspective as to how far technology has progressed (or not!). Sometimes, we might be surprised at how little difference modern advances have made; for example, I think many old CD players from the '80s/early-90's still sound very good despite dramatic claims otherwise by some. Other times, like with speakers such as these, advances in transducer technology and manufacturing have been more significant.

At the US$250/pair asking price in the early to mid 1970's which is about $1700 these days (2023), we can quite easily find less expensive speakers from the likes of ELAC or Wharfedale or Emotiva or KEF or NHT utilizing modern materials, sport more attractive and capable cabinets, and electrical designs that will outperform the Dynaco A-35 at a fraction of the price.

Shigeo Shingo - well known for factory management &
the Toyota Production System.

As I've suggested over the years, technological improvement should be a highly "deflationary" force. Better fidelity, smaller size, more efficient power utilization, faster speed, at lower prices should all be qualities we see in a competitive marketplace. To illustrate, in 1972 when the Dynaco A-35 was introduced, Atari's video game Pong also was released! We can clearly see what technological advancement looks like with improved graphics, story, play mechanics in each generation since the primitive Pong into today's modern 3D gaming. In comparison, have speakers advanced in leaps and bounds like video games and the computer hardware that run them? Of course not! This is an expected sign of technological maturity in audio hardware. While there have been many incremental steps gained between the Dynaco and today's bookshelf or moderate-sized box speakers, it's at a much slower, way less dramatic pace.

The fact that so many of the products in audio have done nothing but inflated in price without concomitantly greater benefits to consumers ("High End Audio" - here's looking at you) is also to be expected for products catering to a "non-utilitarian" luxury market. It's quite possible that the performance of many of these luxury products may no longer strive for high-fidelity. Some products might not even conform to reality testing (here's looking at you "High End Cables")! "Diminishing returns" is a real thing, right? Keep an eye on sale prices in the used market - this is typically more indicative of actual value/sentiment than MSRP.

--------------------

To end, I wanted to double check just how closely my "qSpin" measurements were able to predict the sound of these speakers in the room. It's one thing to be measuring speakers with quasi-anechoic technique, but let's also make sure that this correlates to the "live" experience when listening!

Let's bring the measurement mic into the sound room with the A-35 set up like this; not optimal but good enough for a quick test (speakers around 4.5 feet apart, mic about 6.5 feet away):

You can see the Surface 3 computer on the floor used to capture the room frequency sweep.

Using VituixCAD2, I imported the qSpin data and calculated the "Predicted In-Room Response" (PIR) curve (weighted average of Listening Window, Early Reflections, and Sound Power curves), blended that with the low-frequency nearfield measurement. Note that the PIR is typically quite similar to the "Early Reflections" curve, and should be reflective of the sound in a good small listening space. Then I measured the speakers as per the picture above in the room - all EQ and room correction turned off, of course.

In REW, I aligned the predicted vs. measured curves based on the levels around 2-4kHz to see how well they correlated:

Nice correlation between the calculated qSpin PIR with actual measurements. In fact, I'm impressed by just how closely the two speakers measured which speaks to good manufacturing tolerance at least for this old Dynaco pair after all these decades!

As you can imagine, the frequencies <1kHz are increasingly determined by the room and the furniture. I suspect if I removed my Paradigms and put the Dynaco closer to the back wall, the low frequencies would fill out. Also, the qSpin was done with grille off, which would explain the accentuation around 10+kHz compared to the measured response with the grilles on.

As usual, it's important to be mindful that the sound room is of major importance regardless of how well your hardware measures, costs or actually performs; given a set audio system budget, it's important to be smart about how you spend the money. Hope you're all enjoying the music, audio friends.

Make sure to add your voice to the currently-running Audio Playback System and/or Streaming 2023 Survey if you haven't done so already. Over 1000 detailed responses have been submitted worldwide so far which is great.