AI-1 Dreadnought - Preliminary Results

[DarqueKnight]

I'm working through my project list and got around to making an AI-1 non-common ground amplifier interface. For those who don't know what I'm talking about, the AI-1 was developed by Polk to allow SDA speakers, which require common ground amplifiers, to be used with non-common ground amplifiers.

Depending on the associated equipment, the original AI-1 could cause a slight loss of detail. Some users heard a loss of detail with the AI-1 while other users heard no difference at all.

A couple of years ago, Matt Polk provided specifications for an AI-1 based on a toroidal isolation transformer. I sent those specifications to transformer manufacturer Avel-Lindberg and they recommended two of their models, neither of which were exact matches for the specifications. I chose the Y236906 since it was the more robust of the two.


Figure 1. AI-1 Dreadnought prototype trials. Such Good Sound all the way around.


Figure 2. Dreadnought interconnect cables.

The dual primary and secondary windings were connected in parallel to lower the DC resistance and inductance. Four 10 foot interconnect cables were made from Monster Cable Z2 Reference speaker cable. The two 12 AWG conductors in each cable were soldered together on each end to make a 9 AWG cable, which is the same wire gauge as my speaker cables.

Prior to the Dreadnought, I was running my JC 1 monoblocks in a strapped configuration with the negative terminals of each amp tied together with a 9 AWG cable. The SDA interconnect was also a custom 9 AWG cable.

The best I was hoping for was that the Dreadnought would not take anything away. The sound was actually better with the Dreadnought. The sound stage grew a foot on the front, sides, and rear and there was increased detail.

Immediately after installation, I heard some upper midrange graininess and the bass was a bit slower and less defined. However, those aberrations disappeared after 10 hours of play.

I ran the volume up to 110 dB-C and heard no stress, strain, grit, or grain in the sound. Of course, at that level, I was listening from the next room. I hung on until 105 dB-C then my ears gave out. As I increased the volume from a comfortable listening level of 85 dB-C, the image weight and tactile sensations increased, but there were no deteriorations in any aspect of the sound.

Noise measurements taken with an oscilloscope showed the noise in the dimensional signal to be significantly less with the Dreadnought than with the strapped configuration.

I initially planned to construct the Dreadnought and keep it around for future higher powered amplifier evaluations, since all the amps I am interested in with greater than 400 watts per channel are bridged designs that cannot have the negative terminals of each channel strapped together. Since the sound quality was much better with the Dreadnought, the strapped configuration had to be retired. By the way, I am interested in more power for more realism (detail, imaging, sound staging)...not for more loudness.;)


Figure 3. Former SDA SRS 1.2TL interconnect and JC 1 amplifier strapping cable
prior to going into storage. Sleep well old friends.

The prototype has been bolted to a smaller wooden board and tucked behind one of the 1.2TL's pending selection of an appropriate enclosure.

Full report to come later.:)


The monoliths are delirious with delight about what the future holds.

[NJPOLKER]

DK
I am not sure what to think about all the information but I gotta say that is one beautiful looking system and a beautiful home too.

[schwarcw]

Interesting read and nice writeup (as always). I toy with the idea of higher power amps, and then realize that at the level I listen to them, I'm not sure I'd gain very much. My 2.3's are bi-amped with a pair of Parasound Halo A21's. At a 4 ohm load, I think each amp is rated at about 440 wpc. I'd have to double check on the Parasound specs. I haven't heard the JC1's. They cost about 80% more, and I think I got a lot of power and the good things that go with high power. Detail, image, nice smooth sound with plenty of weight. When I am listening late at night, I don't think the volume level ever goes above -35 to -28 dB depending on the source. I'm intrigued by the JC1's. I think Polkie "Read-a-lot" also has a pair.

I have a brand new A1-1 cable that I never have tried. It still coiled in the original box. I keep it for the time that I might try a big pair of monoblocks;)

Thanks for the writeup!

[BIZILL]

you can always use that negative strapping cable as an extra ground on your vehicle. it never hurts to have good grounding. or hell, use it to replace your main ground wire from battery to chassis.

[MillerLiteScott]

DK Can you disclose the color combination you are using for the pin/blade connections for both left and right speakers. I have been using the exact same Avel Isolation transformer for about a year now and I want to make sure I have mine hooked up correctly.

Scott

[DarqueKnight]

The Avel-Lindberg website has a chart for parallel connection of Y23 series transformers showing the primary positive terminal attached to the gray & brown wires and the ground terminal attached to the blue & violet wires. The dots indicate the corresponding + or - terminals on the primary and secondary sides to maintain correct polarity and phase.

The red and yellow wires on the secondary side should be attached to the positive terminal and the black and orange wires to the negative terminal.

My connections at the speakers are:

Right Pin (+): Red and Yellow wires.
Right Blade (-): Black and Orange wires.

Left Pin (+): Grey and Brown wires.
Left Blade (-): Blue and Violet wires

[MillerLiteScott]

Thanks DK,

I have mine connected the same way. My initial concern was the different/smaller gauge wiring for the secondary winding. The following picture shows my set up with the Pin connection on the top for the left and right speaker. The blade connection is on the bottom respectively.

I guess I can get rid of those nasty crimp connectors and solder the wires to the binding post.

Scott

[DarqueKnight]

When I received my transformer, I wondered why the primary leads were thicker than the secondary leads and I went to the spec sheet to see if this was mentioned. The spec sheet said:

"Our unique double-insulated primary leads eliminate an assembly step required by many safety standards."

I have no idea what they mean by this.

After I read that, I noticed that the metal conductors in all the leads appeared to be the same size (I did not measure them). The leads for the primary just had much thicker insulation.

[riglehart]

A while back I built an AI-1 using an A/L torroid. I had used a tiny Stancor intitially, that had been recommended by folks and didn't like the result. The torroid I used was smaller than yours, but sounded much better than the Stancor.

I sent it to F1nut and he did an AI-1 "shootout". You might want to look at it....
http://www.polkaudio.com/forums/show...light=shootout

I'm back to a straight cable with a commond ground amp.

[zingo]

Where is a good place to get the Y236906? I have been looking on the internet, but all I seem to come up with is spec sheets.

[DarqueKnight]

A while back I built an AI-1 using an A/L torroid.

I sent it to F1nut and he did an AI-1 "shootout". You might want to look at it....I'm back to a straight cable with a commond ground amp.

I remembered your and Jesse's AI-1 threads and will cite them as references when I post the full report on the Dreadnought.

I'm sorry your toroid based AI-1 didn't work out. For me, the installation of the Dreadnought was the beginning of a second SDA honeymoon.

From your AI-1 thread:

Also, I used 16ga wire (not phone wire). So that shouldn't be an issue.

I found that cable size for the SDA interconnect was an issue: Improved SDA Interconnect Cables

Where is a good place to get the Y236906? I have been looking on the internet, but all I seem to come up with is spec sheets.

I ordered mine by phone by calling the number listed on Avel Lindberg's website. They shipped the same day I placed the order and I received it six calendar days later.

Remember to go big on the interconnect cables.:)

[zingo]

Thanks Raife.

[michaeljhsda2]

Where is a good place to get the Y236906? I have been looking on the internet, but all I seem to come up with is spec sheets.

Go to http://www.avellindberg.com I bought mine a few months ago for 96 bucks. Considering how much the SRS 1.2's retailed for, I wonder why Polk chose not to include the AI-1 cable for these speakers? That way you could use common ground or non-common ground amps with the AI-1 in place.

[DarqueKnight]

They may not have wanted to bump up the retail price another $100 for something most of the buyers of these speakers wouldn't need.

These were already fairly expensive speakers for their day:

$3400 in 1989 $ = $5963 in 2008 $.
$3500 in 1989 $ = $6138 in 2008 $.

[michaeljhsda2]

Darn right they were expensive. I wanted to buy them back then but the wife said "no way"! I was only allowed to buy the SDA 2's. Got rid of the wife (divorce, not death) and now have the 1.2 TL's. Life is good.

[DarqueKnight]

My initial concern was the different/smaller gauge wiring for the secondary winding.

When I received my transformer, I wondered why the primary leads were thicker than the secondary leads and I went to the spec sheet to see if this was mentioned. The spec sheet said:

"Our unique double-insulated primary leads eliminate an assembly step required by many safety standards."

I have no idea what they mean by this.

After I read that, I noticed that the metal conductors in all the leads appeared to be the same size (I did not measure them). The leads for the primary just had much thicker insulation.

Avel-Lindberg confirmed that the leads for the Y236906 transformer are the same gauge:

Avel Part No: Y236906
Pri terminated using 18 AWG UL 1672 (Double-Insulated)
Sec terminated using 18 AWG UL 3266

In response to my inquiry about the double insulated primary leads:

"Double insulated primary leads are required by some agency standards in
certain applications. If required, it eliminates the step of oversleeving in
assembly."

[MillerLiteScott]

Thanks,

I never even bothered to look at the actual wire. I just assumed it was a smaller gauge.

Scott

[DarqueKnight]

The Dreadnought has exceeded my expectations in every aspect: overall detail, image weight, sound stage dimensions, etc.


Dreadnought prototype prior to receiving its black anodized aluminium enclosure.

I already knew that the size and quality of SDA interconnect conductors had an effect on the overall sound of the speaker. Naturally, I wanted to evaluate the Dreadnought with two other interconnect cables, one far below it in quality and the other far above it.


The Dreadnought takes a spin with Monster XP and Audioquest Volcano cables.

The Audioquest Volcano speaker cables ($206.25/foot) measured better than the Monster Z2 cables ($3.78/foot), but I didn't hear a difference. The Monster XP speaker cables ($1.00/foot) measured worse than the Z2 and sounded much worse. The XP cables were dark sounding, shrank the sound stage and made images clump in the areas in front of the speakers and in the center.



The aluminium shell provides more than just a pretty face.

I expected the aluminium enclosure to provide a little noise reduction benefit. I was surprised to find that it significantly improved the sound. The background noise measured considerably less once the transformer was shielded. I re-evaluated the XP and Volcano cables with the aluminium clad Dreadnought. There was still no audible difference between the Z2 and Volcano cables. The Volcanoes still measured better than the Z2, but the measured performance gap had narrowed considerably. The XP's sounded better, but were still just another shade of bad, even though their measured noise had decreased significantly.


The whole thing. The enclosure measures 12"x12"x4".


Hold on tight. The Cardas CCGR-S binding posts make a secure connection.

Those interested in making an AI-1 should be mindful that the transformer is not the only component that affects dimensional signal quality. The gauge and quality of the conductors and the type of enclosure can significantly affect the dimensional signal integrity.

After I receive my loaner original AI-1's and complete their test and evaluation, I will publish a thorough report of my subjective and quantitative performance analysis.

Scientific inquiry can be fun.:)

[hearingimpared]

I missed the initial write up Dr. Smith but I'm glad you bumped it up today. As usual you've done a spectacular job and write up and of course, I want one to replace my AI-1.:D

[MillerLiteScott]

Nice work and write up as always Raife.

Scott

[george daniel]

Nice work sir,, as always,, you set the bar up a notch :)

[shadowofnight]

Very nice work.....on my sooner list for me to build.

Be really interesting to place the factory AI-1 transformer inside a sheilded enclosure instead of its plastic enclosure as well , and see what kind of improvements are possible in that regards.

[george daniel]

Doest it matter which negative binding post you connect the "blade" to? Specifically 2.3 tl's with jumpers in place,, upper or lower ,, or either?

[DarqueKnight]

It does not matter with regard to electrical connectivity as long as the jumpers are in place. I would prefer to connect the AI-1's negative leads to the lower negative binding posts because the SDA signal would not have to go through the jumpers. If you have replaced the stock jumpers with thicker, higher quality wire, you may not hear a difference between connecting the AI-1 to either the top or bottom negative binding post.

If a bi-amp configuration is used in the future with one amp on the high frequency section and another amp on the low frequency section, then the low frequency negative binding post must be used. This is because the SDA circuitry does not affect the high frequency section.

[george daniel]

Thank you sir,, I'll get my ears on them in about 3 hours.Thanks again.

[george daniel]

First of all, a big thank you to DK,, my brother bought a pair of 2.3 tl's from HiFi Buys in the early 90. Along the way they sounded pretty good, but nothing near their potential.Last year we did the cap upgrade,, earlier this year he bought a pair of Odyssey mono extremes,,,he felt his TFM-45 was a better match for them,,, still somewhat unhappy with the sound,though,,, threw a Dodd ELP in the mix,, I got him some NOS Bugle Boys for his birthday,,still not pleased with the sound,,,got the Avel transformer on friday,, put it together,, finally hooked it up on Monday,, my phone rings,,, he is like a man possessed,,, mono's ,dodd,,and the SDA's---Magic,,, got his SDA groove going now,, he's already ordered new binding posts,, sonicaps to replace the solens,, rdo replacements. I went over this evening and listened to Third stone from the sun,,and a couple of blues cut's,, the soundstage is HUGE,,I have not seen him this happy/excited in years. Just wanted to thanks those on this forum for contributing and setting the Bar up a notch. The AI-1 was built exactly like DK outlined in his thread BTW. I must sa,, they do sound very,very good. Enough of this ramble,, I just wanted to share my brothers excitement,, after 17 years,, he's all smiles. Thanks----george

[DarqueKnight]

My monoliths are still bitter because they had to
wait almost 20 years for adequate wattage and a
proper SDA signal.

[ben62670]

Is this what you're looking for?

Originally Posted by Kens Email
Hello Rob,
We are searching for the original transformer specification and will forward them if we find them. However here are Matthew's comments:
"Ken -

I don't know if we have a spec in the files. However, a high current
1:1 toroidal transformer will work very well. The key specs are:

1. DC Resistance of primary and secondary should be less than .8 ohms and preferably less than .5 ohms. The lower the better.
2. Inductance of both primary and secondary should be at least 10mh.
Ideal is around 14mHy.
3. Current capacity to saturation should be 5 Amps minimum.
4. Of course the turns ratio is 1:1

Multiple transformers can be used in parallel to lower the DC resistance and increase the current capacity. Just remember that the inductance is half for two transformers in parallel and one-fourth for 4 transformers in parallel.

-msp"
Regards, Ken, Polk Audio


I'm not real certain if this is enough info or not. This stuff is waaay over my head....

From another thread.
I have in my stock pile an Adcom GFA-555 transformer which I am pretty is either 700 or 800va. It's primaries and secondaries are both .8 ohms each. .4 ohms in parallel. I am wondering if It would be suitable for an A1? It is setup for dual voltage like the avel counter parts.
Thanks
Ben

[GV#27]

From another thread.
I have in my stock pile an Adcom GFA-555 transformer which I am pretty is either 700 or 800va. It's primaries and secondaries are both .8 ohms each. .4 ohms in parallel. I am wondering if It would be suitable for an A1? It is setup for dual voltage like the avel counter parts.
Thanks
Ben

The Adcom might not meet the 1:1 ratio spec.Its secondaries are what 50-60-ish volts AC?

[ben62670]

The secondaries are 60 when center tapped, but if you run the transformer with the secondaries tied together you get a one to one transformer. I did double check this with live current. The secondaries were putting out 120vac omitting the center tap. The Avel piece uses 4 wires on the secondaries, and the Adcom uses 5(really 3 two are just y'ed). I have the schematic for the Adcom tranny:) The bad thing is there are 2 primary windings on the Adcom, and one winding on the secondary:(