[ TNT | Who we are | HiFi Shows | HiFi Topics | Listening tests | Tweakings | Inter.Views ]

Inter.View with Paul Marks, owner and designer of Thor Audio electronics

by Lucio Cadeddu

[Italian version]

LC >
Thor Audio is a new name in the game of tubes HiFi equipment. When did you start this project?

PM >
Thor officially opened the doors in January of 1996, but we were working full time on this for three years before then just designing the line. Two ideas were to govern everything we did during that time.
One - If we were make "just another" tube line, then what's the point! And Two - You only get ONE shot in this business. With these two thoughts in mind, we had to do our homework.
The line had to be the very best we could possibly make. I must say that surviving with no real income for three years in the nineties and exhausting the savings, while at the same time saying no to two investors along the way, was indeed a "leap of faith".

LC >
Thor Audio makes some pretty original and revolutionary components using good 'ol days technology (tubes, that is).
Why did you choose tubes and which are, according to you, the main pluses of this technology when compared with the traditional solid state one?

PM >
Our first component was actually a solid state line/phono preamp. It was very good, and could have competed well in the market, but again, "well" was not what we were after.
I must point out that at my age, I was raised on tubes with McIntosh, Fisher, and Eico/Heath amp/preamp kits, etc. With the beginning of the tube resurgence in high end audio, I was only too happy to begin designs using them.
I could combine my most favorite sound together with the active and passive components available today, thus eliminating some of the drawbacks of the "over-warm/no-bass" tube sound, while at the same time being one of the first companies in the tube "wave" hitting then.
There was now no reason to believe that we couldn't make a tube based product that was as fast and detailed as solid state, that would also inherently negate all of the "in-your-face" type of detail of some solid state components of that time, while remaining truly musical in it's delivery.
I didn't know then that it would take as long as it did to develop the line, and we were certainly not one of that first wave, but I knew we had made the right decision after I started listening to what we had with even our first circuits.

LC >
Form follows function. This seems to be your credo. Can you explain why a toroidal cabinet works better than a traditional one?

PM >
Eliminating one of the inherent problems of replacing tubes frequently was the first thing we took on after some of the basic circuits were realized. A three minute timed "ramp" of the B+ and filaments would give the tubes a nice soft start after turn-on, but what about the other inherent problem AFTER turn on - heat!
Vertical mounting with as much free air around the tubes as possible was an obvious answer, but what about totally free air! That would mean NO chassis on top to trap the heat.
It also would mean that the tubes would have to be shielded on all sides to prevent any AC/RF fields from getting to them - maybe in a circle around the tubes. This done, we then had tube sockets coming through the bottom of a flat 6 1/2" circular piece of aluminum (with enough % of copper in it to do good RF rejection).
The tubes were then placed sitting upright within a circular tube of the same material shielding it on the bottom and now on all sides. Once these two pieces were welded together for a reliable contiguous shield, it became the tube "chimney" that we have used since.
What chassis would accept this unusual design? A rectangular chassis with a 6 1/2" circle cut out on the top? Interestingly enough as I did more research on circles, I found that from a purely physics standpoint, anything round would be inherently non resonant as all vibrations would inevitably go to the center and begin canceling themselves out.
This eliminated the last large inherent problem with tubes - vibration. It also made the rest of our chassis obvious. It also had to be round.
The first chassis made was essentially what we ended up with in production but we added one more element.
We placed over 10 pounds of damping material affixed to the inside of the top of the chassis, forcing everything down. Not only did it do a better job of vibration rejection throughout the entire chassis, but it solidified the acoustic image and helped bring the bass to the deep, tight and controlled sound we are known for.
Curious to see what this type of chassis was doing for the sound, I removed the entire guts from the toroid chassis, and put them into a rectangle chassis, complete with the same amount of damping, feet, etc., and the difference was astonishing. 10-15% of our sound is based upon our toroid chassis!

LC >
Many modern tubes amps and preamps are designed around old schemes (Williamson and others) and many HiFi designers swear these old schemes are better than anything else...new.
Do you believe is there some room for new schemes and unconventional designs with tubes? Can you tell us something about the schemes you use?

PM >
While the answer to our layered soundstage, imaging and focus, lies in a new and somewhat unorthodox circuit proprietary to Thor Audio and used throughout our line, I will say that the basic circuits used in the "golden age" of tubes, were my first choice.
Use of noise cancellation via differential circuits, low output impedance via cathode followers, etc. have been used for years and I saw no reason to "re-invent the wheel".
However, in my opinion, the overall sound, using these circuits, is something more closely linked as an answer to your next question.

LC >
There are many things that can affect (or improve) the performance of a tube equipped component: passive components, cabling, quality of the tubes, layout, mechanical construction etc.
Which are, in your opinion, the most relevant ones?

PM >
I believe that absolutely EVERYTHING is audible. Although one of countless things we went through at the beginning, one of more interesting stories happened after we completed the first line stage.
To save the considerable time and expense having a PC board house make, and then change the prototype boards every week or so during the development stage, we did them ourselves.
They were of the "home brew" kind made with photographic film positives exposed to a sun lamp, then etched and drilled. We then used an electro-less plating process to keep the copper from oxidizing. After a dozen or so changes to the board, we finally had the sound we were looking for!
When the last change was made, we had the boards done at the outside vender - complete with solder re-flowed traces, and a great looking solder mask. Before I go on, I must point out that our entire line is essentially hard wired. Any audio carrying signal paths are connected using separate individually shielded wires, and all capacitors are soldered tube pin to tube pin, leaving less than one inch total pc board trace that actually carries signal.
Therefore, we were not expecting any significant sonic differences using the new "pro" boards. Unfortunately with the new board, it was immediately obvious that ALL of the magic had left!
I must also point out that the components used on the new board were not new components, from another batch of caps, etc., they were the SAME ones taken from our home brew board. This was a true A/B test.
Fascinated by this, we then had the vendor make more boards in two different ways. The first had the solder re-flow but no solder mask. Surprisingly, the air returned immediately, but the sense of delicacy was still not there at all.
In fact it sounded smeared, veiled and lifeless, with a muddy upper mid-range and somewhat strident high end. It wasn't until we had them make a totally bare board and put the same electro less plating on it that we had on our home brew one that the magic returned!
Whether or not this would show itself in another less revealing circuit I have not taken the time to research, but from the start we have had our PC boards shipped to us with bare copper and put them through our electro-less plating process.
We also have many boxes full of very expensive and extremely good capacitors, wire, switches, solder, connectors, etc. that did not make the cut. The final selection of our custom made input switch alone was over eight times the cost of most used today in very pricey and well established lines, and took over two years to find. (Once again, it was nice to have the time in the beginning!)
I believe therefore, that it is a cumulative effect of active/passive component choices going through the design process that makes or breaks a circuit, not just the circuit schematic itself. You could change only one component, (a coupling or pi filter cap maybe) and have a totally different sounding preamp or amp.

LC >
Any advice for tubes lovers willing to get the best out of their equipment (tweaks, tips to extend tubes' life etc.)?

PM >
One of the few sonic liabilities of using tubes is the vibration problem. Our tube sockets are mechanically coupled to the chassis, and do not move.
If you notice that when you replace tubes in your component that the pc board moves when you push down or pull up, the use something like "Top Hats" would reduce the inevitable vibration that an approach like that gives. I have seen them work very well As far as my tweak/opinions - Two things I have spent enough time with in enough rooms/systems to comment on are:
1) Use of Black Diamond shelves and racks. They have been the most significant and beneficial to our sound that we have come across so far. So much so that while we use "Audio Points" tiptoes instead of rubber feet on our chassis, we have a reverse triangle of 1/4-20 threads drilled on the bottom to accommodate the Black Diamonds (and any others a customer wishes to experiment with).
The Audio Point feet are slightly shorter than most other tiptoes/feet, so they do not have to be removed to experiment with other cones - another reason to use them stock.
2) The second is AC cords. They have a great deal to do with the overall sound of a system. Experimenting with a few of them is the only way to find out what works in your particular system. We ship a "Discovery" power cord with each unit. This is not to say that our products will sound the best using only this one, and we would not want to be accused of "pre tweaking" anyone, but with all the R&D we did, we felt that sending an inexpensive bargain brand with our components would compromise those efforts.
On the technical side for those building from scratch, or able to do some serious changes to an existing component - If you run a heater on a 12.6V tube at 12.3V, you can expect to double the life of the tube. That is what we did throughout the line.

LC >
Could you tell us something about new products from Thor Audio?

PM >
There are quite a few items that are new and are best served by going to our web site, they are the TPA-30 (30W mono blocks), the TPA-150 (150W mono blocks) and the "100" series of front end electronics (line, line/phono, phono only, and DAC).
Our TPA-30 made their debut at the CES (Consumer Electronics Show) show in Las Vegas this last January and are in production now. They are built around the EL34s, and are designed into our "preamp" size toroid chassis with all transformers, etc. mounted inside the chassis. Features include fuse replacement, bias adjust controls, and a built-in bias meter. All controls are accessed from the top of the unit for safe, user friendly operation.
The TPA-150 mono blocks are designed into a large 22 inch round chassis weighing about 100 lbs. each. Scheduled release time is mid to late summer of 99'. Unique to the TPA-150 is the use of a cumulative hour counter. This meter measures the total time the amp has been on for accurately tracking the normal 2000 hours of expected output tube life. Also unique to these amps are something we are very excited about.
We are calling it "Compu-Bias". Once the customer finds out the overall sound that they like (under biased/over biased, etc.), they program the amp for that particular setting. Once the bias switch is thrown, the input gets shunted to ground, and all bias controls automatically clean themselves be turning all the way left, then all the way right. Each control then automatically biases it's own tube in turn (to the user preset), and repeats the process, going around and around until all tubes are perfectly biased!
This biases the amp in a fraction of the time it would take by hand, and is far more accurate! You can bias the amp easily anytime "on-the-fly" without turning the amp off, unplugging it, putting shorting plugs in, etc.
The new "100" series are the first price pointed components in our line. We still use exactly the same circuits and component manufacturers, but use their second best components as our choice for active and passive components. The thought was to make our line available in the more conventional rectangular chassis, and at the same time, making it more affordable. They are 75% of the sound of the Signature "1000" series for about 1/2 the retail price. Release time is staggered over the next few months, again seen in full on our web site.

Thank you for taking the time to read this, and a big thanks to Lucio for asking us to be a part of his wonderful web site!

Courtesy Paul Marks for TNT-Audio.

Copyright © 1999 Lucio Cadeddu

[ TNT | Who we are | HiFi Shows | HiFi Topics | Listening tests | Tweakings | Inter.Views ]