Tubes in Audio Research Equipment: 6550

Wolzst- I think you'll find that most current tube amps are far more neutral than tube amps of the past. Better passive components have allowed designers to obtain far more detail and control from recent tube equipment. Of course, most tube products still benefit from modifications and tweaks. From improved parts quality, NOS tubes, increasing/improving the power supply, etc... one can radically alter the sound of any tube product. As for the sound of the tube, it is the circuit combined with the tube that produces the sound. Although certain characteristics can be attributed to a particular tube brand and type, it's always best to let your ears be the judge in the context of your system to determine if you like one tube over another. In general, I prefer the sound of NOS tubes in my equipment over current Chinese, or Russian tubes. Just my .02

Very clear and neutral with a little warm glow around the instruments. Seems like these are clear in all ranges whereas a lot of tube amps I've heard are ultra-clear in a small midrange window. Check out the svetlana home page and thetubestore.com for some tube sound interpretations.

I find that the 6550 tubes, while having good dynamics and deep bass, are very thin and cold sounding. Improvements are the cousins to these tubes - KT88, KT90, KT91, etc. While a bit of power is sacrificed, they tend to be more musical. The only reason the 6550 tubes are used so much, is due to the wide avilability of them and the number of suppliers. There are better tubes that produce much better "musicalility" such as the 6C33s. The problem is there is only one source that I am aware of. These tubes are used in the better products, such as the Lamms and the BAT equipment.

I have experimented with the current ARC 6550-C's and the Upscale Audio supplied Type 3 KT-90's in my VT100MKII,and before that,a VT100. As I understand it, the current 6550-C's are an upgraded tube from the ones provided in the earlier VT100's. In my system, the ARC 6550-C's sound quite nice: transparent, good bottom, good top-end extension. In the end, though, I prefered the Type 3 KT90's. These KT90's have more mid-range bloom and present images with a much more 3-D presentation. If you do try the Type 3 KT90's in a VT100, the auto-biasing circuitry in the amp requires matched quartets. Also, I was advised to bias the KT90's to the low 40's verses the 65ma with the factory 6550-C's. The tubes run much cooler and do, in fact, sound better at this lower bias setting. Upgrading the input/driver tubes was also a very effective enhancement.

I have had an ARC VT130 for almost 4 years now. It was wonderful with KT88 (slovakia tubes). The midrange bloom was awesome. But the reliability with these tubes was terrible. One would die every month or so and with the goofy ARC design, a 100 ohm power resistor would blow too. So I was busy with my soldering iron every time a tube died. I had enough of this and went with 6550 tubes. Some of the magical bloom was gone but I got more extension on the low end. There did not seem to be much difference with power rating or high freq. extension. I think ARC goes with the 6550 tubes mainly for reliability reasons. I have had the 6550s for 2 years now and not a single problem. I understand now there are KT88 tubes from Russia so maybe this would be the way to go.

DEM, You are mistaken about an "auto-biasing circuitry " in the ARC VT100 and VT100MKII. The tubes need to be matched quads, because if they are too different, it will take out one or more resistors, reduce the sound quality, or reduce the tube life by burdening the other tube in the push-pull circuit. In regards to the bias figures quoted, it should be 130ma for stock 6550Cs and around 80 to 85ma per channel for the KT90s across the bias test points (resistor). In regards to replacing the driver tubes (6922 and equivalant) -- make sure the tubes are a matched octet and you obtain and perform an alignment procedure. If this is not done, you will have reduced fidelity, increased heat, possibly blown components, and reduced tube life. The procedure takes about 2 hours and requires at least one 3 1/2 digit DVM (2 is preferable), a schematic diagram, tube electronic circuit knowledge, as well as the removal of the top and side covers. In addition, the installation of KT88, 90 & 91 in VT100s could blow the screen grid resistor(s) upon power up. This can also happen with a 6550C replacement, but less likely. I know this is off track on this thread, but important information to know when considering tube swapping in the VT-100. I do agree with post regardin the KT-88s can be unreliable - I have also had bad luck with those... The KT-90s are better reliability, but they are not my first preference. The Tungsol 6550s seem to perform the best IMHO if you can find / afford them. BTW, I have NEVER had a problem with the 6550Cs that came directly from Audio Research.

Does anyone have any experience with the Sovtek 6550's?

J_k. Thanks for your helpful feedback regarding the requirement for matched quads in the VT100 and VT100MKII, as it related to my post. In my post, I was relating cautionary information I was told so that anyone (not experienced with ARC gear) reading my post would not confuse this recent ARC design with earlier designs which had individual tube bias points. Thus, they may not know of the need for matched quartets. Also,I may be wrong on this, but I understood that the factory output bias setting for the VT100 6550C's is 65mA. However, for adjusting across the test points for each channel of the VT100, ARC recommends a reading of 130mVDC (0.13 Volt DC) for the 6550C's. Best Regards!

DEM -- You are also correct -- ARC does bias EACH 6550C at 65ma. The VT100MKI & II uses 6550C x2 in each side of the push - pull circuit, making the bias setting across the TP 130MV, equating to 130ma of current running through the resistor. I heard VT100MKIII has individual adjustments..