Well, when he received the unit from me he said it was damaged and didn't make a sound. Turns outr one of the EL-34s had worked loose in it's socket and blew a fuse. Of course, being a reviewer he didn't check the fuse and sent it to Jim White for "repair". At my expense.
I spend more time at my workbench now and less paying attention to mainstream high-end, so the unit's reputation for unreliability was news to me. My one problem was my fault, I believe, because I failed to turn the unit off and the failure was either thermally related or due to line spikes (or both).
As for sound, I said it was quite good, perhaps the best commercially available - I haven't has the opportunity to hear the Wavac, assuming it remains available. But there are DIY circuits out there that improve on the Io.
One of the unavaoidable weaknessess of commercial tube units is that they must rely on current production tubes. The 12AX7 and its variants aren't very linear but are chosen solely for the ease of use (size) and amplification factor (100). The tube was developed in the early '50s in response to what was perceived as a market trend toward minituization in home electronics. Two triodes in a miniature package with a high mu is a bean counter's dream. The 6922 is marginally a better tube, as is the 6SN7. The use of the popular cathode follower circuit and variants for the low noise and gain offered ignores it's tendency to wander all over the map in terms of operating points means that additional steps in the forms of voltage regulation, active plate loading and a current sink must be used to clamp the tube's OPs with a narrow, desireable range.
None of these tubes, however, exhibit the low level linearity of the directly heated single triodes from radio's early days. Unfortunately, these tubes are no longer made and are therefore not feasible for a commercial venture.
In addition, these are low-mu triodes with amplification factors ranging from 3 to 9. So the problem is where to get the extra gain. The solution is in the use of high-quality interstage transformers and plate chokes. This produces a stage-to-stage stability that can't be equalled by cap coupling and it gets the cap out of the signal path. It also allows employment of the simpler anode follower topology, and even topologies like parallel feed which removes the chore of handling DC current from the interstage primaries. This nallows more permeable core materials to be employed and the result is better clarity, better low level detail, and an overall ease of presentation. Of course, your replaced a $1 part with something that will cost $60 on up to several hundred dollars, depending on circuit requirements and core materials.
This makes it difficult to turn the necessary profit since the market is small and the high-end rule-of-thumb sets MSRP at 5x to 8x the cost of parts. But it undeniably a superior approach if cost is no object. See the schematices at Sakuma-san's Direct HEating site (http://www5.big.or.jp/~dh/) for both amps and preamps that employ these techniques.
All of these circuits are invariably less complex - a good thing - and make liberal use of high-quality transformers and chokes that replace highly variable (in a QA/QC sense) resistors, caps, and silicon devices. As Einstein said, "Make is simple as possible but no simpler." The Io's complexity, coupled with increased demand for a product that is essentially the designer's second job, may have something to do with the reliability issue.
