VTL Tetrode/Triode

Mine don't even have the MultiCaps (MIT caps), they just have the yellow "VTL" caps. I'm trying to decide whether to go with the InfiniCaps, as VTL uses, or the DynamiCaps, both of which are made by TRT. I described where I was sonically and where I wanted to go to both Peter Moncrieff (the manufacturer) and Michael Percy (the retailer) and they unanimously recommended the slightly more expensive DynamiCaps, but that was before I switched to the KT-88EH's and made some changes regarding the input tubes (including adding tube dampers) which went a ways toward addressing the brightness I was hearing in tetrode. VTL officially prefers the Infinis, while Moncrieff's comments seemed to imply that they were flatter in their response than the DynamiCaps, which from his description sound like they're probably a bit more 'tailored' in their response, but in a way that he apparently thinks better serves the music in many circuits and systems...scary choice ;^)

VTL doesn't specify if or how NFB is effectively changed between the two modes, though from what I've gathered about the method in which the switch is implemented vis-a-vis the output transformer and the tubes it does seem plausible that this parameter will not remain exactly constant. But I'm not technically qualified to know or judge about this, and since VTL only specifies "6dB" NFB without further elaboration, I'll assume any difference is not large or intended to be sonically consequential in and of itself (as opposed to the change in the tubes' mode of operation). FWIW, 6dB is considered to be quite a low amount of NFB, and many electronics designers - in contrast to many audiophiles - would not be so quick to condemn the usage of *any* NFB, citing tradeoffs both ways, in which design skill and execution in NFB application are more predictive of resulting sonic quality than pronouncements of absolutes. However, it can't be said whether your listening experience in tetrode is a confirmation of that philosophy, due not only to our not knowing what the actual NFB difference - if any - may be, but also because of the other factors involved (obviously) like output power and tube mode.

Don't know if "it's that simple" as you say...I've never found tetrode, whatever its faults, to suffer from being "a little hazy and distant" compared to triode (whether before or after replacing the tubes). I also doubt that amps really lose all that much effective (as opposed to maximum) power in normal use as the tubes age, at least compared to how the sound changes simply from changes in the tubes' response characteristics. I do wonder if all your new tubes are broken in yet, which I find takes several days. But I'm not surprised you've already noticed improved sound overall, and I'm glad you've gotten back in good with triode, which is part of what switching to the KT-88EH's did for me. Keep rechecking the bias over the next week or so, it can take a while to settle in at a more constant level.

I own MB-185's, powering Thiel CS 2.2's (not the demanding load that some associate with the brand - 8" woofer and fairly flat 4 ohm impedance with well-behaved phase angle, average sensitivity of about 87dB).

My general observations about triode vs. tetrode - the inevitable variations in recordings that cause disk-specific preferences aside - go like this:

Triode: Superior in only three areas, but three of the most important areas - tonal balance, harmonic structure, and audible distortion (within its dynamic limits) on transients.

Tonal balance: Triode has a more neutrally-balanced tonality throughout the heart of the audio range. Tetrode is is brighter and less evenly-balanced in comparison, with a relative lack of lower-midrange/upper-bass fullness combined with an overabundance in the 'presence' range of the upper-midrange through the low-treble.

Harmonic structure: Triode operation reduces odd-order harmonic production, while the push-pull operation takes care of reducing even-order byproducts. Tetrode produces more odd-order byproducts, which push-pull cannot cancel, so they will heard predominantly over the more pleasing even-order series. This skews the amp's harmonic signature to yield a 'harder'-sounding emphasis with tetrode, while triode gives a more natural descending harmonic distortion with asceding harmonic order, and lower total distortion overall. This is entirely independent of the tonal balance issue noted above, which can be addressed to some degree through choice of input tubes, cables, etc. Triode will always have the advantage over tetrode when it comes to natural sweetness of harmonic presentation.

Transient distortion: Although triode won't give the dynamic impact of tetrode, and its limits are lower due to the reduced max power, transients are cleaner distortion-wise, even though they are slightly compressed by comparison to tetrode. This means that triode will deal better with sibilants and sound smoother-textured overall, while tetrode will give some false emphasis to fast-rising impulses that produces a slighly 'edgier' sound, with more emphasis on 'detail' than is strictly natural. BTW, this is just my listening impression, not something I know from a technical standpoint, and I surmise it could be the result of either intermodulation or temporal distortions.

After those three areas, I believe tetrode holds the rest of the cards, including those relating primarily to all qualities physical:

Greater extension: Tetrode has more extended response at the top and bottom of the frequecy range.

Bass definition: Tetrode maintains greater control over woofers for more bass-range detail and tonal differentiation with less overhang, although the bass often won't sound as full-bodied in tetrode as in triode.

Dynamics: As mentioned, triode will compress earlier and more than tetrode on dynamic peaks. But tetrode is also better at eludicating subtle dynamic shadings of expression on the micro end.

Soundstage: Tetrode presents a clearer, and more defined and separated soundspace than triode, which by comparison sounds somewhat homogenized and foggy. Tetrode also has greater scale.

Imaging: Tetrode presents better-focused images, triode more amorphous ones.

Resolution: Tetrode is better at illuminating decays, and combined with its clearer atmosphere and better micro-dynamic tracking, the result is more transparency to fine detail and more precise articulation.

In sum, I think the amps pass more information in tetrode, but don't necessarily present it as nicely as in triode. (Of course, all my observations are drawn by me in my system, and are therefore not to be taken as universal truths.) I do end up using the switch as a tone control sometimes, dependent on the recording. If a recording is overly-aggresive, I'll likely go to triode regardless any impact or definition I give up. Likewise, if a recording is muddy or hooded-sounding, I'll use tetrode to help pry it open, harmonics be damned.

With well-recorded material, its a little more disconcerting to me to have an amp with two distinct sounds, neither of which I know is absolutely correct. I'd love to be able to combine the purity and naturalness of triode with the control, power, extension, and precision of tetrode. But since no amp is totally correct, I guess it can't hurt to own that at least gives you an option. So on decently-recorded disks that won't penalize you too much either way, my preference is mostly determined by the type of musical material, where its inherent priorities lead me.

BTW, I often do my comparisons on the fly, sometimes with music playing. Over the past year, switching with impunity has possibly cost me two output tubes and one output fuse, but I can't be entirely sure of the causes. One thing I know is that if you are going to do this, it is vital to flip the switch as quickly and positively as you can - no half-assed indecisive or scared-acting switch-flipping. Just do it. Or don't do it - this has got to be a personal call. I've flipped hundreds of times with no problem, but who knows what'll happen next time? Oh, and to make valid comparisons, you MUST compensate for the 2-3dB volume differential between the two modes (something that's easy with my Levinson preamp's programmable muting function).

As a final thought, it's worth saying the change in modes is not a night-and-day affair - it can be pretty subtle, or pretty insignificant with certain sources. After all, most of what makes the amp sound like the amp that it is, remains the same in either position. Right now, I'm considering a coupling cap upgrade and some experiments with input tubes in order to try and bring tetrode closer to my ideal tonal balance, which for me is its biggest flaw at present.

So far, I'd say that I run the amps 80% in tetrode over triode, but when I first got the amps, the switches had been removed by a previous owner and left set on tetrode, so for almost the first of a year I had them I never heard triode. As I've gotten used to the restored switch and its effects, I find myself increasingly willing to forego some of the qualities I had become accustomed to before the repair and listen in triode.

Interesting result on the outlets - are they separate dedicated lines too? I guess your 450's in tetrode were pulling too much from the wall to share nicely. My 185's probably don't - I have them plugged into the same Power Wedge Ultra PLC I use for all my gear combined, with its lone cord feeding from the single available outlet. I haven't minded, but I may find an improvement if I ever re-situate to allow plugging into multiple lines or outlets (then again, maybe not - hum can often become worse in such installations).

Freaky that you have the same preamp as me - how many guys could be running Levinson pre's into VTL's? Our slightly different impressions of the relative tonal balances between triode and tetrode could be because of our different speakers, and also because of different tube and cable choices. I've recently done some switching in the tube department, you can read about it on this archived thread.

I think I need to slightly amend a couple of the (mis)impressions I seem to have left from above.

Hgabert refers to agreeing with me that "with tetrode...the frequency spectrum is flatter". Yes, tetrode could be called 'flatter' in that it sounds more extended, both top and bottom, but I said above that triode sounds more neutrally-balanced to me through most of the audioband, including all of the midrange. To me, tetrode is less flat than triode here, which is where it counts most. Questions of ultimate tonal balance, however, are quite dependent on partnering gear and choice of input tube.

Piezo says he agrees with me that "tetrode...has more purity and naturalness". Actually (except for the part I just omitted about tetrode having greater dynamic range, which is true, and which he stipulates to), that is what I said about triode, not tetrode - which generally seems a bit less 'pure' and 'natural' than triode, for reasons I speculate some about above - and in that respect we really do agree.

This also reminds me that we audiophiles are using a particular definition when it comes to the word "natural" - one where we are mostly talking about matters of texture and tonal and harmonic balances, and here the word applies best to triode. If we expand on that definition a bit to consider other aspects that could also be described as being more so, then I would say that tetrode has the more 'natural' spatial presentation, dynamics and articulation, and ability to resolve fine detail. But we normally use other words to denote realistic performance in these areas.

I'll also mention an analogy I've used before to sum up the different impressions I get from comparing the two modes: With triode, I often feel as if the performance is less 'electronically' rendered than with tetrode, but also that I'm listening to it at some remove, sort of as if I were in the next room - while in tetrode, the performance sounds slightly more like an 'electronic' artifice, but also as if it's a performance that I'm more in the direct presence of, in the same room with if you will. Or put another way, with triode I more feel like I'm hearing the real thing but also that I can't 'touch' it, and in tetrode I feel more like it's embodied right there in front of me, but that it's less quite like the original thing and more of a slightly 'off' clone. The choice is like looking through a window at your very own mother (triode) vs. exchanging a hug with her replicant (tetrode). I guess people buy even better amps (presumably - hopefully, for me - at a higher cost) so the two qualities can be combined to an even larger degree than these amps can manage strong glimpses of in the respective modes...

What are your speakers/cables? To reiterate, I'm listening through Thiel 2.2's and either Satori or Au24 cable (though these impressions were formed only with the Satori, as I'm still evaluating the newer-to-me Audience cables).

Anyway, I'm not sure I can put a tag on which presentation is 'most realistic' so easily. I think it depends a lot on the recording and type of music. As I said above, in general I find tetrode more realistic in terms of qualities I would call 'physical'. But I think triode is more 'accurate' in its tonal and harmonic presentation. I've had some recent success in transforming the tetrode tonal balance more to my liking by both changing input tubes (the previous ones had become microphonic, and were brighter) and by placing tube dampers on them. However, I'm not quite done sorting this all out yet, because I've also gone from Svet 6550C's to EH KT-88's, and on top of that I haven't yet carefully evaluated again the tetrode/triode question with all the new tubes in place. Not to mention that I'm auditioning these speaker cables, which sound perhaps more transiently accurate but less warm than my reference, and I also just had a tweeter and crossover resistor replaced and that speaker is still breaking back in...it never ends :-)

Anyway, yes, my physical impressions are pretty much the opposite of what you describe in your second paragraph above, so apparently our 'clones' are 'slightly off' in different ways. Where you say triode gives you more detail, room ambience, palpability, walk-around transparency, and is more pinned-down in space, I feel tetrode does all those things better, with triode being the mode giving more body/fullness (if less extension and tautness). It's enough to make me wonder if our mode switches are wired-up the same way or if one's reversed! But I guess I envy anybody who's come to the comfortable conclusion that one mode is usually best for them...at some point here I'll revisit this question with the new tubes and post my findings.

I was kidding about the switches being reversed - that's pretty unlikely with two monoblocks, and besides you can't mistake which mode has less power. Last night I did my first listening tests between the two modes with the KT-88's installed. I'm not ready yet to report with certainty, but preliminary results suggest that my impressions will be different with these tubes...

Hgabert: I have been posting my ongoing findings regarding the output tube question on this archived thread, but my comments there do not touch yet on the differences between tetrode and triode with the KT-88's installed. VTL has told me that they've auditioned their amps with KT-88's and liked the results, especially in the bass, but preferred to stick with the 6550C's for the vocal-range reproduction. I don't know what KT-88 versions they've tried to date, but I suspect the EH's were not yet available at the time of the trials Bea described to me, and assume they were Svet's. In my auditioning with the KT-88EH's, I have detected no penalty on vocals compared to the 6550C's, and in fact am now hearing more vocal detail, but as I will shortly go into more specifics about on the other thread, I am reluctant to draw any final conclusions since my 6550C's are over a year old and my KT-88EH's are almost new. Nevertheless, I am tempted to wonder whether VTL's preference for continuing to employ the 6550C might not have something to do with the cost factor - though it's a good enough tube, I am thinking the KT-88EH is better in this application (particularly in the bass), at least in my system and to my ears. I also am currently investigating whether it may possess comparitively greater significant advantages over the 6550C when the two are compared running in triode... :-)

Cedhz: RE your experience with your audiophile friends - In this hobby I have found that the stongest opinions often seemed to be formed in the absence of listening... ;^)

Update: I left off this thread around the time I switched from running Svet (SED) 6550C's to Electro Harmonix KT-88's in my MB-185's. The latest report is that I am listening more than ever in triode mode these days. The KT-88EH's seem to largely ameliorate some of the relative weaknesses of triode vs. tetrode that I outlined above.

Tetrode definitely still retains the greater bass control and wider macrodynamic envelope. But now triode suffers much less by comparison in terms of soundstage transparency and extension at the frequency extremes. Triode still retains its advantage over tetrode in naturalness of midrange tonality and the harmonic structure of the overtone series. (Let me not leave the impression that the switch to the KT's didn't also improve the sound of tetrode mode - it did.) At this point, I think the advantages that tetrode does retain are probably due mostly - maybe even entirely - to that mode's higher power.

Owners of higher-powered VTL models than mine, like Alberporter and Mejames (who also use bigger speakers and rooms than mine), have reported their unwavering preference for triode mode. Soon I plan to audition a pair of MB-250's, the one amp VTL makes which is permanently wired for triode mode only. It roughly doubles the triode power available from my 185's (meaning it equals their power in tetrode mode). If I get these, I will fit them with the KT-88EH's and test them against the 185's in tetrode to see if macrodynamic range, microdynamic detail, and bass control/punch are comparable - hopefully combined with triode's superior tonality and harmonics (at a higher cost, of course).

Oh well, my highly anticipated (by me) MB-185/tetrode vs. MB-250/triode showdown will have to wait...I decided not to buy this pair of the latter amps for other reasons. (Didn't get to audition them in the seller's system either - just his replacement Halo JC-1's.) Don't cry for me though; my amps still sound great! (In either mode :-)

Hgabert - This is what I wonder: could tetrode mode be inherently superior in some ways, or are your observational results mostly a manifestation of the power issue? I have a feeling that all inherent superiority may actually lie with triode mode, but that for certain music - played back beyond a certain volume, through certain speakers in certain sized rooms - the power advantage of tetrode will out, even with 450's. The comparison in your case would presumably be your 450's in tetrode vs. a pair of 750's in triode. I realize that the 200+ watts available from your amps in triode seems like plenty of power, but I can sometimes reach the limits of those same watts with my amps in tetrode, and I think there's a reason VTL makes the 750's and 1250's.

As mentioned above, I was more of your frame of mind before I swapped the 6550C's for the KT-88EH's. The 6550C's didn't quite seem to thrive in triode the way the KT-88EH's do. And as of very recently, the amps are now powered through an ExactPower EP-15a voltage regulator/waveform correction unit (along with the rest of the system, followed as usual by the balanced power isolation tranformers in my API Power Wedge Ultra for the sources and pre), and there's even more physical specificity and consistent clarity on tap in either mode.

I'm sure Mejames is right about the 750's, and the comparison is a little unfair not only because of the added power, but also because the Reference amps have an improved input amplifier circuit topology and power supply, not to mention the better parts. Upgrading the coupling caps in my 185's is probably the next step, although going to 250's or 450's could also be in the picture eventually (the Ref amps are too big and too much money for me - and besides, I like to think I can and should be happy without having to have the best :-)

The tube swap to KT-88's requires nothing more than installing the new tubes and biasing as normal. (This is not the case for KT-90's, which need resistor changes.) The driver 6350's are the same as always (there's not much to choose from with this tube, NOS being tough to come by and limited to just a couple of options I believe). I've played with several different input tube varieties of 12AT7's - maybe I'll do a thread on the results sometime.

About your comment that "small-scale" music shouldn't present very taxing power demands, just today I did some comparisons between tetrode and triode using Sonny Rollins' "Way Out West" MFSL CD. Jazz doesn't get much more intimate than this: just a sax, bass fiddle, and a small trap set with a little percussion. No chords at all, and the tunes are taken at easy tempos and volumes, with plenty of space left in between the notes. Still, there are differences to be heard when listening between the two modes - even at moderate levels - that could relate to power, such as bass control and dynamic range. As ususal, those catagories go to tetrode, while overall naturalness of timbre, particularly from the low midrange up, goes to triode. Which presentation sounds more 'real' is a bit of a toss-up, depending on what sonic qualities you prioritize.

I think that to approximate reproducing the true sound of any acoustic instrument or voice, no matter how delicate the music or spare the arrangement, still presents an enormous challenge to any power amplifier, and I tend to agree with VTL's position that a lot of the inevitable compromise is simply related to a lack of available ultimate power. The task, of course, is to make a power amp that can deliver the needed level of clean power and yet do it instantaneously and with good musicality. I guess VTL probably does as good a job of trying to give us this as anybody, but there's a price to paid both monetarily and in size/weight/heat/appearance, and besides most speakers can't take full advantage of dynamically unlimited amps anyway. Most audiophiles will never attempt this, and the SET/horns crowd would pooh-pooh the notion, but my suspicion is that a home 2-channel system utilizing several thousand watts of multi-amped power would not necessarily be overkill strictly from a sonic accuracy standpoint. (You here Sean?... :-)

That's what VTL says, although I couldn't find out exactly how the parts are supposed to differ. I also couldn't get TRT to either confirm or deny this assertion of proprietary specification for me. VTL seems to charge around twice as much for these parts as they're generally available for - assuming they'll even sell them to you at all; I've found the company is very loathe to assist in any way owners wanting to do mods on their own or hire outside tech assistance that's qualified and local to them. That VTL charges $200 more for the InfiniCap upgrade than for the MultiCap upgrade raises my eyebrows; that suggests a parts cost increase of over $16 per cap for my amps, way more than the actual retail difference - in fact, it's slightly above the InfiniCap per-piece retail price itself!

They are insistent that the cap mod be done at their CA factory, which I can understand for reasons of quality control and warranty impact, not to mention their own business. Plus, they will perform other updates if needed - such as to the fusing and certain resistors - while the amps are in, and retest them generally and provide you with updated test documentation. But for those of us who don't have a warranty with VTL anyway and live on the opposite coast (and who've already had the factory updates performed), this means shouldering 2-way shipping charges considerably north of $200 (you must ship the way VTL says) for the privilege paying the factory almost double what is commonly charged for this same job done independently. Tell your local tech that the manufacturer wants $1K (not including shipping) to replace 12 (in my case) capacitors that cost about $15 apiece retail (at $50 an hour, this would imply around 16 hours of labor, but since the job should only take about half a day for a skilled tech, well, you do the math) and see what kind of reaction you get... ;^) Add in the associated expenses, risk, and hassle of shipping, plus the extended time the amps will be away from the system, and you can understand why I'm inclined not to worry about VTL's vague claims concerning the InfiniCaps they use.

One thing I've been doing lately is to use different input tubes for each mode with certain types of recordings. For triode mode, the brighter, more incisive Telefunkens make a good match, while in tetrode the darker, fuller Mullards complement nicely. I agree (I think it's agree) that the Shelley Manne intros to those SR tunes generally sound more natural in triode, but when Ray Brown kicks in the pendulum swings toward tetrode...space is also presented differently in each mode, but there I tend to prefer whichever mode I'm listening to at the moment ;^)

Even if your input tubes display no audible microphonics, I urge everybody to try some sort of tube dampers here (I'm using Herbie's); for a nominal investment, you'll probably hear the amps cope better with stuff like vocal sibilants, plus truer tonality and more stable imaging, with 'quieter' spaces between performer images. Everything gets a bit less edgy and a bit more listenable, especially at higher volumes (not that transient impact is smoothed out - actually, HF transient tracking fidelity is increased, because there's less tizz and overhang). The improvement isn't huge, but neither is the cost or hassle.

Yep, I leave'em off, if for no other reason than I like to check the bias more often than I want to deal with the covers. But playing with input tubes is a more fun reason.

I'm also just paranoid enough that I prefer an unobstructed view of, and access to, the tubes in case one should give a hint it's beginning to act up; a few times in the past (not always with these amps) I've caught stuff out of the corner of my eye which likely saved me bigger headaches within moments - including smoke wisps emanating from around transformers - that you're not going to notice as easily with covers on. Have you ever done the jump-up and across-the-room-sprint when a tube started in with a light show? :-) And although I think the amps look better with the covers in place, I assume that heat dissipation - and therefore tube life - probably gets slightly improved running naked.

Larger woofers to control at the least, and maybe a more difficult impedance, I don't know. The 802 was always my favorite B&W, over the 801, although I haven't really auditioned the N-series. But I never heard one with tubes. Your experience is another bit of evidence suggesting that where tetrode is superior, the extra power may be the main or even the only reason.

Eldee: I've wondered about that before myself. In fact, VTL amps always operate with at least one element disconnected, since there is no pentode mode. I think the decision not to run in pentode is probably a good one from a sonic standpoint (I say this not from hifi system pentode experience but with guitar amps), but I don't know how a similar circuit would sound with EL-34's or 6L6's instead. Triode wiring of 6550 types has also been exploited by amp makers like ARC and CAT, so VTL is not alone in considering the unused screen elements to apparently be no handicap. But to answer your first question, of course there are "real triode amps" (and power tubes, which were around before tetrodes and pentodes were invented) - for instance, not even counting all the SET designs, VAC offers high-powered push-pull amps using paralleled 300B's, and the 6AS7 output tube used by Atma-Sphere is a true triode, along with the Russian 6C33C-B used by BAT.

Sorry, my rather obscure point about the guitar amps was supposed to allude to the fact they are pentode (or beam) designs (using EL-34's or 6L6's) which don't sound as natural in fidelity as my (6550-equipped) VTL's run in triode, but do seem to share some of the less harmonius characteristics shown by the VTL's in tetrode. But as you correctly suggest, for a variety of reasons this probably isn't very good evidence for anything though.

Anyway, here is an interesting webpage on the main subject...

Eldartford: I was prompted to do further research by this question of tetrode vs. pentode catagorization. What I noticed is that beam-tetrodes, lacking the suppressor screen of a true pentode, but with the beam-forming plates (which are not electrically-connected elements of the tube, otherwise known as electrodes) between the cathode, grid, and screen on the emission side and the anode (plate) on the collection side, are nevertheless often described as pentodes or 'beam-pentodes' - even in their original manufacturer's technical description papers. This finding, though I don't know the explanation for it, does help me understand why I've seen certain popular audio output tubes variously refered to as both tetrodes and pentodes in magazine reviews and the like. (Here is a page showing helpful diagrams of the differences.)

I also had forgotten that the British "KT" designation in fact stands for Kinkless Tetrode (duh!) - in other words, beam design. So obviously my KT-88EH's do not have an unused electrode element when run in tetrode in my VTLs, and the same would go for 6550's. Further confusing matters (for me) was the fact that my previous tubed audio amplifier used EL-34's (a true pentode) but ran them in ultralinear connection. The RCA 6L6 was apparently the first production beam-tetrode - see this link for the backstory. Though both suppressor screens and beam-formation plates are intended to deal with secondary emission from the anode, it does seem likely to me, as possibly alluded to in the article I linked in my post of 12/7, that the two constructions might behave somewhat differently when triode-connected.

To the extent the switch may act as a bit of a de facto equalizer, that's exactly what I *don't* like about it. As I probably stated somewhere way up above, the fact that an amp can have two subtly different sounds at the flick of a switch just points up the reality that neither presentation can be trusted as accurate. Of course we already knew that, but with amps that don't offer two different-sounding modes, at least the subject isn't put on a platter for us.

I once owned a DAC, briefly, that offered externally-defeatable 'upsampling', in a choice of three different frequencies chosen from moveable jumpers inside the unit. Objective testing by a rather involved methodology I worked out to help me evaluate the effects showed that performance was most faithful to the signal fed the DAC with the upsampling switched off. However, I couldn't deny that there were disks whose sound I subjectively prefered with one or another of the upsampling rates switched-in, even as I could hear some areas of fidelity being compromised. But I wasn't happy that the machine editorialized the sound at all, and luckily my (non-'upsampling') reference DAC outperformed it anyway and I sold the thing.

With the VTL's, I think the ultimate answer would be to get one of the very high-powered Reference models and run it in triode. Answer, that is, if you're a richer man than I...

On a related subject, I spoke with one the technicians at VTL not too long ago and learned that it's perfectly OK to run the amps with tube pairs subtracted, equally from each bank (same number of output tubes removed respectively from both the left and right sides of the chassis, though it's not important that the tubes be from particular corresponding positions within each of the banks, due to the way the amps are wired). This will simply have the effect of reducing power output, with everything else, like the power supply capacity, held constant.

In the case of my 185's, with their three tubes per bank, I can run them at full power (all six tubes, three pairs), 2/3 power (remove one pair), or 1/3 power (remove two pairs, leaving just one pair in the amps). And I have tried this with no problems at all. But for 450 owners, with four tubes per bank, one of the available choices is 1/2 power (two pairs removed, leaving two pairs in) - a close approximation of the same power reduction you get when switching to triode mode. So a 450 owner could compare tetrode mode with only two tube pairs installed, vs. triode mode with all four tube pairs installed, and the prevailing differences would presumably be entirely due to just the different operating modes, without being confounded by different output powers. Just in case any of you 450 owners are feeling curious and experimental...

Exactly right. I got over the 'unsettling'. The 'useful' remains.

If I'm not mistaken, the knob is only used on the double-decker Reference amps, the rocker on the single-deck Signature models and below. I would assume that the rotary switch is heavier duty rated for the higher power; maybe it's a 20 amp part. But I don't know this, and I suppose it could also just be a cosmetic difference. I don't think I recall them always using it on those models, but that observation doesn't prove anything, as they might have recently upgraded the part, just as they've upgraded their fusing in recent years. Call and ask. BTW, VTL has recently upgraded the input stage on the Signature monoblocks to a differential design as in the Ref's, but haven't seemed to advertise the change or its benefits. (Unfortunately, for owners of previous-gen Sig's like me, there is no upgrade retrofit possible.)

I agree that running XLR into the previous gen VTL amps sounds no better, and maybe just a tad worse than running RCA, even if the preamp is fully-balanced. It has been my experience that between true-balanced components, running XLR usually does sound better - certainly no worse - and I would expect this to hold true for the new VTL's.

It is possible to design a balanced input stage that does benefit from lower noise and distortion even if an unbalanced input is used on the RCA jacks. Converting a single-ended input to balanced before amplifying (not needed for the XLR jacks) can still lower both internally-generated noise, and even to some degree noise picked up by the interconnect, depending on its configuration, although this may not be as effective as if a balanced connection is used. I believe this advantage applies in the case of the new VTL input stage design. It did not apply in the case of the previous gen VTL's XLR jacks, which were included only for the sake of convenience.