At what vol level do you get rumble / flutter?

There is a mixing of terms here.
Wow actually refers to longer term speed variations in the turntable. Flutter is short term speed variations. Neither has anything to do with what you're experiencing.
Rumble used to be the term used to describe noise contributed by the mechanics of the turntable - bearing/ motor noise etc.

Excessive woofer movement is often caused by low frequency feedback to the turntable (acoustically generated or footfalls etc) or warp induced low frequency signals being picked up by the cartridge. Since you have a wall shelf I suspect the latter.

Often these signals will be accentuated by a mismatched cartridge resonant frequency - ie if cartridge resonance occurs too low (below say 7Hz) this can emphasize these low frequency problems.
A heavier cartridge will lower the resonant frequency. A higher cartridge compliance (all else equal) will also reduce the resonant frequency.
Some cartridges just seem to be more susceptible to this issue, perhaps due to poor damping?
Not that it helps you, but I simply haven't had this issue at all for the last 20 odd years I've been using Graham unipivots.

It would be preferable to find the source of the issue rather than apply a bandaid like a low frequency filter. Not only does the latter add another device into the signal path but not dealing with a resonance problem may have other tracking and sonic compromises.

Actusreus, I don't have any personal experience with Lyra cartridges. I did notice in looking at the specs of the Delos that the compliance is quoted as "12 x 10-6cm/dyne at 100Hz", indicating that the compliance would be considerably higher at the cartridge resonant frequency - but this in itself shouldn't push the resonant frequency too low with a typical medium mass arm. It's actually difficult to get a really bad match. Even the highish compliance (for an MC) AT PTG-II posed no issues for me in the medium+ mass Phantom - though, anecdotally, any issues I've had in the past with other arms were with higher compliance cartridges.

I have read that Lyra cartridges are demanding and put a lot of energy back into arms. What you are getting may be the combination of the Delos' slightly higher compliance (= lower resonant freq), strong transmission of energy and the perhaps the VPI arm is not ideally coping with this. Have you played with the oil damping on the VPI?

Assessing the resonant behavior with a test record can help you see what you're dealing with. I've found there can be a marked difference between what the specs suggest and practical results.

Adding mass to the cartridge/tonearm will lower the resonant frequency. Note that warp induced signals are in the 0.5 - 7Hz region, so you don't want to drop the cartridge resonant frequency too low if your cartridge/arm/table already has a problem in this area.

FWIW, I can't recall any cartridges I've used that have fallen outside the broad ideals for cartridge resonance, say 8-14Hz, when a test record is used. However there have been large differences in the amplitude behavior at resonance - no doubt due to the different construction methods of the particular cartridges and arms. For instance I'd expect a rigidly constructed Lyra cartridge to produce a higher Q resonance that a cartridge built on a plastic frame/body. The latter type of cartridge will probably be easier to accommodate, though of course that doesn't make it better.

07-07-12: Actusreus
....if you have very low frequencies cut into the record itself during the cutting process, which is quite common and a fact, your speaker should flutter without a filter unless your system is incapable of reproducing these low frequencies. Certainly a mismatch will make it even worse, but scientifically, how would a full range system suppress subsonic frequencies cut into the groove?

Are subsonic frequencies - below 20Hz - cut into records? AFAIK, there are good reasons cartridge resonance is generally positioned around 10Hz. Firstly to keep it above warp induced low frequencies, footfalls etc and secondly to keep it clear of recorded signals. It would cause serious tracking issues if recorded material coincided with cartridge resonance - witness what happens on a test record which has deliberately been cut with these low frequencies.

Cartridge resonance can't be avoided, it's how the cartridge/tonearm/table deals with this resonance (and other low frequency artifacts) which makes the difference IMO. I agree that having resonance in the ideal range doesn't mean you'll avoid issues. Some woofer movement due to low freq artifacts is perhaps to be expected in a vinyl system, but it needn't be excessive or problematic.

07-07-12: Actusreus
I understand it's marketing as well on KAB's part, but I certainly noticed differences in the degree of woofer excursions between even two perfectly flat records where everything else was constant. How else to explain it other than the subsonic frequencies were part of the recording rather than generated by the playback system.
Actusreus (Threads | Answers | This Thread)

I'm not disputing you have woofer movement, nor the fact that the KAB high pass filter removes these subsonic frequencies. The sub-sonic frequencies are generated by the playback process but probably not by music content. It would be asking for trouble to deliberately record music signals at around 10Hz and I doubt this would be done intentionally. Nonetheless I suppose there probably is unintentional low frequency signals recorded. However if most of your records have an *excessive* woofer pumping effect I'd say the problem lies with the arm/cartridge/turntable.

Whenever I have had this woofer pumping issue, the pumping was usually cyclic and continued regardless of music content on the record - more often than not linked directly to (even minor) irregularities in record flatness. I've also observed a contribution by 'bouncy' flooring - where I could see large pumping as I walked up to change the record. I don't get either of these issues with my 103R/Phantom/TNT in my current room.

Some cartridges I've tried in the past, the Shinon Red and Grado TLZ come to mind, have greatly exaggerated low frequency pumping to the point where they weren't practical to use in that particular vinyl system. Other cartridges, the Audio Technica Art 1 and Garrott P77 for example, were very well behaved on the same records/same system, with only minor pumping effects. It's worth noting that both the Shinon and the Grado were somewhat higher compliance than the AT, though theoretically both should have been ok in a medium mass arm (I was using either the Sumiko MDC800 or the SME V back then).

I've seen a huge variation in this effect over the years with different cartridges/tonearms. Some cartridges also behave wildly at resonance (observed with test record) while others have only mild movement. FWIW the lowly Denon 103R is one of the best behaving cartridges I've used in this respect. Whether this is because of its 'lossy' mechanical construction I don't know. I can tell you that even when this cartridge is made much heavier, 14.5 grams in the Al body I use, there is still no issue - even though the resonant frequency has dropped to around 7-8Hz in the Phantom (observed via test record).

By all means get the rumble filter if you must use a particular combo and you have issues you can't solve with damping etc - personally I'd rather avoid the issue (change combo), but that's just me. Either way I hope you can sort out your problem.

Actusreus,

I do see your point, I probably didn't read your post above.
Like you say, it is probably a combination of factors at work here.
From memory the cartridges i mentioned above produced 'nervous' woofer pumping even on nominally flat records. They were verging on out of control with any ripple like warps near the start of the record - hence I found them unusable in that setup (and I returned them).
I suspect, for whatever reason, they had very pronounced resonance which was 'set off' by any subsonic content.

BTW, thanks for your kind words on my audio page, not sure its as deserving as you make out, but thanks.