I seem to remember somebody (Tony?) who was using Shaun's amp relate that when he's tested deep coins with the amp off they might only null, but when he switched the amp on the coin would sound off properly. I would assume that would then mean that a very deep coin that the machine can still "see" yet is past being "fringe" might only make the threshold null. From my experience the coin doesn't even have to be all that deep if the ground has some level of higher mineralization. In that case two things can cause the null- Too fast OR SLOW of a sweep speed, or too high OR LOW of sensitivity. I keep hammering away at this but at least for me in my higher mineral sites even if the machine is stable at a real high sensitivity setting it will overload or choke on the ground minerals and null the coin out in certain situations. As said, going too fast or even super slow can often cause this as well. It all depends on the conditions at that site and your sensitivity setting as to how fast or slow you need to go to get best ID. I'm finding with the 12x10 it wants sensitivity via my buried dime test higher than typical settings for the stock 10" coil or 15x12, yet this setting is still a good bit lower than what is max stability for the machine. I'm sure in low mineral sites max stable sensitivity will also be where the buried dime test wants it.
I don't normally dig a null. Instead I do either fast/tight wiggles or more often I find that short 3 to 6" slow sweeps seems to derive best ID/audio for me. The lower the sensitivity the fast it normally wants and tight wiggle, the higher the sensitivity the slower it wants and wider short sweeps instead. If I can't produce some form of "iffy" coin signal from the target doing all the above then I move on when land hunting. Often too the target will only sound off if I crawl the coil to one exact spot. Some times this ends up with the target being directly under the coil, but more often than not for whatever reason (probably minerals or trash) I find that the tip of the coil is the best part to use to try to pull good IDs out of targets. If I go over it with the center of the coil it will null more often than not, and I'm talking about targets that did indeed turn out to be deep coins. Not exclusive to the SEF coils either, the stock 10" coil also prefers the tip or tail for some reason when I'm trying to get a deep null to sound off. It doesn't have to have trash around it either. Not sure why that is.
Some of the deep coins I've dug (mainly wheats, since I've of course dug more of those than silvers at worked out spots) will only sound off from one direction and say with the tip. The target might be a total null or a very bad signal from other directions, yet you'll know by just how perfect the response is that one way that it's got to be a coin. It's too perfect most of the time to be iron or other trash causing a false coin hit.
The hard part on many of these deepies is pinpointing them. Often discriminate is the only one that will work. Other times PP mode is the only one that will sound off in the same spot from two directions. And still yet other times I can only PP it from one direction. When two directions are telling you the target is in two different places then I go by the direction that gives the best ID. Don't assume it's got to be iron if it's PPing in PP mode in a different spot than what discrimination says. Yes, often iron will do that, but I've found deep coins do this as well some times. I more go by just how good the audio/ID is from the one direction and dig those even if it's a complete null or trash signal from others. Once I find that I'm digging it and do not concern myself with if PP mode tells me it's not in the same place and could be iron. More often than not it turns out to be a coin. I'd say I dig more deep one way coins by using the audio/VDI as my deciding factor than I do iron now that I've learned that. The perfect coin ID one way is often a very tight spot and the coil must be kept at a specific speed and place and with the right amount of wiggle or short sweeping. When you do that you'll see if it's a perfect 10 in terms of audio/ID versus iron or other junk. Also, watch the negative numbers. Coins will hit the basement here and there but there is a pattern to it and they normally will start steadily climbing again in a less random fashion than iron. Iron will hit negative numbers more and has less ryme or reason to it's climb most of the time. I've been fine tuning the above methods more lately and am amazed at how little iron I'm digging while I'm digging a good bit of deep coins (mostly wheats since they are more numerous).
Do not rely on PP to tell you if it's dig worthy. If it sounds great and perfect one way then ignore that and PP it in whatever way possible. I often mix PP mode and discriminate mode to compare where I should dig. Some times I have to PP in discriminate from one direction but then use PP mode in the other because it's nulling or bad all over the place in discriminate from the other. My main choir or problem to tackle on real deep coins is not getting fooled by iron any more, it's figuring out exactly where I need to dig the plug when they are giving me issues from different directions. The tip of the coil is the key, as well as comparing discriminate to PP mode.