RJT, Aside from guide alignment and quality of setup, spindle rpm and dielectric pressure are the two primary factors that will affect how much the tube will "walk" off location as it progresses through the work piece. These are almost equally important, since it is a combination of the two that makes the whole process work. If the flushing pressure is too low or the dielectric viscosity too high (Believe it or not, water is almost too viscous for this application. The .0030" diameter opening in a .012" dia. electrode is about 11% of the cross-sectional area of the resulting .0015" spark gap.), there is not enough volume of fluid coming back up around the outside of the electrode to keep it centered in the hole, and it cannot burn straight. Spindle rpm is equally critical since even with a "perfectly straight" electrode, variations in wall thickness will cause flushing to exit toward the thinnest wall section of the tube as the electrode wears. The resulting flush exiting at an angle to the rotational center axis of the electrode will cause deflection at the tip. This will cause "walking" if the spindle rpm is insufficient to compensate for this effect. This combination of factors is the reason you will often see "bouncing" when the electrode begins to break through at the bottom of the hole, since flushing pressure is no longer holding the electrode centered in the hole, and the deflection caused by flushing pressure makes the tube essentially "bore" its way through the bottom of the hole with no real benefit from flushing at the point of discharge. Your application has a depth/diameter ratio of 83.3/1, which makes this even more difficult. As a single burn, this application is hovering at the limit of "hole popper" capability, and even the most skilled operator would need to get everything "just right" to perform this operation consistently and predictably from piece-to-piece (keep in mind that at .015" dia. with "perfect" position at entry, a mere .0025" drift at exit gives you a bad part). As Dave suggests, you might find better success in burning half-way from each side, but none of us really want to make two setups for one start hole, do we? Ideally, to do this with a single burn, this application would be best suited to a machine which can provide a STEADY supply of a low-viscosity (ex: Iono-Plus 3000 ET) dielectric at around 1200+ psi and around 1000 spindle rpm for the most accurate and most consistent result. Granted, a machine with this capability is in an entirely different class than the average "hole popper", and is commonly referred to as a "fine hole machine". These machines are typically priced between $150K and $300K, depending upon the builder and machine configuration.