S-Core Firmware 0.95 - Home bolt with flywheels turning at low speed; more robust and snappier speed-based feed control.

Link

Changes:

• Home bolt with flywheels turning (at default low speed)

The previous selftest routine reset the bolt at startup with the flywheels stopped. Occasionally, such as in the case where a stoppage or debris is present or the bolt is not homed but a loaded mag is present, this would mash something into stationary flywheels.

Now we wait until after the flywheel drive rotation check and run the wheels at low speed while homing to spit out anything that is accidentally fed. If something is fired inadvertently during this procedure, it is barfed out at about 36fps, so this is safe - assuming you have an ESC firmware with default minimum speed, which you should.

• Improve speed-based feed control

A series of consecutive in-range tach readings (the number of which is still a build-time setting) are now required to initiate firing. This removes most possibility of enough randomly in-range pulses being received during the timeout window to fire when a drive is not actually at speed. This might happen during a stall condition while tach pulses may be clipped by start timeouts and/or there may be motor vibration which is capable of creating in-range tach pulses.

Offset margins and such have been revised slightly and control is a good bit snappier. Some perturbing of these settings may be called for to fine-tune velocity consistency, but it shoots pretty damn nicely with my Emax equipped primary.

03-04-20 Update: I have hotfixed some configuration parameters in this release. Posted as a Google Drive version, so should be transparent.

• Set maximum flywheel RPM back to 25,510 since 26,000 is a deleterious overspeed for the Hy-Con and resulted in a decrease in mean velocity and an increase in velocity spread.
• Make STC settings a bit more conservative since while it got consistent velocity at 50/25, how I had it set may have been edging toward not getting max velocity in lieu of snappiness in a few cases. This does expose the SimonK mid-speed transient response bug more as a slightly more delayed followup shot at certain speeds, but that's OK.