have you thought about using a magnetic apogee detector like
"zeptobit" [the web site is down for a couple more days] it uses the Earths magnetic field to detect orientation, and can also be used as a timer.
I have both a rocket with a
zeptobit and a couple of rockets with only a single sampling port and everything works well.
You stole my fire as I was going to suggest that. The only thing that would be unknown to me would be the presence of the metal all-thread and perhaps someone
more knowledgeable could comment. I have a zeptobit but haven't flown it yet. I've used the old Galejs MAD unit from Aerocon for 14 flights but it I believe is out of production.
The zeptobit has a lot of safeguards to prevent an accident whereas strict discipline was needed with the Galjs MAD. If the Galejs was turned on with a live charge and say one changes one's mind to launch, they had to make darned sure the unit was turned off (and I removed one leg of the ematch charge from the bulkhead terminals) before tipping the rocket over off the rail.
It would be easy to test in the above rocket by installing a unit and tipping it over. If an LED were attached to the terminals say of a Galejs unit it will give an indication of the ejection angle. The zeptobit I believe has a G switch and will activate the channels (there are two a primary and a backup) when the rocket starts
a flight and there is an onboard LED that will light whenever the unit is tipped over in flight or not in flight. If the G switch has not detected acceleration, the deployment LED will still light up to indicate the angle the charge will fire but no power will go to the ematch terminals. Bottom line is the device arms itself in flight which is a nice safety feature. I tested it with live ematches and simply tipping it over it didn't fire. The indicator LED still lit up indicating the angle the rocket needs to be to fire if armed.
You might get away with a Raven using accelerometer deployment although Adrian doesn't recommend it anymore and suggests baro only apogee deployment.
The accelerometer deployment may still work as he did mention some time ago that a Raven could be sealed up in an ebay, use the accelerometer for apogee deployment and if in the act of deployment, a static port is opened, the Main event would occur nominally. Of course you're just looking for apogee deployment here. The zeptobit would be more economical.
The other thing to remember with the Jolly Logic chute release, since the main chute isn't snapping open at apogee, one has a little more
leeway when the apogee deployment of the package occurs. If the event is a bit early or late the reefed chute would be less likely to cause a
zipper. Kurt