I've been working on a new capability for the Blue Raven firmware over the last few days, which is to limit the current that flows into the output in the event of a short circuit in the load.
Load short circuits aren't that uncommon. Anyone who is firing a head-end igniter for an airstart will almost certainly get a low-resistance short circuit once the motor comes up to pressure, because the glowing gas inside a firing motor is very conductive. I have also had short circuits for other reasons, like an ematch head that had both sides touching a conductive hole. The guidance I have provided since the start of the Raven altimeter has been to only use 9V alkaline batteries or lipo batteries that are small enough that their short circuit current capability is lower than what the output MOSFET transistor can handle (about 20 Amps). This has meant an upper bound on a lipo battery of about 180 mAhrs. That's enough for about 10 hours on the pad for a full discharge, or about 5 hours before the battery gets down to a questionable voltage. Larger batteries can overhead the output MOSFET and cause a permanent failure, and the most common failure mode in this case is to fail short (i.e. the charge fires as soon as power is applied to the altimeter). This has happened to more than one customer over the years who used a lipo battery that was too large.
I designed the Blue Raven to use hardware timer channels to control the outputs, so that the firmware could duty-cycle the output switch at a fast enough rate to cut back the current in a controlled way if the measured deployment current gets too high. I believe I have this capability working now, and I bought some larger RC batteries to test it out. I started the development with a hard short connected across one output, and got it to work when the short was in place at the start of the firing. Then I realized I also needed to cover the case of a short that takes place partway through the output firing, as would happen during an airstart. Here's a test of that, where I start with a 3.3 Ohm load resistor, and then short it out partway through the firing. This is a 400 mAhr lipo that arrived from Amazon yesterday:
When the load shorts out, the current shoots up to 34 Amps. The Blue Raven responds by cutting back the current within 2 msec and then controls the output current to a target of 3 Amps for the rest of the output period by reducing the duty cycle.
To push the system even harder, I used a 2S 1000 mAhr battery fully charged up to 8.4V. The Blue Raven survived a short with that battery unscathed also.
I'm gathering up some other firmware bug fixes along with this new feature into a new build that should be available in the next few days.
Load short circuits aren't that uncommon. Anyone who is firing a head-end igniter for an airstart will almost certainly get a low-resistance short circuit once the motor comes up to pressure, because the glowing gas inside a firing motor is very conductive. I have also had short circuits for other reasons, like an ematch head that had both sides touching a conductive hole. The guidance I have provided since the start of the Raven altimeter has been to only use 9V alkaline batteries or lipo batteries that are small enough that their short circuit current capability is lower than what the output MOSFET transistor can handle (about 20 Amps). This has meant an upper bound on a lipo battery of about 180 mAhrs. That's enough for about 10 hours on the pad for a full discharge, or about 5 hours before the battery gets down to a questionable voltage. Larger batteries can overhead the output MOSFET and cause a permanent failure, and the most common failure mode in this case is to fail short (i.e. the charge fires as soon as power is applied to the altimeter). This has happened to more than one customer over the years who used a lipo battery that was too large.
I designed the Blue Raven to use hardware timer channels to control the outputs, so that the firmware could duty-cycle the output switch at a fast enough rate to cut back the current in a controlled way if the measured deployment current gets too high. I believe I have this capability working now, and I bought some larger RC batteries to test it out. I started the development with a hard short connected across one output, and got it to work when the short was in place at the start of the firing. Then I realized I also needed to cover the case of a short that takes place partway through the output firing, as would happen during an airstart. Here's a test of that, where I start with a 3.3 Ohm load resistor, and then short it out partway through the firing. This is a 400 mAhr lipo that arrived from Amazon yesterday:
When the load shorts out, the current shoots up to 34 Amps. The Blue Raven responds by cutting back the current within 2 msec and then controls the output current to a target of 3 Amps for the rest of the output period by reducing the duty cycle.
To push the system even harder, I used a 2S 1000 mAhr battery fully charged up to 8.4V. The Blue Raven survived a short with that battery unscathed also.
I'm gathering up some other firmware bug fixes along with this new feature into a new build that should be available in the next few days.
Last edited:
.
