To get some experience with board-layout and surface-mount designs, I'm making a new timer circuit (something simple to start with). The prototype is done, and I'm working on the layout.

So, here's what I am thinking: I can get a minmum of three boards from the manufacturer, and in reality I only need one. I considered trying to sell them, but for such small quantities the price would be too high, and the boards as of yet have no track record. Instead, I can give the other two away to people that could use (and would not mind evaluating) a high quality timer. So let me know if anyone is interested.

Features: The timers count in 0.25 second intervals from 0 to 64 seconds for recovery and staging applications. --Can be modified to count in 0.5 second intervals to 128 seconds. They are G-switch activated at 4G's (also modifiable), and will not retrigger until the timing interval is complete. They can switch up to 20A to fire regular ignitors, currently using one or more 9V batteries (newer 9V's with AAAA's inside can flash up to 3A). The high-current capability is in part because of the dwindling supply of e-matches and oxrals. Because current is more important than voltage, I am evaluating using switching-regulators to get 50% more current out of a 9V by regulating it down to 6V so that a single 9V can ignite a "firstfire" or "copperhead" igniter. The timer has small LEDs for continuity checking, battery charge checking, and functional testing. The whole thing runs on about 12mA when not trying to ignite something. I've considered making the circuit fully redundant, but maybe later. The batteries will more than likely be electrically isolated so that different brands, series, and ages of 9V can be used together without causing problems with each other. There are DIP switches for setting the time interval, reset, and to safe the charges (contacts to solder in your own external switch too). By the way, the design is not set in stone yet, so suggestions are welcome until May.

I'll probably have these ready for the June launches. And of course, I'll demo them in my own rocket first. :)

Update: Due to some built-in capabilities of the chip, I've added some features and fixed some problems.

1) The timer can be triggered by launch, or burnout. When triggered by burnout, the timer is "armed" by the acceleration exceeding 4G's, but the doesn't start counting until the acceleration drops below that (negative transition in voltage). For my rocket, the acceleration drops below 4G's 0.2-0.5 seconds before there is no more thrust. --A more sensitive accelerometer could be used for more precision. I'll probably use a
jumper to select between the two.

2) A jumper connection will change the timer's interval to 15 seconds, allowing it to count to 64 minutes. This allows high precision for most applications, including staging, and long delays for other applications (e.g. while drifting under parachute, on the ground). The division for seconds to minutes was built in, it was a very small change to make this available.

I've been tinkering with the circuit a lot because the chip has some quirks. As a one-shot, the timing is restarted if it is triggered again (e.g., second stage firing). This is not what I wanted, so a workaround was needed. It also had the problem that if the timer had been triggered, and you reset it, it would behave as though it had timed out (firing the charge). There is now a simple workaround to disable the firing circuit whenever reset is asserted, so even if you don't safe the charges before inserting batteries or doing maintenance, you'll be okay as long as it is held in reset. There were some other quirks; the timing interval pins (8-way DIP switch) is only sampled before the timer is triggered, and cannot be changed while it is timing. The pulldown/pullup resistors for some inputs precluded using BJT's for switching.

The max current output will be 15A, and the normal consumption is around 12mA (42mA for a fraction of a second while firing).

SIDENOTE: If using Rayovac 9V batteries, I suggest looking for the ones with metallica bottoms. Those are the newer AAAA-variety, and can flash more current than the older button-cell variety.

I haven't decided yet what to do about the continuity checking. A current of 1mA will not ignite anything (I think it takes 50mA to ignite a flashbulb, 500mA to ignite an e-match, or something like that). However, I don't want to consume that kind of power for just that, or add more switches. It will already have an 8-pin DIP switch for the time interval, 2 jumpers for configuration, and switches for safing and reset. I might see how little current I can get away with for the smaller LEDs.

I will have these ready for the July launch. Got all the parts, just need to layout the board.

Got the board layout almost squared away. The two-layer version is just a bit larger than I would like it to be. Going to do a 4-layer version to see how much area this saves (larger setup cost though).

Added the continuity indicator. All told, there are three LEDs on the board. One green one by the two battery terminal blocks indicates if the voltage is sufficient (it wants around 8V). A second red one indicates when the timer circuit has been triggered. --This is good to know if you're testing it or installing it. The third indicates continuity for the ignitor it is connected to. Probably going to make it go off when the charges are safed (so that you know that they are for sure).

Changed the crystal frequency to 64kHz; the 128 kHz crystal took too long to stabilize, didn't have the temperature range, and didn't appear to be too sturdy a package. I plan to use these timers for staging later, so just in case I'm trying to use the better components that go up to 125C.

Added some other features. There is a connector so that two or more of these boards can have common power, ground, and reset lines. They can have their own batteries for redundancy, but this falls in line with my philosophy that there shouldn't have to be several batteries just for functionality. There is a capacitor on the board to prevent brown-outs if the power lines get pulled on (e.g., another circuit is firing a charge). The capacitor has enough charge to guarantee one second of operation, and can provide 1.5 seconds best case. The power inputs and reset input have transient voltage suppression to protect against surges and static electricity.

Will have these ready for July 24th launch. I'm going to need them to work because there aren't motor-based charges with a long enough delay for the flight. --Will probably connect two together for good measure.