Test Stand Ignition Circuts

The IREC team I'm working with, well, they just cobble together things and I'm not comfortable with the amount of safety built into their system. They had multiple pieces of speaker wire run from the test stand to their command center, probably 18 gauge. They used the old trusty touch the stripped ends to a 12V battery method. Plenty of places for wires to get crossed.

They have three devices they need to actuate - fill disconnect (pyro-mechanical), preheater grain ignition, propellant valve actuation (pyro-mechanical), in that order.

I've made a relay actuated system that was a single channel - it used a 9V battery that at 150 feet of speaker wire that actuated a relay that used a 12V car battery near the motor. This is in the same train of thought, though I'd use 18/4 sprinkler wire for the cable and 12V batteries at both ends.

I've attached a schematic - drew it online via Digikey.

View attachment Relay Controller.pdf

S1, S5 and S7 are toggle switches with a cover.
S9,S10 and S11 are momentary push button switches.
These are at the command box.

S2, S6 and S8 are toggle switches with a cover, at the pad box.

S4 is a key switch at the command box and S3 is a key switch at the pad box.

The lamps represent the igniters.

In operation there would be an approximately 25' 14 guage extension cord that plugs into the pad box, then to each igniter.

You would toggle all the switches on, then with confirmation the Command key switch is off and toggles are off, you would turn the pad master arm. That way if anything is welded shut you'd be ~25' from the action. I'll probably also have an actual lamp or LED across each output so that you can test for that condition by turning the key and toggle as a diagnostic before a test fire.

Then, you would arm the command box, then flip each toggle and momentary button as required.

Anyone see any glaring issues? I'm trying to keep this as simple and as robust as possible because things tend to get abused

Edward

Hi Edward

In keeping with your desire for safety with double switching, I would add a 4th relay circuit at the pad box in series with the pad battery. This relay would be switched with a momentary action switch in series with the S4 key switch at the command box. This would allow you to arm the pad box remotely from the command box. I would probably remove either the S1, S5, S7 group or the S9, S10, S11 group, as you now have a common arming switch.



You could probably use 8-way network cable between command and pad boxes as this is a relatively low current circuit. I would probably also use illuminated switches at the command box.

All the best!

I like that idea better - make sure that you have to press two things to fire a circuit rather than flip a toggle and fat finger it. Illuminated - whoa, do you think I'm made of money?

Edward

Wooh! IREC! Good too see the frantic cobbling continues (and that cooler heads are around to suggest otherwise lol)

What disciplines are your students from? If they're straight mechanical or aero, they may not be comfortable designing such a controller, so they opt to go simpler. Or they're running out of budget.

The team I'm working with is straight ME. They have appealed for two years to have 2-3 EE's for the controls system, but haven't been successful. They are out of time and didn't put it in their schedule to use any sort of controller.

I just priced out everything except wires and enclosures and it comes to ~$175. Ugh.

Edwar

I guess everything worth doing is worth overdoing. There are 8 contacts that need to be closed to fire a single igniter; that's probably enough to impress a nuclear weapons designer.

Seriously though, that is well beyond the point of diminishing returns and may already cause reliability issues. In particular, there are many switches that, if they fail unnoticed, will get the test sequence stuck after some pyro actuators are already expended.

My 2 cents:
1) The extension cord is a good idea. Make it a mandatory part of the procedure, that it becomes unplugged at the pad box before anybody is near the engine. Compared to the switches S2, S6, S8, S3 this can also be trivially verified visually from a distance.
2) Add a loud buzzer, that gets activated as soon as the arm relay gets closed. This will warn people before they connect the extension cord, helping not only against stuck relays, but also against certain kinds of user error.
3) If there are no scenarios were it is really important to only arm particular channels, S9, S10, S11, S2, S6 and S8 will probably do more harm than good. Imagine forgetting S8 in the wrong position when executing a test. Currently there is no check in the schematic that will reveals the state of this switch to the user. I would remove them.
4) Safety wise, the arm switch and S4 are redundant. In some systems it is desirable to be able to power it up without arming it, but this doesn't appear to be the case with your system. Therefore S4 could be removed too, but it is less problematic then the switches from 3).
5) Consider adding a continuity check. In your case, that's not only a convenience feature, but also helps to prevent expending pyro actuators when the system is not completely ready.

Reinhard

Reinhard - thank you for the feedback! Do you have an example of a continuity circuit that would work? 99% of the igniters that they use are e-matches, so it would be good to make them low current. I'm envisioning an LED in series with the e-match and a resistor to limit current and a momentary push button.

I'm going to revise the schematic based on your suggestions - I appreciate the feedback.

Edward

Hi,

here is a minimum example with continuity check. Not shown are the connectors of the extension cords and between controller and pad box.


Some notes:
The resistors R1-R3 should be in the pad box, not next to the LEDs. Otherwise the e-matches could be fired by, for example, a short on the continuity signals as soon as it gets armed. Avoid using tiny resistors that can be shorted easily by even a small metal shaving. If you do (e.g. SMD construction) consider making the resistors redundant. In a 12V system, the resistors should be around 600-2400 Ohms, depending on the type of LEDs used.
The amount of wires between the boxes is limited to the minimum (1 common ground + 1 wire per signal). This could be built with a cheap Ethernet cable for example, although the connectors are kinda flimsy. If you can spare another wire, the battery in the controller can be eliminated. D-Sub connectors are examples for widespread connectors with 9 pins. 100ft RS232 extension cables can be had for less than 20$.
You can add additional redundant switches if you wish, but I wouldn't. The arming switch already provides redundancy. Any additional effort is imho better spent at focusing on the operating procedures (e.g. testing equipment before use).

Reinhard

Remember the equipment is only part of the story. You could also allocate one person as Safety Officer who enables the master switch. They do not need to have intricate knowledge of sequencing or other requirements, just double-check that the required area is clear and safe. If that is their only job then things will be less likely to go wrong.

I do like your idea of having a separate key at the launchpad end of things. Last person leaving the pad area arms it.

Just thought about it a bit more. Being a test stand you might be able to do away with some of what I mentioned (not such a long distance from the pad to the launch instigator, and I suspect a wall or something in the way). I suspect it is a good idea though. Just have a think about what is really necessary, considering off-nominal actions by the people involved

All - I've now expanded this to five (5) circuits to help future proof it. I don't want them to invest this much and then kludge something together. If I'm bringing back a common ground for continuity to the LCB (Launch Control Box), I can now eliminate the battery at that end, I believe.

I'm going to edit the circuit and then start choosing parts. Before I was choosing ones that had screw terminals because I really like the flexibility, but with the complexity of this, I'm probably going to make two PCB's at each end and solder all the components in to the boards, and then use terminals where I can for connections.

I'll let you know how it goes.

Edward