So. The whole "two e-matches hooked to one altimeter output" thing, and redundancy.

[cvanc]

I've been following kjkcolorado's thread with some interest https://www.rocketryforum.com/showthread.php?133199-Just-torched-my-new-RRC2-what-did-I-do-wrong and I wanted to chime in here about a concept touched on, but not really fleshed out, in that discussion.

"Why do you want to connect two matches to one altimeter output, anyway? What's the point?"

The short answer, I believe, is you never want to do that.

What are you trying to achieve? Redundancy in case one match has a problem? If you're putting both matches into the same deployment charge that seems the only possible goal. But as the other conversation pointed out well, you're just swapping potential failure modes around. You are not making the list of potential failure modes any shorter! It really doesn't look like a clear win can be had, containing new benefits but no new risks.

Besides, this level of redundancy isn't deep, given all the other things that remain as single points of failure:
1- The shared altimeter output channel (as kjkcolorado sadly found out - thank goodness it was on the ground)
2- The programming of that shared output channel (what if you set it up wrong?)
3- The rest of the single altimeter
4- The single battery
5- And the shared wiring harness from battery --> altimeter --> deployment charge

I think a far better answer is this: "Always one match per output channel. Always." Followed with the corollary "Make all items on the single point of failure list truly redundant"

Sounds kinda familiar, doesn't it? Level three regs. The requirements for level 3 certification are true and full redundancy. It's a mandate. Two altimeters sharing nothing - no common wiring, no common battery. Primary and backup charges for apogee, and also primary and backup charges for main. I think this rule exists for a good reason, and I think it is the right way to go for ALL birds above a certain size/weight/complexity level.

I'm interested to hear what others think about this. Feel free to tell me I'm crazy, or to point out some special case where 2 matches on a single output actually make sense. Me? I don't see it.

Be safe out there, everybody!

 

[dixontj93060]

I have had both Quest igniters and commercial e-matches fail to fire when using non-redundant dual deploy (typically in a rocket <4" dia I will not use redundant altimeters). In that situation I now use series e-matches in drogue and main canisters.

 

[CzTeacherMan]

I've been doing independent, redundant altimeters since I started doing DD. Griffin convinced me to, and I never really need it, except for the one time I launched a K300 to 12000+ and when recovering, realized that only one altimeter had functioned. If only for safety, it should be done all the time, IMO. If not for safety, then at least for the economic reason that the extra charges and altimeter pay for themselves after a single incident like mine.
As far as two matches on a single output, it's not something I would consider being redundant. I'm no electrician, but doubling the e-matches means doubling the output required, correct? Why add the stain to an altimeter? If size or cost are holding someone back from doing full redundancy, the Eggtimer Quark $20 solves the need for a cheap backup.

 

[rcktnut]

No doubt the big birds or the ones you really value should have redundancy. My 4" and under DD are all made standard to accept the same sled/alt. setup, non redundant. That I wish to change for some anyway. I never had a failure yet just using one alt./ one e-match per charge set up. Years ago altimeters were pretty expensive, and large, and e-matches went for a buck or two a pop.

Now with the influx of cheap "foreign" e-matches which I tested extensively without any failures I figured why not use 2 per charge as long as the alt. can fire them. I trust my alts., feel that the e-matches are the weakest link, but really have had no reason what-so-ever not to trust them, but feel better having 2 to a charge. I use 9 volt batteries no lipos. Altimeters are cheap enough now too.

So to conclude I personally have been very lucky, and consider that 2 e-matches are better than 1, when not using a redundant system.

 

[jderimig]

It seems like from anecdotal experience that ematches which passed continuity tests but fail to fire is a common failure recovery failure mode. Probably more common than altimeters, batteries or switches failing.

I would offer that the much of the reliability improvement gained by redundant electronics is achieved by simply having 2 matches per event.

So yes, redundant altimeters is probably best (and good for us altimeter businesses), but if you can't do that, then 2 matches per charge is certainly worthy of careful consideration.

 

[dixontj93060]

It seems like from anecdotal experience that ematches which passed continuity tests but fail to fire is a common failure recovery failure mode.

Yep. I've even had one Q2G2 test good beforehand, burn the "pyrogen", and then still test good afterward--without spark to the BP canister and, needless to say, subsequent repairs. So never again.

 

[Worsaer]

It seems like from anecdotal experience that ematches which passed continuity tests but fail to fire is a common failure recovery failure mode. Probably more common than altimeters, batteries or switches failing.

I would offer that the much of the reliability improvement gained by redundant electronics is achieved by simply having 2 matches per event.

So yes, redundant altimeters is probably best (and good for us altimeter businesses), but if you can't do that, then 2 matches per charge is certainly worthy of careful consideration.

John,

Let me ask you the question differently. If I have fully redundant altimeter systems (Marsa54's for example) is there added benefit to also using redundant ematches for each event channel? Or does this actually add additional failure modes?

 

[mpitfield]

It seems like from anecdotal experience that ematches which passed continuity tests but fail to fire is a common failure recovery failure mode. Probably more common than altimeters, batteries or switches failing.

I would offer that the much of the reliability improvement gained by redundant electronics is achieved by simply having 2 matches per event.

So yes, redundant altimeters is probably best (and good for us altimeter businesses), but if you can't do that, then 2 matches per charge is certainly worthy of careful consideration.

I would echo what John says. I believe we all agree that failures can occur due to human error as well as hardware failure. We can also agree that power supplies can fail, altimeters can fail and e-matches can fail for both the aforementioned reasons, human error and hardware failure. What I would like to see is a poll or statistics as to what components fail and the root cause, human or hardware.

My guess would be human is the biggest cause of failure, but this may also vary from component to component. In the case of batteries it may be arguable that it is humans due to not testing, selecting or installing/operating incorrectly. In the case of the switch it may also be human error in that the switch installed was a poor choice or installed poorly based on the flight profile, which is likely also a factor, altimeters I would guess to be human configurations first, then hardware failures second, although this is one area that I would like to see some stats on. However when it comes to e-matches I would guess that it is the one case that may show higher statistics on the manufacturing/hardware failure side.

Bottom line is seeing some factual statistics would be helpful, maybe we can gather some for the purposes of this discussion.

Re. the thread mentioned in this post, I was a contributor to that thread and I am one of those look at everything objectively and try to cover all the bases. I use fully redundant systems, dual altimeters, dual wiring, dual non-mechanical switches, dual power supplies, and dual e-matches in parallel per event.

I realize this is controversial however I am comfortable with my approach as I am inclined to go with human error as the leading cause, and becasue I am essentially extremely anal/OCD I pay a lot of attention and put a lot of thought into my pr-flight prep and choices. Not to say I am perfect, far from it, and have made mistakes, but the point is the answer may not be black and white. Assuming human error is the leading cause there are ways to mitigate issues. Pilots, Astronauts, Engineers are two close examples where they go though checklists or employ protocols to mitigate risk.

Glad to see this thread posted cvanc

 

[jderimig]

John,

Let me ask you the question differently. If I have fully redundant altimeter systems (Marsa54's for example) is there added benefit to also using redundant ematches for each event channel? Or does this actually add additional failure modes?

No at that point I think the reliability return on complexity is not worth it.

Lets say the probably of a dud ematch is 1 per 1000. Using 2 per event would drop the failure probability to 1 per million. Once you are that low it will be difficult to reduce the failure probability by adding complexity. More likely you will be less reliable by doing that.

 

[dixontj93060]

No at that point I think the reliability return on complexity is not worth it.

Lets say the probably of a dud ematch is 1 per 1000. Using 2 per event would drop the failure probability to 1 per million. Once you are that low it will be difficult to reduce the failure probability by adding complexity. More likely you will be less reliable by doing that.

+1.

 

[Worsaer]

No at that point I think the reliability return on complexity is not worth it.

Lets say the probably of a dud ematch is 1 per 1000. Using 2 per event would drop the failure probability to 1 per million. Once you are that low it will be difficult to reduce the failure probability by adding complexity. More likely you will be less reliable by doing that.

Very well put. Thank you John.

 

[Buckeye]

I use fully redundant systems, dual altimeters, dual wiring, dual non-mechanical switches, dual power supplies, and dual e-matches in parallel per event.

Geez, while you were prepping all this, I flew my fully-non-redundant rocket two times! :wink:

I started flying altimeters and DD in 1999 or so, and I can't quite put my finger on when this redundancy craze started.

The few DD failures I have had were my fault, not the altimeter. My only faulty ematches were the ones I tried to make/solder myself.

It never even occurred to me to use two ematches per event until I read a post from Crazy Jim. If CJ does it, then it must be good! (Crazy Jim also inspects the lighters before he sells them to you, which makes me even more confident in those commercial lighters.) So, I ground tested matches in parallel on the Perfectflight MAWD, then flew them once or twice, and all was good.

I'll try the serial thing, or the MARSA54, which has 4 output channels. Or, maybe not. Single altimeters/charges with motor backup has served me well for a long time.

 

[JohnCoker]

It seems like from anecdotal experience that ematches which passed continuity tests but fail to fire is a common failure recovery failure mode. Probably more common than altimeters, batteries or switches failing.

I am skeptical. Of all the things we use, commercial electric matches the the things with the strictest ratings. Sure anything can fail, but these are used by the millions in industry where reliability is critical.

I've burned hundreds of DaveyFire 28Bs and never seen a failure. In fact, I've never seen one not measure 2&#8486;.

I'll bet that improperly prepared charges and nicking wires account for essentially all the "e-match failures."

 

[jderimig]

I am skeptical. Of all the things we use, commercial electric matches the the things with the strictest ratings. Sure anything can fail, but these are used by the millions in industry where reliability is critical.

I've burned hundreds of DaveyFire 28Bs and never seen a failure. In fact, I've never seen one not measure 2&#8486;.

I'll bet that improperly prepared charges and nicking wires account for essentially all the "e-match failures."

Do you think the 5 cent ematches that are popular today are made with the same quality control standards as the 1 buck Daveyfire 28B's?

Data from from the pyro guys who have the biggest sample size and probably use the most consistent quality matches suggest the failure rate around 1 per 1000. Kevin O has data on this, I hope he is reading this thread.

Edit: Real world experience here: https://www.rocketryforum.com/showt...-Electric-Match-at-Amazon&p=725562#post725562

 

[mpitfield]

Geez, while you were prepping all this, I flew my fully-non-redundant rocket two times! :wink:

That is a good point and my biggest source of pain with this one rocket, prep time. If I can I prep the igniters the night before it helps a lot, then when I get to the field it's just a matter of BP, batteries, packing, checklist. This is the config on my 54mm MD CF Tomach, which was my L2 bird that I launched to 16,000 @ 1400'/s and recovered 1.36 miles away. I have 4 launches on this rocket with max numbers of 16,000 @ 2,100'/s.

 

[mpitfield]

Do you think the 5 cent ematches that are popular today are made with the same quality control standards as the 1 buck Daveyfire 28B's?

Data from from the pyro guys who have the biggest sample size and probably use the most consistent quality matches suggest the failure rate around 1 per 1000. Kevin O has data on this, I hope he is reading this thread.

Doing nothing more than unwinding an e-match and firing it off with a 9v battery, out of less than 100 I have had 2 fail. I have no doubt these are the cheap and cheerful ones you are referring to but if the reliability is supposed to be that high on the better ones then I need to source some decent igniters. Recommendations?

 

[jderimig]

Recommendation only based where large sample size reliability statistical data is available. MJG J-Tek, 1/10000 failure rate based on fireworks guys experience.

Others may be fine but I am a hard data guy.

 

[mpitfield]

Recommendation only based where large sample size reliability statistical data is available. MJG J-Tek, 1/10000 failure rate based on fireworks guys experience.

Others may be fine but I am a hard data guy.

I just read that an ATF license is required, https://electricmatch.com/products/see/9/j-tek-igniter is this the same for all e-matches?

They even have a YouTube video [video=youtube;ccqoBPG97go]https://www.youtube.com/watch?v=ccqoBPG97go[/video]

 

[Titan II]

There is more than one type...

https://electricmatch.com/products

The jury is still out on the ones that do not require a license.

 

[bobkrech]

The FET in an altimeter doesn't care if the restive load is is an e-match, a resistor or a light bulb as long it draws less than the FET current rating. Some altimeters use a 5 amp-rated FET which seems a bit low, but it's not that hard to live with as long as you always obey Ohm's law and do not exceed a 5 amp FET load rating.

For an alkaline transistor battery, V = 9 volts, but the internal resistance of an alkaline battery is high, ~1.8 ohms and acts like a current limiting resistor. This resistance is added to the e-match resistance so for with single e-match, the effective load is R = 1.8 + 1.5 = 3.3 ohms and the current thru the FET is 9/3.3 = 2.72 amps. With 2 e-matches in parallel the load resistance is reduced to 1.8 + 0.75 = 2.55 ohms and the current thru the FET is 9/2.55 = 3.5 amps which is still ok for the FET and puts 3.5/2 = 1.75 amps thru each e-match which is also ok. 2 e-matches in series makes the load resistance R=1.8+1.5+1.5=4.8 ohms. The current thru the igniters is I = 9/4.8 = 1.9 amps which is well below the 5 amp limit.

The internal resistance of a LiPo is very low, and will source 10's of amps, so they don't limit current. To keep the current thru the FET to 5 amps, for a 1S LiPo where the V = 4 volts, and a 2S LiPo where V = 8 volts, R >= 4/5 = 0.8 ohms for a 1S LiPo and R >= 8/5 = 1.6 ohms for a 2S LiPo. Since a typical e-match has R = 1.5 ohms, the 2S configuration draws I = V/R = 8/1.5 = 5.33 amps which slightly exceeds the 5 amp limit but is most likely ok due to a slight voltage drop in the FET and altimeter that I didn't include. However put 2 e-matches in parallel and the resistance drops to 0.75 ohms which blows the 5 amps FET in the 2S configuration with 8/0.75=10.67 amps and good bye FET! With 2 igniters in series with a 2S battery pack, the current is I = 8/3 = 2.66 amps which is perfectly safe. In fact, 4 e-matches in a 2S2P H configuration would sonly draw 5.33 amps with 2.66 amps thru each igniter which is probably ok.

For a NiCad or NiMH transistor battery, V = 7.2 volts for a 6S, 8.4 volts for a 7S and 9.6 volts for an 8S battery. Depending on the internal construction of the cells and battery pack, there may or may not be internal current limiting. All versions are safe with 2 e-matches in series, and likely safe with just 1 e-match, however they may not be safe with 2 e-matches in parallel depending on the internal resistance of the battery.

Also a point to remember is that commercial e-matches are designed for serial use in mining where you want to set of a number of charges at once. If your e-match string has continuity you know that all of the e-matches have intact bridge wires so virtually all of your explosive charges will ignite when the when the fire button is pressed. This is not true for parallel connections.

Bob

 

[mpitfield]

The FET in an altimeter doesn't care if the restive load is is an e-match, a resistor or a light bulb as long it draws less than the FET current rating. Some altimeters use a 5 amp-rated FET which seems a bit low, but it's not that hard to live with as long as you always obey Ohm's law and do not exceed a 5 amp FET load rating.

For an alkaline transistor battery, V = 9 volts, but the internal resistance of an alkaline battery is high, ~1.8 ohms and acts like a current limiting resistor. This resistance is added to the e-match resistance so for with single e-match, the effective load is R = 1.8 + 1.5 = 3.3 ohms and the current thru the FET is 9/3.3 = 2.72 amps. With 2 e-matches in parallel the load resistance is reduced to 1.8 + 0.75 = 2.55 ohms and the current thru the FET is 9/2.55 = 3.5 amps which is still ok for the FET and puts 3.5/2 = 1.75 amps thru each e-match which is also ok. 2 e-matches in series makes the load resistance R=1.8+1.5+1.5=4.8 ohms. The current thru the igniters is I = 9/4.8 = 1.9 amps which is well below the 5 amp limit.

The internal resistance of a LiPo is very low, and will source 10's of amps, so they don't limit current. To keep the current thru the FET to 5 amps, for a 1S LiPo where the V = 4 volts, and a 2S LiPo where V = 8 volts, R >= 4/5 = 0.8 ohms for a 1S LiPo and R >= 8/5 = 1.6 ohms for a 2S LiPo. Since a typical e-match has R = 1.5 ohms, the 2S configuration draws I = V/R = 8/1.5 = 5.33 amps which slightly exceeds the 5 amp limit but is most likely ok due to a slight voltage drop in the FET and altimeter that I didn't include. However put 2 e-matches in parallel and the resistance drops to 0.75 ohms which blows the 5 amps FET in the 2S configuration with 8/0.75=10.67 amps and good bye FET! With 2 igniters in series with a 2S battery pack, the current is I = 8/3 = 2.66 amps which is perfectly safe. In fact, 4 e-matches in a 2S2P H configuration would sonly draw 5.33 amps with 2.66 amps thru each igniter which is probably ok.

For a NiCad or NiMH transistor battery, V = 7.2 volts for a 6S, 8.4 volts for a 7S and 9.6 volts for an 8S battery. Depending on the internal construction of the cells and battery pack, there may or may not be internal current limiting. All versions are safe with 2 e-matches in series, and likely safe with just 1 e-match, however they may not be safe with 2 e-matches in parallel depending on the internal resistance of the battery.

Also a point to remember is that commercial e-matches are designed for serial use in mining where you want to set of a number of charges at once. If your e-match string has continuity you know that all of the e-matches have intact bridge wires so virtually all of your explosive charges will ignite when the when the fire button is pressed. This is not true for parallel connections.

Bob

I just love answers like this, strictly factual and very little opinion. Thanks Bob, I am convinced.

 

[Banzai88]

.......and I can't quite put my finger on when this redundancy craze started......It never even occurred to me to use two ematches per event until I read a post from Crazy Jim. If CJ does it, then it must be good!

I think that there are three key circumstances:
1. Encroachment of housing on the launch fields coupled with rocketeer paranoia.
2. A ready supply of cheap, reliable, small, easy electronic altimeters
3. Widely available information from internet forums mixed with a little bit of hero worship

 

[cerving]

It seems like from anecdotal experience that ematches which passed continuity tests but fail to fire is a common failure recovery failure mode. Probably more common than altimeters, batteries or switches failing.

I would offer that the much of the reliability improvement gained by redundant electronics is achieved by simply having 2 matches per event.

So yes, redundant altimeters is probably best (and good for us altimeter businesses), but if you can't do that, then 2 matches per charge is certainly worthy of careful consideration.

Saw one just yesterday with dual altimeters and dual matches/charges... near lawn-dart, main fired just in time to save the rocket. BOTH drogue charges failed, and so did the primary main. Post-mortem, the altimeters reported firing, and all four ematches showed open (yes, they were checked before the flight). Apparently they burned out before the pyrogen lit, or there wasn't any pyrogen in the first place. If it only had one altimeter and two matches per charge, it probably would be buried in the desert... the one that blew open might have cut power to the one that actually lit.

 

[mpitfield]

I think that there are three key circumstances:
1. Encroachment of housing on the launch fields coupled with rocketeer paranoia.
2. A ready supply of cheap, reliable, small, easy electronic altimeters
3. Widely available information from internet forums mixed with a little bit of hero worship

I think 1 and 2 are accurate for me and the first part of 3 is certainly true but the last part of 3 is just a bit silly.

 

[Buckeye]

I think that there are three key circumstances:
1. Encroachment of housing on the launch fields coupled with rocketeer paranoia.
2. A ready supply of cheap, reliable, small, easy electronic altimeters
3. Widely available information from internet forums mixed with a little bit of hero worship

4. Electric matches are cheap and easy to acquire again
5. Thinking that overbuilding means good engineering

 

[BDB]

Can I just ask a question about technique? I'm getting ready to ground test my first DD system this week (I'll be using a Eggtimer TRS with a 9V battery to fire both the drogue and main channels), and you guys have convinced me to use two e-matches in series. Do you simply tie the two e-match leads tighter (possibly with a little solder) or is there some reason to solder a short lead wire between the two?

 

[Dave A]

Unless a rocket is less than 4" in diameter I would never not have a minimum of dual altimeters. With all the expensive FG kits and lots of time building, an extra $50 is a small price to pay for another altimeter. My 6" and larger have dual alts plus an AltAcc2c for 2nd backup drogue and a WRC2 for 2 channels in the nose: That nose is coming off no matter what!
I would not use 2 matches on a channel. If have a failure with 2 alts, it is more likely undersized charges, too tight clearances, bad wiring, etc.