Cluster Launch Power

Discussion in 'Ground Support' started by delta22, May 17, 2010.

Help Support The Rocketry Forum by donating:

  1. May 17, 2010 #1

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    I plan to build a battery+relay electrical launch system to boost the effective delivered current and voltage to fire big clusters. This would be a relay box triggered by the club launch system.

    Looking for input on design.

    When I was firing clusters of 8 or less engines, was achieving about 99% success rate igniting engines with club launch systems.

    Even using creative methods with ematches to reduce total current, the 8 flights I have attempted with 11 to 30 motors have had much lower success rate. Details at www.bpasa.com.

    Voltage sag due to 16 awg launch leads, connectors, and anything less than a very strong battery becomes very significant. While most ematches still seem to fire, some are delayed until the rocket has pulled away from the pad.

    The relay+battery system needs to be able to deliver very high current bursts without major voltage sag. Some of the flights I have done have attempted to draw 60 to 70 amps from a 12V system.

    Components I am considering:
    > 200A automotive relay
    > 8 awg leads
    > Lipo battery pack: 14.8V (4 cell) 5Ah (40C continuous, 50C for 10 second bursts)

    Proposed battery:
    http://www.unitedhobbies.com/hobbyk...Product_Name=Turnigy_5000mAh_4S_40C_Lipo_Pack

    Questions:

    > 200A automotive relay: One reference (at rocketflite.com) advises using a relay with a clamping diode to dampen the voltage created by the relay coil when it is deactivated. If I only trigger my relay box with the club's relay launch system, I would not expect this to be an issue? I also assume that if I did use a relay with a clamping diode, that I would have to determine the polarity of the club's launch leads each time I use them? and would prefer not to do this.

    > Lipo battery: Rated for 200A continuous (5Ah x 40C) it should be able to give me all the current I'll ever need in a 1.3 lb battery. With 14.8V at no load, using extremely heavy gauge leads (jumper cable wire) I would expect to put 12 or 13V into the rocket's ignition setup even for very big clusters.

    > Suggestions about the battery? Relay? Anything else?
     
  2. May 17, 2010 #2

    Rex R

    Rex R

    Rex R

    LV2

    Joined:
    Apr 21, 2010
    Messages:
    5,764
    Likes Received:
    42
    you had better put a fuse/circut breaker between your pack and the relay just in case you get a dead short(I've seen 8ga wire go POOF when someone shorted a pack). 2) what sort of temps will present at your launch? if you expect to see temps below 75F then I would highly reccomend getting 2(or more) 14.8v* (4s) and run them in parallel, lipos don't like running at max rated output when below 75F(they tend to 'bloat')
    *a 14.8v(4s)pack has a fully charged voltage of 16.8v...I would suggest a pair of 3s packs in parallel if your components can't handle 16v.
    rex
     
  3. May 17, 2010 #3

    SCE to AUX

    SCE to AUX

    SCE to AUX

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    903
    Likes Received:
    0
    The damping diode is a VERY good idea if your club launch system incorporates any solid-state electronics. The high voltage spike generated by a large DC relay coil can easily wipe out semiconductor devices like transistors or IC chips.

    If your club launch controller is just a bunch of switches and relays, you are probably OK leaving the diode out. Installing one WILL require you to determine proper polarity for the control signal.




    I personally avoid using LiPo batteries in applications where the weight/size constraints don't call for them. They can be finicky to charge properly, and they have a tendency to fail somewhat spectacularly (fire/explosion) if abused. Good old lead/acid gel-cells are really the sweet spot for this type of application. Cheap, readily available, easy to charge, and they can source plenty of current. Suitable ones are available at big box home centers, as replacements for alarm panels and emergency lights.
     
  4. May 17, 2010 #4

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    Thanks for the quick input.

    I was on the fence about using a fuse. Many launch systems don't use them to remove one failure mode, but shorting is a risk. I'll add one.

    With regard to current loading, I probably won't load system more than 100A for 1-2 seconds at a time, maybe 10-20 times per year.

    Will the collapsing magnetic field create a voltage spike of more than the original 14 or so volts?

    I suppose I could use a much smaller relay to trigger the large relay. Club system triggers my smaller relay, smaller relay closes circuit with known polarity on big relay that has clamping diode.

    This would allow the club launch leads to be attached with either polarity, but still protect club system from sizable voltage spike.

    I do understand that lipo's are more difficult to handle than lead acid batteries. The plusses with lipo's are 10-15 lbs less weight for the cluster box, and 2-4 more volts under load. As long as it works, those are pretty big advantages.
     
    Last edited: May 17, 2010
  5. May 17, 2010 #5

    Rex R

    Rex R

    Rex R

    LV2

    Joined:
    Apr 21, 2010
    Messages:
    5,764
    Likes Received:
    42
    lipos also require chargers that are setup to charge lipos along with a balancer to ensure that ALL the cells have the same voltage. also lipos are not overly robust mechanicaly. while it has been some time since I've heard of a catastrophic failure...they can happen, add to your launch supplies a bucket of dry sand in case of fire(most fire extiguisers are not designed for lipo fires).
     
  6. May 17, 2010 #6

    SCE to AUX

    SCE to AUX

    SCE to AUX

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    903
    Likes Received:
    0
    Absolutely. The amplitude of the spike depends on coil inductance and switching speed, but can easily measure hundreds of volts.

    That should work well.
     
  7. May 17, 2010 #7

    Rex R

    Rex R

    Rex R

    LV2

    Joined:
    Apr 21, 2010
    Messages:
    5,764
    Likes Received:
    42
    personally if I were to go to the trouble of making an electrically isolated launch circut...methinks a 24v setup would be less work. half the amps means you can use lighter wire, ever try to solder 8ga wire? I have, I sometimes use a micro torch.
     
  8. May 18, 2010 #8

    DANLK66

    DANLK66

    DANLK66

    Well-Known Member

    Joined:
    Jan 19, 2009
    Messages:
    151
    Likes Received:
    0
    This is what I made. I got the plans of the net. It has always worked for me.

    Dan
     
  9. May 18, 2010 #9

    Andyd103

    Andyd103

    Andyd103

    Active Member

    Joined:
    May 15, 2010
    Messages:
    28
    Likes Received:
    0
    to launch my 6 engine cluster i use a lawn mower battery 12v, alot of juice in it, real reliable, made my own launch box and it has never failed to launch them all on a full charge
     
  10. May 18, 2010 #10

    Johnly

    Johnly

    Johnly

    Well-Known Member

    Joined:
    Jan 19, 2009
    Messages:
    597
    Likes Received:
    3
    The good news is that the voltage spike produced by the collapsing field is of the opposite polarity of the energizing voltage. The simple fix is to install a high voltage, high current "catch" diode in parallel with the coil so that if becomes forward biased when the spike is generated.

    I suspect that a starter relay would switch 24 volts without a problem, which would help to mitigate the lead losses. That or use Quest Q2G2 igniters in series. 12 Q2G2 igniters would present a total resistance of ~24 ohms, and a 12V system would easily deliver the current needed to have then all fire in about 0.040 seconds.

    John
     
  11. May 18, 2010 #11

    SCE to AUX

    SCE to AUX

    SCE to AUX

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    903
    Likes Received:
    0
    Yes, that's what we have been talking about. Unfortunately, installing a diode across the relay coil then requires the proper polarity be used when connecting the relay coil to the launch controller (so that the diode is reverse biased while trying to energize the relay). In most cases, the pad clips at launch sites aren't labeled for polarity.

    Possible work-arounds for this problem include the use of a small relay (minus protective diode) as a "first stage", which then energizes the larger relay coil (protected with a suitable diode). The downside here is an additional relay to fail.

    You can also use a small fullwave bridge rectifier to develop the proper polarity. Connect the pad clips to the 2 AC input terminals, and put the relay coil (and protective diode) in parallel across the + and - output terminals, with the cathode of the protective diode to the + end of the bridge.

    With this setup, the bridge will "steer" the output to the proper polarity, regardless of the polarity of the input signal. The only downside here is that the output voltage to the relay coil will be about 1.4V lower than the input voltage, due to forward voltage drops in the rectifier bridge. Most relays have a wide enough "pull-in" voltage spec that this doesn't pose a real issue.
     
  12. May 19, 2010 #12

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    Suggestions about higher voltage are very interesting.

    6s lipo battery packs supply about 22V. Two in series would supply 44V.

    Series strings of 4 to 8 Rocketflite MF ematches, or series strings of 3 to 5 ML igniters would fire well at this voltage.

    Current load would be reduced allowing for thinner lead wire, and possibly a single simple relay.

    >>> Any suggestions on a suitable relay that could be triggered by 12V without so large a coil as to cause a troublesome voltage spike on magnetic field collapse?

    It would need to handle 50VDC up to 80A. Relay coil trigger circuit would have to be completely separate from higher voltage ignition circuit.

    >>> What about a solid state relay?
     
    Last edited: May 19, 2010
  13. May 19, 2010 #13

    SCE to AUX

    SCE to AUX

    SCE to AUX

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    903
    Likes Received:
    0
    Going much beyond 24V actually REDUCES reliability, in my experience.

    Once the voltage goes much higher, the igniter bridgewire can sometimes burn through so quickly that the pyrogen doesn't burn, and the motor fails to ignite. The igniters don't so much burn, as explode.
     
  14. May 19, 2010 #14

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    This would be purely for clusters of 7-30+ motors.

    Igniters/ematches would always have to include series wiring so as to provide a 4v to 12v drop per igniter.

    I wouldn't attempt to fire a single 12v igniter with 44v for exactly the reason you specify.
     
  15. May 20, 2010 #15

    SCE to AUX

    SCE to AUX

    SCE to AUX

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    903
    Likes Received:
    0
    SSRs are nice, but you need to use one which is specifically designed for switching DC loads. They are available, but nowhere near as common/cheap as the ones meant for switching AC. Switching DC with an AC type SSR generally results in an output that turns on OK, but then latches ON until the load current is interrupted.

    For "abusive" service like a launcher, you'll want to oversize the relay a bit in terms of current handling. And don't forget the need for a good heatsink!
     
  16. May 29, 2010 #16

    rocket9005

    rocket9005

    rocket9005

    Well-Known Member

    Joined:
    Jan 19, 2009
    Messages:
    85
    Likes Received:
    0
  17. Jun 28, 2010 #17

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
  18. Jun 28, 2010 #18

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    I built the cluster box.

    Started with a small tool box I had not been using, about 14" long. Cut and sanded a 3/8 piece of plywood. Mounted all the electronics to the bottom and stored the wires on top of the board.

    Power is a Turnigy 5.0 Ah Lipo 4S 14.8V battery, picked up from HobbyKing for $65. Battery is rated 40C (200A) continuous, 50C (250A) for 10 second bursts.
    As I tracked the order I discovered they ship from Hong Kong.

    Power wires are Turnigy 10awg, very fine stranded wire with silicone insulation. Low resistance while being remarkably flexible.

    Used a small 40A rated relay to activate the big 200A relay. Big relay purchased from partsgeek.com was 1990 Ford LTD Crown Victoria Starter Solenoid $26.39.

    Bottom line:
    1) complete cluster box weighs about 6lbs (battery 1.3lbs)
    2) club system activates the small relay and is isolated from the rest of the system
    3) a LOT of power is available on the heavy duty black power lines when needed

    1.jpg

    2.jpg

    3.jpg

    4.jpg
     
    Last edited: May 3, 2011
  19. Jun 28, 2010 #19

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    Comparing performance - Actual measured values using 1ohm resistors in parallel.


    Launch system similar to a good club system using relay. 12.6V 22Ah AGM battery with about 10ft of 16awg power leads.

    ohms...volts....amps...watts
    1.00...10.75....10.8.....116
    0.50....9.55.....19.1.....182
    0.33....8.60.....26.1.....224
    0.25....7.90.....31.6.....250


    Cluster box. 14.8V rated, 15.38V measured at no load, 5Ah 40C Lipo battery, with about 8ft of 10awg power leads.

    ohms...volts.....amps...watts
    1.00....14.88....14.9.....221
    0.50....14.40....28.8.....415
    0.33....14.00....42.4.....594
    0.25....13.60....54.4.....740
    0.16....12.8......80.0....1024

    Crunching these numbers shows that a good typical club type system has an internal resistance of 0.15ohms, while the cluster box has an internal resistance of 0.03ohms.

    Comparing systems with a 0.25ohm load, club system has dropped to 7.9V and delivers 250W, cluster box is at 13.6V and delivers 740W.

    When the cluster box is putting out over 1000watts its output voltage is about the same as a lead acid battery at no load.

    I had to measure voltages quickly because the 10w Radio Shack resistors started smoking almost immediately due to the excessive power they were getting.

    (had to put the periods in to space numbers apart in the post)
     
    Last edited: Jun 28, 2010
  20. Jun 28, 2010 #20

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    The high power circuit is protected with a 175A fuse.

    I have no doubt the cluster box can go a lot further than the 80 amps noted above, just ran out of resistors at the local Radio Shack.

    Depending on how I wire the igniters, might draw as much as 100A to fire the 30 motors in the TOGinator pictured in my avatar. Plan on sending that up for its fifth flight in a month.
     
  21. Jun 30, 2010 #21

    Solomoriah

    Solomoriah

    Solomoriah

    Well-Known Member

    Joined:
    Feb 12, 2009
    Messages:
    814
    Likes Received:
    0
    Can you upload a full-res copy so I can read the text?
     
  22. Jul 3, 2010 #22

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    Found a few more 1ohm resistors at another Radio Shack. Tested cluster box output into 10x 1ohm resistors in parallel as pictured.

    Actual:

    With 0.1 ohm load, output was 11.4V.

    Calculated 114 amps current and 1300 watts output power.


    Projected:

    With 0.034 ohms internal resistance for the cluster box as wired, a 0.07A load is about as low as it is practical to go. The voltage would drop to about 10V, output current would be about 150A and power would be around 1500W.

    Lower resistance loads could draw more current, but the voltage drop would be so severe that total delivered power would never get much higher, and system reliability would suffer.

    For example into a 0.34 ohm load, output voltage would drop in half to 7.5V, current 213A and maximum theoretical output power of 1738W is achieved. But the fuse likely blew and the battery's life may have been shortened.

    IMG_3092.jpg
     
    Last edited: Jul 3, 2010
  23. Jul 3, 2010 #23

    JRThro

    JRThro

    JRThro

    Forum Supporter TRF Supporter TRF Lifetime Supporter

    Joined:
    Jan 18, 2009
    Messages:
    4,014
    Likes Received:
    0
    If I remember from college, you get maximum power transfer into a load when the load resistance is the same as the source resistance.

    Even though college was 30 years ago, I seem to have remembered correctly:
    http://en.wikipedia.org/wiki/Maximum_power_theorem
     
  24. Jul 4, 2010 #24

    Handeman

    Handeman

    Handeman

    Forum Supporter TRF Supporter

    Joined:
    Jan 18, 2009
    Messages:
    7,302
    Likes Received:
    29
    Location:
    Stafford, VA
    I used motorcycle/garden tractor batteries for a lot of years, but I recently got a car starter from Home Depot. For $50 you get a 350 Amp starter that will fire any cluster I ever heard of. It charges with at standard extension cord, has charge level meters, an emergency light and USB port to charge anything that charges from powered USB ports.
    http://www.homedepot.com/h_d1/N-5yc1vZ1xh7/R-202019917/h_d2/ProductDisplay?langId=-1&storeId=10051&catalogId=10053

    Works great, is easy to transport, and can start your car if you let your battery run down playing tunes at a launch all day.
     
  25. Jul 9, 2010 #25

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    I stumbled on this outcome as I was modeling outputs into different loads in Excel.

    When I saw this, as it did for you, it stirred memories from electronics courses in college decades ago.
     
  26. Jul 9, 2010 #26

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    Car starters can be a very convenient power source for rocket launches. That is what I am using to power the launch system described above as "typical" of a good club system.

    One factor that is VERY important is the power wiring. In a low-voltage high-current system, voltage drop in the wires is huge.

    MOST of the voltage drop in the values I quoted in previous posts is due to power wires. In both systems the wire resistance is more than half of all the launch system's internal resistance.

    Please note that in post 4, the 12V battery is capable of over 200A output, but the system with 10ft 16awg leads will not deliver more than about 8V at 30A.

    With the 10awg leads, huge relay, and 14.8V Lipo battery, have been able to deliver over 110A at over 11V. I will be firing a 30 motor cluster with this next week, the TOGinator in my avatar and at bpasa.com.
     
    Last edited: Jul 10, 2010
  27. Jul 9, 2010 #27

    RimfireJim

    RimfireJim

    RimfireJim

    Well-Known Member

    Joined:
    Jan 20, 2009
    Messages:
    988
    Likes Received:
    0
    For the kind of current you are running, I don't understand why you don't go bigger than 10AWG with your leads, at least for the majority of their lengths. Flexibility doesn't seem nearly as important as low resistance. Compared to the rest of your system, they strike me as a significant "weak link".
     
  28. Jul 9, 2010 #28

    cjl

    cjl

    cjl

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    12,505
    Likes Received:
    0
    I have to agree here. Although the ratings aren't that relevant for short current pulses, 10AWG wire is only rated for 30A, so I would be strongly tempted to go up to something like 6 or 8 AWG if I were building such a box.

    Even with the 10AWG though, it still seems to be able to source a tremendous amount of current, so it shouldn't be too much of a problem.
     
  29. Jul 9, 2010 #29

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    When talking about short bursts of electrical power in a system where electrical codes do not apply, the key is the resistance of the wire.

    16awg, used in many a typical launch system, has 4ohms of resistance per 1000ft.

    10awg has 1ohm of resistance per 1000ft.

    For 10 ft leads there is a 20 ft round trip for the circuit.

    16awg, 10ft lead, 20ft x 0.004ohms=0.08ohms
    or about 1V drop per 12.5A current due to leads

    10awg, 10ft lead, 20ft x 0.001ohms=0.02ohms
    or about 1V drop per 50A current due to leads

    I was concerned that heavier leads would be very prone to pulling out the rocket's igniters. Wanted wires that could follow the rocket for 3-4 feet as it lifted off.

    The Turnigy (hobby RC) wire I used has very fine strands and silicone insulation, making it far more flexible than any other 10awg wire I have ever seen.
     
    Last edited: Jul 10, 2010
  30. Jul 14, 2010 #30

    delta22

    delta22

    delta22

    Well-Known Member

    Joined:
    Jan 21, 2009
    Messages:
    1,015
    Likes Received:
    0
    As I am building experience with Lipo batteries in general and this one in particular, I discovered that the battery charger had stopped charging previously due to an arbitrary timer and not because the battery was fully charged.

    Ran the charger until the battery was fully charged at 16.8V and ran three more tests:

    Cluster box. 14.8V rated, 16.75V measured at no load, 5Ah 40C Lipo battery, with about 8ft of 10awg power leads.

    ohms...volts.....amps...watts
    1.00....16.0.....16.0.....256
    0.50....15.5.....31.0.....481
    0.16....13.9.....86.9....1208

    I will be flying projects this Saturday that will have igniter resistance loads in this range. One is here: http://www.rocketryforum.com/showthread.php?t=8817
     

Share This Page