MAWD crash with strange data

Hello folks,

Technically this is my first thread, so bear with me on this post. I crashed one of my good rockets on a supersonic flight to 9200 ft this past weekend and I'm trying to troubleshoot. The thing is, the data I got back from the unit is a bit strange. First, the peak altitude reported is 20,000 ft - I know I didn't break 10,000 ft. Also, at the 39.25 second mark the data makes a big jump in altitude (while descending) and looks exactly, and I mean exactly, like the previous flight from that point.

I know that the MAWD fired the apogee charge, but the main failed to deploy. After the hard landing, the unit was "off" and not beeping. I opened what was left of the ebay, and it started up. I know you are thinking "loose wire", but I was unable to find any and the switch (aerocon PCB style) was still tight, 9.07V on the battery too.

Prior to the crash I had 53 dual deploy flights with this unit, and only 3 hard landings (13 successful flights since the last hard landing). I'm well experienced with dual deploy, with 130 of my 153 HPR flights being dual deploy and I'm L3.

I've attached the files of interest (.pfd changed to .txt). I'd appreciate any help - I suspect that the unit lost power mid-flight, but I'm not sure why.

Sparky

View attachment Critical Velocity Crash K805G.txt

View attachment Critical Velocity K185W Moon Burn.txt

Unfortunately I know what the 20000 means to me....I have a ebay that doesn't seal too well, when I don't add a piece of tape to help it, it tends to let some of the ejection charge pressure in which usually gives me that number.

I think your idea about loosing power is correct. if the wires are not loose, how about a internal broken wire? This could be the culprit especially if you use solid wires - if it's broken, the ejection charge (or even just the bouncing around our rockets do on recovery) may have been enough to allow it to power down. When you picked it up and opened up the ebay it moved the wire in a position that the internal broken wire was touching again...

just my thoughts, I could be completely wrong.

I can't view your file at the moment, but the first thought that comes to mind is a hard deployment that caused the battery to disconnect internally.

What brand of battery did you use? I'm assuming it was a conventional 9V?

Unfortunately, the MAWD doesn't have an accelerometer, so we cannot see the G load on the system at deployment.

Many batteries use spring-loaded contacts internally, and enough force can cause the contacts to flex, breaking contact. An intermittent power blip can sometimes cause electronics to act strangely.

-Kevin

What was the mach inhibit delay set for?

Bob

troj said:

I can't view your file at the moment, but the first thought that comes to mind is a hard deployment that caused the battery to disconnect internally.

What brand of battery did you use? I'm assuming it was a conventional 9V?

Click to expand...

You might want to append to your post some information about brands of 9V that are particularly prone to or immune to this problem. Even if the original poster is aware of issues, others reading this could benefit from the expertise, since the topic was raised.

At or about 39 seconds, your altimeter power was interupted, and the unit shut-off in a non-standard manner. From that point on, you are looking at the previous flight data, and other garbage that was in the memory. The evidence is below.

1.) The mach transitions are clearly visible in the boost phase at ~0.5 seconds, breaking mach, and at ~1.6 seconds, coming down from mach.

2.) An apogee charge fired at ~21.6 seconds into the flight.

3.) Power to the altimeter dropped out a few seconds before 38.4 seconds when the altimeter stopped working. Power is sustained for a few seconds by the anti-brownout capacitor.

4.) After that point, you are looking at the previous flights data that was not overwritten. The clear evidence for this this that at ~121.5 seconds, you see a main deployment, and you clearly state that this did not occur.

5.) Since the flight computer did not control the powerdown as it would at the termination of a normal flight, your simply seeing whatever was left over in the data memory, and by pure coincidence the previous flight went to approximateley the same apogee.

Bob

FWIW, got (vertical) velocity and (vertical) accelertion graphs from the good data set.

Acceleration is for entertainment only. Rendered it with and without low frequency noise.

-LarryC

Larry

Very nice. Is it possible to do this for the failed flight for the first 38 seconds?

I'm wondering if the airframe sections had a violent collision after apogee that dislodged the battery?

Bob

Thanks for the help. I pretty much figured that it powered down mid flight. Its plausible that it had a violent collision while descending, but usually a 20ft streamer of caution tape helps keep it separated. However, the upper level winds were 50kts+. I did notice a significant mark on the airframe, but that easily could have happened on the hard landing. I used the batteries they sell at Radio Shack now - does anyone know if these are prone to shock failures?

Also, all my supersonic flights seem to start underground (negative altitude values). Is that the Mach transition or something caused by the excessive G's?

I'll post pics of my altimeter setup. Thanks for the help everyone.

-Sparky

SparkyVTFlyer said:

I used the batteries they sell at Radio Shack now - does anyone know if these are prone to shock failures?

Click to expand...

Last time I checked a RS 9V battery it would only source about 2A. That made me think it was press-fit and did not have soldered cells.

SparkyVTFlyer said:

Its plausible that it had a violent collision while descending, but usually a 20ft streamer of caution tape helps keep it separated.

Click to expand...

I think that it is more likely that the power was disconnected during the apogee shock and the altimeter continued to operate from the capacitor. How long will the MAWD run after disconnecting power? The stated figure is 2 seconds but I bet it is closer to 20.

bobkrech said:

Larry

Very nice. Is it possible to do this for the failed flight for the first 38 seconds?

I'm wondering if the airframe sections had a violent collision after apogee that dislodged the battery?

Bob

Click to expand...

I'll do it tomorrow first thing.

Photos attached. The sled and electronics didn't sustain much damage, but the rest of the rocket did. Either way, I haven't touched any of the hookups other than the main in the photos.

-Sparky

SparkyVTFlyer said:

Also, all my supersonic flights seem to start underground (negative altitude values). Is that the Mach transition or something caused by the excessive G's?

-Sparky

Click to expand...

Mach Transitions are pressure fluctations that occur as shock waves move to toward the NC as the rocket accelerates from subsonic to supersonic speed and from the NC aft as the rocket deccelerates back down again. They have nothing to do with high g loading.

When you accelerate from subsonic to supersonic speed, there is a compression wave indicated a a negative attitude followed by a lower pressure rarefaction wave. When you deccelerate from supersonic speeds the inverse pressure trends are observed, first the rarefaction wave is observed as a pressure drops as indicated by an apparent increase in altitude followed by a compression wave indicated as a drop in altitude.

The attached altimeter trace from your data shows the mach transition on your supersonic flight and the lack of them on your subsonic flight.

Bob

Thanks. I had always wondered about those negative values. I thought that maybe the G's were adversely affecting the baro sensor right at liftoff and causing the initial negative altitude values.

I'll definitely take the advice on the batteries. Which ones are the most resistant to shock? I had been using Duracell's prior to this, but they had a bunch of 9V batteries at Radio Shack for pretty cheap. I guess I get what I pay for.

-Sparky

SparkyVTFlyer said:

I used the batteries they sell at Radio Shack now - does anyone know if these are prone to shock failures?

Click to expand...

Duracells have welded connections. Energizer is pressure fit with padding under the end cap. I haven't disassembled a Radio Shack 9v. I use the Duracells.

Here are the curves for the crashed flight. I had avaoided posting them because, as expected, most of the features are artifacts of the Mach transitions. The two peaks in the velocity curve come from that source. The whole shape of the first part of the acceleration curve (before ~5.3 seconds) is also bogus.

Hi Sparky,

On review of your file "Critical Velocity Crash K805G.txt" the text at the top of the data shows;

Apogee: 20000
NumSamps: 04800

Per the MaWD's firmware this indicates a power loss in flight. Copy of the firmware revision history attached for info - last dot point in Firmware 1.20

As to why power was lost, that's a more difficult question.

I had the exact same scenario take place, power loss in flight. When I reached the rocket there was silence and the switch was still on.

My issue was we ran a single chute and I had inadvertently connected the one chutes ejection charge to the "Main" terminal and not the "Apogee" terminal so had a very late deployment at high speed on descent. I believe that this imparted a large force back into the rocket and caused the power loss.

The data reading really confused me as it too read 20,000 Ft and 4,800 samples. When I plotted it, it also showed past history. Bobkrech discuss this in points 4 and 5 of his message above - not normal powerdown.

Note: 4800 samples = 240 seconds

Next to the firmware revision below are graphs of your and my MaWD data, very similar. I believe if you look back in your flight history you'll have graphs to match the spikes captured in the memory at 150 seconds and 175 seconds.

I replaced the switch, battery and flew my MaWD later in the day and it worked perfectly !

I do suspect my switch to be the cause for my power loss. However, if I had not have incorectly wired my chute, I would not have had the event.

Hope the above helps.

Regards,

RobW

Larry Curcio said:

Here are the curves for the crashed flight. I had avoided posting them because, as expected, most of the features are artifacts of the Mach transitions. The two peaks in the velocity curve come from that source. The whole shape of the first part of the acceleration curve (before ~5.3 seconds) is also bogus.

Click to expand...

Thanks Larry. It looks like there was some rotation after the apogee ejection.

SparkyVTFlyer said:

Photos attached. The sled and electronics didn't sustain much damage, but the rest of the rocket did. Either way, I haven't touched any of the hookups other than the main in the photos.

-Sparky

Click to expand...

Sparky

I see 2 issues with the way your battery is mounted.

1.) Your battery connector is mounted pointing forward. On high g-rockets, the connectors should be aft, so as the rocket accelerates off the pad, the connections get compressed and are made tighter, not looser.

2.) You used only one wire tie to secure your battery to you sled but you didn't use a battery holder. Your battery needs to be mounted in battery holder so that it will not rotate. The battery rotated and pushed against the bulkhead and disconnected one contact.

A battery holder that contains the battery and prevents rotation is shown below. Coupled with 2 wire ties it won't more. Mounting the battery perpendicular to the direction of motion against a bulkhead in the rocket will also eliminate any major force attempting to dislodge the battery.



https://aeroconsystems.com/cart/products/Nylon_9_Volt_Battery_Holder-241-4.html


Bob

1.) Your battery connector is mounted pointing forward. On high g-rockets, the connectors should be aft, so as the rocket accelerates off the pad, the connections get compressed and are made tighter, not looser.

Click to expand...

Bob,

My ebay is setup to compress the connections at liftoff. My fault for not labeling which end was aft. Either way, I agree with your assessment of the battery setup - I definitely could secure it better. I'll look into mounting the battery holder you suggest.

I'm pulling that 9V battery apart tonight and inspecting it. I'll post what I find.

-Sparky

SparkyVTFlyer said:

Bob,

My ebay is setup to compress the connections at liftoff. My fault for not labeling which end was aft. Either way, I agree with your assessment of the battery setup - I definitely could secure it better. I'll look into mounting the battery holder you suggest.

I'm pulling that 9V battery apart tonight and inspecting it. I'll post what I find.

-Sparky

Click to expand...

I recommend Gorrilla Tape for keeping the harness on the terminals. Whatever other stuff you're doing, it adds to the security.

bobkrech said:

1.) Your battery connector is mounted pointing forward. On high g-rockets, the connectors should be aft, so as the rocket accelerates off the pad, the connections get compressed and are made tighter, not looser.

Click to expand...

I use connectors of the same type as SparkyVTFlyer so this post makes me worried. I dont rememeber which way i mounted the batteries in my currect project now but I remember thinking to myself - This must be the ultimate battery connector, how can it possibly come loose. Being so light it will easily move along with any possible movement of the battery even if it is mounted with the batteryterminals pointing forward.
Please dont tell me Im wrong but is it not harder to keep a heavy battery firmly connected to terminals that are bolted to the sled?

B.r. Niklas

Keystone makes a number of battery holders. They are available from most electronic distributors. https://www.keyelco.com/index.asp Accessories include battery holding clips for the battery holders.

Bob

WillMarchant said:

Last time I checked a RS 9V battery it would only source about 2A. That made me think it was press-fit and did not have soldered cells.

Click to expand...

The ones from radio shack vary, but make sure to ask.
Dont use energizer batteries either, they are press fit.
Tim From apogee says that duracell batteries have all their conections soldered