L3 Build - Harakiri

[degreaser]

One down. I used an xacto knife to trim. A hose clamp acted as a guide at the fore and aft end. There are a few small divots and pits that will need to be filled. They aren't dry, there was more than enough epoxy to fill them. I'm pretty sure that vacuum bagging would have solved that problem. The fin surfaces are pretty clear of any pits.

 

[Eat, sleep, and Fly]

One down. I used an xacto knife to trim. A hose clamp acted as a guide at the fore and aft end. There are a few small divots and pits that will need to be filled. They aren't dry, there was more than enough epoxy to fill them. I'm pretty sure that vacuum bagging would have solved that problem. The fin surfaces are pretty clear of any pits.

That looks really good! I would think that there would not be a problem putting that up on an N5800:duck:.

 

[degreaser]

That looks really good! I would think that there would not be a problem putting that up on an N5800:duck:.

Should I change the name of the rocket to Hara Kiri now or later? After all, we all know the N5800 is an evil plot by CTI to sell motor hardware

 

[grouch]

Should I change the name of the rocket to Hara Kiri now or later? After all, we all know the N5800 is an evil plot by CTI to sell motor hardware

With or without the N5800, I still think you should call it Hara Kiri. That has to be the coolest name for a rocket I have heard. If not that then Seppuku. The carbon looks great by the way.

 

[CarVac]

Write it on the flight card in kanji and watch the LCO try to read it...

腹切

 

[degreaser]

Two down. Some finishing will be required but this set turned out better than the first.

 

[degreaser]

With or without the N5800, I still think you should call it Hara Kiri. That has to be the coolest name for a rocket I have heard. If not that then Seppuku. The carbon looks great by the way.

Ok, done. I was thinking of Ichiwawa (Hot Lady) but Harakiri is good too.

I've been naming my 3" rockets Alpha, 4" Beta, 5" Gamma, 6" Delta, etc. Alpha 1 was my L1, Alpha 2 was my L2. Beta 1 was actually my AMRAAM 4 build. Gamma 1 is actually my HoJo scratch build which is still in a partial vaporware state.

 

[DRAGON64]

That Carbon Fiber tip-2-tip turned out really nice... and Harakiri is a unique name.

 

[Eat, sleep, and Fly]

Two down. Some finishing will be required but this set turned out better than the first.

That looks good, but the one word of advice I have is don't CF the bevels on the fins. End the layup at the beginning of the beuel, and coat the bevel in a high temperature epoxy (proline 4500, contronics). This works as an ablative, and helps prevent delamination.

 

[TheSamurai]

That looks good, but the one word of advice I have is don't CF the bevels on the fins. End the layup at the beginning of the beuel, and coat the bevel in a high temperature epoxy (proline 4500, contronics). This works as an ablative, and helps prevent delamination.

Given the meaning of hara-kiri are you sure that's what you want its name to be? lol

 

[degreaser]

Given the meaning of hara-kiri are you sure that's what you want its name to be? lol

That thought went through my mind as well.

 

[degreaser]

That looks good, but the one word of advice I have is don't CF the bevels on the fins. End the layup at the beginning of the beuel, and coat the bevel in a high temperature epoxy (proline 4500, contronics). This works as an ablative, and helps prevent delamination.

I think that was what Tony was alluding to as well. If I had xray eyes, I would have taped the bevel. It's a little late though as it's 90% cured.

Edit - between the name change and this suggestion, now I'm paranoid

 

[degreaser]

3 and 4 down. The first image is from last night, the second is tonight's layup.

 

[TheSamurai]

That thought went through my mind as well.

I was a missionary in Japan. Love Japanese and the Japanese people. We always joked in the mornings where we were tired that we would just commit hara-kiri that morning and be done. lol. In my head I can only think of your rocket committing suicide and gutting itself. lol.

 

[degreaser]

I was a missionary in Japan. Love Japanese and the Japanese people. We always joked in the mornings where we were tired that we would just commit hara-kiri that morning and be done. lol. In my head I can only think of your rocket committing suicide and gutting itself. lol.

Hopefully so - once at apogee and again at around 700AGL

 

[degreaser]

Now that I can find the mill again, I'm back to a more familiar medium - aluminum.

It's time to begin to concentrate on the avbay builds. I'm getting started on the Contour mounts. I'm going to try a couple of different designs starting with the model as shown in the first image. This one mounts on the interior of the avbay. I'm also contemplating one that would mount on the exterior of the airframe. TBD.

I'm starting with a 3.5" length of 1/2x4" 6061 stock. In preparation for mounting to the milling fixture, I've drilled the standard pattern used for this avbay. A quick test fit with a bracket I machined earlier verifies that I turned the knobs properly. The center portion between the two center holes will eventually be milled out allowing the camera to be properly positioned longitudinally within the avbay.

I need to work up the gcode and do some tooling changes. This milling operation will require a ball endmill to mill out the radius for the camera. The 1/4" endmill I was using for the earlier machining is showing some wear so I'll probably change that out as well. Both will require updating the linux EMC2 tool table configuration.

 

[degreaser]

Spent the day at Mudroc yesterday. I presented the fincan to Tony for inspection. He stated that it's bigger than he thought. He referred to it as "large boned" which I take as meaning it's fat (it weighed in just under 20lbs). However, I take this with a grain of salt from someone who named his primary rocket after a hyperactive mouse (Speedy Gonzales). He said he thought it should handle an N. I've updated the Rocksim model and I have better than 1.4 calibers of stability regardless of the motor I plan on using. That should improve as I add more fat to the rocket.

I didn't fly anything but I did go shopping and supported the local vendors. I got an 84" Iris, 18" drogue and a couple of nomex blankets from Gene for the Gamma L/Harakiri and some EX stuff from other vendors. Tony supplied me with some raw 38mm and 54mm motor casing materials to make some EX motors. He also loaned me a couple of gorgeous 38mm and 54mm nose cone mandrels. I think this is a subtle hint....

I've worked out the basic gcode for machining the first Contour mount but I need to work up some coordinate transforms and get some more fasteners to hold it on the fixture. It's 3 roughing and 4 finishing passes.

 

[TheSamurai]

Hopefully so - once at apogee and again at around 700AGL

Hopefully so. lol

 

[degreaser]

I've made some progress on the avbay assembly. I drilled a bunch of mounting holes in the brackets which took the better part of the evening to complete.

I've also received my boards back from the fab. They look great but I'm going to have to do a turn on it. I found during breadboarding that the hall effect device comes up in a state upon connecting the battery such that the altimeters are powered up. This is a major safety issue and must be fixed. Normally I breadboard before committing to a layout but I was in a hurry... I'll probably populate one of the boards for testing but this rev won't fly. 'Nuff said.

I mounted a few of the components to get a visual idea of what it will look like when completed. The blue connectors connect to the main bulkhead on the top of the avbay and the drogue bulkhead on the bottom. They are locking connectors so there's little risk of anything becoming unplugged during flight. The telemetrum is on the top of the sled, the raven on the bottom. Each has it's own battery although I've only strapped in the LiPo for the telemetrum. The sled can be moved easily to any 3" or larger diameter rocket by removing 4 screws and disconnecting two plugs.

The Contour will mount toward the bottom of the avbay.

 

[dixontj93060]

I've also received my boards back from the fab. They look great but I'm going to have to do a turn on it. I found during breadboarding that the hall effect device comes up in a state upon connecting the battery such that the altimeters are powered up. This is a major safety issue and must be fixed. Normally I breadboard before committing to a layout but I was in a hurry... I'll probably populate one of the boards for testing but this rev won't fly. 'Nuff said.

Can you share a bit on your PC board approach, such as supplier/vendor, typical costs, turnaround, required input files, etc.? Or maybe there is a link you can point us to?

 

[JP Morgan]

:drool:
AWESOME work dg!

JP

 

[degreaser]

Can you share a bit on your PC board approach, such as supplier/vendor, typical costs, turnaround, required input files, etc.? Or maybe there is a link you can point us to?

Sure. How much detail? Would you like a design process overview?

 

[dixontj93060]

Sure. How much detail? Would you like a design process overview?

No, I understand basic steps from doing that type of work years ago. More interested in software, handoff requirements, vendors and costs.

 

[degreaser]

No, I understand basic steps from doing that type of work years ago. More interested in software, handoff requirements, vendors and costs.

Ok, for schematic capture and layout I'm using Eagle from https://www.cadsoftusa.com/. They have a free version available. The Altus Metrum guys are using gEDA from https://www.geda-project.org/. The latter is totally free and is under GPL. It's been a few years since I've used gEDA so I'm sure it's matured. There's also Kicad at https://www.kicad-pcb.org. For verification, I'm using the ERC and DRC tools available in Eagle configured with the specific rule sets from the fab house. Once my design is complete, I export it to Gerber 274x using the CAM tools in Eagle. Each fab has slightly different requirements w/r to DRC rules and what Gerber layers are required.

For fab, I use OSH Park at https://oshpark.com/. Batch PCB was ran by the Sparkfun guys but it has closed (probably due to competition from OSH Park). You can get the CAM and DRC files from their site.

Submission is simple - create a zip file containing your Gerber files and upload it to the OSH Park site. They also accept Eagle layouts but I prefer Gerber. During the submission process, they'll run a DRC on the fly. Then do a final sanity check and pay for it with paypal. Their rates are $5/square-inch for 3 copies of a two layer board, no setup fees, free shipping. Panelization takes about a day. They submit around 3 panels per week. Turnaround is two weeks or better. Their boards are purple and they are produced domestically.

Batch PCB had slightly better rates but had a set up fee which made it impractical for small batches. They also had a 4 to 5 week turn as the boards were produced in China.

 

[dixontj93060]

Great, thanks!

Ok, for schematic capture and layout I'm using Eagle from https://www.cadsoftusa.com/. They have a free version available. The Altus Metrum guys are using gEDA from https://www.geda-project.org/. The latter is totally free and is under GPL. It's been a few years since I've used gEDA so I'm sure it's matured. There's also Kicad at https://www.kicad-pcb.org. For verification, I'm using the ERC and DRC tools available in Eagle configured with the specific rule sets from the fab house. Once my design is complete, I export it to Gerber 274x using the CAM tools in Eagle. Each fab has slightly different requirements w/r to DRC rules and what Gerber layers are required.

For fab, I use OSH Park at https://oshpark.com/. Batch PCB was ran by the Sparkfun guys but it has closed (probably due to competition from OSH Park). You can get the CAM and DRC files from their site.

Submission is simple - create a zip file containing your Gerber files and upload it to the OSH Park site. They also accept Eagle layouts but I prefer Gerber. During the submission process, they'll run a DRC on the fly. Then do a final sanity check and pay for it with paypal. Their rates are $5/square-inch for 3 copies of a two layer board, no setup fees, free shipping. Panelization takes about a day. They submit around 3 panels per week. Turnaround is two weeks or better. Their boards are purple and they are produced domestically.

Batch PCB had slightly better rates but had a set up fee which made it impractical for small batches. They also had a 4 to 5 week turn as the boards were produced in China.

 

[Derek]

That is a really neat idea to have a pcb fabbed for your av sled.

I use oshpark too and so far I've been really happy with the results. I didn't realize batchpcb has closed. They were the other hobbyist/small run pcb vendor. too bad.

 

[Green Jello]

Amazing build. I hope I get to see this at XPRS.

 

[degreaser]

Thanks guys

I use oshpark too and so far I've been really happy with the results. I didn't realize batchpcb has closed. They were the other hobbyist/small run pcb vendor. too bad.

Yep. One of the boards had a short. Out of the one of the three I selected to populate, it was the one with the short (Murphy). I thought it was my screwup but it wasn't. I tracked it down by using a lipo to cause the trace to overheat. Sure enough, it started to smoke on the bottom of the board between a trace and ground pour at a 90 degree bend. This was of course after I had populated the board with parts. This rev definitely won't fly I've had this happen once with Batch PCB and once now with OSH Park. For this volume I/they can't afford to do bed of nails testing but I will definitely probe all future boards prior to soldering parts. I'm sure OSH Park would stand behind it if I sent a defective board back unpopulated and untouched.

Yeah it's a shame about Batch PCB. Competition is always a good thing.

 

[Derek]

wow. I guess that is one reason they say don't use 90 degree bends! how many mils of spacing did you use?

So far I haven't had any issues like that with 12 mil spacing. I did have an issue with the silk screen shifted, but it didn't cause a functional problem.

 

[degreaser]

wow. I guess that is one reason they say don't use 90 degree bends! how many mils of spacing did you use?

So far I haven't had any issues like that with 12 mil spacing. I did have an issue with the silk screen shifted, but it didn't cause a functional problem.

I've been using Sparkfun's design rules which allow for 8mil spacing and 8mil traces. OSH Park allows 6/6. Stuff happens. As I said, from now on I'll probe all pads to ground and Vdd as well as adjacent pads on ICs before I apply paste, parts and solder.