Pressure Trace II - finding a suitable barrel

[Benchrst]

Want to sell the Encore barrel?

[Geeman]

Nope... Too purdy



The gage...



The wires/connectors are fairly delicate.  They gave my big mitts fits.  I wish they would have used plain old phone connectors, positive and about 10 times the size.

Greg

Greg

[Benchrst]

WAY too purdy!

[Ramjet]

Oh boy that is really really nice......

[REDLINE]

Gorgeous!

What barrel length is that?

[Geeman]

Thanks everyone.

That is a custom 9" barrel.  Even the lowly PPU ammo hits 1200 fps.  2" at 100 yards is possible with well assembled ammo, but crimping in a separate step is necessary.

Greg

[sqlbullet]

Those strain gauges are $5/each on ebay in 5 packs.

I think you could add a MCP3008 A/D converter and a wheatstone bridge and add in some python code on a Raspberry PI and have one going for less than $100.

Gotta figure out a way to have time to play with these kinda things.

[Geeman]

Those strain gauges are $5/each on ebay in 5 packs.

I think you could add a MCP3008 A/D converter and a wheatstone bridge and add in some python code on a Raspberry PI and have one going for less than $100.

Gotta figure out a way to have time to play with these kinda things.

A few considerations...

Something around .6 ms beginning of ignition to out the barrel. 

There needs to be quite a few data points if you are going to attempt to plot the curve. 

There would need to be a trigger to begin the event so data didn't overflow the ability to keep track of it.

Greg

[sqlbullet]

That math is spot on for a 180 grain at 1300 fps from a 5" barrel.

The worst case scenario is short barreled magnums.  A 44 magnum firing a 170 grain bullet from a 2" barrel only gives .33 ms sample window. 

In the code I am envisioning I will use a noise floor function.  Once a value exceeds the noise floor, I will capture a timestamp, and then on each subsequent sample, the clock value and magnitude until the value descends below the floor.  The entire array will get dumped to database once the "event" is over.  I have written similar code for another project, just at a much, much more glacial sample rate.

The first issue though is the MCP3008 only runs 200,000 samples per second at 5V.  That means for a .6ms window, I can capture 128 samples.  In the 44 magnum worst case, I will get only 67 samples.

The second issue I am going to have to deal with is the MCP3008 is only a 10bit AD converter.  If I assume a 100,000 PSI ceiling, then I will only be capturing to the nearest 100 PSI.

Here is a mock trace from data I pulled from a 6.8 PPC load adjusted to 67 samples over 1ms with a resolution of 100 fps.  This is still very usable.



I just need to get some sensors, build the bread board, write some code and get to the range.  Good news is I can power the whole thing pretty easily with a battery.