Chronograph for bullets

[wannabegeek]

hey thanks for the reply...

I thought about foil and did some proof of concept testing with it.

The trace I want needs to be very tight, perhaps less than 1 mm, so I don't think foil is an option.

The die idea is good too. I'm thinking of having a template made out of brass, that I can use to paint on ( spray or brush depends on resistance of trace ) silver/copper conductive paint, onto a sticker or the target itself. There's also a conductive glue I can use for attaching lead connection points.

cheers,
wbg

 

[CocaCola]

The trace I want needs to be very tight, perhaps less than 1 mm, so I don't think foil is an option.

That would not be an issue with either the etching method or a proper die cut piece of foil...

 

[wannabegeek]

I think I might as well throw in a bi-directional uni shift register, to copy what's in the counter.

After the measurement is taken, the operator presses the counter clear button. I could use a two pole clear button, that also clocks the shift register by momentarily connecting it to Vcc.

Then the data gets backed up so to speak.

This could come in handy when first learning to use my counter for actual data collection. When getting used to what 'good' data are one can become confused.

Imagine, the operator gets a number and thinks it's OK, but really it's not. They repeat the experiment, and get another number a whole order of mag different than the first. They wonder, 'did I write it down right...?'

Having the register auto shift when hitting the counter clear will save some head aches. As well there would be a shift-back back to review the old counter state.

 

[wannabegeek]

That would not be an issue with either the etching method or a proper die cut piece of foil...

Ok, I will definitely include the die idea in my discussions then.
I'm hoping to meet with a friend who is a great machinist and has made all kinds of precision parts.

As well, one of my professors is an electronics guru and she makes thin films. She offered to meet with me and discuss a chemical method.

Assuming I get the timer-counter working and precise to spec, then I'll likely still use the
silver pen trick to proto-type and trouble shoot until the rest comes together.

Thanks for your input.

PS What kind of press do you think I could use with the die? There's a small one in my old Physics shop at school that would do. Hopefully I could rig home made so I don't have to ask too many favors...

 

[CocaCola]

PS What kind of press do you think I could use with the die? There's a small one in my old Physics shop at school that would do. Hopefully I could rig home made so I don't have to ask too many favors...

For thin foil you won't need much at all, even a basic hand press will work find, the die cutters used for scrap booking will work, and heck you could probably sit on it to cut thin foil just fine The real issue is the investment in having the die made...

She offered to meet with me and discuss a chemical method.

That would be etching, not much too it, and still quite popular for fine details even though laser cutters have become very accurate at doing it as well...

 

[duke37]

Could you make a rectagular frame and wind fine wire round it? The spacing can be as close as you wish, the double thichness may not be a problem.

 

[wannabegeek]

Hi Duke,

You I did think of this idea for proof of concept and testing...
It may be too much effort to do several times....

I'd like to start experiementing with the practical stuff....this is not a trivial part of my project...

At this point, I'd prefer to use a silver pen and free hand the mesh, or have a template made that allows a pretty quick trace. Or spay on conductive copper paint.

cheers,
wbg

 

[(*steve*)]

Yeah, there are pens that have conductive (silver?) ink and drawing a pattern on a piece of film may well be the quickest way to do this.

A template and spray on paint would be even faster.

I was thinking about suspending strips in a frame, but that would be delicate and fiddly.

With film and conductive ink you could probably re-use the film by drawing around the holes (if they're small and neat)

 

[wannabegeek]

slightly off topic of the printed circuit idea...

I thought I'd share this and get shot down by pple who know more than me

Reusable detector ( multishot):

Make a capcitor out of the target. Dielectric paper with an aluminum center target, separate with plastic wrap and make another dielectric paper sheet with aluminum target.

The paper can be shot through, makes no difference. The cap is dominated by the Al foil.

Using a high freq (MHz), create an LRC circuit where the C is the 'target' in series with a well matched discrete capacitor, C_tot = C_var + C_static.

C_tot ~ k/C_var which is sensitive to small changes in C_var.

C_var ~ Area of Al target

Send in F_ref to the LRC circuit, and integrate to get RMS output ( must be super fast inegration time like 100 ns ).

Use a parallel array of comparators to detect a distinct ( reliable ) change in LRC output for fixed F_in ( F_res ) as C_var changes with reduction of area, due to bullet holes.

Two of these setups together could form the dectors that then go to the clock trigger.

Total waste of time...eh? But fun to think about...

Gonna get ready for an optical system too.....

 

[wannabegeek]

decoder versus display driver....

I order 7442 decoders when I think I wanted 7447 display drivers....

I want my 7490 counters to display numbers on 7 seg displays...I thought a 'decoder' was the right part but am realizing that 'display driver' is what I need....

Correct ?


thanks...

Also, having issues testing counter. My f generator is cheap and slowest freq is 10 Hz...I made a debounced button but I think the capitor is too small, so I'm still getting bounce...

wbg

 

[(*steve*)]

I think the catalog name is "BCD to seven segment decoder/driver"

 

[wannabegeek]

yeah...I locked onto "decoder" and order the 7442 which is a demuxer/decoder

They look cool for channelizing...?

I ordered the right parts to driver the display....more shipping charges....I also got some universal shift registers to play with since I'm paying for shipping any ways...

I did figure out how to get my new function generator to pulse TTL at 1Hz...and the counters work as expected....making a debounced button to required spec is harder than I thought...


Does anyone care to share a schematic on button debouncing...I read a college EE lab pdf about it and thought I had it figured it.

UPDATE:http://http://www.eetimes.com/discussion/break-points/4024956/Solving-Switch-Bounce-Problems

 

[wannabegeek]

Hi all...back after some bread boarding...

I almost have a 4 digit working counter just need to wire everything up on BB.

I would like to set up this project to run two complete experiments before a human hand is required to do anything.

My detectors oprate in pairs and are static at LO or HI. Recall that my counter-gate controll is an XOR gate that takes detector signal that is connected to an AND gate which has the clock connected to it.

I think I could simply connect two pairs of the above setup to the counter gate, since each turns off after the second detect. This assumes that the clock can fanout to two AND inputs . Then, I need a way to store the set of first counter ( all 4 counters ) bits. I was thinking of using a universal shift register. However, I'm not so sure this is the right thing to do.

Speakng here in terms of one counter/driver/display unit: The shift register would have to be connected in parallel with the 7 seg driver, from the counter. The driver doesn't appear to care whether it gets a stream of bits that stop, then displays the final number. But can the register have the same varying signal and then just hold the last set until a clock shifts them somewhere...?

I am thinking that the user would actively see the second experiment results on the display and press a button to see the results of the first.

ls there a fundalmentaly better way of doing this...?

Thanks !

 

[wannabegeek]

Latches, detcotors and counter gates...

Hi all,

I've been spreading my project questions around different threads, sorry about that, but it seems to attract more attention than my older threads.

I am learning about bounce and the dreaded TTL logic land of 0.8 - 2 V...
I have two detectors which are really just 'wires' that get broken, thereby breaking the path to ground, and allowing voltage to show at the output.

These 2 detectors feed into a counter gate ( XOR out and clock to AND to counter input).

I just tested my break wire detector and got a really fast time and realized I'm probably timing the voltage rise time and not the event that breaks the wire.

I think that I need a latch in this case...the only one I see that might work is the JK latch. With this solution, I'd still need to rest the latch with an RC debounced circuit ( which I haven't had luck with yet, maybe my button sucks ).

I'd love to hear what the experts have to say about simple latch design, as I suspect there's a simpler way to do what I want.

I need this truth table, just like my original XOR setup, but with feedback to keep the bounce out:
0 0 0
1 0 1
1 1 0

TIA,
wbg

 

[Rleo6965]

Can you post your bounce circuit. It's a good start to help you.

 

[wannabegeek]

I've tried the classic RC debouncer on page 12 of the PDF below. I read the article which is pretty cool.

Since my application only uses the clock for the timing of the event, I can reset with just a button. As I understand it, that means I could have a really long RC time constant on my debounce circuit ( 1 s ).

I used a normaly open push button from Jameco, which I think might be the problem. The article auggests that button differences can make debouncing difficult. I don't have my invoice right now to get the button part number. I think I chose the wrong button the more I think about it.

Perhaps a rocker switch would be better than a momentaty button ?

I think the JK latch would work for me if I understand it. I would then need to reset it manualy each time the experiment is run and a fresh detector is connected.

I'm open to other ideas about how to set up this logic.

TIA,
wbg

http://www.eng.utah.edu/~cs5780/debouncing.pdf

 

[CDRIVE]

Schematic is the language we speak. Without it we only interpret what you're asking.

 

[wannabegeek]

My detector's are a thin trace as the switch, which is broken by a projectile. I am measuring the speed of the projectile as it break the two traces.

During a recent test, the time to traverse the distance between the two traces was an order of magnitude too fast. I think that I was measuring bounce of some type.

Thinking about it after the fact, it seems pretty clear that the detectors have to move between 0.8 V and 2 V. I am correct that this is less than ideal ?

EDIT: The time for the voltage to propagate down the short wire is at worst 1/3 the speed of light right...so I'm not reading the rise voltage time.
But, it's possible, that as the trace breaks, the break is not clean on a microscopic level and weird stuff happens that acts like bounce.

Would it not be best, to have to the voltage snap to HI ?


TIA,
wbg

 

[CDRIVE]

Assuming that you're measuring the time between the first event (S1) and the second event (S2), wouldn't an RS Flip-Flop be in order? Contact bounce would then be ignored.

 

[wannabegeek]

Hi CDRIVE...

Flip-flops are pretty new to me so I'm basing my statements here from the wikipedia article on them.

I'm confused about the difference b/w using a clock signal to do stuff and just the inputs. There's latches and flip-flops and the terms are used loosely it appears.

Both the SR NAND flip-flip and SR NOR flip-flop do not appear to meet my detector status. Both have restricted inputs 1 1 or 0 0 respectively.

Inputs of 1 1 and 0 0 are my exact states I need...

The wikipedia artcle mentions the JK latch but doesn't elaborate. My research has lead to confusion about this latch.

I'm pretty sure that I don't want a master-slave situation, but a transparent design. I don't care to have a clock signal do anything for me at the moment. I feel like I just want the setup I already have but with feed back to make the transition more discrete.

thanks again,
wbg