Chronograph for bullets

[wannabegeek]

Hi all...back in at my bench this weekend....

Got all my parts...I have a decent collection of 74LS and 74HC chips now...

Started with my 10 Mhz oscillator.
I have been having trouble getting above 20mV swing out of it....well, I knew that my scope wasn't really cut out for 10 Mhz, but the time base amplifier has a 0.1 us setting. After building an re-building three different 10 Mhz oscillators and reviewing the theory of oscillators, I put my function generator on the scope at it's highest freq 3 Mhz and wouldn't you know, a 10 V output looked like a 40mV output on the scope. I'llbe borrowing a scope at work this week.

I'm still unsure what the load cap should be for the crystal I have. There's no data sheet just a listing in a catalog and the cap says 'Series'....yeah..I'll try a few and see what happens in a Pierce configuration and 74HC inverter.

SR FF
Yay...this fun...I just set up the basic circuit that Steve gave me using the modified detector with rising and falling cap. Gonna test after I take a break.

Then I'm gonna work with Rleo's design.

CDRIVE 's detector idea
Soon I'll be able to test CDRIVE's idea about using a Al foil sandwhich as a detector. I hope that my .22 Cal air rifle pellet will accurately simulate the effect.\

This is a nice break from trouble shooting my computers....just switched from Ubuntu 10.10 to Mint 13 Xfce...wiped my SSD using my W7 desktop machine and broke the boot loader...and discovered a HD with a lot of 'bad sectors'....not as fun as electronics....

cheers,
wbg

EDIT: Page with crystal...the catalog link has the minimal info in it.
http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_14381_-1

 

[(*steve*)]

You should get a full (rail to rail) swing from your oscillator.

The reason may be that you have a problem with it, or that the bandwidth of your scop is really limited, or that you're using a x10 probe by accident (or a combination)

 

[Rleo6965]

You can test your probe by clipping it on scope provided test signal output called CAL. or Calibration.

 

[CDRIVE]

A 10MHz scope will be down -3db at 10MHz. This is also based on a sinusoidal wave. A digital wave will roll off far below 10MHz.

 

[(*steve*)]

So in summary, a 10MHz square wave signal on a 10MHz scope may look like a sine wave, and may have an apparent amplitude about 1/2 of the real signal.

So it might appear to be 2.5V

A x10 probe would make it appear to be 250mV, and a x100 probe, 25mV

 

[wannabegeek]

I'm glad you are all still watching this thread ! I couldn't do all this without you...

I've been careful to use my probe with with the switch on x1 and not x10 .
I recall being told in a class to always use the x10 switch and set the digital scope
accordingly, I thought this was for higher impedance...??

I think my bandwidth is just very small. I forgot the model number gonna go look at it in a minute. I have a recollection that that the model is a 2 MHz scope, even if the modular time scale amp has a .1 us setting...

Will bring the clock circuit to work and use 300 MHz scope there.
Still a little unsure about the load capacitance for the crystal and the feed back resistance, which should match the crystal too....

I'll post more on that later. Need to dig up the pdf I was reading on theory and ask some detailed questions....

Now I want to go down and try some trigger experiments using function generator for clock !

will post back..

wbg

PS I've included some pics of the project with the old gating circuit. I had my clock in the upper right corner which now lives on it's own board. The wiring could be better. In the future I will put the displays up higher leaving room for the gate logic under the counter chips. I am low on wire

I made a sandwhich of foil and plastic wrap to test CDRIVE's idea.

 

[Rleo6965]

If you encountering intermittent problem with your protoboard circuit ( 2nd picture ). Place some 1,000mfd and 0.1 mfd on +5V supply rail. This is to eliminate noise on +5V.

 

[wannabegeek]

If you encountering intermittent problem with your protoboard circuit ( 2nd picture ). Place some 1,000mfd and 0.1 mfd on +5V supply rail. This is to eliminate noise on +5V.

Thanks...I know that at 10 MHz capacitive coupling is an issue. Also, I am using an old Lambda 10 A power supply that is huge and weighs a ton. I think the power is pretty clean but I haven't looked at it on the scope, but I can hear it humming away. I'll scope the power soon.

Do I want put the caps between the positive supply and the rails...? Or, try to bleed off noise with shunts from the protoboard to ground...? Or both ?

Test using CDRIVE's and Steve's Idea
Using one SR FF, and a detector with 1K Resistor and 100pF cap, I tested CDRIVE's idea that the bullet ( .22 lead pellet in the test ) would trigger the gates.

And it did ....!! This is promising since if it works with the real thing it makes the detector design much simpler. I tested the resistance of the lead pellet and found it to be ~7 Ohms.
So I think that the pellet is actually shorting the two layers of aluminum foil momentarily as it flies through. I was able to shoot 3 times and there was still no lasting continuity.

If designed right, I might be able to reliably fire several shots into one detector.

I'll have to work with Rleo's design later this week.

Sorry Steve that I haven't had a chance to get back into the physics theory of the velocity. I have been thinking about it though....

Side Thoughts:
I think that the rotation of the bullets might cause a lot more drag than one might suspect. Vorticity typically steals energy away from systems. Looking forward to getting data on the 22 cal pellet and modeling it's velocity in x and y and making/ testing some predictions. I'm starting school again soon for one class so I need to dust off my mechanics books and get to work deriving the equations I pulled out a few posts back.

Having a mathematical model should enable me to avoid needing tons of data for statistics. I'll only need one shot and I can model ALL shots as a small perturbation of the model if I assume that the deviations are random and the model is complete.

 

[CDRIVE]

I've been careful to use my probe with with the switch on x1 and not x10 .
I recall being told in a class to always use the x10 switch and set the digital scope
accordingly, I thought this was for higher impedance...??

You were instructed correctly. Have you ever noticed that scope sales literature usually state "Supplied with 10:1 low capacitance probes"? This because a 10x probe also greatly reduces capacitive loading by the scope too. A typical scope will have an input resistance of 1Meg paralleled by ~ 100pF. This is a combined capacitance of the input circuitry + the cable capacitance. Using a 10x probe will increase the effective input resistance to 10Meg and reduce the effective input capacitance to ~ 10pF.

There are exceptions to this rule. Labs that work with RF devices typically provide 50 Ohm input and output impedances for all their ports. In these cases the cable connection to the scope is usually 50 Ohm transmission line with a (50 Ohm) termination at the scope end using a BNC "T" connector.

Some high end scopes have a front panel switch to provide a 50 Ohm load. Others have a dedicated 50 Ohm input port. In cases like these the scope probes are not used in either case.

Chris

 

[CDRIVE]

If you encountering intermittent problem with your protoboard circuit ( 2nd picture ). Place some 1,000mfd and 0.1 mfd on +5V supply rail. This is to eliminate noise on +5V.

Speaking of Protoboards. Are you testing the 10MHz osc. on the protoboard? Not a good idea.. As handy as they are they present quite a bit of capacitive coupling to adjacent conductors. Working with HF requires adherence to accepted techniques. This excludes long and parallel conductors, alligator jumpers, etc.

 

[wannabegeek]

Speaking of Protoboards. Are you testing the 10MHz osc. on the protoboard? Not a good idea.. As handy as they are they present quite a bit of capacitive coupling to adjacent conductors. Working with HF requires adherence to accepted techniques. This excludes long and parallel conductors, alligator jumpers, etc.

Yeah...for the moment I have a second protoboard with my clock components on it. I have worked to keep all the leads super short, flat and at right angles to each other...but I suspect there's got to be coupling ocuring.

I used a ground plate pre-made board for a kit before with 5 MHz clock. But since I'm still learning about clocks, and need to experiment, how can I avoid this coupling ?

Will I have to solder to a copper plate every time I want to test something...?

Thanks...this is a big question on my mind.
wbg

 

[CDRIVE]

Yes, it's a bummer but it is what it is. Even on a copper board you can't experiment with component values with lead lengths as they come out of the package.

 

[wannabegeek]

... with lead lengths as they come out of the package.

I didn't catch your meaning here...meaning the components have to be surface mount ?

btw, are you happy that your idea appears to work so far...?
I think the plastic wrap is a great help to keep the foil layers from shorting after a hit, since it is rubbery and stretches a little instead of shredding.

Would the PICAXE be a good way to store and manipulate multiple trial data ?
I'd use my existing circuit for the high speed timing capabilities, then use the uC for storing each trial into a register to spit out again later when collecting the device.

 

[CDRIVE]

I didn't catch your meaning here...meaning the components have to be surface mount ?

btw, are you happy that your idea appears to work so far...?
I think the plastic wrap is a great help to keep the foil layers from shorting after a hit, since it is rubbery and stretches a little instead of shredding.

Would the PICAXE be a good way to store and manipulate multiple trial data ?
I'd use my existing circuit for the high speed timing capabilities, then use the uC for storing each trial into a register to spit out again later when collecting the device.

No, I didn't mean SMT. I was referring to the tendency that many of us have when test tacking components to see how a value works out. At these frequencies the leads must be cut short.

Yes, very happy the foil worked.

Yes, I love uCs, especially Picaxe. However, If I were doing this project it would be all uC.

 

[wannabegeek]

read this http://new-ham-radio.blogspot.com/2010/08/ground-plane-circuit-construction-what.html

I have used a ground plane printed board but never prototyped with a ground plane.
The link makes it sound easy. I ordered some Dri Wick, Rosin Paste and copper foil sheets to protoype with. I have a fair Weller temp controlled iron.

I'm considering drilling/punching small holes on the copper for ease and keeping all non-ground connections less than 3 mm. Looks kinda ugly but good enough for testing...

How do you make final durable nice circuits when you have everything worked out ?

 

[CDRIVE]

Keep in mind that hams work with some extremely high frequencies. For your 10MHz clock you should be able to get by with copper clad bread board.
http://www.radioshack.com/family/in...um=RSCOM&utm_source=CAT&utm_content=CT2032230

You can easily bridge multiple pads to create broad ground planes but at 10MHz I wouldn't be loosing sleep over it;. ... you're not building microstrip.

 

[wannabegeek]

Please correct me if I misunderstand...

The proto board has a high capacitance by construction and enables the tendency to longer jumpers and leads which makes circuits vulnerable to capacitive coupling.

The ground plane treatment I linked to is a high form of noise/C.Coupling control for RF.

PCB is just a middle ground where the capacitance is reduced but there's no ground plane.


I can use PCB for my 10 MHz clock.

I'm unsure if I need to use some shunts and/or ripple caps when connecting a working clock on PCB to my counter circuit currently on the protoborad.

TIA !
wbg

 

[CDRIVE]

plane.
I can use PCB for my 10 MHz clock.
TIA !
wbg

Yes, of course.

 

[wannabegeek]

I read more about crystals...I was only thinking in terms of parallel tank circuit types, which is fair since they are pretty old school and show up everywhere.

The crystal specs in the catalog had 'series' written for the load capacitance, meaning, it doesn't need a special tank cap.

I tried to read about series crystal oscillators and found some discussion. I connected my crystal in series to two CMOS inverters with a 22pF ( guess ) between inverters and 500k Ohms for feedback on the first inverter.

My scope just ain't cut out for it and appeared to poop out tonight; after over 30 years I guess it deserve to retire. I'll have to bring my proto board to work. I won't solder anything until a get at least an output in the ball park. The electronics bench was removed from the job when they changed buildings so there's just random stuff hanging around, and no real way for me to experiment w/o a proto board. We don't take fMRI scans very much just process them....

cheers,
wbg

EDIT Setup Rleo's FF and it works great too....for a reset I just used one button and the extra AND gate to send that voltage to the other reset terminal required. I could sum it all up in clear display button too in the future.

Still thinking about ways to make the mechanical detection reusable and robust.

 

[CDRIVE]

Still thinking about ways to make the mechanical detection reusable and robust.

It's hard to conceptualize a mechanical (infinitely reusable) target. This issue always brings me back to a photoelectric solution.