USB camera for powder check

How does the food packaging industry do this?

jsw

Weight.

Many industries are utilizing machine vision. A check point may have 6 or more small cameras to image a part or measure of volume in a container.

I believe the actual go/no go (pass-fail) is done by the software.

I don't fully understand the capability of using a certain amount of pixels in a CMOS imaging device, but someone here has implemented an imaging device for a measuring or comparator device.. I forget who it was.

There are a lot of surplus RFE removed from equipment industrial machine vision components for sale, many on eBay. One problem with the surplus odds 'n ends is getting good cables that fit the uncommon connectors on industrial video components. Not BNC, but various miniature Hirose, JAE, LEMO and other high quality connectors on high quality Japanese cables.

I dunno what Tom's particular situation would require, but if it's a matter of the amount of travel a part makes, opto interrupters could probably handle that.

Yes. It should be a lot easier to setup to weigh each completed round automatically and alarm on out of tolerance ones, even log all the weights.

"Pete C." fired this volley in news:4bf1475d$0 $31890$ snipped-for-privacy@unlimited.usenetmonster.com:

You'll have to tare the weight of the empty casing first. Even minor variations due to sizing and trimming can affect your total cartridge weight.

Another weigh (pun) to do this is to measure the mass of the cartridge, first empty, then with the powder load. A clamping collar at the neck of the cartridge attached to a piezo-electric transducer can determine the resonant frequency of the whole assembly with the empty, then re-tune after filling to determine pretty precisely the mass of powder added.

LLoyd

Assuming that you use "squib" to mean a load with less then the correct amount of powder, this was common when loading shotgun shells. Probably because many people had progressive machines for shotgun shells but used a single station press for anything else.

The "problem" was so common that I have shot on Skeet ranges that had a ramrod hanging on both the high and low house.

But being less facetious, the problem was almost always that the powder ran out and a few primed cases were loaded without powder and firing them would blow the wad and shot about half way up the barrel. If you shot a second round you had an instant short barrel gun :-)

In reference to your camera idea I have seen an article about how to rig a security camera and compare each frame with the previous to detect entry. the same sort of thing would certainly work for checking powder levels.

I'll look around tomorrow and see whether I can locate the article and perhaps I can give you a reference to how it was done.

John B. Slocomb (johnbslocombatgmaildotcom)

A simpler method is already out there, an additional die body with a movable rod in it. Rod checks powder level. Google up "powder-level check" or "powder check die". Sometimes high-tech isn't the best.

To the other poster that asked, yes, a "squib" is a round with little or no powder in it, invariably sticks the bullet in the bore. The classic sequence is a no-charge round followed by a double-charged round in rapid fire, followed by an orbiting revolver top strap. Mostly happens with dense powders and light target charges and on semi- progressive and progressive tools. The tools have been improved over past units, it's harder to half-stroke and advance the case without getting a powder charge, but CAN happen. One of those deals where you need your eyes on the tool, not on a screen.

Stan

Have you looked at a USB Microscope?

It seems like either of those ideas would need an extra station or two, which Tom wants to avoid using since all the stations of his five station progressive press already are busy; otherwise he would use a powder-check die, like he mentioned, and as Stan also suggested in his post, for some reason, just after quoting Tom's post that said, "a powder-check die would do just that but I would not be able to use my bullet feeder or my separate crimp die".

It might be practical to mount an ultrasonic distance transducer looking down into the cartridge, instead of a webcam. This would probably be good enough to distinguish no-powder from powder. Of course accurately knowing how much powder would be a good thing, but he apparently has loaded thousands of rounds without needing to measure it for each one, and just wants to make low-powder loads much less likely.

Measuring the cavity resonance frequency (with a tiny speaker and microphone) would be another alternative. Of course not as accurate as the clamp-on piezoelectric transducer Lloyd suggests, but probably good enough and wouldn't use up a station.

Would it be practical to weigh each round after it's done? I don't recall what cartridge Tom is loading, but presumably the weight difference between powder/no-powder would be a percent or more (a few grains out of a few hundred), which ought to be easy to detect, unless the brass varies by that much. Of course, if the rounds were weighed in the same order that they come off the press, it would be even more likely that the "classic sequence" Stan mentioned, a no-charge round followed by a double-charged round, could be detected.

I mostly load .38 but followed closely by .45 then 9mm. Unfortunately, my sweet load is 2.7 grains of 700x which is not a bulky powder and my cast bullets vary by more than that, cases aren't much better as I use mixed headstamps with questionable pedigree. I use the .38s in a .357 Smith that would handle a double load. The auto loaders scare me the most.

I got one, it won't focus any farther than about 2". I need at least 4"-6". I DO like the scope but It won't do it. Too bad, it has two levels of LED lighting in a ring config.

Different models have different focal length, but the sellers rarely post the full specs.

has NTSC output for $19.98. It is a spherical design, and runs on AA batteries in a separate box. It's a toy, but could have some use around a shop or production floor.

The only problem I can see is that they recommend that the user be eight, or older. I suppose you could lie... ;-)

"Buerste" wrote in news:hssckq$enn$ snipped-for-privacy@speranza.aioe.org:

I have had a squib in 223, but the neck was tight enough that the bullet stayed put (thank goodness). With rifle cartridges, the powder charge is big enough that an empty is easy to find by weight. In addition to adding an extra inspection step, I can weigh small sets of cartridges & tell if anything is amiss.

I've got a new Dillon 650 press with a powder level checker. I will feel much better once I get that up & running.

Doug White

"Buerste" wrote in news:hsscvc$f9t$ snipped-for-privacy@speranza.aioe.org:

How about adding a lens in front to stretch the focus a bit? Not sure what sort would be required, but it wouldn't have to be perfect.

I know with new cases & a flashlight, there's quite a bit of reflection that is very distinct from a case with powder. A lot may depend on how well you clean your cases. That's why watching the powder fall through a clear drop tube sounds better to me. It also gets the camera in a more favorable orientation.

Doug White

Why wouldn't you just use a sensor on the piston (light gate, whatever) and not advance to the next step unless the piston has cycled down to the proper level? What is the advantage of video and image processing?

Kevin Gallimore

Tom, I thought I'd hold back a while then present my simple solution. I just weigh every completed cartridge on accept inspection. Set your scale to 0 on an average bullet and go. A gross mischarge is quick to spot. I do every bullet for wieght and primer seating. check about every fifth with a case guage. Maybe I'm an anal engineer, but zero defects make it seem worth while.

Karl

I used to work for a machine vision system integrator and did some projects for Olin for their military cartridges. One project was for the 30mm or

35mm (don't remember) cartridges for the A-10 tank killer and another project was for 25mm cartridges.

You would either need some machine vision software and then a camera that would be compatible or a stand alone machine vision system. I worked with Allen Bradley, DVT, and Omron vision systems, you might be able to get an eBay bargain. If I were going to use one at home I'd prefer the DVT, plug it into your ethernet to configure it and the processor, on some models, is in the camera.

A good lighting for seeing inside a cartridge might be a beam splitter type of device that reflects a light from a partially silvered mirror, the camera looks through the mirror and the light is directly in line (by using a 45 Degree mirror) with the camera lens. That setup would let you align the camera to see the primer through the cartridge neck and also shine the light in line with the lens, through the neck and to the primer. For the 25mm cartridges the opening was big enough to use a ring light made of LED's.

RogerN