Hacking the Crosman Vigilante

[George Schmermund]

The barrel heating experiment is pretty straight forward. The heater will be the same type that is now being used on the powerlet. It will be attached on the breech end of the barrel. The temperature will be set to somewhere between 100º F to 150º F to start. There will be a temperature gradient down the length of the barrel toward the muzzle. This will be an easy bench setup.

As far as using direct measurement of volume as a metric for the test, it will be much easier and more accurate to get the numbers from the time/pressure curves.

On another note, I've been giving some thought to the area around the valve exit. There is no real seal between the detent cylinder and the cylinder plate. There is an ID/OD difference of ~.004". This could be a source of considerable(?) leakage when the valve opens. The plan at the moment for measuring this potential leak sight is to use the small probe attachment on a lab ultrasonic translator. I've got a vintage HP 4918A up in the attic somewhere that can can be pressed into service for this job. This instrument is very sensitive for finding small leaks in pressurized systems. Its directionality is also remarkable. Another use for this instrument is to locate the source and magnitude of friction. I had forgotten about the friction location capability. It will be interesting to survey the moving parts inside of the pistol frame. This could also be used with the pellet pushing experiments when I get back to them.

Anyhow, my for leak testing I'll measure the leak magnitude of the detent cylinder with the breech pressed against the cylinder's cone without having the pellet clip sitting between them. This will eliminate extra interfaces. The tests would be run with and without a pellet in the forcing cone. It looks like I'll be back on the test bench for a while.

[George Schmermund]

Here's a photo of the instrument and gun that will be used in the detent cylinder leakage testing. I'll also use the translator to evaluate the various potential points of friction in the cocking and firing mechanisms.

[George Schmermund]

While the bench testing is going on I thought I'd do a fit-test with a rifle scope on another build. I'm going to use this scope on the 18" carbine that's in the works.

[George Schmermund]

I did some target shooting at 33 ft. this afternoon using the new scope set-up and a rest. I didn't take any pictures, so the results didn't really happen!(?) When I get some photos, though, I think you'll find them impressive. The new 18" barrel and a new extended fore-stock should be ready by Monday.

This project has moved me far away from wanting anything to do with break barrel guns. The Benjamin Titian that I bought last year to deal with rodents has killed more scopes than rats! With my latest Vigilante carbine designs I'm ready for a new season of infestation in the trees and ivy. Bring it on!!

[sfcnewy]

Very impressive from the start!!!

[WhatUPSbox?]

George,
I just found this great thread yesterday. I'm in awe at the breadth in expertise and equipment you've applied to help us understand airguns.
I'm also in awe that someone in Carlsbad has an attic.
Getting back to your latest experiments, It will be interesting to see if you can get any heat from the barrel into the CO2 in the short timeframe. The heat transfer coefficient may not be high enough. It would also be interesting to see if there is a change in the aimpoint as you heat the barrel. Since the sights attach to a plastic rail, this may drift with temperature. A laser pointer could help track that.

Thank you again for sharing with us the fun you are having with the Vigilante

[George Schmermund]

I've been experimenting with various approaches to doing muzzle crowns. I've found that cutting and polishing compound facets seems to greatly reduce the number of fliers. Today I'm using a Temo 4mm green rubber burr to both cut and polish the facets. The Dremel motor is held in the tool holder and rotates at ~ 20,000 RPM while the lathe chuck turns at 1000 RPM. This keeps the facets concentric with the bore. The Dremel can be adjusted to any chosen angle. The facets are cut deeper than is customary. At this point it appears that the increased depth helps the knurled choke to be more effective. There's plenty of room for speculation at this point. I'll just put out the information as it accumulates.

The photos are of the 18" barrel that will be fitted into the next version of the Vigilante carbine. The results of various muzzle treatments will eventually be evaluated using schlieren photography when it gets its day on the test bench. The use of several short barrels with different muzzle treatments will be used for these tests.

[George Schmermund]

After giving the 18" barrel the new 'grind and polish' treatment it was tested out in the backyard range at 33 feet. The barrel housing of the gun was clamped in a vise and the sighting was done by looking down the bore to get a first approximation of where the center of the target was. I just wanted to keep things on the paper. The temperature was 70º F and no powerlet heating was done. A photo was taken this time to keep things real. What you see is the result of 8 Crosman Destroyer pellets fired in fairly rapid secession. There was no attempt to do anything quantitative in this test. It was the end of the day and I just wanted to shoot a few pellets at a target that was further away than the 15 foot range that I use in the garage. The group was so small in the garage that I had to check a couple of times to see if the clip actually had pellets in it!

All of this is testing the barrel before the choke is applied. Things are getting very interesting again unless some one points out that this is typical of the results from most Crosman 1077 rifles. Maybe it would be useful to buy a 1077 and do some testing on it. In the mean time I'll just gloat over the results.

[Rob M]

lol,I cant shoot that group with my 1077 at 5 feet.. but, thats the trigger.. the barrels could be really good for all I know.. very nice results.

[sfcnewy]

Wow!! Everything in this thread is impressive! The research that you are doing is way beyond anything I have seen before. And on a Vigilante!!!

[George Schmermund]

There is a lot of information about airgun pellet muzzle velocity out there. If the chronograph that you're using doesn't also give you the muzzle energy, it's an easy matter to plug the velocity into any of the calculators available on the internet and get an energy number if you know the pellet's mass.  What there seems to be little information about is the true pellet energy at the point of impact (POI). I've gotten temperately distracted with POI measurements lately and decided to set up an experiment to make those measurements.

Earlier, I had posted some info about using a force transducer as a timing marker to measure the transit time for a pellet in a barrel. This approach worked well and some good numbers were generated. Now it can be used for the job it was intended to do and it should make an excellent way to easily get POI numbers.

The original transducer was mounted on a block of aluminum and a PVC pipe end cap filled with duct seal is used as a trap for the pellet. That was an OK target when it was placed at the muzzle but I'm interested in making measurements at 5 to 10 yards now. This will require a larger target If I'm the one shooting at it! To that end I'm replacing the small PVC cap with an ABS one. It will also be filled with duct seal.

The measurements will start with the transducer at the exit of a Combro chronograph which is attached to the barrel. This will be ~ 4" from the muzzle. The Combro can output both velocity and energy. The output of the force transducer will go into a Kistler 5026 Dual Mode Differential Amplifier. The 5026 amp can then be calibrated to give an output signal that matches the Combro energy numbers. This output signal will go into the LeCroy scope where the pulse can be captured and stored. The scope can then calculate the energy/time integral between two cursors.

My interest in this experiment is to make sure that a pellet from my Vigilante carbine will be able to humanely dispatch rodents at 5 to 10 yards. This information may also be of interest to any spreadsheeters that might be watching........

[George Schmermund]

Extra photo.

[George Schmermund]

The new force transducer/target stop combo is working very well. The ability to get the force/time integral now opens a new chapter in doing some of this bench testing. I'm going to go back to the pressure/time curves and investigate those experiments using the same type of analysis. There doesn't seem to be any end to this. I find that to be very satisfying.

[George Schmermund]

There has been some discussion from time to time about the sound levels of airguns. I've taken some steps in that direction, but keep retreating because of the ambiguity involved in making and interpreting the results. There is no standard method of doing the testing as far as my reading has been able to come up with. Therefore, I've continued to find other things to fritter away my time with on the test bench besides sound levels.

The typical quagmire involves variations in methods, instruments, distances, weighting factors, etc. The problem that I've found most perplexing is that virtually everyone who uses a sound level meter (SLM) to measure gun noise tends to just aim the SLM toward the barrel and assume that the numbers it gives you are real and valid. The easiest way to avoid wayward discussions about 'proper' SLM measurements is to not get involved.

Well, I've collected too much sound and vibration instrumentation over the years to abandon the measurements at this point. The only way out of the speculative arena of these tests is to design my own approach. This comes about because I was recently asked if, and how much, the barrel mods I've been doing change the noise level of the Vigilante. Embarrassed, I had to admit that I couldn't give a number, only an impression. I've decided to resolve the issue and try to redeem myself.

Here's what I'm going to do. First, I'll consider the that the only real number that relates to the gun's noise level is what exits at the muzzle. I'll further assume that this noise is an 'impulse'. Any other noise in the measurement is directly associated with the environment in which the measurement is taken. The environmental noise is always going to be unique to sets of reflections, time delays, and the instruments' responses to them. Let's also assume that I'm only interested in measurements made on MY guns using MY setup geometry and MY instruments. This will always give me reproducible results if nothing in the experiment changes. Now I can give reliable impulse numbers that relate directly to any changes in the barrel mods.

I'll expand on this as the measurements progress.

[George Schmermund]

The time has come to get some real numbers relating to how loud the Vigilante pistols are from the factory and and from the longer barrels that I've been working on. I've chosen from a collection of General Radio and Bruel & Kjaer instruments that I've been collecting over the years and have narrowed the preliminary test down to mostly the B&K devices. The GR instruments are also excellent, but when you have a choice it's hard to get past the overall exquisiteness of how B&K builds things.

Anyhow, What I've chosen to start with is a 1/4" laboratory microphone capsule type 4136 coupled to a type 2639 microphone preamplifier. These are run through a dual preamplifier power supply type 5935. The output of the 5935 is then routed to the input of a measuring amplifier type 2636. The 2636 processes and conditions the signal and then sends the output to the LeCroy scope for further evaluation.

Most of the measurements that I've used this equipment for has been audio testing and measurements for designing and building planar magnetic loudspeakers. Those were the days! Well, membrane speaker diaphragms are fast, but nothing compared to what I'm measuring now with pellet guns. I'm seeing things that are not intuitive and have clearly shown that good impulse measurements simply can't be done by any commercially available sound level meter. I though this was going to be easy, but I was quite wrong.

I'll proceed slowly in the following posts so that those of you still watching this thread may also be able to enjoy the very interesting (to me) things that are beginning to materialize from the tests. If you doze off, that's OK, too.

[George Schmermund]

Since I said that I'd start slowly, let's look at what the limitations are of the measurements. The photos show what a fast positive pulse from an HP 3310B Function Generator looks like. As can be seen the rise time on the pulse going straight into the scope's channel 1 is 40 nanoseconds. The channel 2 trace is what the 2636 measuring amplifier output looks like from the same signal. The rise time is 1.06 microseconds with some overshoot, but no ringing. This represents a bandwidth for the 2636 of over 300 KHz. B&K rates the bandwidth conservatively at 200 KHz. Most of the SLMs are rated at 20 KHz.

Next we'll deal the microphone and its preamp.
 

[George Schmermund]

An important part of making impulse measurements is to understand the way in which a Type 1 SLM  actually manipulates a very fast signal. It takes some amount of studying to get to the bottom of measurement standards, so I'll just take you on a shortcut here. A signal coming from the microphone is passed through a selection of various filters before the meter presents a value to be read. The filters are called "weighting" values. The most common weighting is "A". They are designed to allow the incoming signals to be contoured to the way our ears perceive sounds. These weighting filters are suitable for things that would be encountered in most everyday circumstances in the typical workplace.

The problem is that reports from gun muzzles and explosions are not usually part of the workplace environment (well, maybe not so uncommon now, but....). When the SLM is set to measure "impulse", for instance, the rise time and filters alter the signal to the point where the values read out are quite meaningless. As an example we can take the previous pulse signal that came through 2636 unmolested and switch in the identical weighting filters that a Type 1 SLM would use to represent the "bang". The photos show how the signals look as the SLM would represent them. I'll slowdown the time window so that we can see the full effect of the weighting. The traces are offset vertically for clarity.

The first photo is the pulse as seen in the last post.

The second photo is with an "A" weighting.

So, my question is, what are those people on YouTube really telling us about how loud a gun is when they use an SLM to measure it?

[George Schmermund]

The 1/4" microphone's preamp has been changed to a B&K 2633 type. The system's next addition will be a B&K Band Pass Filter Type 1617. This completes the requirements for a Type 0 sound level meter. It's time to start making some muzzle blast measurements. As stated before, the numbers that come out of this setup are going to be very reproducible, but unique to this testing arrangement. The initial measurements will be with the microphone placed 3" in front of the muzzle and 2" above the axis of the barrel.

Preliminary measurements indicate a sound pressure level of 152 dB. This is a big number, but well within the microphone's upper limit of 185 dB. The closest reflection surface has a time delay that is well outside of the time measurement window. The photo shows the basic arrangement for doing these starting tests.   

[George Schmermund]

I finally got around to doing a couple of experiments to confirm that my time has been well spent putting together a hot rod impulse measuring system. This first set of experiments is definitely ground breaking info (to me) about airguns , but may seem ho-hum to others with more experience in these matters.

I've been doing a lot of reading of late about the best ways to do impulse measurements on firearms. This arena is where the real testing quagmire is getting sorted out. There is starting to be some agreement as to how the testing should be done. There is basically nothing of interest that I've found that relates to airgun impulse measurements. I won't belabor the subject here, but I would like to share some stuff I've captured with the new instrument setup.

The photo shows three different shots. I'll try to keep this simple. The top trace is channel 1 and represents a typical pellet shot in a 1 mS time window. This same time window is used for all three shots. The first pulse starts and finishes in about 50 microseconds. In firearm parlance it's called the 'N' wave. Keep in mind that the display of these pulses is inverted. Don't worry about that for now.

From everything that I've read so far, the N wave is the air in the barrel being compressed into a supersonic shock wave by the advancing bullet traveling up the barrel in excess of Mach 1. There is always a time interval between the clear N wave and the start of the 'real'(?) muzzle blast. Since the N wave is of the same magnitude as the blast coming from behind the projectile, what do you think the ear does with the first shock wave? Ignore it? If we use A weighted impulse measurements then, yes, it is ignored by the METER!. That's enough for now on the top trace.

The second trace, in the middle, is a shot with some canned spray duster sprayed into the barrel just before the shot. The duster gas is about 2.2 times as dense as air. It was sprayed in to displace most of the air in the barrel.

The third shot is a quick followup after the middle shot to see how much of the duster gas was removed from the barrel. There's something interesting going on here that I'll just throw out there for speculation. Keep in mind that we're only looking at 1 mS of information.

This looks like the entrance to another rabbit hole.

 

[WhatUPSbox?]

George,
Did you ever set up the high speed microflash unit you showed us in Jan? Could that be triggered by or at least synchronized to the audio signal and show the pellet position compared to the audio time trace?

Also, is the baseline (trace 1) relatively repeatable?

This is all great reading. Thank you