Open Source Ultrasonic Whistle: Difference between revisions

Device data

Manufacturing files	https://www.thingiverse.com/thing:1181877
Hardware license	CERN-OHL-S
Certifications	Start OSHWA certification

This page was part of an MTU course MSE4777 OA and MSE4777 OB/MSE5777/EE4777/EE5777: Open-source 3-D printingPlease leave comments using the talk page. The course runs in the Fall semester. It is not open edit.Read more...

Open Source Ultrasonic Whistle

Project developed by Troy Johnston Tmjohnst (talk) 06:21, 17 November 2015 (PST)

Template:Statusboxtop Template:Status-design Template:Status-prototype

You can help Appropedia by contributing to the next step in this OSAT's status. Template:Boxbottom

Abstract

1. Wolves are a problem with reindeer husbandry. This 3D-printed ultrasonic whistle is a cheap method to be used as a wolf preventative. As problems grow worse, this solution loses it's use; however, as an initial preventative the whistle can provide some warding effects.
2. Wolves have a much higher range of hearing than humans. Humans hearing peaks at about 20kHz, while most dogs can hear up to 54kHz. The higher end of that range can be irritating to them and thus is applicable for warding (also used in dog training).
3. I imagine the most savings for this project are in the time it takes to get a working whistle. If you order it you can get it from China for just under $4, but it'll cost you about two weeks, sometimes even months.
4. Since the whistle is so tiny, the final model includes a loop to attach it to a bracelet or string tied to your coat zipper.

Bill of Materials

1. 1g PLA
2. ~10 minutes at 0.1mm layer height
3. FreeCAD file and STL are on thingiverse.

Tools needed for fabrication of the OSAT

1. MOST Delta RepRap or similar RepRap 3-D printer

Technical Specifications and Assembly Instructions

1. None needed. Simply print the part.
2. Print time: 10 min.
3. The model is made so that by editing the LENGTH variable in the VARIABLES spreadsheet you can adjust length and therefore pitch. The 25mm whistle provided in the STL is the one I've experimentally determined to be the most effective.

Common Problems and Solutions

1. Printing slower may give better results. Such a small part can't cool very well before the next layer is placed, so printing too quickly will smear the whole thing.
2. Lowering temp to print may also be beneficial; I printed at 180C.
3. I already tested 4 different lengths to determine the best length. While you can change the length to whatever you want, the intent of this is to ward wolves and to do that we want to be in the 20kHz-50kHz range. 25mm is the best I've found.

This is a spectral graph of 2s worth of each whistle being blown (40mm,30mm,25mm,20mm). You can see clear bands (harmonics) in the spectral domain. Using this I found that the 25mm whistle produced beautiful bands in the range we want (clear harmonics at 20kHz, 28kHz, 35kHz, and 45kHz). The thickness of the bands depicts the magnitude of the frequencies in the band.

You can see the largest portion of our 25mm whistle's frequencies are in the 0-10kHz range. This area just sounds like white noise and without the spectral analyzer you might not know there's harmonics at all.

You might notice that the 20mm spectrum has more frequencies in higher ranges, but we want clearly defined bands to maximize the wolves' annoyance. Just by visual analysis you can tell that the bands in the 25mm whistle are the clearest and should, theoretically, have the most effect.

Cost savings

1. Cheapest alternative is a $3.89 whistle.
2. Total cost: 1g at $23/kg = $0.02
3. ~$3.50 savings. That's 99% savings.

References

The build was constructed entirely from trial and error since the size was so small.