My Own Electric Ship's Whistle
03.19.03 Restoration of a freighter's electric piston horn
Click any of the photos to see it larger in a new window

With the current political crisis, I'm taking a break from sounding large sirens at random. March 17, 2003 I took delivery of my latest side project, a large electric ship's "whistle" from a bulk freighter. The tags on the motor and horn housing are worn to the point of being unreadable except for "2.2." visible in one spot, indicating that this is a 2.2kW motor (3HP). All other specifications are to be determined experimentally and with help from the folks in rec.crafts.metalworking. UPS dropped this on my front porch, and the tag reads 135 pounds. Fortunately, a large part of the weight was the shipping crate, the horn itself is just a bit over 100 pounds. Just the same, it took a good half and hour to manhandle this up the two flights of stairs and into my attic (where all my serious electrical hardware is, for the sirens). It's March 19 now, and I'm still feeling it a bit in my lower back.

Here is a shot of the horn, freshly moved into my attic. The only thing I have done to it thusfar is remove the clear acrylic cap over the motor's rear bearing, to keep this from being cracked as I bounced the horn up the stairs. The black ring you see on the motor is the neoprene gasket that was under the plastic window.

The next thing I did is take a peek at the motor. I did this not only out of curiosity, but because one of the motor's rear cap bolts was headless. I wanted to start disassembly with this tough part first. Luckily, it turned out that the headless bolt was only frozen into the non-threaded hole in the cap itself. Once I pried the cap off, the rest of the bolt turned freely out of the tapped hole in the motor body.

This is the inside of the motor. Standing on the end of the motor shaft is aforementioned headless bolt, now free. At this point I know a bit more about the horn; I can see clearly that it has a 3-phase, 2-pole, squirrel-cage induction motor. But the specs are as of yet unknown.

BLAM! The horn explodes into a billion small parts. Well, actually it took more like 2 hours of hardcore SWA (socket wrench action) to get this beast down to its basic parts. In this photo you can see the rear horn section and piston cylinder (upper left by the rubbermaid box), the motor's speed-increasing gearbox (lower left), the piston (dead center), and the remains of the body casting (right).

At this stage, I have made the decision not to attempt removal of the front section (the mouth). The bolts holding the horn on are in there really tight, and the top two are threaded into blind holes in the body casting. A blind hole is one that you cannot access from the rear, i.e. the hole is drilled into but not through the casting. The disasterous risk of snapping off the head of a stainless-316 bolt that is in a blind hole outweighs the only advantage of removing the bolts, which would be easier cleaning and painting of the front horn. Removal of the rear horn "elbow" has already given me a straight view through the front and middle horn sections, so I can clean their insides adequately with my 4' flex-shaft and various diameter wire wheels.

Here is the motor, separated now from the body of the horn and from its speed-increaser gearbox. The motor sports a 39-tooth helical gear on its threaded shaft. I have now done some simple tests of the motor. Measurements on the windings show that each winding is 26.3 ohms and 130 millihenries. This doesn't mean much to me in terms of guessing voltage or horsepower, but it does tell me that the windings are intact.

This motor turns out to require 480 volts 3-phase. I hooked it up to my 240V 3-phase converter and it only spun up to about 1000 rpm under load. I've just ordered some power hardware to try to get this running, including a big 240V variac and a 240-to-480V 5kVA step-up transformer. I'm going to step the voltage up to 400+ and run my phase converter's idle motor wired for its 460V setting. Then I should be able to get a honk out of this thing.

This is an inside view of the speed-increaser gearbox. This gear has 16 teeth. Thus, the ratio of the gearbox is 16:32. Now, I know from the size of the horn mouth and the length that this is approximately a 140 Hz horn. What this gear ratio now tells me is that this motor is designed for 60Hz power (not 50Hz). A 60Hz motor spins at about 3450 rpm, and: (3450 / 60) * (39 / 16) = 140.15 Hz A 50Hz motor would run at about 2875 rpm, which would move the piston at 116Hz with this gear ratio. So it's a 60Hz motor--hooray again!

So here is the piston itself. Surprising to me, this looks just like an engine piston, not at all like the thin discs in the newest generation of piston horns. This older horn is also oil lubricated, which you've probably guessed already from the enclosure and oil viewing ports. The piston is 10cm wide, and has a stroke of only 7mm! When I took the horn apart, I found that the piston pin had slid all the way to one side, and both retaining rings are missing. I just ordered stainless replacements.

April, 2003
The electrical problem

April 5-6: This weekend I worked on putting together a 480V 3-phase converter to power this horn. The mess shown at right is what I slapped together on a board. Not pretty, but it works!!! I found out through experimentation that I really need to have a balanced rotary converter to get full output from the horn. With no run capacitors on the converter's idle motor, the manufactured 3-phase outputs were low, around 420 volts, and this got the horn within about 2% of max speed. But the horn is highly resonant, so falling a few Hz short on piston speed resulted in greatly reduced output. After balancing the idler motor with capacitors to get at least 480V on each leg, I got wonderful full-volume honks right at 140Hz. All that is left for me to do at this point is assemble the finished converter. I'm waiting now for my new start capacitors and NEMA 12 enclosure to arrive via UPS.

At the end of April, I aimed the horn out the window for some test blasts. Here is the view from inside the attic, of the horn pointed out my attic window overlooking a town common grassy area (not aimed at anyone's house).

Here is the view from across the street from my house, looking up at the horn, which you can see in the right attic window.

The horn works great from my homemade phase converter, and I've given it a few outdoor honks. Here are the recordings:

Play pistonhorn140hz.mp3 (Outdoors, recorder 12 feet below horn, aimed in same direction as horn)

Play pistonhorn_april20.mp3 (Outdoors, recorder 200 feet straight out, aimed back at the horn)

The first file sounds muted because the microphone was not pointed towards the horn, and was not even located within the coverage area of the horn's output. The second file is more representative of how the horn sounds in person, though the extreme volume obviously cannot be appreciated from a recording alone.

I will be making more and longer recordings in the future. The volume from this horn is extreme and startling to anyone nearby, so I have honked it for only a few quick blasts so far. I plan to do a proper long series of honks on July 4th, when neighbors will be more understanding of loud horn honking. I'm going to try to rig up a klaxon horn to sound prior to the ship horn, to give people a heads up.

Here is the schematic I'm using for the converter. It is based on a design by George Carlson,I have just scaled it from 240V to 480V and reconfigured the output contactor to switch all 3 lines.


Click to enlarge -Adam