Month: October 2019

Yesterday, the new, correct nozzle was delivering a 105°C measured thermistor temperature rise above ambient (actually, the vacuum intake, which includes a bit of CO2 exhaust). It had been running for over a day, and seemed stable. We are running out of testing days. It was time to replace the higher capacity regulator with the IGT type used on the Mosquito Magnet Patriot and test that one.

After doing so, the trap wouldn’t start, twice. I said “here we go again,” attached the gas pressure NodeMCU, and tried again. This time it started right away. With an atmospheric pressure of 30.07 inHg, the IGT regulator put out almost exactly 11 inWC relative pressure. Perfect.

I noticed, however, that the temperature rise was only about 95°C at an ambient of 20°C, more or less the same as the incorrect WDA nozzle. Alarmed, I looked at the logs, and noticed that a few hours before, the temperature differential had, for the first time, dropped from 104°C to 95°C on the more expensive Marshall Excelsior regulator, which delivers less gas pressure. What was going on?

With the IGT regulator gas pressure measured, I disconnected the Schrader valve connection and turned off the NodeMCU. This time, however, there was no jump into the noisy incomplete combustion mode previously observed in these circumstances. This marks a great improvement in stability.

I don’t know what to make of this. I do know that prior to entering the rough incomplete combustion mode, the old nozzle typically exhibited an unexplained few degree temperature rise before falling down to the 55°C range. Could it be that the initial hours of the newly cleaned nozzle had a combustion mode that matched the rise before the fall in the WDA nozzle?

I do not want to investigate this now. Instead, I want to qualify the IGT regulator for use on the Mosquito Magnet. The extra long double hose doesn’t look so bad wrapped around the tank with the tank rotated 180 degrees. I hope I can recommend this device.

All this means I was probably too hard on the ill-fated Doz yant regulator that failed in the initial tests. As I said earlier, the Defender at that time was too sensitive for this world. The Dozyant regulator may work just fine in a normal trap. It does have a more attractive polished finish. But at the time, it wouldn’t work at all in the mis-configured trap. A minor negative, but one that bears some mention when discussing the Defender and its many quirks.

How embarrassing. Since June, 2017, I have been using what I thought was the correct nozzle for the Mosquito Magnet Defender trap. In 2013, I was delighted after replacing the clogged original nozzle with a 1.5 gph 45° hollow nozzle ($5) “resulting in the old trap catching hundreds and hundreds of mosquitoes each day, like when it was new” (see the wiki). But then, after more research, I mistakenly combined information regarding the purported Mosquito Magnet Liberty trap nozzle with the Defender, ordered 1.5 gph 45° hollow WDA nozzles, and have been using them since.

Anyway, the fall is upon us, and the temperatures on the decline. A few days after the regulator “bake-off” (below), with the Mosquito Magnet Patriot IGT regulator still installed, the Defender had an episode of abnormal combustion, and stopped. I restarted the trap several times, and within just a few minutes, the abnormal combustion re-started. I then replaced the IGT regulator with the more rugged Marshall Excelsior device, but it did the same thing!

Well, this trap is just too sensitive for this world. What could be wrong? I had just replaced the nozzle this summer with a brand new WDA after measuring the pressures using a tire pump connected to the Schrader valve. But what was left? Could I be using the wrong nozzle?

Earlier this summer, I was using the auto tire pump to blow out the Defender on a regular basis, but then that stopped working, so I decided I needed a new nozzle. I took the existing and all the old nozzles I had, screwed them into the Defender, measured the pump air pressure, and recorded the pressures (in the Used column):

Markings	Used PSI	Cleaned PSI
Open (no nozzle)	25 (Defender Valve)	20 (Spare Valve)
1.5 45°A WDA I14	77	40
1.75 45°A WDA H05	86	45
1.5 45°A I09	90	35
1.5 45°A I09	97	43
1.5 45°A WDA A04	103	43
1.5 45°A WDA d14 (new)	68	43*

The list is ordered by time of measurement. Although the 77 psi pressure was not all that high, there was a need to do something, e.g., replace the nozzle, again. The first entry is the replaced nozzle, followed by whatever was next in the bag, ending with a new nozzle that had much lower pressure, which ended the test. This nozzle was installed and was in service until today, when it was removed but not cleaned.

The test setup is an automobile tire pump with pressure gauge connected to the Schrader valve of a Defender propane valve assembly. The “used” measurements were made using the assembly from the trap with the original regulator and hose connected (but the valve closed). The “cleaned” measurements were made using the spare, with a male type “F” RF connector screwed into the gas port (the closed valve leaks some).

The cleaning was done yesterday, and consisted of 1/2 day soaking in mineral spirits with several hours of ultra sonic agitation, followed by 1/2 day soaking in carburetor cleaner (mostly acetone with some toluene) and additional agitation. I got the soaking solutions from the now defunct Mosquito Magnet Forum, some of which is available on the Wayback Machine, and added the ultrasonic cleaning myself, mostly because I had a cleaner. I believe it made a large difference.

This cleaning and replacement was motivated by me re-reading an old post, where I read a definitive spare parts report that listed the Delavan 15045A as the correct nozzle, and not the WDA version, which that same post said was used in the Liberty in a 1.5 gph size. I had conflated the WDA from the Liberty list with the Defender, a grievous error a la the 737MAX. The much more expensive ($20) WDA nozzles are designed for humidification. Why they are used in the Liberty, if they are, is a mystery for another time. I immediately changed the wiki. Also, it was time to change the nozzle, this time with a cleaned original type, a cleaned trap, and an empty catch basket.

Looking at the table above, cleaning made a large difference. We want the lowest pressure using air, which would mean an unobstructed nozzle orifice and fuel filter. Generally, although the filters look very dirty, they clean up more or less perfectly. The nozzles themselves seem to be the issue, although I didn’t test any dirty filters with clean nozzles.

I selected the lowest pressure cleaned nozzle of the correct type, and installed it a few hours ago. The Defender started and has been working since. The nozzle performs differently, though. Instead of a ~95°C rise from ambient, this nozzle is running at 105-107°C rise (123.6C at 5:7984 T=18.1C H=44%), a large increase. It is adding more heat energy (presumably by burning more fuel). Perhaps this will isolate the system from minor fluctuations in gas pressure and reduce the likelihood of the dreaded incomplete “abnormal combustion” and keep mosquito-repelling unburned propane out of the exhaust.

So that’s it. Another day, another mistake corrected. Let’s hope the Defender continues to run during the cooling nights. At some point it will be time to shut the system down. What a shame.