Liberty MM No Gas after 20 Minutes

Hi. I've been following your MM project for some time. I have a liberty mm and when it worked, it worked very well.
It failed to stay lit last season. I tore it down.

The combustion chamber has been cleaned.
The nozzle has been replaced.
The ignitor has been replaced.
The thermistor has been replaced.
The fan (and entire unit) has been blown out with compressed air.
The regulator has been replaced.
The tank is fresh.
The line into the unit to the gas valve and out of the valve to the nozzle have been cleaned.

It will light and go to green each and every time.
About 20 minutes into operation, it faults, indicating no gas.

You have put a lot of work into yours, and I mean a *lot* of work.

The only thing I can think of is that the gas valve is mechanically letting go and closing without the unit detecting that.
Do you have any ideas or pointers? Did you ever get your liberty going again?

Thanks.

Well, that was fun setting up this forum...

A few questions:

1) Have you read through the schematic and voltage and timing description on
the wiki?

2) Have you read the blog, (which talks about the Defender)?

3) Do you have an oscilloscope to read the voltages on TP16 (Temp/) documented on that page? If so, what happens and when?

If at all possible, measure the TP16 Temp/ voltages with a scope. If the 2nd stop voltage does rise to 3 volts, then combustion is failing (or the temperature is dropping). If not, something else is going wrong. Measure the other test points to see if these other sensors are tripping. Regrettably, all this testing needs to be done with the case on, otherwise perhaps the trap will not heat up enough.

According to the timeline on wiki, the unit is warmed up and stable in about 10 minutes. But then, yours faults after another 10 minutes at 20 minutes. That sounds a lot like what the Defender was doing last year as the case temperature in the vicinity of the valve and nozzle heated up and somehow caused the gas to stop flowing. The case temperature was measured using another temp/humidity module attached to the NodeMCU controller used in place of the blown up PIC. Some problem with the valve was discovered by bypassing the valve mechanism by removing the "plunger" and sealing the hole with tape, then using the tank valve as a substitute, which is a dangerous move that cannot under any circumstances be recommended.

The Liberty valve is different from the Defender. It has a connector in the valve electromagnet assembly. This and the mating connector should be thoroughly examined, ideally under a microscope. You may have to unsolder the wires to the PCB. In any event, spray a contact cleaner like Deoxit after mechanically removing any contamination. You can measure the solenoid current by unsoldering a wire and inserting a milli-ampere meter (i.e., a VOM). Bypassing the valve even for testing would be dangerous and therefore not recommended.

The June 19, 2019 blog entry showed another anomaly which was not fixed by bypassing the valve. That was fixed by blowing out the system using a tire pump, or the "Gas Reset Tool," or both. If it was the high pressure blowing that did it, perhaps there is something somehow stuck in the path between the valve and the nozzle. This will be examined following mosquito season (the basket last filled entirely after about 2 weeks). Although you checked and cleaned the gas lines to and from the valve, the path inside the valve may somehow being blocked. High pressure air may remove this or jam it even more, but perhaps you will know from the tire pump pressure gauge what's happening.

High temperature may be causing your PCB to fail in some way. That's why it is important to measure TP16. Normally, it is low temperature along with high humidity that causes problems from contamination (dust) on the PCB conducting and giving false signals. It may not hurt to remove and thoroughly clean the PCB assembly, washing all the dust and mosquito parts off, etc.

That's it for now. Good Luck!

Wow.

Ok, I read the wiki but didn't see the blog. I'll read it now.

No scope at the moment, but I have a good dvom for test voltages. I'll have to pull the cover and clip on some wires so I can read voltages with the cover on.

In my travels, I'm going to take a close look at the valve. I have other "parts" units and I know they used at least 2 different types of valves. Perhaps the one in there was problematic? I do have DeOxit, so I will be using that. It's excellent for restoring contact operation.

Thanks for the update and the help. I'll post what I find!

Yes, you will need to borrow a scope unless cleaning the valve contacts works for you. The TP16 voltages are visible as a stairstep signal, that have different values each for a few milliseconds, and then they go off (to zero?). However, the overall average as read by a vom will perhaps change enough to give you a clue as long as it can average satisfactorily for this purpose.

Well, I think I'm on to something...

If I leave the cover off, the unit runs nicely. Once I put the cover on, something is getting hot and doing something to cause this. That would also explain the way it lights, runs and then shuts the gas valve off after a short time.

My unit is outside, cover off and running for over an hour now.

I notice that there is no real facility for ventilation of the electronics. As a matter of fact, there's no gasket on the out-facing fins of the combustion chamber, so once the outer screen is put on, that adds resistance and I am sure some of the heat which should be exiting the trap is diverted inward. That being said, I put some foam weatherstrip around it and of course, more heat comes out the side, but not enough to keep the trap going.

It either needs ventilation internally or there is something going into fail mode from the heat.

What do you think?

The temps inside the traps get pretty hot. Here is the instrumented Defender just now:

Temp=128.5C (0) at 5:7674 F=1 I=0 G=1 S=1 E=0 T=58.1C H=9.2% M=10296 (204/1014:4741)

The combustion chamber is 128.5°C, the inside case temp is 58.1°C (9.2% RH), pretty hot, and it is very warm 32°C outside.

I think you need the case on to catch mosquitos, no? In any case, you have strong evidence for case temperature being a factor in the failure. The electronics (PCB) are supposed to work at fairly high (for us) temperatures.

You have to do some measurements to tell if the electronics are failing. The simplest is to insert a VOM to measure the solenoid current, run wires outside the unit and run it with the case on. If the current goes to zero before the unit displays an error message, then something in the solenoid circuit is opening up at high case temp (a problem with the connector, or perhaps some intermittent trace on the PCB). Otherwise, the electronics is shutting the valve off because of some perceived fault (e.g., the temperature dropping because of some mechanical blockage in the gas line). One problem with this approach is that the liberty measures the solenoid current at TP22 (Gasl). If this signal goes down, it means the solenoid current is too low, and it should trigger the "solenoid not working" error code. So, the solenoid current going to zero will trigger a fault, but I would hope it would not be the "Gas" fault. According to the fault codes, the "solenoid not working" code is solid red, flashing yellow & green, which is different from flashing green, solid red, yellow off.

(I don't understand "... I put some foam weatherstrip around it and of course, more heat comes out the side, but not enough to keep the trap going." The temperature that matters is measured by the thermistor, which comes from inside the combustion chamber, which is much hotter.)

By the way, your new regulator, does it provide 11 column inches of pressure like a standard device? Is this the same as the original regulator? Perhaps you could measure both using a gauge. I don't know the spec. The trap may require more or less pressure.

Did you try the troubleshooting steps in Mosquito Magnet Depot?

Wow. Lots of stuff...

In a nutshell... I think, or rather now I know that this is a mechanical issue that influences an electrical issue.

In its stock form, the fan has a weatherstrip or gasket which goes between it and the cover. That can shrink and the gaps cause 2 issues; 1 is that it loses vacuum from the catch area and 2 is that it creates a draw over the pc board which is pulling heat back from the combustion chamber.

It is not too scientific but, I just got back and my unit has been running without a hitch.

I drilled a 3/4" hole in the bottom case behind the schrader valve. Then I purposely opened up a small area of the gasket around the fan closest to the pc board so it can pull fresh air from the bottom hole, across the pc board and back through the net and out the combustion chamber fins.

I also put closed-cell foam weatherstrip (may have to revise the material later...) around the combustion chamber where it meets the upper and lower casing so that the fan can't draw back any heated air.

Not scientific, but it worked. The unit is operating. We went to dinner and it's still on the back porch, fully-assembled and a solid green light on.

The regulator I used is adjustable and has a pressure gauge on the magnet-side, not the tank-side. I set it for about 1 psi. I worked down to the minimum it would fire at. Without pulling out the manometer, I'd say this is a good setting. The stock regulators get oxidized inside and then they clog things up and fail. Exact replacements are hard to come by so I just used an adjustable. It works.

Based on the size of the heatsink on the regulator in this original version versus the much larger heatsink on the newer version (which also has a different style gas valve), I'd say that regulator heat may have been (and may still be) an issue.

This is why when I repaired it and bench tested without the cover it worked reliably. I used my IR thermometer to check the chamber output temp and temp around the thermistor. They were always spot-on, but when I assembled it and put it out to work, it would fail in 20 mins. I thought it was air in the propane tank or just a bad tank/valve or something in the line, etc., but this does seem to have fixed it.

So it's a kludge to be sure. A band-aid, but I'm heading outside now to move it into position with a fresh attractant cartridge....

I guess if anyone has a unit that has all known-good items like a nozzle, igniter, thermistor, tank, power supply, clean chamber and lines, etc., then perhaps running without the cover may be revealing.



Edited 1 time(s). Last edit at 07/20/2019 10:23AM by liberty1.

Just an update...

Ran all night. Catch in the bag. I am a happy camper.

(addendum)
I just got done reading your blog. I think you were definitely on to something with the internal case temp causing issues. Something affected by the case heat causes the solenoid to release, whether it's the solenoid itself of some other item in the chain. The interesting thing is that none of the electronics has to be located inside the unit, so an external mounting box (vented?) is an option too. My hypothesis, since these work ok out-of-the-box and fail over time, is that the fan seal and lack of sealing around the combustion chamber leads to heat being drawn to the electronics as the unit ages. Then we see the failure. By adding seals around the combustion chamber where it meets the case, I've significantly reduced the amount of heat that can be drawn back into the unit, resulting in an increase in net (bottom-line) heat output through the fins. Drilling the hole in the bottom of the unit creates a cooling air intake, so any vacuum present due to the fan operation will no longer pull heat back from the combustion chamber, but will draw fresh air through the case section where the electronics are located and then send that air back through the net (catch) area and out through the combustion chamber to exit the trap.

This is a restatement of what I did, which really just builds on your concept of the internal case heat causing an issue. Once the internal case heat is taken out of the equation (on my Liberty), it seems to work like it used to.

If your Liberty is still assembled with known good parts, then I'd say that if you have some free time, try it with the cover off, just to see if you get the same result I've gotten.



Edited 1 time(s). Last edit at 07/20/2019 11:07AM by liberty1.

Great that you got the trap working. For me, it is more important to catch than investigate. However, in the likely event it fails again, or for future reference...

If you are using 1 psi, which equals "27.6799 inch of water column," that would be almost 3 times the typical regulator output pressure of 11 inches. I then have to wonder whether this means more fuel is being used for combustion, and whether there is sufficient air being supplied by the fan into the chamber to completely burn the fuel. I remember that the MM patents say that unburned propane is a mosquito repellent, so they wanted to make sure that combustion was complete before exhaust. Also, that might make the chamber temperature too hot, which would yield an over-temperature error, and shut down the trap. This is another thing to look into. Finially, how long does the tank last? Perhaps it is time to measure those regulators and see exactly what pressure they are delivering.

You did these changes in the evening. Let's see what happens when the outside temperature goes up today. Also, I found that sunlight hitting the unit caused a large increase in case temperature. You might want to keep your trap in the shade under a tree, or an umbrella. I believe the mosquitos like that better, too.

I remember I had to resolder some wire on the Liberty controller PCB -- it might have been the thermistor lead. The manufacturing technique was not the best, with contaminants causing unwanted conduction in the high-impedance thermistor circuit. I had to clean (with water and alcohol) and reseal the Defender circuit board a few times. Plus, I resoldered several connections that looked suspicious (deteriorated). These issues could be causing the controller to fail. Remember, it is required to work even at pretty high temperatures. New, it works. Later, it starts to fail. Intermittents only get worse with time. The next time it fails, consider inspecting the PCB for these problems.

Temp=130.3C (-1.8) at 5:33708 F=1 I=0 G=1 S=1 E=0 T=61C H=7.4% M=9608 (197/1016:4528) R=45~140 B=0.0 V=3.1

Just now, the instrumented IOT Defender case temperature is at 61°C in the sun, not in the shade under a tree. Even at that location, it make its best catch ever.

After you catch some mosquitos and get some free time, it would be great if you could upload a how-to on your modifications, with a few pictures.

Do not get discouraged if the trap fails again. Just fix it and keep catching. Mine has failed perhaps a half-dozen times. It's part of the fun of being a Mosquito Magnet owner.



Edited 2 time(s). Last edit at 07/20/2019 03:02PM by dev.

Just an update...

'You did these changes in the evening. Let's see what happens when the outside temperature goes up today. Also, I found that sunlight hitting the unit caused a large increase in case temperature. You might want to keep your trap in the shade under a tree, or an umbrella. I believe the mosquitos like that better, too."

You called it. As soon as the temp went up yesterday, it failed again. I also renewed my understanding of the psi-wc ratio and you are 100% correct. I have to change that.

I'm all about the catch too, believe me. I live next to wetlands. The sad part is that when this worked, it worked very, very well and I advocated their use and showed people pictures of the catch. No pesticides. I'd feed the catch to fish in the local canal. They went fast.

I'll review everything, including the solder joints, etc.

Thanks. :-( Who knows... perhaps soon it will be :-) . I definitely want to get these going so they work reliably.

Sorry for the not-unexpected news. Did you get any mosquitos? Actually, when I was having this problem every day, the trap would not work during the heat of the day, but I could restart it around dusk and it would run until late morning. This was sufficient to catch some mosquitos. Some are better than nothing.

FYI some temperature readings

Temp=37.7C (-0.1) at 7:12693 F=0 I=0 G=0 S=1 E=7 T=35.4C H=49.5% M=5744 (869/1014:112811) R=45~140 B=0.27 V=3.1
Temp=121.9C (1.2) at 5:38345 F=1 I=0 G=1 S=1 E=0 T=49.8C H=17.3% M=9784 (235/1014:5678) R=45~140 B=0.0 V=3.1
Temp=129.2C (1.7) at 5:50486 F=1 I=0 G=1 S=1 E=0 T=56.1C H=10.8% M=9784 (201/1014:4654) R=45~140 B=0.0 V=3.1

The first entry is after 12693 seconds of a fault, with everything off. The case temperature is similar to the combustion temperature, and both are high (the unit is in the sun). The case temperature should be the more accurate because it comes from a temperature/humidity module with a specified accuracy. The combustion temperature is a calculation based on the specs of the thermistor. The accuracy is sufficient to detect combustion.

The second entry is the lowest case temperature measured after a remote restart. 49.8°C is still pretty hot. The third entry is more recent indicating overcast conditions. I expect the case temperature to reach the mid 60's today, which, who knows, may cause trouble. We shall see. I should put the trap in the shade, but that would increase the distance to the wifi, and the wifi connection is not 100% where it is now.

It gets hot inside the case. You can test what happens with the case open by using a hair dryer and an oven thermometer to heat different parts of the unit to see what fails at high temperature (I wish I had thought of that earlier). Don't over-do it, but you should be able to go up to 70-80°C and see what happens.

Good luck, and feel free to keep me posted using this forum. The daily struggle logs make an amusing read later on during the winter.

Don't forget blowing out the lines and valve with the tire pump if you can, and that "gas reset tool," which is what worked for me the last time.



Edited 1 time(s). Last edit at 07/21/2019 05:43PM by dev.

I just had a problem with the Defender not starting (low temperature, same as shutting down), and it turned out to be, arguably, an obstruction in the gas line. It was fixed for now by attaching an automobile tire pump to the unit and running it for several minutes. See the blog. This might apply to your situation.

I got sidetracked, but I'm working on it in spare cycles. I read the blog update (thanks). I use my compressor at 150psi and it doesn't seem to make a difference. It will also light every time, so I don't think I have a blockage. I found the 12v regulator had almost no thermal paste and what was there had dried up to a powder, so I replaced that, but no joy.

I think that this would benefit greatly by replacing the microcontroller with a pi so I can see everything and write my own control software. I'm not sure where I could find relays and enough adc's on one board though. That would surely make these much more reliable and it would simplify debugging.

In the meantime, I'm leaning towards something that heats up and fails and lets the 12v to the gas valve drop, unless the chamber temp is just going too high. I am taking the first step, which is changing nozzles to get to the minimum size. The variable regulator was replaced with a 11" wc version. I'm at 1.25 45 degree extra solid and it still does the same thing. Runs to green and works fine (with or without cover) in cooler temps. As soon as the air temp comes up, off it goes.

I am also wondering if this is part of the detection mechanism thinking the trap is overheating. With a constant supply of flame from the gas and higher air temp, we're going to see a varying temp on the chamber itself. Gas is binary if I'm not mistaken, so it can't throttle the temperature. I can't make it fail with the top off, so it is also possible that the transistor that controls the gas valve is failing from heat or the fan doesn't cool the chamber off sufficiently in warm air temps. It is difficult to debug voltages since it will not fail with the top off. I'll have to make some sort of harness so I can read them.

I also thought, rather than bypassing the gas valve as you mentioned before, I can bypass the transistor and enable the gas valve directly with an external power supply and then see if the trap still fails by watching the 12v on the plug coming off the board, although I'm not sure of the logic when the trap goes into fail mode, does it drop the ignitor voltage too? I bet it does, but again, not sure.

I wish I had a scope and more time. I am envious that you are getting full-bag catches. That's awesome!

[Full bag catches are awesome sure, but in a strangely shocking and appalling way. I don't want revenge (much), I just want to be free of mosquito bites when outside. For this, the best catch is small and steady, indicating that the local population has collapsed, and just a few outliers from surrounding areas are being caught. These would be outliers that could otherwise start a new local population. It is so important to keep running the trap at the end of the season, and also at the beginning even though the catch is small.]

A few notes:

The igniter is on only until the trap combustion temperature has reached a certain value, or for a certain time, e.g., 1 or 2 minutes. The combustion sustains itself without the igniter. Igniters have a short lifetime, like a light bulb: some number of hours and some number of cycles, then they fail. So the igniter will be well off when your trap fails.

This may be the time to try the heat gun/hair dryer on each part in succession and see which part of your trap fails with higher temperature. Shoot for 80°C, that should be enough.

Are you sure that you are getting the Gas is empty / Insufficient gas flow out alarm and not the High Temperature Reading alarm? Of course, the software may not blink the correct code because of some bug. This is where a scope would come in handy to measure the temperature. After your previous message, I looked into ultra-cheap scopes for sale. There are a few miniature scopes in the $100 range which may be enough to troubleshoot this problem.

To measure voltages with the top on, solder an insulated wire to a test point and bring the wire outside the trap to a convenient location. Use a small flexible wire so that wiggling the wire does not put any force on the PCB.

As far as the valve circuit is concerned, bypassing the valve transistor does not catch problems in the valve or valve contacts. You could unsolder one of the leads to the PCB and attach two test leads, one to the unsoldered solenoid lead, the other to where the lead was connected. Then put an analog VOM in the circuit and measure the current. If the drive or the solenoid is electrically failing, the current will go to zero before a fault condition (it takes about a minute for that to happen after the valve closes).

All the other measurements would be voltages to ground, so don't forget to attach a ground lead.

Another topic here is from a Defender user who used a board to replace the controller. I just added a $4 wifi computer chip module to mine attached by 9 wires just hanging off the original controller board, from which I had removed the PIC micro controller (which I had blown up because of carelessness with external wires as I described above). I just had mine become intermittent after about a year of service, it was contaminated with mosquito body parts, plus general dust and awfulness. Plus the module pins were corroded somewhat. I will analyze the intermittent unit later when time permits. My point is that the trap is a hostile environment for electronics. The controller was sealed with a PCB sealing coat, which protects the circuitry from the elements, to some extent. Fortunately, I just de-oxidized the connectors, then plugged in the backup unit, and was off and running.

Good luck. I had a terrible day yesterday myself, with a failed trap that I finally got working. I think it was the nozzle, or the igniter. I replaced both, now it is working. I will write a blog post on the topic soon.

"As far as the valve circuit is concerned, bypassing the valve transistor does not catch problems in the valve or valve contacts. You could unsolder one of the leads to the PCB and attach two test leads, one to the unsoldered solenoid lead, the other to where the lead was connected. Then put an analog VOM in the circuit and measure the current. If the drive or the solenoid is electrically failing, the current will go to zero before a fault condition (it takes about a minute for that to happen after the valve closes)."

That right-there is a great idea! Yes, it would take time for the gas to be exhausted after current flow stopped (and valve closed). I also like the blow dryer idea.

Whatever it is that is causing this, I think I have the same issue in my other Liberty as well.

Perhaps I can adapt an arduino like the other fellow used on his Defender. If I could, then I'd be foolish not to take him up on his offer for source so that I can get mine working again and have better diagnostics for future potential issues.

As far as scopes go... wow. I just looked at some of those. I guess I've been asleep for a while. They look pretty cool.

Answering a question posted to the Mosquito Magnet Defender thread.

"Read it all. So should I be using the MEGR-230 on a Liberty, and what actually is the correct nozzle for a Liberty?"

Regarding the Marshall Excelsior MEGR-230, this regulator is designed for 140,000 BTU/Hr, and is larger than the IGT regulator discussed in the blog. I am worried that it may not be as suitable for low-flow applications as the IGT regulator and others of its type. On the other hand, other purchasers seem to report that the quality of the device is higher than the cheaper regulators, so maybe it would last longer. I have no way to tell. Regrettably, it requires two adapters, one 1/4" to ACME nut to connect to the tank ($10), and the other to adapt the 3/8" outlet to 1/4" ($6) to match the existing hose, so it ends up costing $30. The hose sticks out of the tank a few inches (see the blog photo). But it might be somewhat adjustable, which might be important for your Mosquito Magnet Liberty. The one from my liberty was about 11 inWC, so that might not matter.

All in all, I prefer the IGT regulator because although it results in a too-long combined hose, it wraps nicely around the tank and does not stick out, and the single $4 adapter results in only $14 total expense. It may be reliable enough (the Mosquito Magnet Patriot uses it). In short, no clear answer here from me.

Regarding the nozzle, one important step is for you is to first measure the air pressure via the Schrader valve, then disassemble the trap, make some photos, and the nozzle by soaking in brake cleaner (acetone and toluene) and / or mineral spirits for a while, then blowing it out again to achieve a lower pressure. This may be all you need.

The reference I quoted earlier stated it is a 1.50 GPH 45° "A" WDA. I purchased some and used them mistakenly in the Defender, but I didn't like the results, as it resulted in instability, going into an incomplete combustion mode when there was a pressure disturbance. It took a long time to find this problem. If MM did indeed use the WDA nozzle, it might not be the best choice. And the WDA nozzles are 4x more expensive. The source suggests using the Defender nozzle, and seeing how it works. Again, no clear answer.

Hope this helps. Good Luck!

Thanks very much!

I saw your latest update on the defender project. Amazing progress.

Will any of your work be applicable to the Liberty? I now have 2 non-functional units and 3rd for parts, and plenty of mosquitoes to eradicate (hopefully without pesticide). Thanks.

I saw your latest update on the defender project. Amazing progress.

Will any of your work be applicable to the Liberty? I now have 2 non-functional units and 3rd for parts, and plenty of mosquitoes to eradicate (hopefully without pesticide). Thanks.

I think it should work. The Liberty has several sensors that would not be used. The Defender uses just the combustion chamber temperature for all of its input, but that one measurement tells whether the trap is working or not, and what to do about the fan, propane valve, and igniter, which is all you need. Refer to the wiki Liberty schematic. Click on the schematic, download the original version for a full resolution image, and open the image.

The Liberty measures the igniter current to detect an open (or shorted?) igniter. The Defender stops after a few minutes if the temperature fails to rise. Emitting unburned propane for a short time was apparently not a concern as long as it was used only outdoors, as labeled.

The Liberty measures the fan rotation speed. The Defender uses an open-loop approach of providing 1/4 power to make a rich mixture during initial ignition. The fan seems to be reliable, but, if the fan is insufficient, the combustion temperature gets too high, and stops the Defender.

The Liberty measures the propane valve solenoid current, stopping if the solenoid fails. The Defender stops if the combustion temperature doesn't reach a minimum after a few minutes.

The Liberty measures combustion temperature as being in one of 4 temperature ranges only. The system cannot calculate a temperature rise or fall trend. The Defender's 4 range temperature circuit is a great improvement over the Liberty, which is important given that the Defender lacks the other sensors, and must make decisions based on temperature changes only. Both units have provisions for catch basket sensing, but the Defender does not use a sensor.

Otherwise, the Defender and Liberty are quite similar, except that the Defender has an illuminated on/off rocker switch, and the Liberty uses those awful membrane pushbutton switches. If the power goes off for a second, the Liberty will stop and not restart, a huge negative. The Defender, on the other hand, relies on an initial cool down cycle upon power application to restart the unit, which I prefer. The Liberty uses a larger PIC controller with more pins to handle the extra inputs, and drive the 3 LEDs.

For me, the Defender's approach to safety and efficacy is adequate. The lack of sensors is more than made up by the 1 per second telemetry, from which I can diagnose a problem. One could even set up an alarm on your PC, but there are other things to do, such as enjoy the outdoors.

In order to use the Defender Add-on, you would have to add a weatherproof on-off switch to the unit, and remove or disable the PIC. I believe there is adequate space for the add-on, so the project looks feasible. I have a currently unused Liberty, but am reluctant to add the switch or modify it. Sounds like a lame excuse, doesn't it?

If you want, you can make a Liberty version of the controller add-on. Start "Liberty Add-on Project" and "Liberty Add-on Construction Details" wiki topics, and a corresponding new topic on this forum. You can begin by following the Defender Construction topic up until "Build the Wiring Harness."

You will find issues and differences. For example, to disable the Liberty's PIC16C55, carefully unsolder and lift pin 28 (MCLR/Vpp), and connect a very thin wire to pin 4 Vss. You should install miniature in-line connectors for the devices that are now soldered onto the controller board.

You need access to soldering and other tools, a meter, an oscilloscope (very helpful), and a suitable magnifier or microscope.

Ask for support using the new forum topic, or the Defender Add-On topic. Put what you know into the new wiki pages. Good Luck.

That sounds like a great project. It is going to take me some time to free up to get to this though.

BTW, your not wanting to start on your Liberty is completely understandable. You have been able to shed a lot of light on the Defender and how to get it working well and with a very intelligent setup. Why bother with the liberty, except just to have fun exploring a great product idea destined to spiral down to its death while frustrating the hell out of the owner.

I wish there was one that just worked and stayed working.