Early Ice on the western shore of Lake Erie 11/24/14
Short Story: My 1999 Mercury Marquis's fuel pump died. Since the fuel pump is in the gas tank, the tank has to be drained as much as possible to get the pump out.
The standard procedure to replace the pump is to drop the tank, replace the pump, and re-install the tank. Since I'm lazy I am going to try and get the pump out without dropping the tank. (Supposidly it can be done - I'll report on this)
So back to removing the gas from the tank; You would think that this would be a non-issue but the car makers have created issues for us DIYers.
Two things come to mind immediately.
1. Remove the drain plug and drain the tank. Problem: There is no drain plug. (they probably saved 50 cents by doing this)
2. Insert a hose and siphon out the tank. Problem: The Ford nitwits decided that you should not be able to do this so there is a device in the tank to prevent this.
The Ford nitwits overlooked one thing. Although a 3/8" id fuel hose will not make it into the tank for siphoning; a 1/4" outside diameter poly/nylon tube will get through the antisiphon device and allow a pump to be used
to remove the fuel from the tank.
Carter Pump attached to extension cord for easy hookup to battery (every motorhead must have one for stuff like this)
(That is not gas on the pavement, that is water as it was raining on and off.)
Battery with extra wire - note that I reused the plug and socket so I could disconnect power quickly without pulling off the battery clips.
The only way to get the gas out! See the 1/4" diameter plastic hose running into the tank. Its stiff enough to be able to snake it past the antisiphon valve (an idiots idea).
Below is the 3/8" ID rubber hose that connects to the Carter Pump and the connection to the 1/4" OD plastic line. I wrapped some electrical tape around the 1/4" line to make it bigger so I could slip
the 3/8" ID hose over the taped end and secure it with a hose clamp. Not a permanent want to connect a 1/4" OD line to a 3/8" ID hose but it works for a temporary jury rigged setup.
Transferring 18 gallons of gas out of the tank (it was full) took about 20 minutes.
Changing the Pump
The standard method for changing the fuel pump is to drop the tank, but that is fraught with problems. For one, the tank straps where too short to begin with so reinstalling the tank with the same straps is near impossible. (Gathered from many sources on the web.) So replacement straps and studs have to be purchased. I found a Ford Tech who said the pump could be removed in place. I jacked up the car after draining the tank. Removed the fuel lines that conect to the pump plate on the front of the tank. I cut the wires as the connector is apparently above the tank.. :-( Further evidence of Ford nitwits. I removed the 6 screws holding the pump assembly in the tank and pryed the plate loose with a large screwdriver. The plate, pump and level sensor came out without an issue. I moved the assembly up and over the rear axle, then down and out of the car. Seriously, no big deal. I'm obtaining a replacement pump (not the entire assembly) as the assembly I have is in very good condition and I don't want to risk the sender not working in a replacement assembly. I ordered the parts from Amazon. I ordered an Airtex E2471 and a new infeed strainer along with another fuel filter. The total was about $75.00. Everything should go back together easily.
OK... I received the pump from Amazon along with the strainer (which is not included with the pump). The total bill for parts was about $80. I removed the old pump from the in tank assembly, installed the new pump along with a section of hose and new Stainless steel worm gear clamps. A wiring adapter is included in the kit. I put some gas into a pan and set the pump assembly in it and ran the pump off a spare car battery to make sure it worked before I put it into the tank. I then stripped the wires that I had cut about an inch and a half away from the intank assembly and crimped on insulated butt connectors on the flying leads. This is easier done on a bench than under the car. (Make sure you do a slight tug test to make sure they are secure.) The in tank assembly went back in in the reverse order that it came out. Up and over the rear axle, pickup filter first, then put the float into the tank hole then slide the rest of the assembly into the hole. Secure with the 6 original screws. Then I stripped the wires ends that I cut that were part of the gas tank harness, matched them up with the wires with the butt connectors already installed and crimped the harness back together. I used some liquid electrical tape to seal the crimp joints to eliminate corrosion problems in the future. I installed a new fuel filter, although the old one was probably fine. I put some gas (new gas) into the tank and purged the system - via key on, key off with the fuel line disconnected at the engine injection manifold. When gas came out of the line to the engine I reconnected the fuel line to the engine injection manifold. I started up the engine and checked for leaks. No leaks. :-)
If you know how to do this change out, it is easy. The idiots on the web and Youtube always drop the tank which is obviously a waste of time. (A perfect example of the Lemming effect). The deals wants $750 and up to do a pump replacement. Total cost was well under $100 plus my time and it took about 3 hours, but I was very cautious. (Remember, this is not suppose to be possible! ) If I had to do it again, and I had the parts on hand, I could probably do the swap out in about an hour.
Airtex makes oem pumps for the car companies and I have never had a problem with their pumps. I would also trust Carter and FOMOCO of course.
One more thing. In the past, on other cars I have replaced the entire in tank assembly. What I have found is that the gas tank level sender is oftentimes not the same so that screws with the gas gauge. Using the original in tank assembly along with a new pump eliminates this possible snafu. Also, the in tank assembly on this car is stainless steel and very nice. The replacements I have seen are galvanized steel of lower quality.
Here is a Babington "style" waste oil burner I have been working on and off for a couple of years. The Babington style burner design has been beat to death with little concrete results documented on the web for almost 10 years now. One article even declared the burner design a failure. You can find all kinds of Youtube videos of nice round balls or door knobs that are atomizing diesel fuel at 75 degrees F which is totally unrealistic. If it is 75 degrees then why are you running a burner?? You need to flip on the air conditioner!! The only thing available for burning real waste oil is Turk style pot burners which require maintenance, and a noisy electric blower powered by electricity. This burner makes flames and heat... The dirty oil does not pass through any orfices, so nothing to clog. I didn't try and get the oil film extremely thin on the nozzle after all we are burning waste oil, not diesel, sun flower or preheated veggie oil.
No electricity is required to run this burner if you use a gravity feed. No compressed air is required to run this burner. So your compressor will stay quiet. (These concepts were later revised as development continued.. continue reading).
(SEE ADDITIONAL CONCLUSIONS AT END OF ARTICLE THAT RESULTED FROM FURTHER TESTING)
What is required is propane at 1-2 lbs per hour depending on how hard you want to run this burner. Remember this burner can be turned down! Total BTU output is between 60 and 300K BTU based on actual testing. One lb of propane is about 1/4.5 x 90 = 20,000 btu. So what happens is that you pay for propane at 1-2 lbs per hour which is 20-40,000 btu of propane and along with that you also burn up to 2 gallons of waste oil along with that. So the total heat output is actually a near 300,000 btu when running full tilt. The cost to run this burner at current local low volume propane prices of about $2.40/gallon is 2lbs/4.5lb/gal x $2.40 = $1.06 per hour for 300K btu assuming you get the waste oil for free. If you were to run just propane to get this much heat, the cost would be 300K btu/90K btu/gallon x $2.40 = $8.06 for 300K btu. I can live with just over a dollar per hour for 300K BTU of heat. That is very cheap heat. To put this into standard terms of dollars per million btu: 1million btu/300K btu x 1.06 = $3.53 / million btu of heat. (The final results are even cheaper than this.)
Not to put this into context check out this website: Fuel Cost Comparison So this is less than half the cost of heating with bulk coal (if you can buy it) and less than the cost to heat with wood if you can buy it for about $50 per cord!
I have been testing this design in 15 degree F ambient temperatures. This design is realistic and usable and doesn't require constant tweaking and fiddling to get it to run right. The amount of oil fed to the burner can be varied without creating smoke meaning that you can turn this burner down for use in moderate temperatures or turn it up when it is really cold.
This burner uses propane to atomize the waste oil, not compressed air. Compressed air works ok if the conditions are just right, the oil is preheated, and the oil is not too heavy, blah blah blah.... In other words compressed air is simply not practical! (REVISED AFTER FURTHER TESTING - see below)
Stay tuned, as this burner will be made available as a kit on Ebay at a reasonable price for experimental use. (IE, if you light your house on fire... don't blame me..as you are playing with FIRE! )
The kit will consist of the most of the parts in the picture. You will need to supply a propane cylinder with gas. A 20 lb unit will last for 10+ hours. And you will need to supply a source of waste oil under some slight pressure to feed the steel pipe tube that is above the pipe tee in the pictures. A gravity feed from 3 feet above the burner works fine. Optionally you can use a small motor driven gear pump to meter oil to the burner (recommended) or a peristalic pump.. That will allow you to pump the oil out of a container to the burner and then the unburnt oil can recirculate back to the tank.
Below is a picture of the burner running on REAL waste oil. Most of this is 15W-40 and 10W-30 conventional oil with some synthetic added in. It is throwaway oil from engine oil changes. Those are bright yellow, whiteish flames with virtually no smoke.
Here is the burner preheating on propane only. It takes about 3 minutes to bring the burner tube up to temperature (700+ degrees, which is required) in order to burn waste motor oil without smoke. This is with the propane setting at about 1.5 lbs per hour. (Notice the snow to the right... it was cold! )
Uses for this burner:
Put this into the end of a 55 gallon drum with exhast stack for a space heater.
Weld a large pipe or cylinder into a 55 gallon drum or a 275 gallon oil storage tank and create an ambient pressure water heater. That water can be used to heat your house or shop or both!
Here is a 2" black iron floor flange that I have bored out on a lathe so the burner tube can slide throught it. I am going to use this to mount the burner to a steel bulkhead on a wood burning stove. This could be valuable for many installations such as using this burner with a 55 gallon barrel for a space heater. This could also be used to adapt an existing wood burner or boiler to use this burner. I think the threaded flange was $12 at Home Depot.
A muffler exhast clamp could also be used to mount this burner tube.
The bored out iron flange with the burner tube slid though it.
Some type of flame safety needs to be implemented if this burner is operated in an enclosed space so that if the flame goes out a hazardous situation is not created. I may elaborate on this later, but a flame safety unit from an old oil heater could be used for this purpose.
After doing further reasearch on burners I found that gas assisted oil burners are rather rare. However there is a patent (which has expired) to use a gas assisted burner to burn crude oil in artic temperatures as a means of getting rid of waste crude oil. Waste oil is probably as uncertain in viscosity as crude oil. The patented deviced use a circular burner ring of gas flame around the oil jet of some sort. The patent is actually a little vague. If you do a google search for "gas assisted oil burner" it will pop up.After all who knows what they dump into the waste oil tanks at lube shops. A mixture of oils and antifreeze is not unusual.
I did find a bunker oil burner design patent that used two gas burner jets on both sides of the main oil burner jet but keep in mind that this was for high volumes of oil. Not 1-2 gallons per hour. Probably more like gallons per minute.
One of the challenges with any burner design and control is safety. It is normal to purge the combustion chamber prior to a startup to get rid of fumes that might ignite and explode. However waste oil burners (especially the one described above) is not really adaptable to standard oil furnaces. The flames are very long ( heavy oil ) and the burner tube is also very long (24" in the picture above). While a shorter tube is possible, I think that might cut down on effciency and make the burner more susceptable to smoking and limit the ability to "turn down" the burner. Right now this burner can be turned way down without smoking.
Oil feed systems: A gravity feed system for a burner that will be attended that is using propane to atomize the oil is reasonable, but it must be attended! If the oil runs out, the propane continues to run, but there will be reduced heat output. But no big deal, the person attending to the burner should be able to detect that there is an issue when the heat output is reduced and make corrections. For unattended operation (similar to the Murphy Machine automatic boiler design) a more reliable way of maintaining oil feed is needed. A constant flow system would be best so that it can be set and forgot. The most reliable way of doing that is with a gear type pump such as an oil burner pump! I have done testing with an old oil burner pump made by Suntec. In particular their J type pump which has apparently be made for decades. The pump I have will deliver a constant output flow when fed via gravity. I tried driving the pump with a variable speed drill at about 500 rpm. By varying the speed I can get a nice variation of oil flows. I have a small 3 phase motor that I will couple to the Suntec pump's 7/16" input shaft and then control the motor speed via a single phase input VFD. Another way to do this would be to use a standard 1/2" drill chucked onto the pump shaft and use a speed control unit to turn the drill speed down. An entire setup like that could be bought from Harbor Freight for probably $75. However I don't know how long a Harbor Freight 1/2" drill would last being run continously.
So I am rethinking a couple of things and these are the questions I have..and some conclusions:
1. Can a gas assisted burner be used along with a babbington type ball atomizer be used to get the best of both worlds?? IE, reduced propane usage (below the current 1-2 lbs per hour) with the benefits of more complete combustion by the use of propane.
Conclusion: After some testing: Not easily and not without making the burner MUCH more complex. So I have stopped pursuing that.
2. Can the propane jet be turned down when not running oil in such a way that it can be used as a pilot light? A couple of electric valves and two propane pressure regulators could be used to implement a two setting propane burner. This would facilitate unattended operation. Turn the gas on high, preheat the tube, add oil and burn at full output. Then when temp is reached, turn off the oil, wait for the oil to clear the burner, then reduce the gas flow and maintain a pilot.
Conclusion: (Revised) No, using the burner in a very large stove, the low pressure flame via the main jet becomes unstable at "low pilot level flows". This was a surprise since this was not a problem when the burner was running in the open.
3. Can an induced draft be created in the furnace stack using compressed air to "suck" any vapors that are present out of the furnace? (Think "air jet" in the center of the exhaust stack pointing in the direction of the desired flow.) Using an air jet to purge the combustion chamber is: a. Simple! b. cost effective Using 20-30 cf of compressed air to purge the chamber a couple of times per day is much cheaper and simpler than installing and maintaining a blower do pressurize the chamber to blow out any fumes.
No conclusion yet.
4. Is there any real advantage to preheating the oil to a constant temperature?? As long as the oil can be pumped, so far the answer I have for that is no. I can perceive no benefit But this has been a nagging question. The guys making burners to melt aluminum and iron have decided that preheating waste oil has no benefit for them but this is a different application. Preheating the oil to 150 degrees or so with electricity is fairly simple so it may be worth doing some experiementing.
No conclusion yet.
5. Why does it appear that some people can just use air with their babinton type burner?
Conclusion: They can't! Read and watch closely and you will find that they have cut the oil with diesel or kerosene or paint thinner. Or they are operating the burner in warm weather with warm oil. Not realisitic. Try and find someone who runs a babinton style burner routinely to provide heat. I can't find any. (REVISED - see below)
6. What about Delavan suction nozzles??
Conclusion: Many commerical burners use this nozzle. Also you will find many very frustrated users of these burners. That nozzle can only suction oil up 1/2" below the nozzle level. Any blockage or change in viscosity or antifreeze and you will get a flame out. Not good.
7. What about burning synthetic oil.
Conclusion: (Revised) The key is a hot burner tube. If it is well above the flash point of synthetic oil, then that oil will burn just as well as conventional oil.
3/8/15 - Major discoveries!
I made an adapter plate for this burner so I could fire this burner in my large potbelly stove. This stove is really large cast iron, about 4 1/2 feet tall, and several hundred pounds. It is probably near 100 years old and was used to warm a commercial shop with coal and wood. I used the iron flange shown above welded to a new door insert.
I fired the burner on propane and fed the burner oil via a gravity feed. It took a while to get the stove up to temperature, but after about 1/2 hr it hit 450 degrees at the stove's surface and I decided to see if I could run the burner on compressed air along, no propane. At this point the burner tube was a lot hotter than that, probably 600-700 degrees. I set the air regulator to 10 psi and quickly made the change from propane to compressed air without stopping the oil flow. The flame went out when I disconnected the propane and then relit when I connected the compressed air!
So what was going on?? And why did this not work before?? I believe it is because the burner tube was in a hot stove was the difference. The oil was also somewhat warmer than before also ( about 40 degrees F), but I think the difference was really the hot burner tube and the fact that it was in a hot stove (burner chamber) that was reflecting radiant heat back on the burner tube.
So this burner likes being in a hot stove / burner box. Temperature makes all of the difference. A hot burner tube means efficient combustion of heavy oil and easier firing.
I ran the burner for over an hour on just compressed air and tested it at various air pressures and oil flows.
Most of the testing was with the compressed air at 8-10 psi. At that pressure and with the flow optimized for a clean burn, I recorded an oil burn rate of 2 lbs - 12 oz in 30 minutes or 5lb - 8 oz per hour which is approx 150K btu per hour. At this rate the stove would maintain a nice surface temperature of about 450-475 degrees. However a couple of times the flow into the furnace decreased (being fed via a gravity feed) and the flame would sputter. A quick opening of the ball valve controlling the flow got rid of the blockage and then a readjustment of the valve regained the proper flow. This happened several times over a 1 hour period.
Later I decided to see how much heat this burner could put out in this stove and I increased the air pressure to 20 psi and adjusted the oil flow to make a nice hot flame. I estimate that the heat output doubled. The stove temperature quickly climbed well past 700 degrees and the stack temperature above the stove hit 600 degrees. Too hot! The stove was being overfired. I'm quite sure that this burner can maintain 300K btu output in the proper stove/boiler but this stove is not it! I shut the burner down and looked into the stove to see how hot the burner tube was and it was glowing a dull red color about 8 inches from the pipe tee.
Looking at this color chart: Steel Heat color chart It appears that the burner tube was in the 1100 to 1200 degree range. I never saw the tube get that hot when the burner tube was operated in free air outside of a stove.
Conclusions from testing of 3/8/15
This burner will burn waste oil using compressed air alone. However the burner tube must be very hot (preheated) and it needs to be in a "burner chamber" of some sort to maintain a high burner tube temperature. Propane can be used to bring the burner tube and fire box chamber up to temperature but it takes about 30 minutes which means 1-2 lbs of propane. Kerosene could likely be used as a liquid fuel to do the same thing instead of propane, but using propane at a few PSI of pressure along with waste oil is very simple and virtually a fool proof method of preheating the burner tube and fire box. So I am going to stick with the use of propane and waste oil to do the preheat.
Another way to maintain a high burner tube temperature without a firebox is to use a heat retaining ring around the first 6-8 inches of the burner tube. I plan on adding a square piece of steel tubing about 8" long that will over the first 8-9 inches of the burner tube. This should help reflect radiant heat back onto the burner tube and speed the heating of the tube. This should also allow the burner tube to operate on air only after a preheat period without a burner box or stove enclosure around the burner tube. This would be important since I would like to be able to run this burner inside of a boiler. The boiler inside surface will stay below 250 degrees since there will be water on the otherside of the boiler surface, so having this heat retaining tube around part of the burner tube will be necessary to allow this burner to work in a "cold" burner chamber which is basically what a boiler is when there is no burner box.
Putting the burner into the stove drove home the fact that consistent oil delivery is VERY important. Too much oil and there is a lot of smoke. Too little oil and not enough heat. Too much compressed air and the flame runs lean. Too little air and the burner runs rich and smokes some.
So while a ball valve "could" be used to regulate oil flow, it would be a constant PITA to keep the burner from running rich or lean. Overfiring would also be possible which is not at all good.
A gear pump for consistent feeding is a very good idea. I think the total cost of the gear pump setup I am assembling would be roughly $200 if the motor and drive were purchased from Automation Direct and the pump was purchased off Ebay. Not free, but this is a one time cost to burn a fuel that can be obtained for close to free!
Here are some pictures of the testing today:
Slightly smoky stack when run on low a little rich.. too much oil or too little compressed air.
The burner installed and running in the ancient cast iron stove.
Burner operating on low heat - with 8-10 psi or air and a proper supply of oil so no smoke.
The same burner tube running on 20 psi of air and a good supply of oil for minimum smoke. The heat was so intense that this is as close as I dared put the camera to the stove.
The radiant heat is very impressive.
I think that with the addition of a gear pump this would be a good manual burner setup as it is. Adding some controls would allow automatic operation.
Further development: I have a new oil feed system that is very precise. A Nema 56 frame 3 phase motor is driven by an old Siemens Variable Speed drive. The motor is coupled to a Suntec J type oil pump with a Lovejoy coupling with a rubber insert. It runs very smoothly and is very precise and is easily controlled. A 5 gallon container is being used as the oil tank at the moment. I used the same tank for my gravity feed tests. But now the tank sits on the ground and the J type pump draws oil out of it. Any excess oil over the Babington type nozzle surface runs downward and trhough a J trap and then back into the same oil tank. This setup is really simple and very effective. This is a huge improvement over a gravity feed system since the pump is a positive displacement pump and the flow does not vary with changes in viscosity. A PLC has been added along with a gas management manfold using some 24 V DC valves. A pushbutton box has also been assembled. The controls have been assembled (mocked up) on a peice of white poly. When the controls design is done, everything will go into an electrical box. The PLC has been programmed to preheat the burner tube for 5 minute using propane at about 3 psi. After 5 minutes, the oil pump is turned on via the Siemens drive for one minute and the burner runs on propane and oil. Then after the minute times out the gas is turned off and air is turned on. After that no propane gas is required unless the burner is shutdown, in which case it will go through the startup sequence again.
This setup works well with the VFD running the 4 pole motor at about 8 hz and the air pressure at 12 psi. I also tried running the burner at 10 hz and 15 psi and that worked fine as well but I could see it was going to overheat the stove so I turned it down. (A good problem to have. :-) )
Here are some pictures of the latest setup:
Mockup of the controls on a poly board.
Variable Speed Drive
Motor and Feed pump setup running. Both are surplus equipment. The pump came off an old oil burner and the motor from a surplus store long ago. The coupling parts were bought off Amazon.
This setup with the VFD drive works very well and is very precise. This burner can be operated without something like this but it won't be as "tunable".
So.. what is left?
1. Development of a pilot flame system that includes flame detection (I already have a board to do that in the control system.)
2. Overtemp detection and shutdown safeties. These will be wired external to the PLC for increased safety. I want to put a switch on the stove and the stack.
3. Installation of a thermostat in the heated space. Programming is already in the PLC to accept a thermostat.
4. Installation of the working controls in a real electrical box.
5. Setup of a larger oil tank (which I already have) to allow longer run times without refilling.
6. Setup of a cart to mount the controls and pump on so it is semi portable.
Did more testing last night. The oil was about 42 degrees so it was thick. The startup went ok but I noticed that the oil was quite viscous coming out of the vertical tube over the flare cap.
Here is a pict of the oil globbing on the cap. All of the oil was being blown off the cap and there was no overflow off the cap so the flame was running rich and smokey.
I moved the flare cap forward via the adjustment and then took the next picture. The oil is now spread over the cap and there is some overflow but still the flame was running rich..
The tank is near the stove and it only had a few gallons of oil in it so it warmed up quickly to 55 degrees or so and the oil thinned out and the burner started running right (little smoke). The oil became noticably thinner. In the second picture below you can see the oil appears quite black. Once the oil warmed another 15 degrees the oil color appeared amber and I could see the brass flare cap through it. So the oil was much thinner.
As the oil further heated the oil on the cap thinned and the atomization became better and the flame was even nicer. That process from cold start with temps in the low 40's to a really nice flame was about 20 minutes.
I think this may be sufficient to justify an oil heater for the tank. If the oil was preheated to 90-100 degrees this burner would start with less smoke and be more consistent. Remember the goal is full automatic operation with this type of burner. This burner (unlike siphon burners) can't clog the nozzle so it should be very trouble free. I have a tank and some electric heating elements so I think I will set it up so the oil is heated. I could go the way of wrapping copper tubing around some part of the burner etc, but that would throw in more control variables and I don't need that.
Also, because of the varying viscosity of the oil (the flow was consistent due to the gear pump) I was able to turn down the speed of the motor and reduce the flow during the startup. The VFD (and most do) has an analog input on it so the PLC could slow the pump down during startup to compensate for the thicker oil, but I think that simply heating the tank might be more effective. Still thinking. If I control the oil temp, oil flow, air flow/pressure, this burner should be extremely predictable. Which makes automatic operation fairly simple.
A few years ago after the purchase of my first sailboat, an Ericson 29, I soon realized that there was a limit to how much involvement family and friends desired in my "new to me" sailboat. Not many were willing or wanting to spend an entire day on the boat. A few hours.. that would be fun. A day.. uh maybe.. A couple of days...ugh... Thanks..I'll keep that offer in mind. Also the people I know are busy with their lives, and oftentimes our schedules do not coincide. On top of that my first mate strongly dislikes sailboats. And I'm being kind in saying it like that. So as it turns out the boat that can sleep 4 or 6 usually has one on board. That means single handing. Anyone who has sailed a decent sized sailboat knows that the rudder requires a lot of attention especially in a an IOR designed hull like the Ericson 29. Tying off the rudder is not an option on most IOR type boats. They need to be actively steered.
My Ericson 29 did not have an Autopilot and the expense of buying an autopilot was prohibitive. However as a controls engineer I had access to controls. Hence my Homemade Autopilot project started in 2011.
Here is a picture of the control box which is a 8x8x4" plastic water tight box. About $15 at Menards.
What is in the picture?
The PLC mounted on a din rail that is screwed through the back of the box
To the right is the perf metal housing for the 12 volt to 24 volt DC inverter power supply. I think it can output about 1 amp at 24 volts which is plenty to power the PLC and the two relays and drive the PLC inputs.
The little perf board wire tied to the 12 volt to 24 VDC power supply is a 5 volt regulator chip since I needed 5 volts to power a RS232 to RS485 converter which is tucked under the terminal strip. I think I wrapped the converter with
electrical tape to prevent shorts. The converter is used to adapt the flux gate compass to the RS485 serial ports on the bottom of the PLC. The bottom serial port on the PLC is used to communicate with the LCD display that serves as a tuning/setup/display device for the autopilot.
The two relays are used to drive the electric actuator. They are easily replaceable since they are socketed.
The current incarnation uses an Automation Direct PLC which is very inexpensive as PLCs go and the software to program it is free.
The PLC used in this project was this one:
This plc has 4 - 24 VDC inputs and 4 - 24 VDC outputs. It also has two analog inputs and two analog outputs. The project was started before analog I/O cards were availabe for the Click family of PLCs. I also used a add on card for the PLC to gain additional inputs.
The additional inputs were used on a cabled waterproof pendant which has a joystick and two buttons on it. A resume button and a tack button.
The joystick is nice since if you are in Man mode you can steer the boat with the joystick. Pretty nice if you have to steer but want to plant your butt in the corner on a pillow for comfort.
Here is a picture of the pendant:
I also used a display unit to show the compass heading, deviation from desired heading, and tuning parameters, and mode settings.
The display is a standard Automation Direct display which communicates with the PLC via a cable that looks like a phone cable, but is not.
This is the screen that was used.
Here is a picture of the actuator I used. This is a 12" stroke model and I used the available end fittings. This was purchased from Surplus Center in Nebraska. http://www.surpluscenter.com/
The actuator has a built in potentiometer that is used as a rudder position feedback device to the autopilot control.
Here is a link to the PLC program.
Here is a link to the Screen/display program.
Here are wiring diagrams. I just realized that these diagrams are incomplete. I will update them.
The compass I used was a military surplus fluxgate compass purchased off Ebay for about $100. It is in a nice cast aluminum watertight box and it works very well.
But since then newer compass chips have come out for less money and they work just as well. The CMPS10 chip is a good compass and I have tested it.
However the CMPS10 has been discontinued and the replacement is the CMPS11 which I have not yet tested but it looks good despite the price increase from the CMPS10.
What would I do differently?
The Automation Direct screen is nice in that it is viewable in daylight. And it is nice to be able to tune the autopilot with the little screen.
However the screen is really not required. Newer PLCs are now being sold that have built in web servers. If I am going to use a Tablet for navigation at the helm, then pulling up a webpage on the tablet to tweak the PLC autopilot controller makes a lot of sense. It also cuts about $150 out of the cost of the system.
When I first selected this PLC, Automation Direct did not have analog I/O cards, they only had PLCs with onboard Analog I/O. Also, I thought it would be important to be able to tie the navigation computer to the autopilot so the navigation computer can control the autopilot. With a Nav computer link, a compass link, and a screen link to the PLC, three serial ports are required. In hindsight, I think a Navigation computer link is really not required and may not even be desirable. Sailboats do not move fast. There is plenty of time to think about the desired heading for the autopilot.
When sailing close to the wind, having a wind indicator so the autopilot can sail to the wind direction would be a great advantage. So a second analog input would be desirable for a analog based wind indicator.
With these things in mind, a different PLC could be used. I'm thinking that the Siemens S7-1212 or 1214 would be a very good fit. That PLC has a built in webserver. Tie the PLC to a small Wireless AP for communication to a Tablet and that simplifies the hardware further and gets rid of some wires and it eliminates the need for a dedicated screen/MMI for the PLC.
If you make any improvement to this design please let me know and I will post them on this blog.
Happy Autopiloting ! :-)
I grant this software/design a GPL3 open source license.
PLC based Autopilot design Copyright (C) 2015 Dave Cole
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 of the License. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>
The marina where my boat is at does not like boat stands and they charge a premium (about $300 per winter) to have a boat stored on stands. I paid the premium last winter but this fall I decided to make a cradle for the boat. The boat's basic specs are 10 meters long - about 33 feet, 11 ft wide, draft of about 6 feet and a empty weight of about 12,000 lbs. The boat is stored with the mast up. The boat yard transports the boat on the cradle with a hydraulic yard cradle trailer. So the cradle frame must be robust. A lightweight cradle frame will not work.
People have asked me why some other marinas prefer cradles and the only reason I can come up with is that for some marinas using cradles improves their efficiency. They can place the boats closer together in storage since narrow hydraulic trailers are used to set the cradle/boat. Also using cradles can use less labor. Once a cradle is made for a particular boat, placing the boat into it's cradle with a travelift is fairly simple, whereas moving stands around, placing the stands and then chaining them together can be some work.
I watched Craigslist and Ebay for cradles that might fit this boat and found nothing that was even close, so I started looking for basic cradles that I could rework.
I found one in Sandusky for a lot less than I could purchase the steel. So I drove their with my trailer and picked it up just before the weather really turned cold (highs of about 10 degrees in early November!)
The cradle base I purchased is about 17 feet long and 7 feet wide. I cut off all of the uprights with a plasma cutter and then ground them close to flush with the top of the channel steel framework. The cradle frame is made of 8 inch channel on the outside with a mix of 8 and 6 inch cross members. The existing frame welds were adequate.
But where to place the uprights? At what angle? And how tall?
Before I launched the boat in 2014 I used a plumb bob, an inclinometer (Harbor Freight), a few long tape measures and some boards (for use as straight edges) to determine the location and angles of the existing stands which were holding the boat securely. The boat was supported by 7 stands. Three on each side plus a bow Vee stand. I referenced the port/starboard measurements off the centerline of the boat, and for a fore/aft location reference, I used the point that is formed between the centerline of the boat and the waterline of the boat at the bow. From these references I measured to the center of contact point for each stand support pad. (You have to visualize where that point is and make an educated guess since the reference point for the pad is actually at the pad/hull surface.) I also measured from the ground up to these pad contact points. In this situation, the boat's keel was sitting on several 6x6 wood blocks that actually measured 6" square. So putting all of this together resulted in 7 support locations in 3D space over the framework (which was already welded). Keep in mind that you need to be able to get travelift or crane straps under the boat while the supports are still in place.
Here you can see my hand written notes when I sketched the locations of the support pads in relation to the centerline of the boat and the front reference point on the bow.
With this used cradle base that I purchased, I got lucky as I had support steel in the base where I needed it so I could attach the stands uprights. The rear two pads were shown in the drawing as being placed tilted in from the side but when doing the welding I also decided to tip them forward slightly. While that didn't hurt anything, I tipped them towards the front further than I should have by a few degrees.
Here is a shot of the working drawing I used to place the uprights.
For the new vertical support tubes, I purchased some seamless structural 1 1/2" OD tubing from a local steel surplus yard. It is surprising how much steel tubing is used for the uprights and then the diagonal supports for the uprights. I believe I used about 60 feet of steel tubing. I reused the pads and screws that came with the used frame.
For the bow I decided not to use a Vee type stand top. I decided to go with a ladder type support and use two laminated 2x6 southern yellow pine boards as cross members that was carpeted to give the bow better support. With the swept keel, there is a lot of weight on the forward boat support stands and I was concerned that a V type stand there would have to much point contact pressure and possibly damage the boat. The flat 2x6 (times 2) beam covered with carpet distributes the load better. The ladder support moves up and down via two scaffold jacks. Scaffold jack feet are bolted to the 2x6's which are laminated together with epoxy and the scaffold jack screws fit into the vertical "ladder" members. A cross member stabilizes the "ladder" and fore-aft support is via a diagonal steel pipe that clips over the ladder cross member and attaches to the crossbeam in the base. The ladder support sits on short steel pins that are welded to the top of the 8" channel. In this way, the entire ladder support can be lifted off the base so the boat keel can be slid in between the uprights with the travelift without lifting the boat over the front bow support. The travelift operators seems to really appreciate this feature.
Here is a picture of the boat in the cradle. As you can see it still needs a paint job which will have to wait for warmer weather.