Automatic Leveling for the Oliver Trailer — DIY style!

[Snackchaser]

As far as I can tell, I’ve made the first automatic leveling system for the Oliver’s Barker stabilizing jacks!  Maybe even the first 3-jack leveling system on a travel trailer!  It’s simple to operate, safe, and works great!

I built the leveling system as a novelty for my own entertainment, so I’m only posting it as a interest item not a recommended project.  In reality, manually leveling with the electric jacks is already easy, particularly with a LevelMate, so automatic leveling isn’t really needed.  There are even warnings that the jacks are stabilizers, not to be used for leveling.  However, I believe the warnings are more about liability than capability, as long as the jacks are used within reasonable limits.

I'd been thinking about an auto-leveling project for a long time, but was deterred by the thought of modifying the jacks with fiddly proximity sensors or revolution counters as used in conventional leveling systems. I didn't want to alter the jacks for a DIY project that might not even work.

But then it dawned on me.  The only time I really needed to know the position of the jacks, is when retracting them to their parked position.  I wasn’t interested in returning the tongue jack to the truck hitch hight like some systems offer.  So this led to a revelation that automatic leveling could be done with nothing more than current sensors mounted in a control box rather than on the jacks themselves. This was only possible because of the Barker jack’s mechanical clutch.  When the jack reaches the end of its travel, the clutch activates with a distinct clack-clack sound. It was likely that clutch engagement would produce a unique current signature that could be used to detect when the jack is fully retracted.

A plan was developing, but I needed actual amperage values for proof of concept. Fortunately I had a data logging ammeter, and the chart below illustrates the results for one of rear jacks starting from its fully retracted parked position. The jack was extended through free air until it touched ground and began picking up load, maxing out when the wheel lifted off the ground. Then it was retracted to the parked position until the clutch actuated:

I didn’t include the tongue jack chart because it's far less dynamic — it's always under load, and the clutch doesn't normally actuate during leveling.  The clutch only comes into play after trailer is hooked up to the TV and the tongue is manually retracted.

The amperage results confirmed that current sensors could work, and beyond detecting the parked position, they could also be used to distinguish the different leveling phases: rear jack extension until firm ground contact for stabilizing, and then transition to "roll" leveling if needed.  Completion of the rear jack phase could then trigger the tongue jack phase to handle "pitch" leveling.

The amperage chart became the basis for the circuit and software design. No proximity sensors. No revolution counting. No permanent modifications — just two wires to each jack, a component box, and a control board.  Sounds easy, but this was only just the start of a long process to perfect the system.

I hope someone finds this interesting!   Cheers!  Geoff

[OffWeGo]

Impressive!  Thanks for sharing.

[DanielBoondock]

With a 7k trailer, if it was held only by the jacks (or if one jack lifted the tires) they'd be at 2,333 lbs each and the 2026 Barker 3000 VIP jack is rated to 3k lbs. So the limiting factor is the rear jack mounting. It's not nearly the same as the the front jack I think, but without looking (I'm not at the trailer) I think it might be attached to the frame? Oliver tends to over engineer at any rate which I approve of. Also unknown long term effects. Anyhow that's a crazy nifty solution, nuts to have that level of automation in a small trailer like this, but I think its great. 

What I like even better is that water valve panel, as beautifully done as it is the under bed solution is a PITA. 

 

[John Dorrer]

A 2' level does the job. Set on the rear bumper for side to side. Add levelers as needed, chock and disconnect.

 

I take the 2' level up against the belly band and raise or lower to get level

 Actually a good idea to have the tongue raised slightly to allow flow in the tanks toward the back.

 

Very simple

[CRM]

1 hour ago, John Dorrer said:

A 2' level does the job. Set on the rear bumper for side to side. Add levelers as needed, chock and disconnect.

 

I take the 2' level up against the belly band and raise or lower to get level

 Actually a good idea to have the tongue raised slightly to allow flow in the tanks toward the back.

 

Very simple

Exactly how I do it. 

[BigTexas]

This is brilliant!

Oliver already makes it pretty easy to level the trailer.

But if I wanted to make my camping friends (even more) jealous, I would build this and watch them grimace with envy. 

 

[Snackchaser]

Custom Automatic Leveling for the Oliver's Barker Jacks — Part 2

Here is some more detail of the automatic leveling project as it transitioned from an idea to a working prototype.  

The project evolved from concept with safety and trailer protection as core design priorities.   Limitations were built into the software including current-based motor cutoffs and maximum leveling offsets to prevent over-stressing jack capacity and minimized frame stress. On the mechanical side, the design incorporates fail-safe circuits, redundant shutdown methods, and four levels of overcurrent protection.  A side benefit is the electronic monitoring of jack condition where abnormal current can reveal lubrication issues, mechanical wear, or circuit degradation before they become catastrophic failures.

First step was writing a program script with open source Arduino software, and then uploading it to an ESP-32 microprocessor.  The ESP-32 is basically a $10 miniature computer with programmable memory and integrated WiFi module (I might develop a Leveling App later).  It receives inputs from a Inclinometer and two 50 amp current sensors, and it sends outputs to a status LED, a piezo buzzer for audible tones, and a 8 channel Darlington driver that converts digital inputs to 12 volts for relay coils.  Six 30/40 amp Bosch style automotive relays are wired in an H-Bridge configuration to run the jack motors in forward and reverse.

With a basic design plan and software, a test board was needed for proof of concept before anything could be installed in the trailer.  The test board included potentiometers to simulate jack current, and LEDs to indicate the jack's up and down movement.  It had a status LED, an isolation relay LED, a piezo electric buzzer, and a circuit board mounted with the microprocessor, inclinometer, drivers, and power supply.  These standalone plug-in modules are wired together with soldered jumpers on a generic printed circuit board.

Testing resulted in many design and software changes as the bugs were worked out.  During this time it became evident that the tongue only needed the inclinometer for leveling, no current sensor required.  I also realized that the jacks should be run in decreasing intervals with a limited number of leveling attempts before timing out.  Another improvement was to run the jacks uninterrupted in free air until they were firmly seated on the ground for stabilization — before starting incremental leveling movements as needed.  Current values are easily adjusted if more or less stabilizing pressure is desired.

Testing found a flaw with the inclinometer.  Thankfully they came in three-packs and I replaced them because they wouldn’t stay calibrated.  The inclinometers have accelerometer and gyroscope functions, and it turned out that the accelerometer has known drifting issues.  After turning off the accelerometer in the software, it worked great!  In the end, the Arduino script grew to over 1,000 lines of code.

With a lot of trial and error, a sequence of operation was developed:  When the Level button is pressed, the rear jacks lower without interruption until they are seated on the ground in a firmly stabilized position (phase 1).  If the trailer is level at this point, then there is no further action and an audible “happy tune” indicates successful level was achieved.  If the trailer is not level after phase 1, then the left or right jack is cycled in intervals starting at 1.5 seconds (~1/2” travel) depending on how close to level it is.  As the trailer gets closer to level, the interval shortens to account for coasting and fine adjustment to within 0.2 degrees of level.

During the roll leveling sequence, a pitch limit of 5 degrees prevents the rear jacks from inducing too much pitch.   It's a safeguard that temporarily interrupts roll leveling while it moves the tongue jack to bring pitch back within limits.

Phase 2 pitch leveling begins, if needed, after roll level is achieved.  When roll and pitch are level, the “happy tune” is played.  The system allows up to 20 level correction attempts per axis before declaring unsuccessful leveling with a “sad tune.”  The operator can stop the leveling at anytime with the On/Off switch or by re-pressing either momentary button.

The “Retract” function proved difficult to program because of the mechanical clutch.  Originally it was programed to shut off when the mechanical clutch actuated in the parked position, which is 13 amps.  This didn’t work because the jacks had a current spike when  coming off load, and there were oscillating current spikes during the clack-clack of the clutch that caused shutdown before before a clack was heard.  These spikes hadn’t shown up in the time compressed current chart, but they were picked up by the current sensors.  This was resolved in the software with a current delay when coming off load, and a longer delay when the clutch actuated so a few reassuring clack-clacks could be heard for audible confirmation that the jacks were fully parked.

The leveling system was designed so that the jacks manual switches could remain fully functional.  However, the H-Bridge relay configuration reverses the motor polarity for the up or down direction, and that could cause a direct path to ground when using the manual switches.  To prevent this, I used a normally open relay to isolate the ground when the leveling system is not being used.  I used a single 250 amp generic lawnmower starter relay on a common ground bus for all 3 jacks.  It seemed reasonable. .. what could possibly go wrong?

Hope this is still interesting…Cheers!  Geoff

Please excuse poor photos, I just didn't take very many!

[Mike and Carol]

Interesting!  Are you starting a waiting list for your production model?  😁. Mike

[HDRider]

Some people are really smart.

 

@Snackchaser You sir, are really smart.

[Snackchaser]

Custom Automatic Leveling for the Oliver's Barker Jacks Part 3:

I skipped over a lot of detail to keep this project interesting, and it’s probably still too much information.  But some might be interested to hear about some of the background work that was involved.

After the test board phase was complete, and everything was working well on the bench, the system was installed in the trailer for real-life testing.  This was an anxious time because it had only been tested with simulated current, and I had no idea how it would react under real current conditions.  Things went south immediately!

I started testing with rear jacks because they were easier to wire and much more complex than the tongue jack.  I didn’t want to run the tongue jack wires if the whole project could end up in the trash bin!  I came up with a test procedure to verify each function in different scenarios of operation, and that required manual jack operation.  That’s when the problems started!

I was blowing the 30 amp jack fuses and couldn’t figure out why.  I first focused on the H-bridge relay wiring as the problem.  Then I suspected that the ESP-32 or driver outputs were miss-wired during transition from the test board.  They had been soldered together one wire at a time under magnification, and back tracing the bundled wires was a nightmare.  I couldn’t find the problem and came close to scrapping the whole project.

I was so confident in the design, and overly focused on wiring, that I completely overlooked the obvious. There was a major flaw in the way the auto-leveling system was isolated from the manual jack switches. I had used a high-amperage relay that closed to a common ground only when the H-bridge relays triggered, which broke the short circuit path when the manual switches were used. However, I hadn't considered that the pole-reversing function of the switches created a different ground path through the hot side of the leveling system.

The solution required major re-work, replacing the single ground isolation relay with six 30/40 amp relays that isolate the manual switches from both the ground and hot side.  Although there were two spare Darlington driver channels capable of handling two relays each, there weren't enough channels for six new relays and a second driver was added. The system now runs 12 relays from 9 driver channels — six H-bridge relays and six isolation relays, two per jack.

All the components were installed in a 12” project box that was limited in size by the 3-D printer capacity.  So the relays were stacked 2 high on a stepped base to save space, and a elevated terminal block helped with wire congestion.  I was able to mount the current sensors directly on the PCB with the ESP-32 and other modules.  It’s tight, but nicely compact too!

The switches for the leveling system are mounted on a 3-D printed control panel inside the outside storage garage.  The panel includes flush to the surface push-buttons to safeguard against accidental activation, one for “Auto Leveling” and the other for “Retraction.”  There is a master On/Off switch, status LED, and a Piezo Electric buzzer for the status tones.  The entire panel is also recessed to protect it from cargo.  The system can be shut down with the On/Off switch, or by re-pressing either momentary switch.   The systems parasite draw is near zero when on, but it’s normally off when not in use.

The existing freshwater control system was incorporated into the lower part of the control panel due to  space constraints.  It has a 4-way water pump switch for the outdoor shower, with indicator LED.  There is also a 4-position, 6-contact, rotary switch that controls motorized valves for the four freshwater configurations; Normal, Freshwater Tank Drain, Boondocking, and Winterizing.  The LED indicators show configurations other than normal.

A printed weather proof hood was used for wire penetration into the tongue jack head, and a terminal board was used to splice into the rear jack wiring.  This was a useful addition for quick disconnecting and troubleshooting.  Those were the only modifications made to the trailer, other than enlargement of the existing switch panel cutout on the garage wall.

All the custom components used in this project were designed with a free on-line CAD program called TinkerCad.  Some items were 3-D printed from PLA filament, a plant based plastic.  Other pieces were printed from ABS plastic for better durability, or TPU — a rubbery filament.

Automotive Deutsch style plug-in connectors were used to connect the switch panel and control box, and they allowed me to connect a remote switch box that was used to operate the system from inside the trailer during testing.  There is also a panel mount USB cable port for a laptop connection to view the serial monitor or update the software.

The system needs a one time calibration that is done with the trailer perfectly level.  Both momentary switches are held for 5 seconds until the Status LED quickly flashes three times and the audible calibration tone indicates satisfactory level.  The software determines the pitch and roll off-sets from absolute level, and sets them as the new target level +/- 0.2 degrees pitch and 0.3 degrees roll.  This equates to about 1/4 to 1/2 inch tolerance.

The end result works great!

There will probably be more software fine tuning in actual camping conditions, but it will be easier to do that in the future.  I’m just finishing up a Bluetooth App for the automatic leveling controls, and it will include setting adjustments for tolerance, ground contact pressure, clutch runout, leveling step durations, and more!

If you made it this far, I hope you enjoyed the project!

Cheers!  Geoff