H-Bridge Caravan Motor Mover Controller Design

I'm planning to build a replacement caravan motor mover to replace the commercial option which has failed after about fifteen years.
Lewis Loflin Power Mosfet H-Bridge: • Power MOSFET H-Bridge ...
Items used in this video:
❗️These links are affiliated and I may earn a small commission if you purchase❗️
Notepad: ebay.to/2Ee4T2L
Pen: ebay.to/2tGl7wt

[@AdamWelchUK]

Thank you so much to everyone who has pointed out errors in this design - I really do appreciate your advice. I hope to address them soon, but in the mean time, anyone who is tempted to copy this circuit design - don't!

[@AdityaMehendale]

Dear Adam Welch, NO The circuit will not function; in fact it wild create a massive fireworks as you watch bunches of four FETs blow up. Here's why: When the gate-voltage is transitioning between 12V and 0V it will momentarily be 11V, then 10V, then 9V and so on. For all gate-voltages between say 3V and 9v, you'd have _BOTH_ FETs in the pertinent leg of the H-bridge turn on, essentially shorting out the battery through both FETs. What happens next? The P-FETs fail first, but FETs tend to short-on-fail, so they keep conducting. Next the N-FETs fail - and now you have no control whatsoever on what happens next. As current builds up, limited first by the inductance of the cables and copper-tracks, then by resistive-drops in the circuit, the wiring shall heat up in a matter of milliseconds, and shall spectacularly release the magic smoke. Trust me, this makes for FANTASTIC KZbin clickbait-content - just watch electroboom; Just make sure you have the proper means to video-record this with multiple camera when you turn it on ;) You only get one shot per H-bridge stem. Cheers!

[@AdamWelchUK]

Thanks for this. I can see what you are saying and it does make sense. I might try and see this effect on my bench with a limited power supply and the scope - rather than missing my one chance!

[@pirateman1966]

Why did you have to go spill the beans like that? I was hoping to see a fireworks. "Little knowledge is a dangerous thing" encapsulates this design.

[@peterdkay]

I'm glad someone pointed this out! Consider the case with gates at +6V. Both P-Channel an N-Channel are fully ON. This puts about 10 milliohms across the supply for a short time every time the MOSFETS change state. That's over 1000A transient current --- BANG!!!

[@AdamWelchUK]

Just to let you know I have found this comment. :-) I always planned to place the first version on a limited power supply - so I'm afraid you would have never seen fireworks or smoke.

[@AdamWelchUK]

Yes thank you. I've clearly not done enough research here and hope to rectify that very soon (rather than juts pull the video and pretend it never happened - which is tempting). I'm learning from my mistakes and getting some great advice from comments here as I so often do. Cheers

[@RyanHeaney42]

Sweet! Looking forward to seeing this come together!

[@sarkybugger5009]

I bet you finish it before Julian gets his Muppet 2 done. Or his synth. Or...

[@zaprodk]

Sounds like a good plan to replace those old clunking relays!

[@michaelknauth943]

Looking forward to a follow up video that might be in the works.

[@jaymiddleton772]

I'm working on a similar project at the moment and hadn't considered opto-couplers to drive the mosfets, but I might incorporate them for extra protection. Thanks!

[@AdamWelchUK]

Like I said - not my idea but it seems sensible.

[@jontscott]

No dead time between the high/low switching? I would suggest checking how fast the mosfet turn on/off times are in this setup because I suspect that without dead time between when the high side turns on and the low side turns off you will see some big power spikes. Dare I ask why you don't use the existing PCB and work out the PIC pins that control the relays, then wire them to a nano? Not that I do not like the idea of building a custom PCB, just seems you have a nice set of switching power electronics already.

[@AdamWelchUK]

I’ll be interested to see if this is an issue - I’ll look on the scope when I test this on the bench. The value of the pull up might be key. There seems to be a number of issues on the original PCB. Something caused the voltage regulator to fail - I’m also aware the relays have been known to get corroded inside on these old units. It’s always a risk putting up videos like this. I might fail - but hopefully I’ll learn some stuff on the journey. :)

[@AdamWelchUK]

Really useful advice thank you. I always intended to try this on the bench first with a limited power supply - I might still do that to see exactly what happens. But I can now see the major issue with this design. I've also got a couple of DC breakers at different rating which I intend to put in line with the final version if I ever manage to build a final version... :-)

[@newburypi]

Looks great. My only thought is the trustworthiness of the MOSFET body diodes. I've blown up DIY H-bridges in the past and would be tempted to supplement with addituonal diodes. Maybe boulder it into the PCB and only stuff them if you find them necessary.

[@AdamWelchUK]

Thanks for this. I’ve already got a pcb coming without additional diodes and enough parts to build a few of them. I guess I can redesign or bodge on the diodes if I have issues. Again, thank you for the suggestion.

[@newburypi]

@@AdamWelchUK it's likely just my paranoia speaking out of turn :)

[@AdamWelchUK]

Well I’m willing and able to prove or disprove the theory! Time will tell. Cheers. A

[@TheDefpom]

I would be focusing on the power supply fault, that is likely the root reason for it not working even after powering the pic, possibly a dying tantalum, or simply a bad solder joint from the lead free solder. You could always post it to me to see if I can figure out what’s wrong with it.

[@damo1023]

Starting similar project... plan to use DBH-01B DC Motor Driver 50A + arduino + RF

[@kapegede]

Brilliant idea! I hope It will work.

[@AdamWelchUK]

Fingers crossed.

[@NiHaoMike64]

Have you looked at just replacing the board with off the shelf R/C hardware?

[@electronic7979]

Very good 👍

[@Sheevlord]

I wouldn't trust the body diodes to handle inductive spikes. Adding external beefier diodes in parallel might be a good idea just in case.

[@AdamWelchUK]

You’re the second person to suggest that. Sounds like I need to rethink my pcb. Thanks.

[@zaprodk]

@@AdamWelchUK You can just solder some diodes across the FET's on the board. No need to redesign the board for that.

[@DWiskow]

How about using an HC05 Bluetooth serial module to facilitate the remote control? . . . using a pair of these you could have a battery powered Arduino based remote where you are using the two bluetooth modules as a simple ‘virtual’ serial connection between it and the motor controller. The HC05 modules are very cheap ($3-4$), secure (paired with a user defined PIN) and short range (10m). If you want to get more sophisticated, it is also possible to write an application on your mobile to function as the remote.

[@Chupacabras222]

Are you sure those tiny mosfets are capable of handling 100A? How are you going to handle heat dissipation?

[@AdamWelchUK]

No - not at all. They are rated for that but their legs aren’t going to take that current for any real length of time. Heat wise I’m going to see how that goes - it may become an issue, but I hope not. There will be a fair bit of copper around the fets which will dissipate some heat. I’m also not going to be running this for long periods.

[@Chupacabras222]

@Undefined Lastname Your math is optimistic. It is not 0.5 W, it is 50W at 100A.

[@Chupacabras222]

@Undefined Lastname ohms x amps = volts, not watts; 0.0048 ohms * (100A)^2 = 48 Watts, according to my calculator

[@trickyriky1]

Sounds like your having more fun with arduino's mate. My arduino solar monitor is now well on its way to being a solarmonitor and 7s bms, you have the link if your curious how its evolved. :)

[@AdamWelchUK]

Ah yes - I do need to do some catching up. I’ll take a look soon. Thanks.

[@kawasakiinitful]

Hi Adam I have the same board , one of my motors does soft start but the other one just kicks in straight away causing the van to turn slightly when going forward or backward until the soft start one has spun up . any ideas where to start looking please ?

[@AdamWelchUK]

I never found a circuit diagram for this and the project didn’t really work out for me. However, my guess is that the four fets sticking out of the side of the pcb are used to control the soft start. Probably a pair for each motor - one for forward and one for reverse perhaps? If soft start isn’t working I’d look at those and see if one had failed short.

[@kawasakiinitful]

@@AdamWelchUK thanks for ur help

[@peterseddon8363]

I had two of these in my twin axle caravan and have since upgraded them to a single AWD controller. The old controllers did have soft start so the motors sped up slowly to prevent high currents on starting. Also at it was Four Wheel Drive, the opposite wheels to the driving wheels were pulsed to prevent the tyres from scrubbing. If your handset is the old white one then it used AM transmission to send the signals to the control unit. These are now obsolete and the newer units are FM and the had controller looks more robust. All spares including new control units complete with a new handset are available here, www.powrtouch.com/

[@AdamWelchUK]

Yeah my unit is an old model - it’s probably over ten years old. They do have a soft start - that’s what the mosfets are for I believe. I might be able to incorporate this into my code - but I haven’t actually written any yet! I’m trying to do this on the cheap if I can. Hopefully it’s not wasted money and I end up buying a new one anyway! Cheers

[@tengelgeer]

9:43 No you did not. You connect the emitter to the cathode so bye bye isolation. Now it's just a transistor with low gain and low input impedance ;)

[@AdamWelchUK]

You’re right I was going to investigate isolation a bit more but there are so many other flaws in this design I think that’s the least of my worries!

[@davidluther3955]

WHERE ARE THE ARROWS ON THE MOSFETS.

[@Mark1024MAK]

Bit disappointed ☹️. (1) Adam says he was always going to bench test it using a current limited supply, no bang or magic smoke, and, (2) various people have pointed out the problems with simple two control line switching systems ☹️. Spoils all the fun... But anyway, just to help with this... Switching high current DC supplies, especially electric motor feeds is always a tricky bit of engineering. Unlike AC, there is no zero crossing point to help kill an arc. So everything has to be very much overrated current wise compared to what you would expect. Best if everything switching (or connected to the motor, directly or indirectly) is rated at least to 24V (32V is better). High speed high current diodes are recommended to suppress the back EMF. As is some 100nF capacitors in series with a 100 ohm resistor (across the motor connections). As was said earlier, it’s best if there is a delay between switching the 0V and the 12V side switching devices. And it is essential to include fuse protection. At my workplace, we use 24V electric motors in hydraulic systems. They are fused at 20A, but the starting current can hit 30A. The 24V control valves have neons across their terminals to help suppress the back EMF. The positive wiring/tracks and the “ground”/negative wiring/tracks have to be thought about carefully. Keep the control circuits and the high current circuits completely separate as much as possible. As with high current spikes, this can create interference that could potentially crash or disrupt a microcontroller (it certainly messes up 74HCxxx and 74HCTxxx logic). Using opto-isolators will help. But don’t connect the common for the control side to the high current “ground” where the motor feeds are. Connect to the logic/microcontroller ground. Think star earth.

[@AdamWelchUK]

Sorry to spoil your fun but thank you for taking the time to pass on this useful advice. I’ll definitely think about what you’ve said and take it on board - you’ve obviously got a bit of experience in this area. Perhaps I’ve bitten off more than I can chew this time.

[@Mark1024MAK]

Adam Welch - If you don’t try, you’re never get anywhere. Don’t give up. I’m sure the PCB that you have on the way could still be used for principles testing at the very least.

[@fourzerofour7860]

Caravan... *googles British speak* Wait wait wait... You mean to tell me that your RV has electric hub motors so you can MOVE it with a remote, to 'fine tune' the parking? Wait... what the every loving shit is the American RV industry doing? I've never seen this... 0.o Please clarify, Adam. Does your RV really have electric freaking hub motors?

[@AdamWelchUK]

That’s right. I can engage an electric motor to each wheel and remotely control the movement of the caravan without the car. Very useful for moving it in a tight spot or when you need to move it just slightly to endure it is level. Do a search on KZbin for ‘Caravan motor mover’

[@twotone3070]

Not sure that RV is the correct translation for caravan? My understanding is an RV is what we may call a motor home. I haven't done any research to substantiate this.

[@piconano]

wth is a caravan? your h-bridge looks bad my dude. if your pmos won't turn off before your nmos turns on cuz your mcu froze... boom.

[@user-ug9nn]

bulshit talk ... you ware not able to repair a simple power supply on an electronic board ... you are not able to build a new controller