Mobipus Controllers

[spinningmagnets]

Mobipus is selling a huge volume of controllers to dealers globally. There is the rare customer that owns a soldering and knows how to use it, but...I'm guessing that's less than 1% of their customers.

If given a choice, I would source one with no potting, but...my main interest in Mobipus is that...I have never seen a higher percentage of customers who were happy with the performance and reliability of a controller. That caught my attention.

Easy programming is a HUGE benefit. Google "programming a Sevcon, Adaptto, Kelly, etc", and see what builders say about them...

 

[ElectricGod]

The Mobipus controller is OEM level quality and reliability that's much more accessible than others.

I'm not implying anything about the Mobipus controllers at all.

1. They are potted. AKA 100% non-servicable...which I really don't like!
2. They have not been around long enough to see what kinds of long term reliability they may or may not have.
3. They are "more accessible" which is a good thing. Most FOC controller makers make it hard for DIYers to get their products.
4. See number 3...if they really are "OEM quality" then they will hold up well to DIYers...we will see.

Potting...from the perspective of someone that does electronics repair since the 1970's:

Potting (burying the electronics in soft epoxy) is a great way to make any electronic device un-repairable. If you need to secure some large component, a dab of silicon at that part will do the job. If water incursion is a problem, use conformal or good seals or both. If there is something I've done in more controllers, it's blowing mosfets. Sometimes they die for no obvious reason. Sometimes it's my fault. Electrolytic caps go bad from time to time. Heat is a capacitor killer. It is possible that they may need to be replaced after a while. Electronics fail...it happens. I MOD just about everything. There's always "something" that can be done to an electronic device to enhance it's functionality!!! I could say the same for my car or the kitchen knife I use. About 80% of the time when I blow up a controller, I can fix them. That's becasue usually mosfets blow or some other relatively easy fix can be done. I'm 100% against potted electronics becasue when things break, all I can do is throw it away or spend many hours picking that crap off the board. Conformal serves the exact same purpose as potting and does not keep you from repairing things if needed. Potting is done 100% for the purpose of generating future sales. GRRR!

I like the idea of the Mobipus controllers and these controllers are producing good results so far. Honestly, the potting flat turns me off! I will eventually blow up the controller. That's a foregone conclusion in my mind. If I'm going to drop serious bank on a controller, then when I do break it, I will want to be able to find out if I can fix it again. Potting will make that quite difficult.

I'm seriously looking at getting the biggest model Mobipus makes, but it will set me back Some $1200 and I need 2 of them. There is a pretty good chance I will pop something after a while. Maybe It just died on it's own...it happens. Maybe it's me pushing it hard and I caused the failure. I'll want to find out what my $1200 really went into and why it broke.

Regen limits...lol...don't worry folks...someone will write a work around for this. Hack the firmware...that's my first choice. It's just software and this IS an open source product.

I couldn't disagree more with every point. You are the polar opposite of the target market, I'm confident mobipus aren't going to change their opinion on potting based on your preference for repairing things you've exploded. Please do not buy there's controllers if you do enjoy fixing your fuckups, you'll be rather bored.

You are entitled to your opinion.
Please try to be respectful and non-critical of others and theirs!

Attempt to read what people write. Things go wrong and not just becasue people "frock up" as you say. That was rather condescending on your part.

It is true I have made mistakes and damaged or destroyed things. I said as much.

I've been working on electronics professionally since the mid 1980's. Electronics "frock up" all by themselves too. I'll estimate at 80%. AKA...80% of the time it's not a human fault...it just died. I've seen 3 types of typical electronic failures that happen without humans. Infant mortality in electronics is the biggest. Electronics are more likely to fail right after production than any other time. This is why testing completed products is so important! Electronics get old, caps develop internal current leaks, transistors don't switch as fast or fall out of spec over time. Solder joints from heating and cooling over time crack and separate. Non lead/tin solder grows hairs that short adjacent pins. The list is endless! Sometimes freak events happen...lightning strikes, acts of god...call them what you want and the electronics die. 80% of electronics repairs I've ever done were not caused by humans "frocking up".

I think you might be personally invested in the Mobipus products. Good for you! But it also means you are more likely to take things personally when someone expresses a perspective you don't agree with. It was never my intention to be offensive or to disparage the Mobipus product line. Potting specifically makes electronics repairs very difficult.

My perspective is based in years of repairing electronic devices. Call it cynicism if you like, but electronics have a failure rate much higher than 0. The more complex the electronic device, the greater the chances of a failure. Add high current, high voltage demands and the failure rate climbs yet again. None of this has anything to do with good or bad designs. It's just the nature of complex electronic systems. Repairs WILL be needed! I don't care what it is or who makes it. THINGS WILL BREAK.

Remember Samsung? Most people don't. People think "high quality and reliable electronics" when they they think of Samsung. I know Samsung from the 80's. GAWD!!! The WORST products on the planet! My first electronics repair job was fixing Samsung products. There was easily a 25% failure rate within 90 days of purchase!!! Their monitors had a 60% failure rate within the first 6 months! I did loads of component level repair on Samsung products for several years.

Potting is BAD when stuff breaks....AND IT WILL.

This will definitely wrankle you...
I'm not implying anything so don't read into this that the following statement is implying anything about Mobipus controllers. It's NOT!!!

Potting often times gets used to hide poor quality work. I've run across this more than once in repairing things. For a while I was repairing car ignition modules. They cost $300-500 each and so it was compelling to repair them and sell them as refurbished. This was for Ford and Chevy in the early 90's. I'd heat up the module in an oven to soften the potting so it would crumble off more easily. Once I had a bare board, I'd start going over the board looking for the failures. Solder joint failures were super common. Parts insulated inside the potting would get hotter than they would if they were in dead air space and would burn up. This was very common with small resistors and transistors. There would be patch work. An unfinished design or flaws found after production had started and patches added after the fact. There would be little added boards manually wired in or cut traces with wires going to some parts hot glued to the board elsewhere. Just awful quality and pathetic design work! The potting was there to hide the craptastic ignition module design and to make them hard to repair so people were forced to buy another module, not for water protection. They were already in sealed cases.

Potting gets used to hide your design. Want to keep the Chinese from duplicating your work? Bury it in potting.

It's not just there to supposedly protect the electronics from water intrusion. Based on my experience, most of the time that's not the reason at all.

Top reasons to use potting:
1. Conceal poor quality work
2. Conceal the design so it is hard to duplicate and reverse engineer
3. Make repairs difficult or impossible to increase repeat sales
4. Protect electronics from water incursion, but conformal does the same thing
5. Add some level of thermal transfer to some components at the risk of others that it insulates

 

[ElectricGod]

Mobipus is selling a huge volume of controllers to dealers globally. There is the rare customer that owns a soldering and knows how to use it, but...I'm guessing that's less than 1% of their customers.

If given a choice, I would source one with no potting, but...my main interest in Mobipus is that...I have never seen a higher percentage of customers who were happy with the performance and reliability of a controller. That caught my attention.

Easy programming is a HUGE benefit. Google "programming a Sevcon, Adaptto, Kelly, etc", and see what builders say about them...

I've never messed with an Adaptto...can't say about that...how hard or easy it is to configure.
I have 2 Kelly's and they are pretty easy. The app is not always intuitive, but the options are pretty obvious what they do.
I have an ASI and it is pretty difficult to configure with a steep learning curve. Anybody looking for a first controller should avoid ASI.
I've messed with Sevcons a fair bit. GEEZ! Expensive app, expensive programming dongle and a complex set up!
I don't have any experience with VESC based stuff either, but it looks fairly complex too.
Sabvoton is pretty easy, but all that exists anymore are the garbage Chinese knock-offs.
PowerVelocity controllers are easy to configure.

If Mobipus falls somewhere around a Kelly for difficulty, that won't be difficult to configure. I wont know until I have one in my hands for comparison.

Electronics fail. That's a fact...like the sun rises in the east and sets in the west. Cold hard facts. I don't care who makes the product or how careful they design or the parts used. Electronics flat out fail. At some point, there will be OEM's with a stack of DOA or older controllers that have died. Those controllers will end up in the hands of somebody and that somebody will probably be someone like me and they will want to try and repair them. Their boss may even tell that person to fix them!

What happens if 6 months down the road Mobipus has a rash of mystery controller failures. 5000 dead controllers come back to them for replacement and it's costing the company a fortune to replace them. They NEED to investigate the failures and find out what is going wrong. BUT they have shot themselves in the foot and buried everything in potting! LOL. Oh well...I guess we'll just keep going on like we've been going on and hope for the best. Uhhhh...I doubt it! Mobipus WILL be repairing those 5000 controllers and try to recoup their losses!!! Oh yeah, but they can't...FML...buried them in potting. LOL!!!

 

[Samd]

Good lord we've jumped the shark.
Haven't checked in and it looks like a bunch of kids squabbling and ranting.

Its a small controller with high power capability.
They use thermal potting. For thermal management.
I suggest you move on son.

 

[ElectricGod]

Good lord we've jumped the shark.
Haven't checked in and it looks like a bunch of kids squabbling and ranting.

Its a small controller with high power capability.
They use thermal potting. For thermal management.
I suggest you move on son.

Another investor in Mobipus...good for you! I really hope you make loads of money from this venture!!!

Hopefully I can get one unpotted so I don't have to do it myself.

I did list thermal management, but I also listed other stuff and pointed out that potting is good and bad. It makes a great insulator to keep in heat and makes things hard to repair. It protects manufacturers from people that would copy their work. It serves multiple purposes...some good, some bad. How is anything I said untrue, an exaggeration or in any way non-factual?

I have yet to squabble or rant...just stating facts and why I don't like potting...from the perspective of a person with years of electronics repair experience! IF that's squabbling and ranting you have a lot to learn! What's more probable is some people don't like the exposure of a differing point of view so anything I might say is treated like it's said by a whiny 5 year old. I'm only presenting facts.

I get it I really do, the folks that disparage the facts I present do so becasue they are probably investors in Mobipus. Good for them! They want to protect their investment. I can't blame them! And honestly, I hope Mobipus does well and has a long term business.

However that does not change the purposes of potting and the good and bad reasons for using it. If people can't see that and wish to continue being condescending instead of dealing with facts, well that's their prerogative. I get it...attack the messenger since you can't deal with the facts. Lame...but I do understand.

BTW...I'm still interested in these controllers and seriously want to buy a few. However I REALLY dislike the closed minded and condescending attitudes. If you really have a spectacular product, then there is nothing to hide.

Cheers!

 

[Samd]

Good lord.

Please move on. You can genuinely add zero value here with that lecturing approach. Lecturing both myself or the OEM on thermal properties is really not worth the time of any parties for a military supplier.

Please just go back to your other threads. This isn’t helping anyone.


Sent from my iPhone using Tapatalk

 

[ElectricGod]

After this, I promise, I won't mention "potting" again in this thread. I'll use "some other term".

Depending on which one of these groups you fit into, potting is good or bad for you.

1. For OEMs and parts sellers, potting is brilliant. They know the electronic device can't be repaired so they get to sell a new widget.

2. For the end user such as a DIY EV builder, they have to buy that new widget when the old one may be repairable IF it wasn't buried in potting. For you, it's BAD.


I'm talking to end users...NOT OEM's or resellers.

Potting makes your electronic device very difficult if not impossible to repair WHEN it breaks. 100% of all electronics devices will eventually fail. Some fail sooner than others. High amperage electronics (motor controllers) are FAR more prone to failing than low power electronics. Operating electronics at a few milliamps versus several hundred amps has it's advantages. Same for voltage...5 volts vs 100 volts is a BIG difference. Both of these things together make low power electronics tend to last far longer than high power electronics. I have several 30 year old digital watches that still work. I don't have any 30 year old high power electronic devices that still work.

Some percentage of electronics devices will FAIL soon AFTER the warranty runs out. Some percentage will last many years before they fail. In either case, you won't be able to get your expensive controller repaired. If it had no potting inside, that may be easily done. Blown mosfets are super common and easy to replace. Thanks to potting, you won't have that option and will be buying a new controller instead of replacing $30 worth of parts in a $600 controller.

2 nights ago I blew 2 mosfets in a brand new controller. No big deal, I have spare mosfets and that controller only had conformal on it. Replacing the blown parts was cheap, quick and easy. I spent an hour and it cost me $10. IF it was potted, I would have been unable to repair it. I didn't even need the warranty!

This is potting. It's poured in as a thick liquid that flows down into everything. I agree that it does protect from water and chemical incursion and that does have it's place. BUT, protection from incursion can be done other ways too. Conformal is a great solution and so are silicon seals on the controller shell. Potting cures into a soft epoxy that generally has a rubbery feel to it. Once it is used to cover electronics, it's a safe bet that those electronics wont ever be repairable when they fail. It's a giant PITA to remove when you need to repair said electronics. Depending on the formula of the potting, some get soft and crumbly when heated. Others do not. Quite often potting chemistry is designed to be chemical and heat resistant. This also means that common solvents don't remove the potting and that heating won't make it break down. AKA...the electronic device is NOT repairable since you can't get past the potting easily.

EDIT:
I forgot to mention what conformal is. Most people won't know what this is.

Conformal is a thin, clear coating. It's sort of like clear fingernail polish or spray on enamel. Coating electronics with it protects them from water and chemical incursion. It also doesn't get in the way of soldering or desoldering. If you need to do some form of work on a circuit coated in conformal, go for it! At worst your soldering iron tip will get dirty quicker. I buy it in a bottle and brush it on the exposed electronics I want to protect. When cured it usually creates a glossy, clear protective layer over the electronic parts that does not impact repairability at all.

The BIG difference here is conformal protects just like potting does, but leaves the electronics accessible so they can be repaired or accessed later.

Conformal since it is a thin layer, does not thermally isolate parts. If I needed to add a heat sink on top of these parts or blow air on them or whatever the cooling method, that is still possible. It's a much more flexible protective product than is potting and provides the same kinds of protection from water and chemical incursion.

These close-ups are from a 20S BMS that has conformal on the small electronics. Notice how everything looks glossy and a bit bluury. That's not a camera or photo issue, that's the conformal coating. I think on this BMS, they dribbled it over the electronics or brushed it on since in several of the images you can see where the edge of the conformal is and where bare circuit board is. It can also be sprayed on to coat everything.

 

[Samd]

Admins can you please remove the troll


Sent from my iPhone using Tapatalk

 

[trazor]

Admins can you please remove the troll


Sent from my iPhone using Tapatalk

That is rude man. ElectricGod wrote a very informational and neutral piece to potential buyers of these controllers. Move on.

 

[ElectricGod]

Admins can you please remove the troll


Sent from my iPhone using Tapatalk

That is rude man. ElectricGod wrote a very informational and neutral piece to potential buyers of these controllers. Move on.

Thank you!

I deliberately chose to NOT attack anyone, but to present facts only.

I understand the hostility and personal attacks. I present facts that OEMs and resellers don't want you to know.

Never used the P-word...lol!

 

[romelec]

Mobipus is still a new company and for more than a year they didn't answer to anybody, even their distributors.
This could happen again or they can go out of business in a few years, and that's why repairability outside from the manufacturer is important.
Even if their controllers are better than the other chinese ones there will always be some failures:

Anyways Shenta has had a change in life direction and can’t be contacted for awhile. Its a shame as sketch and I sent back a 600a controller each for repair and no one will return our queries. We’re a couple of grand out of pocket so I will keep following it up.

Did you get your controller back ?

 

[larsb]

Yes, can we all agree that EG has a point but it's been stated enough/too many times now.

I'd like to hear some user stories at the absolute max output side of things

And samd i guess you'll want to explain how the controllers died now that the questions out there, I'd like to hear that too.

 

[ElectricGod]

Yes, can we all agree that EG has a point but it's been stated enough/too many times now.

I'd like to hear some user stories at the absolute max output side of things

And samd i guess you'll want to explain how the controllers died now that the questions out there, I'd like to hear that too.

I would have stopped a lot sooner, like the first time I talked about the P-word. However, I got attacked instead of people acknowledging the impact that the P-word has on electronics repair. The end result was I stayed on topic much longer and wrote it up in a much more concise way that is hard to refute. It's hard to attack the message when it is so obvious. AKA attack the messenger instead. It's certainly not the first time that what I said couldn't be refuted and I got slammed as a person.

I mentioned to Ohbse that I thought he was probably an investor in Mobipus and that would explain his bias against what I presented. I said more or less the same to SAMD. I congratulated both of them for their involvement and wished them well. As it turns out neither one of them ever acknowledged their strong ties to Mobipus. I presented plausible reasons for their strong perspectives and why they took what I said personally and chose to attack me. I never attacked either one of them or belittled them in any way that I'm aware of. I didn't even attack Mobipus. I presented reasons why P-ing creates problems.

I edited my last post on the P-word and added a section about conformal...which I forgot to do when I hit submit.

So anyway, back to the controller...
Me too! I'd love to see some real solid results doing 600 phase amps or whatever that specific controller does for maximum continuous rating.

People will read "maximum continuous" and think I mean peak. Peak is a fairly meaningless value IMHO. You can't run around for an hour straight at some peak value. At most, you can do that for a few seconds before the controller burns out. However "maximum continuous" means I can run around for an hour straight at some current level or wattage level or whatever it is and not blow up the controller.

I mod motor controllers. A 12 fet rated for 2500 watts continuous, in my hands can do 4-5kw continuous and do that for many miles in a row. That 5kw is not a peak value. It is a "maximum continuous" value.

That's the kind of info from people that use these controllers that I'd really like to see.

SAMD...
Please tell us about those failures. Whether you do it or not, eventually this will become known. People will have problems and will post what happened.

 

[Samd]

That wasn’t a failure with the controller it was a failure with the rs485 input where Sketch accidentally connected 80v to the 5v input via a USB dongle. You can’t do that with the wifi dongle.

Yes we got them back.

What’s weird and bizarre is an electrics tech repairer lecturing someone with a double degree and five years of thermodynamics on how thermal management is best done inside electronics. Its not like we can have a discussion on specific heat units or entropy. And thats just me.

Moreover assuming that an OEM supplier to military and OEM racecars wants you to open one, scrape it out and offer your view on their thermal design either EG. When there is tens of thousands of these in asia being used daily.

You might as well get a mechanic to call Koenigsegg and offer his views on turbocharger design. Maybe drop Mate Rimac a phonecall and tell him not to use mobipus either?

Sketch’s superbike has been running and racing for two years. The taipei red devil still gets ridden daily.

Why are we even going down this road with concern trolling? Have you just assumed that an OEM is the same as some chinese kunteng or kelly controller?

ES web has jumped the shark. This thread is testament. It used to be a place where admins managed concern trolls. Now its an archive of cool things some of us did two decades ago frequented by people who are out to argue the negative in everything.

I can see why other forums are doing well. Its a world different on Facebook that’s for sure.


Sent from my iPhone using Tapatalk

 

[liveforphysics]

In my testing, conformal coating made only a few minutes difference from uncoated in salt-fog testing. Potting entirely prevents the failure. This is because conformal coating leaves the upper sharp geometry edges thin/pinholed. I experimented with multi-layer sprayed conformals, dipped conformals, and Parlyene coatings, and only Parlyene offered any repeatable useful protection in salt-fog. This is because you can sometimes get something conformal coated defect-free using a multi-layer process, but even 99% being defect free would still be unacceptable failure rates. Potting offers vibe protection, thermal advantages (for some pottings), and most importantly corrosion protection against salt fog.

 

[ElectricGod]

In my testing, conformal coating made only a few minutes difference from uncoated in salt-fog testing. Potting entirely prevents the failure. This is because conformal coating leaves the upper sharp geometry edges thin/pinholed. I experimented with multi-layer sprayed conformals, dipped conformals, and Parlyene coatings, and only Parlyene offered any repeatable useful protection in salt-fog. This is because you can sometimes get something conformal coated defect-free using a multi-layer process, but even 99% being defect free would still be unacceptable failure rates. Potting offers vibe protection, thermal advantages (for some pottings), and most importantly corrosion protection against salt fog.

Thanks for that input on conformal. I also suggested using seals on the controller shell. Conformal isn't a perfect protection, but it is "pretty good" protection. I coated a logic board some time back with it and then stuck the hole thing dangling by it's wires in water while it was running. I left it in the water for about 30 minutes and it never had a problem. I should take a nano and do this again. I don't see what would change, but I bet it would still work while fully immersed in tap water. Maybe try salt water since it's a bit more reactive.

The idea behind conformal is to make the electronic device still repairable and yet fairly impervious to incursion. Full epoxy drowning, if what you want is perfect incursion protection, IS the best option, but forget repairing anything if you do! I guess if you are building stuff for the military or an airliner or a pace maker where you can't tolerate any amount of failure from incursion, then full epoxy drowning may be the best option. For most other electronic devices, it's probably not needed and conformal would add an acceptable level of protection.

Thinking about this a bit more...
I have a digital watch good down to 200 meters. I took it apart about a year ago out of curiosity. There was no coating on the components at all. The ONLY protection was the seals in the watch case. Based on that Casio watch, I'd say that even conformal isn't needed. Just seal up the enclosure properly.

 

[liveforphysics]

It all depends on your risk if the seal fails.

It's another thing that works great 99% of the time when the seal surface interfaces had no defects or debris contamination in assembly. The 1% remainings risk with defects in a motor controller where life safety or fire is at risk, that 1% just got more expensive than the whole margin of the 99%.

I've been doing EV systems for 12 years now, and I've still never found a replacement for potting performance in product robustness mechanically and chemically both. Even in aircraft you pot whatever you want to survive with high reliability.

 

[ElectricGod]

OK...NVM...going to use the P-word. lol!

Potting has it's place to be sure. If it's life or death, by all means make those electronics as hard as possible to damage as can be done. OF course on any kind of commercially made airplane the cost for parts is insane. A simple 12v light bulb that can be found at any auto parts store for $2, costs $30 just becasue of the liability associated with it. Same light bulb, just much more liability. If maximizing safety or minimizing liability is the goal...potting can be part of that answer.

If it's NOT life or death or high liability...I don't see the point. Not when other methods can be used that are sufficient. I've never considered an EV a place where I need that level of protection. Maybe if I had a people carrying electric airplane or quad...maybe. IF it is considered experimental by the FAA, then I'm pretty sure I would want access to every part of that flying craft. IF it was gas powered, I could absolutely take apart every part of that engine and replace worn or damaged parts! If there was sufficient concern about water incursion into the motor controllers for my people carrying quad, I'm pretty sure I could work out ways to seal up the controllers without making them impossible to service if needed.

A friend of mine in Arizona is building an 8 motor human lifter. He has 8 Alien Power 120100's on it that are custom wound. He originally potted all of his controllers, but then from time to time one would burn out for whatever reason. He's on his second set of motor controllers now and they are NOT potted. It has saved him a ton of money being able to repair them instead of replace them like the first 8. I don't know if the boards have conformal on them or not. Last time he talked about it, the shells were water tight. For most people and uses, that's probably sufficient.

How many people drive down the road every day in cars that are full of connectors that are not water proof at all...just pins sliding into open holes. If that's acceptable in a car, how is a motor controller on my electric motorcycle or ebike or whatever in need of sooo much more water protection? IMHO, I don't see the need for this level of incursion and mechanical protection for electronics for 99% of consumer uses. We just don't need it. I bet if I pulled out the ignition module in my 2018 Golf R and opened the case, the board inside would not be potted...just the case sealed up...if even that.

Anyway, enough about potting!

What kind of maximum continuous results are people getting from these controllers?

 

[macribs]

Thinking about this a bit more...
I have a digital watch good down to 200 meters. I took it apart about a year ago out of curiosity. There was no coating on the components at all. The ONLY protection was the seals in the watch case. Based on that Casio watch, I'd say that even conformal isn't needed. Just seal up the enclosure properly.

As I said earlier, potting got a big upside for people living with salted road conditions. That salt mist penetrate things you wouldn't believe was possible. Something even well known car manufacturer's get bitten by now and then. And many of them have been in business for more then 100 years.

Seals are great when they work, but one day they will fail. Like the well known o-ring. Applied correctly it will ensure a tight seal. But every once in a while something eats away that seal prematurely. Like on forks or shocks.

Point is that I don't think there is one universal solution that will fit all needs. But as for most users I think potting is an ok compromise. You give away the ability to fix it yourself, but gain weather proofing and improved cooling. For many of us, we are not even competent enough to do our own controller-work-repair anyway so we might as well don't care too much.

 

[ElectricGod]

For many of us, we are not even competent enough to do our own controller-work-repair anyway so we might as well don't care too much.

And that's why in just about any city anywhere are people like me working for companies that do these kinds of repairs. I haven't done component level repair as a job in a long time...1994? But I still do it for myself and friends.

Remember the old TV repair shops from the 80's and further back? Those guys could fix just about anything at the component level. Now they are mostly board and module swappers. I went to school for an EE and got jobs doing electronics repairs in college. I repaired PSU's, monitors, TV's, motor controllers...just about anything. The place I worked for in the lat 80's had service contracts with Samsung, OKIdata, WYSE and a few others to do this sort of work. They expected us to replace components, NOT boards. Of course with SMT becoming mainstream, more of that went to board swapping since SMT is generally considered too inexpensive to be worth repairing. I disagree, depending on what it is...I've done a good bit of SMT repair work too. The trend now days is towards board swapping as the norm for most repair shops. Still..there's plenty of people like me around working for places where they use a tracker, scope and logic probe to find the root cause.

I learned how to reflow a few years back just so I could do SMT work at home.

Googling for my area, I found 3 shops that do component level work in less than a minute.

If you personally can't fix it, there are people around you that can.

I'm in the USA, Several times I've come across people on ES that blew a controller. I'll ask if they would like me to take a look at it. I don't always succeed and I make sure they know that up front. For example if the MCU is gone, I'm not going to bother at all. If a couple mosfets or caps are gone...that's cheap and easy and worth it. If they have blown all the mosfets...then it gets more questionable if it's worth it. I've seen that a couple of times. Short 2 phases together while the motor is running...LOL...lots of mosfets die instantly. I'd say most of the time I'll find 2-4 blown mosfets and the controller is good to go again. IMHO...that's well worth the repair cost.

AKA...$30-100 to have someone repair a controller vs $300-1200 to buy a new one...totally worth it! An ebay $40 controller...just buy another one.

 

[liveforphysics]

It matters for the EV controller application because it's parked in someones garage/home/work place, so it's undesirable to have it fail in corrosion induced fire. Being a motor-controller you may be using for human-life-safety applications like highway traffic, it's important not to randomly fail.

It's actually fine if you still don't understand why potting is the best choice for a motorcontroller, and it's also fine if you like having failures and servicing things at the component level (I buy much of my $$$ battery lab equipment broken on ebay and repair it myself at the component level). It's still going to be the best choice to pot vehicle electronics which are exposed to environments nothing like benchtop electronics equipment.