Rickys High Power Flexable motor controller

[deVries]

Bumping for any updates... it's been sooo long... (by Ricky_nz » Sat Jan 28, 2012 6:54 pm ) been dreaming about your controller since you started this thread...

[Ricky_nz]

HI All,
Just has a PM from deVries that reminded my I hadn't really been on the sphere for quite a while... The email on PM feature is really handy.

I just got busy with house maintenance and visitors chewing up my spare time.
Now that the weather is getting cooler and its darker earlier I should be able to find more time to work on my controller.

Now I've paid for my kitchen changes and a few other things I can spend on my hobbies again so I'm sending off an order to digikey for some more parts next week so I can make a minor repair to my first board and also get the parts to build up a second controller. I will put it in a box so I can try it in sensored mode on a bike while I get the sensorless stuff sorted.
I haven't had time to work on the software but hopefully I'll get back into that shortly.

I just had to make use of the good weather for general maintenance on my house. Also removing TV aerials and replacing one.
With analogue switch off coming the VHF antenna is gone and the combiner and corroded UHF aerial are also gone. I replaced the UHF with a slightly higher gain one and put up new coax and the original stuff was full of water (I was on the place when I brought it) The new stuff won't get water in it!
While I was on the roof a bit of fine tweaking and my 70cm satellite dish receives NZ freeview and also the Aussie SBS channels both with 0 bit errors! Kind of strange sitting on the roof with a laptop .

I had to get my mythtv box nice and reliable as none of my TV's have DVBT tuners and I don't have a set top box so when analgue is gone I have no backup for the mythtv setup. Mind you I wouldn't want to watch TV without the mythtv box .

I have had a new set of Lipo here to build an 16S2P pack sitting here and after a couple of cycles I had left it to sit for a month and there is no significant self discharge imbalance between the cells so I think its good enough so I will probably build up a pack and then I will take the two year old pack off the bike and use it for controller development. Just need to change the caps on a e-crazyman controller as 67V is probably a bit too high for the 48V one . You heard right my little 300W headline/cyclone/elation is in line for 67V.
If I order some smaller CT's than the 200A ones I have on the first controller I built I could probably run my controller sensored in the place of the ecrazyman which is probably a good idea. I think the 50A CTs should provide sufficient resolution that I can control the current in the headline motor so if they are In stock I'll order a couple. I also need to get a couple of rain proof boxes to put my controllers in. The more I think about it I think I will build up one controller for the GM 1000W hub with 200A CTs and one for the headline with 50A ct's and only 6 FETS fitted.. For the headline it probably won't even need a heatsink . I figure if i tried to even put 35ARMS (50A peak) in the cyclone windings they might cook hence the 50A to handle the peak. If i can't get 50s Ill get 100A CT's the 200's just don't really give enough resolution for small motors.

Anyway I'll try and keep the updates more regular as I get back into the motor controller project.

[Ricky_nz]

Finally ordered the parts I needed to continue from Digikey this morning so hopefully I can start making progress again.
Its great timing as ordering the parts Saturday morning NZ time as it has them out of Digikey before their weekend starts . I already have valid tracking data showing they are moving through the US. Hopefully I'll have the parts before next weekend provided I don't have issues with customs.

As often the case Murphy struck and I left a couple of parts off.
Why is it I always think of something I should have ordered just after finally confirming the order
Seems to be related that the problem of not spotting spelling errors until the print button is pressed

It was just a couple of hall sensors and some mini fit connectors I didn't order which luckily are available and in stock locally from rs components for a slightly higher price .

[Ricky_nz]

small update
On Friday I got my Parts from Digikey and also another package from hobby king so I can start playing with the hardware again .

I pulled out a partially assembled power board I had set aside waiting for parts and completed assembling the power supply section and decided to test it but it wouldn't power up. The output was 0.2V!

It wasn't drawing excessive current.
The under voltage lockout threshold was crossed and the chip on chip regulator was becoming active and providing the correct voltage.
The bottom FET had very short gate pulses.

All the SMD resistors were correct and I was certain the caps are the correct value as I double checked them as I assembled it.
I started probing around a bit more and eventually found the soft start pin stuck at 0V!

I measured a hard short to 0V but I could not see a short with a magnifying glass and couldn't wick anything out from behind the pin on the converter chip.
I finally decided to stop wasting time and just pull the chip So I cut the legs off the NZ$9 chip and then heated the ground plane vias below the chips ground pad and removed it and in the process cleared the short. What I believe happened was I got the chip slightly askew and not flat enough to the board so a small blob of solder managed to flow between the ground and and the back of the pin when I soldered the ground and due the the thermal mass of the ground I couldn't melt the bridge from the pin side!

Heres a photo of the bottom side of the replacement chip. The big pad being ground and the pin spacing center - center = 0.5mm!

LM5116 Bottom

LM5116_bottom.jpg (7.7 KiB) Viewed 362 times

I guess I just need to be more careful mounting these chips as there isn't much room for error. Also I should probably feed a minimum amount of solder through the vias to reduce the chance of shorts also.
At least I know that my board provides good heatsinking of the chip as I had to dump my 70W iron into the ground for a significant amount of time before I could free the chip.

I have got a diecast box to put the second controller in and if It works out alight I'll buy another of the same type and build up a third controller.

Not sure how I am going to cut rectangular holes in a diecst box.
In the past I've done it with a bunch of holes and a file but that is kind of painfully slow. Maybe I should buy a dremal or similar rotary tool to save some time. I'm sure that would allow me to cut holes in the wrong places quicker


EDIT: After a coffee I decided to solder on the new chip and power it up and it fired right up .
Now I need to build up the gate drive section but that can wait till tomorrow.... err um tonight... 30 mins past midnight here oops.

[h0tr0d]

Any updates on this?

Sales thread?

[Arlo1]

h0tr0d you are to impatient. There is many of us working on bad ass controllers but if you want it any faster you need to build your own.

[John in CR]

h0tr0d you are to impatient. There is many of us working on bad ass controllers but if you want it any faster you need to build your own.

That or figure out how to get the most out of cheapies. I'm getting 15-25kw peaks at well under 1.5cents/watt.

[Ricky_nz]

Any updates on this?

Sales thread?

No sorry,
Not enough progress in 2012.
To much house maintenance (did i mention painting roofs sucks). The roof is galvanised iron (steel) and the galvanising was getting thing (40 years old) so I had to kill some rust and then two coats of good primer (oil based one as the water based ones are no good on old roofing) and then a couple of coats of water based top coat. Also replaced a heap of old lead head nails with roofing screws.
All done now and looks good and didn't cost too much thanks to my dads trade account at local paint store .

I also had to spend some time making sure I have a second working bike as my place of work is moving to a new building further away so I wanted a backup. now i have multiple batteries etc so I can mess with one bike and still have the other working etc

Enough with the excuses
I'm just about to start getting back into it seriously now the Christmas visitors have gone.
I'll try and do some more updates when I make some progress.

I think I have figured out why I managed to kill a couple of power supply chips LM5116 ( killed another one handling the board while powered up recently). I have located a via on the board that is a low voltage point of the converter chip that is way too close to a +100V track. I guess I missed a clearance rule in the PCB layout . I'm going to have a closer look at it but I think a small trim to the thick +100V track with a knife and then a blob of epoxy over it should prevent issues.

I have some more control code changes that after a bit more fiddling in the simulator I will transfer to code.

[h0tr0d]

h0tr0d you are to impatient. There is many of us working on bad ass controllers but if you want it any faster you need to build your own.

That or figure out how to get the most out of cheapies. I'm getting 15-25kw peaks at well under 1.5cents/watt.

Nahhh... I'm just looking for a cost effective solution. The needed learning curve regarding controllers is long and steep. If I've got to go in that direction, I'll probably abandon this project.

There's many people here that "almost" have a working controller, or basis, all didn't quite make it. This is very unfortunate I would say.
For them, could make good money (and pay other people to fix their roof, ex. ) and help this community and entire EV revolution.

Since Jeremy's "lobo" controller could, in theory, go up to 100khz PWM, and phase current control, I'll investigate that...

Happy New year to all, may all your projects succeed!

[Lebowski]

my controller works fine (I think ) just got no big motor / battery / vehicle combo to try it on

But I don't believe there's (big) money to be made from a good controller

Batteries and motors, that's where the money is, controllers are just seen as the
bit you need in the middle to get it working. Like with cars, the money is in the
gas (not in the cars anymore, ask GM), not in building the carburetor.

[Njay]

As I see it there's a big difference in there, Lebowski. At least in 2-3 wheel diy vehicles, a good/powerful controller can cost half or more of the motor and 1/4 of batteries.

Just look at how many controller builders there are here; I'm sure it's not just a matter of having fun

[Lebowski]

Just look at how many controller builders there are here; I'm sure it's not just a matter of having fun

? You sure ? That's why I build mine, just for fun and out of curiosity....

[walls99]

Just look at how many controller builders there are here; I'm sure it's not just a matter of having fun

+1 for the fun of building you own controller

Of course, not finding something that met my requirement, also pushed me in that direction. FYI, my controller is also fully functional and open for other DIY-er to build their own controller with limited effort on the SW side.

[Alan B]

Sevcon makes good money selling controllers. But they are a major pain to program for ebike use.

The big problem is that we have essentially zero really excellent ebike controllers. Every one that is out there has shortcomings.

There are some promising developments, but nothing available yet.

There is a huge opportunity for improvement.

But it is a hard product to make money on since the cheap controllers out there set a low expectation of the price.

[Njay]

I have no doubt that the major part of it is having fun. But the thing is, and as was already mentioned here, there are "shortcomings" in the currently available ones that push people into designing their own. And it's hard stuff, so we don't take the decision "lightly". I also thing there's room for a lot of improvement, and it's not always a matter of "better" hardware; many improvements are just sw, but people is locked on the closed "chinese" stuff.

[Alan B]

I agree with that - the hardware improvement is interesting, but the real improvement is in software. The software that is out there is too closed and not suitable for improvement, so it is necessary to start over.

However the manufacturers of the chips provide huge support for software development. If we had some good open hardware using standard processors these software tools would be used to make a variety of good controllers. So we need open hardware and open software and this will take off.

[Lebowski]

I thought about it, making a controller and selling it.. But you have to realise we here at ES are
not the typical person on the street. We here have a way above average interest in electronic
vehicles and the controllers associated with them. And most of us here don;t care about
the controller, most questions here are about batteries and torque arms.

I cannot think of 25 people here on ES that would be interested in my controller, and at a few $
profit per sold controller it's just not worth going into business.... especially, as pointed out
by Alan, the market is flooded with cheap chinese controllers which, lets be realistic, are good
enough for most peoples needs.

Output stage wise, I don't think it as difficult as many of us think it is to make a big output
stage capable of say 10-20 kW. Parts and technology for this are readily available. I'm an
analog IC design engineer by profession so this sort of thing is right up my alley. If you
want to build something like that though you need a lot of equipment to test it, i'm talking
motor, batteries, vehicle or some sort of dyno. Plus, even then for me personally there
would be no point as laws here prevent me from ever riding around on my creation so why
bother ?

I hope I'm wrong though, it would be nice to start a profitable EV controller business......

[Njay]

For ebikes there are plenty cheap and reasonable options... not so much for motorcycles.

[h0tr0d]

Personally, I want to buy a controller not print pcbs...
Lebowski, you provide a chip, not a working controller (Arlo1 is almost there)... What's the max PWM freq your chip can go?

Burtie and Marti are developing a controllers but not there yet...

Wall's controller is looking good also, can you sell me something so I can "bolt" some big fets and caps and go for 20kW?

[Arlo1]

Personally, I want to buy a controller not print pcbs...
Lebowski, you provide a chip, not a working controller (Arlo1 is almost there)... What's the max PWM freq your chip can go?

Burtie and Marti are developing a controllers but not there yet...

Wall's controller is looking good also, can you sell me something so I can "bolt" some big fets and caps and go for 20kW?

What motor do you want to run with it at what voltage? If you settle for 100v or less a 24-36 fet controller with 4110 fets will rip hard. I was able to get the picture from my avatar from a 24fet but its just a spike. It actually only run at 250 amps an 90v under load lol. Man the pour X5 was so hot instantly!
If you want plug and play you will be waiting a while unless you want to spend BIG money.

[Lebowski]

Personally, I want to buy a controller not print pcbs...
Lebowski, you provide a chip, not a working controller (Arlo1 is almost there)... What's the max PWM freq your chip can go?

Burtie and Marti are developing a controllers but not there yet...

Wall's controller is looking good also, can you sell me something so I can "bolt" some big fets and caps and go for 20kW?

Arlo is using my chip, so.... PWM can go up to a few 100 kHz, so it really depends on what
the output stage can handle.

Getting back to selling controllers versus a chip / PCB combo... You have to keep in mind that,
with all the metal work involved building something like the Big Lebowski controller takes
2 to 3 days. Are you willing to pay $50 an hour (my daily job rate) for that ? Having it build in
China for less only makes sense from hundreds or thousands of controllers upwards...

[Njay]

Getting back to selling controllers versus a chip / PCB combo... You have to keep in mind that, with all the metal work involved building something like the Big Lebowski controller takes 2 to 3 days. Are you willing to pay $50 an hour (my daily job rate) for that ? Having it build in China for less only makes sense from hundreds or thousands of controllers upwards...

As a side note, you "should":

1) Re-design the controller (including, or mainly, the mechanics) for manufacture.
Make things simpler to build, less custom work, easier assemble.

2) Outsource the assembly job to a cheaper technician.
You're too skillful (and therefore expensive) to spend time assembling stuff.

I'm not saying it will make your controller "price competitive", but in general terms that's how you do it

[Ratking]

I thought about it, making a controller and selling it.. But you have to realise we here at ES are
not the typical person on the street. We here have a way above average interest in electronic
vehicles and the controllers associated with them. And most of us here don;t care about
the controller, most questions here are about batteries and torque arms.

I cannot think of 25 people here on ES that would be interested in my controller, and at a few $
profit per sold controller it's just not worth going into business.... especially, as pointed out
by Alan, the market is flooded with cheap chinese controllers which, lets be realistic, are good
enough for most peoples needs.

Output stage wise, I don't think it as difficult as many of us think it is to make a big output
stage capable of say 10-20 kW. Parts and technology for this are readily available. I'm an
analog IC design engineer by profession so this sort of thing is right up my alley. If you
want to build something like that though you need a lot of equipment to test it, i'm talking
motor, batteries, vehicle or some sort of dyno. Plus, even then for me personally there
would be no point as laws here prevent me from ever riding around on my creation so why
bother ?

I hope I'm wrong though, it would be nice to start a profitable EV controller business......

I would like you to come to Norway and start a company with me
We will build high performance bikes with rc-drives and hardcore controllers that's bulletproof.
Every downhill guy would want one and we could be rich. After that we would focus on
building super motards with even bigger motors so they could compete with the gas ones.

I am sure that there are a marked for a 15kw downhill bike if we could make it sturdy and
user friendly. I would do this any day if I knew I would succeed in my quest

[amberwolf]

I hope I'm wrong though, it would be nice to start a profitable EV controller business......

see soliton1 saga on diywlelectriccar forum sfor example of such a star tup .

[h0tr0d]

Lebowski,

You seem the true definition of a specialist.
You know a LOT about controllers and the like, but you know next to nothing regarding the potential market for your product.
If your chip is really flexible to drive any kind of pmdc motor, then your market is EVERY ice powered vehicle and convert that to ev. Plus niche market like downhill monster ebikes.

I want to start a company to convert anything to EV, I'm starting with ebikes because of low initial investment. Then bikes, cars, trucks... And all is STUCK waiting for a decent flexible controller that I just change the power stage.