Simple BLDC controller

[Jeremy Harris]

I've managed to etch another pair of boards this morning, I'm just waiting for yet another set of components to arrive and then I should be able to get the final version built.

To recap, here is a quick run through the versions so far, and how I believe the final version should end up:

This is the Mk1 version, that I had to bodge by making a daughter board to invert the high side drive signals. This is the controller that's in the video clips, but it's been robbed of components to build this one:



This is the Mk2, which includes the comparator for kick starting the bootstrap circuit.

Finally, these are the boards for the "brawn and brains" version, using two boards:



The idea here is to have a through hole circuit board for the high power stuff, the FETs, driver chips and commutation capacitors, with a smaller board, that will be mounted at right angles to the main board, to house the controller chip. This smaller board can be either through hole, using legacy versions of the controller, or surface mount to use the current versions. It could also be a uP board, as all the driver board needs is +12V and 6 FET drive signals at logic level. The driver board includes the current shunts, so a current feedback signal is also available on the ten pin connector.

This is how the two boards will fit together:

The new version saves about an inch of width over the original, with no increase in overall height. The copper ground trace is also about twice as wide.

I'm talking with Burtie about combining this controller with his neat timing adjustment circuit, with a bit of luck we may be able to get together to offer something to the community that fills the need for driving these big outrunners well.

There's still a fair bit of work to do, but it's looking pretty good so far.

Jeremy

 

[AussieJester]

I'm talking with Burtie about combining this controller with his neat timing adjustment circuit, with a bit of luck we may be able to get together to offer something to the community that fills the need for driving these big outrunners well.

I had a post all typed out Sundee arvo in which i was going to ask if this would be an easy do, great to
read today your going to be adding the circuit Jermey/Burtie... What sort of time frame are you thinking before
a 'kit' will be available to those wanting to assemble one up for their own use Mr Jeremy? OH... haha if there
are any assembled units on offer/unspoken for I have the $$$ for immediate payment..?

KiM

 

[Jeremy Harris]

It's early days yet, AJ. I need to be sure that the controller is robust enough to take heavy loads from a low inductance/resistance outrunner. Burtie and I need to chat about how best to move things forward together, perhaps with some help from Tiberius on the PCB manufacture front, then we need to work out how best to offer boards/part finished systems to those who want them.

I'd guess I'm a couple of months away from being able to state with certainty what the power limits are for this controller and have a finalised set of boards. I'd not want to let something out in the wild that still has bugs, lest others end up wasting money on parts. Once I have what I hope will be the final version of this controller working well, I then plan on making up a test stand and brake so that I can do some load tests. The motor I'll be using for these tests will be a Colossus 7kW, so similar to the motor you're running.

The weather is hampering things at the moment, we've had a fair snowfall in the past couple of days and it's snowing heavily again now. Added to the fact that the temperature has hardly climbed above zero for the last few weeks and it makes my (unheated) workshop a fairly unpleasant place to be. Luckily, being retired, I can use the kitchen to make circuit boards (during the day, and as long as there's no visible trace left when SWMBO comes home from work..............).

Jeremy

 

[AussieJester]

I'd guess I'm a couple of months away from being able to state with certainty what the power limits are for this controller and have a finalised set of boards.

Okedokies SiR... :wink: i am at that stage on my current build where i need to make the decision on how i am going to go with a controller and make the purchase...I have bought one of Edward Lyens lil tester units so i think i'll have a play with hall/infineon setup and make it work this time until the Jeremy/Burtie offering is available As always wish you the best with it all Jeremy

KiM

 

[texaspyro]

Burtie's timing adjuster was prototyped using a small single board controller which costs a bit of money. I suspect that its functionality would be real easy to get into a small micro chip. I would probably do it with a ATTINY23 which costs around $2. Maybe set the adjustment factor via a pot read by the micro off an ADC channel.

 

[Jeremy Harris]

Thanks for the tip. I'll leave this bit to Burtie, as he's the wizz with the ucontroller stuff! I'll stick to getting the controller part to be as rock solid as I can make it.

Both of the new boards are built up as far as I can until the next batch of components arrive, but they'll probably be delayed by the weather. We're practically snowed in at the moment - bloody crazy the way that Southern England grinds to a halt when we get a few inches of snow and sub-zero temperatures for a couple of weeks. I worked in Canada for a while and nothing much in the way of weather seemed to stop things working - folk just got on and coped with it. Mind you, I never really got used to driving to work in the rain at +8C, seeing it start to snow at around zero at lunchtime, then driving home with it down to -5C. How folk in the Maritimes cope with that crazy changeable climate is beyond me.

Burtie may well be in a worse state than we are - the local news showed that Dorset had around 8" of snow today...........

Jeremy

 

[MitchJi]

Hi Jeremy,

I'm talking with Burtie about combining this controller with his neat timing adjustment circuit, with a bit of luck we may be able to get together to offer something to the community that fills the need for driving these big outrunners well.

There's still a fair bit of work to do, but it's looking pretty good so far.

Jeremy

http://endless-sphere.com/forums/viewtopic.php?f=2&t=19054&start=30#p319312

Burtie,

I'd not been following this thread, but during a discussion on the "fitting Halls to an Astro" thread I happened to mention that there was an option to switch between sensored and sensorless operation and vary the timing. Here's what I wrote on that thread:

"the idea would be to run in fixed timing sensored mode up to a reasonable rpm, to give good start up and low speed torque, then switch to sensorless mode for medium to high speed running. The Halls will give the solid position feedback needed for start up at low speeds, and by switching to sensorless you get the ability to do timing adjustment on the fly, just by varying the threshold at which the zero crossing detectors operate. This is similar to the way that pre-digital electronic ignition systems managed to get advance; they used a sinusoidal sensor and varied the threshold on a comparator with rpm."...

If you guys come up with an Astro Hall Sensor Kit and a switching 72v 120amp controller kit it would be great (I started to say you'd be stars but you already are ).

 

[ZapPat]

Both of the new boards are built up as far as I can until the next batch of components arrive, but they'll probably be delayed by the weather. We're practically snowed in at the moment - bloody crazy the way that Southern England grinds to a halt when we get a few inches of snow and sub-zero temperatures for a couple of weeks. I worked in Canada for a while and nothing much in the way of weather seemed to stop things working - folk just got on and coped with it. Mind you, I never really got used to driving to work in the rain at +8C, seeing it start to snow at around zero at lunchtime, then driving home with it down to -5C. How folk in the Maritimes cope with that crazy changeable climate is beyond me.

Burtie may well be in a worse state than we are - the local news showed that Dorset had around 8" of snow today...........

Jeremy

Sooooo.... that's where all our snow is then?! All we've been getting is rain here in eastern canada - every time it gets bellow zero centigrade the precipitation stops, and it waits until it's warm again to start raining again... yuk! :x I want to go snowboarding!!!

 

[Jeremy Harris]

Sooooo.... that's where all our snow is then?! All we've been getting is rain here in eastern canada - every time it gets bellow zero centigrade the precipitation stops, and it waits until it's warm again to start raining again... yuk! :x I want to go snowboarding!!!

I think the simple answer is over here. I can't remember having much snow before Christmas here, in fact in the last ten to 15 years I don't think I've even seen so much as a single snowflake before January. So far we've had weeks of sub-zero temperatures (when normally it'd be wet and miserable around now) and several fairly hefty snow falls, probably more than we get in two or three years, over the space of the past ten days. Plenty of snow around here for snowboarding...........

I've done as much as I can on the final version of the controller now. The "brains" board is working fine, I've tested it by driving a Hall-equipped motor with an RC ESC and looking at the six outputs:


The four Molex headers on this board are for the controller power (effectively an on/off switch), throttle, Hall sensor and forward/reverse switch. The board picks up the current shunt signal from the "brawn" board and outputs six FET drive signals and the power feed to the drivers.

The "brawn" board will have to wait until the components arrive, as I don't really want to rob the Mk2 yet again - some of the components on it are looking a bit tired from having been pulled and re-used, plus I want a controller I can abuse.

Jeremy

(edited to correct a typo)

 

[texaspyro]

What, no blinkenlights? I thought that all electronics has to have blinkenlights these days...

 

[Jeremy Harris]

What, no blinkenlights? I thought that all electronics has to have blinkenlights these days...

I have to admit to often wondering why on earth some designers feel the need to litter boards with fairly pointless LEDs.

I'm trying to keep this thing to the thread title, "Simple". For those in the UK, this translates into "Simples", when said by a meerkat called Aleksandr Orlov, with a Russian accent.............

Jeremy

 

[johnrobholmes]

It isn't electronic without lots of blinkens!


Awesome work, I look forward to how you progress with the design.

 

[liveforphysics]

Ohhh! I'm loving that brain board! Any chance you would revise it for some little wide-input-range DC/DC converters? I think a pair of them would be about perfect, one just for taking pack voltage and running the controller brain, and a second one that inputs pack voltage (or runs from the output of the other DC/DC), and has 3 isolated ~12v outputs on it for high-side fet power, and an opto-coupler for each gate signal output, so the control for the power stage can be very isolated, robust, and extremely simple to connect (because all the tricky problems are solved on the brain-board).

Sooooo.... that's where all our snow is then?! All we've been getting is rain here in eastern canada - every time it gets bellow zero centigrade the precipitation stops, and it waits until it's warm again to start raining again... yuk! :x I want to go snowboarding!!!

I think the simple answer is over here. I can't remember having much snow before Christmas here, in fact in the last ten to 15 years I don't think I've even seen so much as a single snowflake before January. So far we've had weeks of sub-zero temperatures (when normally it'd be wet and miserable around now) and several fairly hefty snow falls, probably more than we get in two or three years, over the space of the past ten days. Plenty of snow around here for snowboarding...........

I've done as much as I can on the final version of the controller now. The "brains" board is working fine, I've tested it by driving a Hall-equipped motor with an RC ESC and looking at the six outputs:


The four Molex headers on this board are for the controller power (effectively an on/off switch), throttle, Hall sensor and forward/reverse switch. The board picks up the current shunt signal from the "brawn" board and outputs four FET drive signals and the power feed to the drivers.

The "brawn" board will have to wait until the components arrive, as I don't really want to rob the Mk2 yet again - some of the components on it are looking a bit tired from having been pulled and re-used, plus I want a controller I can abuse.

Jeremy

 

[Jeremy Harris]

Ohhh! I'm loving that brain board! Any chance you would revise it for some little wide-input-range DC/DC converters? I think a pair of them would be about perfect, one just for taking pack voltage and running the controller brain, and a second one that inputs pack voltage (or runs from the output of the other DC/DC), and has 3 isolated ~12v outputs on it for high-side fet power, and an opto-coupler for each gate signal output, so the control for the power stage can be very isolated, robust, and extremely simple to connect (because all the tricky problems are solved on the brain-board).

Thanks, LFP.

I could easily add a DC DC converter for the supply voltage, if I can find one that copes with up to 100V input. Getting one that runs at up to around 75V is easy enough, though. For my application, I already have a 20A DC DC converter that chucks out 12V with the bike "ignition" on, so I'm going to use that to feed the controller. It only needs around 50 to 60mA, even when switching all the FET drivers at maximum speed. With optos isolating the FET drives, the high side DC DC converters can stay where they are, on the "brawn" board. My only concern with opto-isolating the gate drives is speed. It probably doesn't matter much, because the gate drivers will be speeding the edges up again, though.

The idea of having a nice robust connection is a good one, but there is still a need to feed the current shunt signal back, plus, in practice, the drive signals are pretty robust, as they have a 12V swing with pretty good noise thresholds.

Jeremy

 

[grindz145]

What, no blinkenlights? I thought that all electronics has to have blinkenlights these days...

I have to admit to often wondering why on earth some designers feel the need to litter boards with fairly pointless LEDs.

I'm trying to keep this thing to the thread title, "Simple". For those in the UK, this translates into "Simples", when said by a meerkat called Aleksandr Orlov, with a Russian accent.............

Jeremy

LEDs are helpful for debugging, but when the controller is this simple... Why bother? I really like the idea of keeping the "Brain" separate from the "Brawn". Ricky is doing this with his "bells and whistles" controller as well.

 

[liveforphysics]

I could easily add a DC DC converter for the supply voltage, if I can find one that copes with up to 100V input.

Little cell phone chargers (from modern USB based charge phones) tend to output a stable 5v (at 500mA-1A), and handle an input range of ~80vdc to 250vdc. Methy found some vendors on ebay selling them for $1 shipped from China... And about 3/4th of the types he tried we NOT isolated, so they shock your face with 110vac if you answer the phone while it's charging. lol He let me find that out for myself when I asked if I could charge my phone.

BUT... a couple types were fully isolated, and $1 each shipped from China. Otherwise you can get known-good-isolated units for like $4-5 each.


You would need Methy to confirm (he has a box of every type offered on fleabay), but I think this model was indeed fully isolated:

http://cgi.ebay.com/1000mA-5v-USB-Universal-Power-Adapter-Mains-Charger-EU-/280572049694?pt=Other_MP3_Player_Accessories&hash=item415365b91e#ht_4108wt_905


This one also was isolated i think... but I can't remember exactly, they all look similar:

http://cgi.ebay.com/USB-AC-Power-Supply-Wall-Adapter-MP3-Charger-US-2P-Plug-/300422883221?pt=Other_MP3_Player_Accessories&hash=item45f2996795#ht_3152wt_1139


Unfortunately, while a genuine apple usb charger is totally isolated, all the units on ebay in the apply charger-cube format were unisolated face-shockers.

 

[Jeremy Harris]

Little cell phone chargers (from modern USB based charge phones) tend to output a stable 5v (at 500mA-1A), and handle an input range of ~80vdc to 250vdc. Methy found some vendors on ebay selling them for $1 shipped from China... And about 3/4th of the types he tried we NOT isolated, so they shock your face with 110vac if you answer the phone while it's charging. lol He let me find that out for myself when I asked if I could charge my phone.

BUT... a couple types were fully isolated, and $1 each shipped from China. Otherwise you can get known-good-isolated units for like $4-5 each.


You would need Methy to confirm (he has a box of every type offered on fleabay), but I think this model was indeed fully isolated:

http://cgi.ebay.com/1000mA-5v-USB-Universal-Power-Adapter-Mains-Charger-EU-/280572049694?pt=Other_MP3_Player_Accessories&hash=item415365b91e#ht_4108wt_905


This one also was isolated i think... but I can't remember exactly, they all look similar:

http://cgi.ebay.com/USB-AC-Power-Supply-Wall-Adapter-MP3-Charger-US-2P-Plug-/300422883221?pt=Other_MP3_Player_Accessories&hash=item45f2996795#ht_3152wt_1139


Unfortunately, while a genuine apple usb charger is totally isolated, all the units on ebay in the apply charger-cube format were unisolated face-shockers.

Thanks for the heads up. It doesn't need to be isolated for this app, but it's scary that they're selling non-isolated ones at all. Bad enough when the household supply is 110, seriously scary here where it's 240............

Jeremy

 

[oldpiper]

What, no blinkenlights? I thought that all electronics has to have blinkenlights these days...

I have to admit to often wondering why on earth some designers feel the need to litter boards with fairly pointless LEDs.

I'm trying to keep this thing to the thread title, "Simple". For those in the UK, this translates into "Simples", when said by a meerkat called Aleksandr Orlov, with a Russian accent.............

Jeremy

With you on that one, Jeremy! Save them for the iPhone app.

Cameron

 

[liveforphysics]

Little cell phone chargers (from modern USB based charge phones) tend to output a stable 5v (at 500mA-1A), and handle an input range of ~80vdc to 250vdc. Methy found some vendors on ebay selling them for $1 shipped from China... And about 3/4th of the types he tried we NOT isolated, so they shock your face with 110vac if you answer the phone while it's charging. lol He let me find that out for myself when I asked if I could charge my phone.

BUT... a couple types were fully isolated, and $1 each shipped from China. Otherwise you can get known-good-isolated units for like $4-5 each.


You would need Methy to confirm (he has a box of every type offered on fleabay), but I think this model was indeed fully isolated:

http://cgi.ebay.com/1000mA-5v-USB-Universal-Power-Adapter-Mains-Charger-EU-/280572049694?pt=Other_MP3_Player_Accessories&hash=item415365b91e#ht_4108wt_905


This one also was isolated i think... but I can't remember exactly, they all look similar:

http://cgi.ebay.com/USB-AC-Power-Supply-Wall-Adapter-MP3-Charger-US-2P-Plug-/300422883221?pt=Other_MP3_Player_Accessories&hash=item45f2996795#ht_3152wt_1139


Unfortunately, while a genuine apple usb charger is totally isolated, all the units on ebay in the apply charger-cube format were unisolated face-shockers.

Thanks for the heads up. It doesn't need to be isolated for this app, but it's scary that they're selling non-isolated ones at all. Bad enough when the household supply is 110, seriously scary here where it's 240............

Jeremy

Chip power doesn't need to be isolated, but high-side power for FET switching does. (unless you're setting up this brain board to boot-strap?)

 

[liveforphysics]

And yes, I agree, 240v homes with this un-isolated charger would be quite a bit more exciting. You should give them as anomous Christmas presents to the folks you don't like much. WHAM! 240v on your face! Surprise!

 

[Burtie]

Burtie's timing adjuster was prototyped using a small single board controller which costs a bit of money. I suspect that its functionality would be real easy to get into a small micro chip. I would probably do it with a ATTINY23 which costs around $2...

Not a hope in hell it would fit in one of those

 

[Jeremy Harris]

Burtie's timing adjuster was prototyped using a small single board controller which costs a bit of money. I suspect that its functionality would be real easy to get into a small micro chip. I would probably do it with a ATTINY23 which costs around $2...

Not a hope in hell it would fit in one of those

I suspected that might be the case, Burtie, judging by your switch to a much faster processor.

What's the snow like down your way? The local news made it look as if your neck of the woods got it a bit worse than we did.

Jeremy

 

[Burtie]

Was snowed in for a while, the only way to get around was by walking, or knobbly tyred ebike!

Main routes are ok now though.

 

[texaspyro]

Not a hope in hell it would fit in one of those

You might be surprised... I built up a device that does essentially the same thing on four channels and it fit/works just fine. Could always use a Mega328 if you need to get fancy pants features.

What is the max freq of the hall signals that you are dealing with?

 

[Jeremy Harris]

What is the max freq of the hall signals that you are dealing with?

Around 1400Hz for 12,000 rpm on a 7 pole pair motor.

It looks like the processor has around 70us to execute however many lines of code are needed to do the stuff Burtie's doing, so I'm guessing it needs a fairly fast processor to bring the instruction time down enough to get it all done in time, especially if it's doing maths in that interval. I suppose a look-up table would be faster, but then you may lose some versatility, although to be honest I'm just guessing here, as it's Burtie's baby.

Jeremy