Simple BLDC controller

[Jeremy Harris]

Thanks, guys.

I'm aiming to try and make stuff that can be built by others without needing mega facilities. I really enjoy the challenge of making something that's hopefully useful to others and that's also as simple and practical as I can make it.

The one snag with this project is that the through-hole versions of the controller chip I'm using are being phased out, in favour of surface mount. I find surface mount stuff harder to work with, but it may be that others are happier with it.

This version of the controller will be tested on a standard Colossus 7kW motor, running at around 56V, 120A (so around 6.7kW max at about 4000 rpm) to see how well the design copes with a relatively low inductance, low winding resistance motor. The final version (assuming that this one works as planned) will use the big Ixys Hiperfet modules to drive a higher Kv version of the same motor.

If I can get the workshop up to a bearable temperature tomorrow then I'll have a go at making up the heatsink bar and copper fittings for the phase and power terminals.

Jeremy

[Kingfish]

Exciting Jeremy! Wish I wuz there to help stoke coal to the furnace
Cheers, KF

EDIT: My bad – I meant to say “Adjust your solar collectors towards your den”

[Alan B]

Jeremy, have you tried that easy to make home etchant instead of ferric chloride? It was a mix of something like muriatic acid and hydrogen peroxide or some such. Hardware store stuff. Here is some info:

http://hackaday.com/2008/07/28/how-to-e ... sided-pcb/

[amberwolf]

Just in case it's useful to know a way *not* to do it, here's the PCB I started trying to work out for the MC33035 chip, which would've used six TO264 FETs NTY100N10. At the time I did not realize I'd need separate gate drivers, and that I could probably not successfully drive them directly from the MC33035. These are from June of 2009. I've sure learned a lot since then.

PCBcomplete1.0.1o.PNG (42.52 KiB) Viewed 501 times

I agree about the manual track layout...but it is very frustrating when the software tries to remap your carefully placed traces to somewhere else, for reasons you can't understand, and when you haven't initiated any actions that should cause it to even try.

[Jeremy Harris]

Jeremy, have you tried that easy to make home etchant instead of ferric chloride? It was a mix of something like muriatic acid and hydrogen peroxide or some such. Hardware store stuff. Here is some info:

http://hackaday.com/2008/07/28/how-to-e ... sided-pcb/

I haven't tried it, mainly because over here getting hold of stuff like hydrochloric acid and hydrogen peroxide is difficult - I'm always envious of the stuff you guys in the US can buy from hardware stores, the safety mafia seem to have virtually stopped the open sale of stuff like that here.

On the other hand, I can buy dry ferric chloride pellets through the post easily enough here and just make up the solution by adding water. It keeps for a long time in a sealed jar, too, the solution I'm using at the moment I made up around three years ago. I just pour it back into a sealed glass jar when I'm done and save it for the next time I need it.

The downside with ferric chloride is that it stains everything it touches brown, even the sink when rinsing boards if you're not careful! It also needs careful disposal, as it can't just be flushed down the drain when I've finished with it, at least not legally.

Jeremy

[Jeremy Harris]

Just in case it's useful to know a way *not* to do it, here's the PCB I started trying to work out for the MC33035 chip, which would've used six TO264 FETs NTY100N10. At the time I did not realize I'd need separate gate drivers, and that I could probably not successfully drive them directly from the MC33035. These are from June of 2009. I've sure learned a lot since then.

I agree about the manual track layout...but it is very frustrating when the software tries to remap your carefully placed traces to somewhere else, for reasons you can't understand, and when you haven't initiated any actions that should cause it to even try.

Nice to know that someone else has looked at using a simple BLDC controller chip. I'm still convinced that the features these have would probably meet the majority of user needs.

UPS have just sent me an "Adverse Weather Event" delay notice on the parts I was hoping would arrive today, so I guess I'll have to get on with more work on the house instead of the controller...................

Jeremy

[amberwolf]

One of the blown controllers I was sent (by Ianmcnally, I think) was an analog controller powered by this series of chip (forget if it was 33 or 35); the FETs were blown (VERY blown) but I think the rest is intact. I did borrow the shunts though, to fix another controller that only had melted shunts and a couple of blown caps. If I remember to do it when I finally have some free time, I will clean up the PCB and take some pics in case their layout helps you any.

Also, if you have any interest in it, I can post the rest of the PCB/schematic/etc from my attempt. It was for that ceiling fan motor conversion, which at the time I had no controller to run. I'd still like to actually finish the design and build it, using those big FETs, someday, but I'm not sure I will ever get around to it (there are just too many things I want to do).

I do not know if I have any more of the MC33033 or 35 in thru-hole, but if I do and you need some, I could send them to you. I am pretty sure I have at least a handful of the 35 in standard SMT, and maybe some of the 39 in thru-hole.

[Jeremy Harris]

Thanks for the offer of the chips. I managed to buy a bunch (via ebay of all places) so have enough to keep me going for a while. If this works, then I might look at getting some boards made that will take the surface mount versions, as basic 'building block' modules that could be used in other projects.

I still like the idea of a modular controller, where the drivers and power section is separate from the main 'brain'. Although this board is an all-in-one layout, that's really just for ease of getting a prototype up and running.

I've just been out an put the workshop heating on, in the hope that I can get some of the hardware made today. It's -4 deg C out there at the moment though, so I guess it'll take a while to get warm enough to do some work.

Jeremy

[Alan B]

Ferric chloride is messy but I think Radio Shack may still sell it here. Seems like there was a good way to dispose of it safely, but I don't recall now what it was.

[liveforphysics]

Jeremy- If you need boards my friend, you know my cnc is at your disposal anytime. It cuts the traces and drills the holes as neat as can be. My cutting area could do at least 2 dozen boards in a batch.

[texaspyro]

Jeremy, have you tried that easy to make home etchant instead of ferric chloride? It was a mix of something like muriatic acid and hydrogen peroxide or some such.

Be VERY careful with that mix, and keep quiet about it. It is one ingredient away from your favorite international terrorist's foot rub and crotch liner. You could have some serious splainin' to do to the federales, provided that they can scrape enough of you off the ceiling and walls.

[Alan B]

Those PCB etchant mixes are well documented on the net. I don't know how popular they are, searches turn up lots of sites and discussion.

But I send my boards out anymore. Have been using expresspcb but planning to try out batchpcb. No fuss, no muss, fantastic boards in a week or three. No chemicals and they have plated through holes, etc.

It all depends on what your time is worth...

[heathyoung]

I gave up on toner transfer after some bad experiences with pinholing and poor resolution - I've been getting the kalex pre-coated positive photoresist boards, much better results - also - a dremel drill press is fantastic for those carbide bits. And I highly recommend MG chemicals Liquid tin - beats using PCB laquer...

Oh yeah - and the fastest way to get on a government watch list is to order concentrated nitric acid in quantity. Every bloody airport (including US airports) I get stopped for 'random' explosive trace tests. Problem was once they actually found traces of nitrocellulose on the outside of the luggage (from a box of fireworks stored in the same cupboard). Try explaining your way out of that one.

[Alan B]

I used to use the tin plating liquid. Until it turned bad and plated my boards with something unsolderable one time. It doesn't keep well.

So I got a small bottle of kool amp silver plating powder. It costs more these days but is still available and a small bottle is likely a lifetime supply. The powder is rubbed on the board with a damp rag after etching and removing resist. It plates silver onto the copper. Really nice. Do this before drilling. Boards look great and solder nicely. Takes very little powder to do a board. Does not plate the fiberglass.

I also use the dremel drill press with carbide bits. Works like a charm.

[liveforphysics]

CNC the boards.

No mask.
No copper clean-up.
Holes drilled, and accurately.

Cost? $0.00 if you're doing something to help the electric revolution (like this project.)

[Tiberius]

I used to do all this etching and drilling. But over 20 years ago I decided it was simpler to send the artwork off and get the boards made. I'm sure that if I took my time into account then it was cheaper.

For me, the real killer was that I was always working with double sided boards and needed plated through holes.
Another advantage was that I would always get a few more boards than needed as part of the deal. So there was always a supply of breadboards that could be modified for the Mk2 or for another project.

Since then, I have found myself working with smaller and smaller components, and the cost of getting boards made has come down, so for the sort of stuff I do, the balance has shifted even more in favour of not doing it yourself.

Jeremy, what resolution can you work down to?
Luke, what's the finest groove you can machine?

Nick

[Jeremy Harris]

CNC the boards.

No mask.
No copper clean-up.
Holes drilled, and accurately.

Cost? $0.00 if you're doing something to help the electric revolution (like this project.)

This is the way I want to go. I have all the parts for two CNC machines, a modestly sized flat bed machine aimed at cutting out alloy/wood/plastic sheet materials and a proper CNC milling machine. The X/Y table for latter is partly built, sitting waiting for me to get time in the workshop and finish it (it's still sub-zero in my workshop at the moment). I have the controller box partially built, with stepper drivers etc, too. I still need to go and get a cheap PC with a parallel port (none of the ones I have lying around have parallel ports) or try and use a Smoothstepper over USB. I'm planning on using Mach3, which means sitting down and learning how to use yet another software package (I get supremely impatient when faced with having to spend hours learning to drive something, just to get a job done...........).

I have a proper high speed PCB bench drill that I picked up surplus for next to nothing, as no one wanted it in the local second hand tool store. It looks like a proper bench drill, in cast iron, but runs at 20,000 rpm and has a nice precision miniature chuck and feed, plus a light that shines down on the bit. It makes life easier when drilling with carbide bits, as long as you hold the board very still whilst drilling and don't let it move sideways even a fraction. Opening out previously drilled holes to a bigger diameter is another sure-fire way to break carbide drills I've found. I keep a few HSS drills for this purpose, as although they wear quickly they are ideal for this job.

The boards are nearly finished now, all I need to do before testing is make up the heatsink spreader bar, mount the FETs and make up and fit the terminal posts for the power and phase connections.

Nick,

The laser toner transfer process seems to work OK down to about 15 thou traces if you're careful. I tend not to go below 35 thou, as it makes everything less critical.

Jeremy

[liveforphysics]

If I make traces finer than about 0.3mm, then you gotta check them on a meter to make sure it didnt have a little wiggle in the tool path.

Traces at 0.5mm and thicker seem to be reliable.

I used to do all this etching and drilling. But over 20 years ago I decided it was simpler to send the artwork off and get the boards made. I'm sure that if I took my time into account then it was cheaper.

For me, the real killer was that I was always working with double sided boards and needed plated through holes.
Another advantage was that I would always get a few more boards than needed as part of the deal. So there was always a supply of breadboards that could be modified for the Mk2 or for another project.

Since then, I have found myself working with smaller and smaller components, and the cost of getting boards made has come down, so for the sort of stuff I do, the balance has shifted even more in favour of not doing it yourself.

Jeremy, what resolution can you work down to?
Luke, what's the finest groove you can machine?

Nick

[Tiberius]

Lol, Jeremy answers in thou and Luke in mm!

In the past I've had some customers who wanted rules like minimum trace 12 thou and minimum gap 12 thou (0.3 mm). That's what you really need to work to for old school through hole stuff if you want to pack it on a 0.100" grid.

But some of the surface mount stuff I use nowadays has a pin pitch of 0.5 mm.

Obviously, its horses for courses. High power and high voltages require different rules. The single thing that made me go completely over to professional manufacture was plated through holes.

Nick

[nieles]

are the fet drivers good enough for 6 of these fets?

IXFN340n06 SOT227
http://ixdev.ixys.com/DataSheet/98751.pdf

[Jeremy Harris]

are the fet drivers good enough for 6 of these fets?

IXFN340n06 SOT227
http://ixdev.ixys.com/DataSheet/98751.pdf

Should be, but those FETs don't have a very good spec when compared to the IRFP4368 TO247 package ones I'm using.

The Ixys FET has a max current of 100A (package limited), a max voltage of 60V and an Rdson of 3mohm maximum.

The IRFP4368 has a max current of 195A (package limited), a max voltage of 75V and an Rdson of 1.85mohms maximum.

The only characteristic that's better for the Ixys FET is the junction to heatsink thermal resistance, at 0.23 deg per watt versus 0.53 deg per watt for the IRFP4368. This thermal advantage is partially lost from the higher Rdson though.

The parts I was waiting for have now arrived, so with luck I may get the board up and running later today.

Jeremy

[nieles]

yes i know they are not the best fets, but i like the fact i can mount them very easy. (isolated base plate)

the 100A limit is probalbly because of the screw connection at the terminals when the fets are in 24 7 use.

[Ricky_nz]

The parts I was waiting for have now arrived, so with luck I may get the board up and running later today.

Jeremy

Good timing.
I've got enough parts to start bringing up my power board this weekend too with the same MOSFETs although I would really like to have the capacitors that are still in transit I think I can make do.

I look forward to seeing how well your simple to build low cost controller works .

I use to do the toner transfer thing myself but with bad eyesight the hole drilling was hard for me. I just used cheap HSS 0.8mm and 0.5mm drills . The HSS drills were cheap enough I brought packs of 10 and you could get about 200-300 holes before the fibreglass killed them.
Still I had better luck with HSS rather than with reduced shank carbide bits. The HSS just flex a bit rather than breaking.
The carbide ones cut really nice but just too fragile, especially with a less than perfect drill press.

Ricky

[Jeremy Harris]

Thanks Ricky. I think yours will be more versatile, but I want to try and get some hands-on experience with the power stuff before trying something more adventurous with the control end.

I managed to get the controller finished this evening and have powered it up to check that all looks well. Quiescent current is around 20mA and all the voltages look OK. I'm in desperate need of a nice glass of wine right now, so will put off the final testing with a motor until tomorrow.

Meanwhile, so you don't all think I'm slacking, here are a couple of photos of the finished unit.

Controller ready for test - 1.JPG (90.15 KiB) Viewed 532 times

Controller ready for test - 2.JPG (83.83 KiB) Viewed 532 times

There's only one shunt fitted at the moment, to limit current to 20A for testing. I'll add more shunts as I get a feel for how the thing works.

Jeremy

[liveforphysics]

Loving it!
Great work Jeremy!