Tired of ESC problems

[churd]

Tired of the cost. Tired of the lack of info. Tired of the lack of options. Tired of the lack of support. Tired of not knowing what's going on when it doesn't work.

Just started a scratch home brew ESC. Target for the moment is 100V & 200A.

I want to test big motors without fighting the ESC.

The parts are in house and construction has started.

Will keep you all posted on progress and smoke levels.

Mr. C

[Mark Wood]

Ummm...I assume you don't use Castle Creations ESCs.
That's a statement, btw.

mw

[Unterhausen]

cool project. there have been posts in DIY electronics and in the motor construction forum too. Good luck with it, I've thought about this myself.

[CA's my friend]

Mark,
I have known a few people that fly CC esc's and I can say they are easy to program, and have amazing customer service. The problem is cost :eek: , their cost is completely justifiable given the amazing quality but that is still a lot more than some modelers(myself included) can afford. A home-brew solution would be a miracle if the cost can be kept fairly low.

[CV580drvr]

Sounds like an interesting project CA, and I'll be watching your progress.

Castle? You betcha! Have you checked the T-Bird line? If they're suitable for your needs, they're a good deal and if you know where to look (hint) they can be had NIP all day long for a fraction of the cost.

Good luck with the project nonetheless....

[CA's my friend]

CV580drvr, this is not my project, this is churds. I am to stupid to build my own esc :D unless I am looking at a schematic. I have a very limited knowledge of electronics and I am limited to simple Boolean logic chips as my most complex sensors. If this was a laser trip wire alarm or a phone tap :D . . . . er. . . never mind. I have looked at the thunder-birds long and hard, and well I am still looking.

Oh and churd, Keep us updated

[churd]

100V 200A ESC

FET's 6 or 12 digikey $2.40 in small q
http://www.st.com/stonline/books/pdf/docs/9522.pdf

FET Driver 6 digikey $1.35 in small q
http://www.fairchildsemi.com/ds/FA%2FFAN7371.pdf

Micro Controller 1 digikey $10.50 for 1
http://www.silabs.com/public/documents/tpub_doc/uguide/Microcontrollers/Small_Form_Factor/en/ToolStick_F362_DC_UG.pdf

[mjsas]

One thing to keep in mind. While most DIY builders think having a safe voltage level is for pussies, there are other people involved.
For example if you take your creation to the local flying field everyone in the club takes part in the liability. If you take your creation to an event the sponsor has liability. And so on. If I was holding an event and some one can along with a DIY 100 volt system I would send them away as fast as possible.
Go ahead and flame me if you want to, but I want nothing to do with a 100 volt model airplane. You can fly it in your back 40 but stay away from the public field.

[rmteo]

At 200A, those FET's are going to be dissipating 400+ watts! Should be fun to watch. :D

[tadawson]

One thing to keep in mind. While most DIY builders think having a safe voltage level is for pussies, there are other people involved.
For example if you take your creation to the local flying field everyone in the club takes part in the liability. If you take your creation to an event the sponsor has liability. And so on. If I was holding an event and some one can along with a DIY 100 volt system I would send them away as fast as possible.
Go ahead and flame me if you want to, but I want nothing to do with a 100 volt model airplane. You can fly it in your back 40 but stay away from the public field.

Who says 100V is unsafe? I sure as heck wouldn't . . . . pretty much everyone on the planet deals with 120 up on a daily basis in their home, and people aren't falling over dead all the time, so why is this any different? And what collateral damage can possibly happen? Worst case is the owner gets a bit of a shock . . . that's it. Club liability? For what, pray tell?

If someone in the club I am in tried to run someone off for running 100V homebrew, I would try to get that person institutionalized for being pathologically without a clue . . . . IMHO . . . .

- Tim

[AndyKunz]

There is a major difference between DC and AC. With AC, the pulses actually help you free yourself from being the short circuit (during the zero crossing). With DC, your nerves LOCK DOWN and you can't shake yourself free no matter how hard you want to.

Sorry, but 100V is not at all safe. UL and all the other regulators use 60VDC as a limit for a reason.

The voltage isn't the real problem - the current is, if it happens to pass through your heart. A hot, sweaty hand is a pretty good conductor. Ever try a 9V across your tongue? You only need a few mA (20 or less) to kill yourself.

I got hit by 600VDC one time and I'm very blessed to be here. I happened to have rubber-soled shoes on (but sweaty feet) and touched a 600V line from one of our pieces of equipment with my right hand. I was stretching to reach the area to install another piece of equipment. When I accidentally completed the circuit, my body lurched and broke the connection. My arm, shoulders, neck, spine, butt, and legs were in pain for a day.

If it had been my left arm, I might not be typing right now. The current may have been through my heart instead of how it traveled.

(Yes, we fixed the grounding on the cabinet too).

Andy

[tadawson]

There is a major difference between DC and AC. With AC, the pulses actually help you free yourself from being the short circuit (during the zero crossing). With DC, your nerves LOCK DOWN and you can't shake yourself free no matter how hard you want to.

Sorry, but 100V is not at all safe. UL and all the other regulators use 60VDC as a limit for a reason.

The voltage isn't the real problem - the current is, if it happens to pass through your heart. A hot, sweaty hand is a pretty good conductor. Ever try a 9V across your tongue? You only need a few mA (20 or less) to kill yourself.

I got hit by 600VDC one time and I'm very blessed to be here. I happened to have rubber-soled shoes on (but sweaty feet) and touched a 600V line from one of our pieces of equipment with my right hand. I was stretching to reach the area to install another piece of equipment. When I accidentally completed the circuit, my body lurched and broke the connection. My arm, shoulders, neck, spine, butt, and legs were in pain for a day.

If it had been my left arm, I might not be typing right now. The current may have been through my heart instead of how it traveled.

(Yes, we fixed the grounding on the cabinet too).

Andy

Well, I still respectfully disagree. When I work on 20 and 30KV, *then* I feel at risk. Model batteries at 100V still don't concern me, and that is an "educated opinion" as an electrical engineer, not just "joe punter" . . .

Besides, how many folks use a single pack to get the higher voltages, versus combining a number of smaller packs in series? Seems like the risk is more in the pack size than the overall voltage . . .

And I still see *zero* potential for risk to others than the operator, so still *zero* reason to restrict at a flying site . . . .

- Tim

[lutach]

See if you can get a car version of the controller done as well :).

[BushmanLA]

200A will require some pretty fat copper traces. :)

Read the thread where they were hammering on Jim Drew for confusing wattage of the whole system vs wattage in a copper trace.


http://www.rcgroups.com/forums/showthread.php?t=868199

[churd]

As far as 100V being a worry: The airplane is an isolated system. It does not have a common ground so you can soak your feet in a salt water puddle and never be in danger. You also have to find a negative terminal and a positive terminal to touch at the same time to feel a shock. I don't see very many opportunities to do that in any installation I've seen.

[churd]

For example if you take your creation to the local flying field everyone in the club takes part in the liability. If you take your creation to an event the sponsor has liability.

I wasn't planning on taking it to a flying field or an RC event. I might be 500' above your field or event. So make sure you don't fly over 500' and we'll all be very safe.

[vintage1]

At 200A, those FET's are going to be dissipating 400+ watts! Should be fun to watch. :D

How so? they are rated at 10 milliohms at 110A.

so using 12, thats 5 milliohms per phase. or assuming two phases are alwys on. 10 milliohms per phase. I make that 20W.

hot, but not super hot for 8 units.. a watt or two each. More obviously when switching hard. But not THAT much more.

[churd]

Right on Vintage1. And most likely run even cooler at full throttle than it would at 90%. At 90% you have the most significant source of heat. The 12khz switching frequency.

[rmteo]

No matter how you cut it, 200A across 10 milliohms is going to be a total of 400 watts (200x200*0.01).

[churd]

No matter how you cut it, 200A across 10 milliohms is going to be a total of 400 watts (200x200*0.01).

I was wrong with the "right on". But I don't think you're quite right either.

Let me try to do better: 2 FET's = .005 Ohm x 200A = 1V. 1V x 200A = 200 W.

200W /12 FET's = 16.66666666 watts per FET but still 200W total.

Do I get an 'A'?

[rmteo]

Getting closer, but a total of 200W is still a lot of heat to get rid off - not anywhere near the original 20W mentioned. And this is just a best case situation - not including switching losses, which as you have surmised, could be significant. Think about a pencil type soldering iron - they are typically 25-30 watts.

Also, you will need to consider what BushmanLA mentioned above. To carry 200A (and allowing for a 40 degree temperature rise above ambient) you PCB trace will need to be about 2.7 inches wide - if you are using 4oz. copper. With the more usual 1oz. copper, the traces will be almost 11in. wide. You will have to come up with some other solution.

[churd]

Are you talking about heat sinking using the pc board or are you talking about current carrying?

We will be using heat sinks. The first purpose of this project is to bench test a HXT 80-100-B 130Kv Brushless Outrunner. So far I've been only able to run it at no more than 3.5HP because of problems with the TURNIGY Sentilon100A HV 2-12S Speed Controller. Anyhow I'm shooting for a static thrust of 45lbs or better with that motor. 3.5HP input did produce a thrust of 32lbs.

[rmteo]

I was referring to current carrying.

[tadawson]

As far as 100V being a worry: The airplane is an isolated system. It does not have a common ground so you can soak your feet in a salt water puddle and never be in danger. You also have to find a negative terminal and a positive terminal to touch at the same time to feel a shock. I don't see very many opportunities to do that in any installation I've seen.

And at any rate, at least not any that are much more than about 1/2" apart . . .

- Tim

[BushmanLA]

Are you talking about heat sinking using the pc board or are you talking about current carrying?

We will be using heat sinks. The first purpose of this project is to bench test a HXT 80-100-B 130Kv Brushless Outrunner. So far I've been only able to run it at no more than 3.5HP because of problems with the TURNIGY Sentilon100A HV 2-12S Speed Controller. Anyhow I'm shooting for a static thrust of 45lbs or better with that motor. 3.5HP input did produce a thrust of 32lbs.


Churd, read that thread that I posted a link to earlier.

Remember, if you have a device, and the total current draw of that device is X Amps, then the wires leading into that device from the current source (the battery) had better be able to handle X Amps.

This is a nice table of what guage wire can handle what amperage.
http://www.powerstream.com/Wire_Size.htm

Just because a device can handle X amount of amps at Y Volts, doesn't mean it can handle handle 2X amps at 1/2Y Volts. The wattage is the same in both cases, but the wattage is spread around very differently to individual components inside the device from one case to the other.

[vintage1]

No matter how you cut it, 200A across 10 milliohms is going to be a total of 400 watts (200x200*0.01).

Shows how tired I was yesterday. You are of course right mathematically.

Mind you 10 milliohms is the absolute guaranteed worst case. Typically its about 1/2 of that at realistic currents.

Well, he will have to use more FETS. Every time you double up, you end up with half the resistance, so half the power,spread across twice the number of devices, so the power per chip is a quarter..and often RDS is lower at lower currents..these things aren't really resistors.

If you take 5W as realistic with minimal heatsinking, per device, that's probably around 30A per device fully on..since each phase s on 2/3rds of the time that nets you 50A per 6 pack!. At an average of about 5 milliohms per device.

So about 4 devices per half phase. 24 in all. Probably total heat around 100W and a bit.

It seems to me that its probably better to go for a somewhat lower RDS (on) device..you can do the sums..

For a one off, you are probably looking at reinforcing the PCB traces with something like some 14swg copper wire, to carry the current. I've seen production PCBS equipped with CNC milled copper bus bars..

The other thing that will need to be pretty hefty is the drivers. You have to cope with about 20nF of gate capacitance for 4 FETS and Lord knows how much drain/gate capacitance amplified via the Miller effect..not trivial.

[Dan Baldwin]

Is this a brushed ESC, or brushless?

Dan

[rmteo]

Dan, see post #22. It is for a HXT 80-100-B 130Kv Brushless Outrunner.

[Dan Baldwin]

If it's a brushless ESC using 12 fets for commutation, there would be 6 sets of 2 fets in parallel. Each of these pairs would have a resistance of .005 ohms (worst case), but 2 sets of fets would be on at a time with current flowing through both (essentially in series) for a total resistance of .01 ohms, so we're back to 400 watt total dissipation at 200 amps.

Dan

[churd]

Got any hot dogs?

[lutach]

Not a 100V, but does the Jeti Spin200 and 300 really put out those kind of AMPs (Given a bettery, connectors and wires can handle that)? After reading what you guys are saying, it makes me think they don't. I have 3 Spin200 that I'm thinking of using on some real high AMP applications. What about the SHV controllers from Castle? They have a 90V 200A monster?

[churd]

This is some of the first turns on a hard supply.

I'm always impressed with the 80-100 Turnigy motor with a 32x10 prop. This is only running at 1/8 the motor capacity as advertised.

Video:

http://www.youtube.com/watch?v=W5Lz-AVEYMs

Stay Tuned!

[Kipp]

Sorry for the repost -- I figured this is a more appropriate location for the video update:

http://www.youtube.com/watch?v=QoZcEAxKw4c

This was about 4kW (5.2HP). Switching rate here is 100KHz.

[churd]

Thread has moved to "300A 100V DIY ESC design goal"

[Ron van Sommeren]

Thread has moved to "300A 100V DIY ESC design goal"At
http://www.rcgroups.com/forums/showthread.php?t=903307