DIY 12F675 Brushed ESC ? - RC Groups
Thread Tools
Aug 03, 2004, 07:45 PM
Martin - AKA mr.sneezy
PLMS's Avatar

DIY 12F675 Brushed ESC ?

Hi all
Has anyone built the fairly recent DIY ESC described here besides me ?

Sign up now
to remove ads between posts
Aug 04, 2004, 01:09 AM
Registered User
MAVA's Avatar

Good link !!


Thank you for the link, the design looks simple.

I have not built one, but I like the 64 step throttle. Will build one soon and lay-out several boards to solder onto the back of the motors.

Step 7 Very good info for the first time builder.

Simple for a first time DIY ESC project.


Aug 04, 2004, 07:18 AM
Registered User
Hi Martin,

Ironsides has just pointed this thread out to me, and suggested I stand up...

Well, I'm the designer of the ESC and I've build quite a few...

I hope the ESC is working OK for you. If there are any questions I am happy to answer them.

I actually have an improved version of the software (not on the web site yet) that also offers:

- glitch counter
- Lost Model Alarm

as part of the design.

One of the reasons I have not updated the web site is that I have been busy with other things. Plus the number of options required too much documentation (ie. I got lazy).

As I said - any questions just fire away.

Aug 04, 2004, 04:08 PM
Registered User
MAVA's Avatar

Thank you for your comments,

It is good you are including a Lost Model Alarm and a Glitch counter.

I realy have no questions, but I do have a dream.. Maybe you can add a programable BEC voltage detector. Where you can use a rotary hex switch(s) or even a simple push button switch. Make the setting on a EEprom. That is it. Again it is a dream.....

I will build one of your controllers soon.

Thank you

Martin Vargas
Aug 04, 2004, 05:12 PM
Registered User

The real problem is that there are not many pins left on the PIC for really complex things. What this means is that certain options can't be used in combination - or in some cases you need the external transistor inverter on the receiver signal input (this frees up some I/O pins).

Because the LVC is programable (when the PIC is programmed, just change the value in the assembler source file) I just have a few different ESCs and use the 'right one'. Of couse this is not so easy with the 'back of motor' mounting.

I don't know how many different LVC voltages you want, however, it would be easy to set the ESC up with two possible voltages (both set at the time the PIC was programmed) and use the jumper to select between them. (with the 'no-brake' design just add a pullup resistor and a jumper to ground to GP3). I presume you want 2S or 3S LVC options?

Just out of interest are you looking at building the 'no-brake' design or the 'brake' design?

Last edited by TugBoat; Aug 04, 2004 at 05:17 PM. Reason: ammended content
Aug 04, 2004, 05:16 PM
Registered User

If you were to post a layout for 'back of motor' mounting that would be great and could help others as well.

Aug 04, 2004, 09:28 PM
Martin - AKA mr.sneezy
PLMS's Avatar

Well done Tim

Hi Tim
Pity we're in different states, we have much in common I suspect. See my RC and Automotive stuff here and you'll see what I mean.
Go to the SPAD/Mugi page for RC tinkering

About the ESC.

I built your full version with brake, using the hex file with the soft start and brake.
Well done on the programming BTW, very smooth bit of code. I love the variable rate PWM on the drive AND the brake outputs. The "Armed" ring via the motor amature is just a thing of wonder. I did it about 10 times in row I was so amused. Good thing.

I've made the BEC cutout somewhat crudely adjustable by including a small SMD trim pot on my layout. My layout is all SMD except for the FET and Reg.

It's etched single sided so it's not much smaller than your vero layout but I did it as test case for spend a few dollars on professional DS and though plated boards (in with other stuff we get made).

I attached a low res image of the PCB. If anyone wants the whole protel files I'd be happy to post it zipped up.

Two things.
One, I had an early failure of the ESC after a couple of minutes running. It seems the Schottky diode in the brake FET failed (IRF9540). It took out the main switching FET also, as well as burning off a track. I have now fitted a 3A SMD Schottky in addition to replacing the FETs. Seems to be going stong now.

Second, if you have any new SW I'd be keen to try it. Email me at is you like.
If you were to include a selectable BEC cutout then this ESC would be ultimate for my use. I go between 6V and 9V cut outs (9 cell NiMh or 3S Lipo)
The I'd be happy to see the brake setting locked on in code, this would free another pin.

Great Stuff Tim, I can't imagine the hours you have spent on this.
Aug 04, 2004, 10:50 PM
Registered User

Re: the schottky diode (loss of FETs)

I suspect that the problem may relate to the use of 9 cells/3S lipo and the rating of the diode in the FET. I have never had a problem with 8 cell and S400 motors - but all my testing is with 8 cells.

I agree that an extra Schottky is a good idea - I will update the diagram and text as soon as I can to reflect this.

Re: Programmable LVC

The current software permits the brake to be permanently on (you need to set the constants in the source and assemble the code yourself). Most of my ESCs are like that by the way - I tend to use the jumper to trigger the 'glitch' counter readout (also by buzzing the motor).

I will update the software to support two LVC settings in the code that are switch selectable using the 'Brake off' jumper. It may take a week or so for me to get round to it.

(I did consider the trim pot solution and didn't particularly like it - mainly for long term reliability reasons. With the 'two LVC switch selectable option' the trim pot will trim both voltage points - which should work fine.)

I will let you know when it is done.


Glad you like it - it did take me a while to get the whole thing working like it does - but the result was worthwhile. All the ESCs I build are now 'power PWM' - makes flying much easier in my opinion.

I should also give credit to Ironsides in Toronto who helped with the testing and documentation proofing. His input has resulted in a 'proper' product of much better quality than if I had just designed it for myself. Many thanks David.

Aug 04, 2004, 11:43 PM
Martin - AKA mr.sneezy
PLMS's Avatar
Hi Tim and David

In my last post the 9 cell pack was a typo, it should have been "(3S Lipo or 8 Cell NiMh)".

I was on an 8 cell pack testing it when FETs failed. I have not actually run it yet with the 3S pack (till I make another, or do a dual cutout).

The only related question I can think of is how critical is the 3.3nF capacitor on the brake FET drive ? I think I used 2x1.5nF (SMD's stacked up).

Anyway, all is well with it now.

Aug 05, 2004, 03:34 PM
Registered User
Ironsides's Avatar

Thanks for the kudos, all I tried to do was break it ;-)

Aug 06, 2004, 11:55 PM
Registered User
MAVA's Avatar

Todays controllers do not include the additional diode to save space or cost!
My opinion is to add them to the Break and Motor Output part of the circuit.

I like your site-The pusher model interest me for some of my friends.

PS you would know a good site for DIY ODB-1 reader for FORDs.

Aug 07, 2004, 06:40 AM
Registered User

A DIY Speed controller


A very good DIY brushed speed controller. The PIC can be programmed incircuit. I buid one and it runs cool at 20A!

Aug 08, 2004, 04:48 AM
Registered User
I found the time over the weekend to update the circuit diagrams, software and web page to include the LMA, glitch counter and dual (jumper selectable) LVC options.

To use the more complicated options you need to grab the source code, configure the options and build the code.


It could be that the reduced value of C6 contributed to the problem with the FETs. C6 ensures that the brake FET doesn't turn on due to noise and stray capacitance - reducing the value of C6 increases the risk that the brake FET can turn on while the motor drive FET is on.

All my measurements and testing show that 3.3nF is OK for use with 8-cells, a S400 motor and a IRF9540 brake FET. However, the noise on the gate does get close to the gate/source breakdown voltage. Reducing C6 below 3.3nF is definitely not recommended, if in doubt increase it.

I was considering increasing the specified size to something like about 5.6nF for C6 - however I have not had a chance to look at this on the bench. However, I can't see why there would be any problem and the effect on the brake turn on time would be minimal. The problem is that too large a value for C6 will limit the ability of the soft brake to function correctly. If larger values of C6 are required then decreasing R12 & R11 may also be required.

If you look at the bottom of the web page there is a .HEX file for the ESC with the brake permanenty enabled and 6V/9V LVC selected by jumper.

Aug 11, 2004, 01:49 AM
Martin - AKA mr.sneezy
PLMS's Avatar
Hi Tim
Martin here (PLMS).
I will add another 1.0nF to the brake circuit. However it (your ESC design) has done a few long flights since I last wrote and I have not blown another fet anyway.
I have had some past ESC experience with running Cobalt motors, and the size of Schottky diodes on them. After burning out the original ESC's diode (GWS ICS-480) I fitted bigger ones and got no more trouble. I think the diode should be on the motor tags, but this is not always neat or practical.

I'll build another one now using the update with dual BEC cutout. I think I'll add a 3A schottky across the brake fet regardless (as I have plenty of SMD ones).

Since we both fly Zagi's (both clones) I'll ask a question about ESC brakes. How much reduction in propeller speed do you see ? My Gunther prop still turns, but I'm fairly sure it's slower than with brake off. My IRF9540 fet has 0.1ohm spec, does a lower On Z give better braking ?

Your ESC is the first one with braking I've used.

Thanks for doing the updates, I apprieciate your efforts. It allows me to spend a little more time doing the flying part of the hobby.

Aug 11, 2004, 08:13 AM
Registered User

I must admit my Zagi and Unicorn both have the standard speed controllers in them - I built both planes before starting playing with ESCs and didn't want to pull them all to pieces to change the ESC.

Just out of interest I will try replacing the ESC in the Zagi when I next fly it and compare the brake with the standard ESCs brake.

The ESCs I built are all on S400 gearbox planes - on these the brake stops the prop (unless you build up a lot of airspeed...)

The standard ESC for the Zagi and Unicorn have a brake and the props stop unless you are doing a lot of speed. But as I said I will do a direct comparison next time I get a chance.

Sorry, I don't know how much improvement in braking you get as the resistance of the brake FET drops. I suspect not much, as far as I could see all the other DIY ESCs I could find circuits for used FETs similar to the IRF9540.

The 'soft' brake was intended for gearboxes to reduce the stress on the gears. I never thought about its use in direct drive applications - but I can't see any reason not to use it. Once the brake is fully on you can't get any more braking than that (without going to some form of active brake).

The diode directly on the motor is more important the longer the leads between the motor and the ESC get, also it reduces the reliance on the brake FET - which as you found is a good idea!


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion DIY Brushed ESC Mosfet garethadams DIY Electronics 6 Oct 16, 2006 12:39 AM
WTB: 15-20 amp brushed ESC Bill Germanson Aircraft - Electric - Airplanes (FS/W) 0 Oct 26, 2002 02:08 PM
WTB - 50 amp brushed ESC, 3000 NiMH Panasonics Bill Germanson Aircraft - Electric - Airplanes (FS/W) 3 Oct 14, 2002 09:33 PM
WTB Brushed ESC for 20 to 30 cells GoodTill Aircraft - Electric - Airplanes (FS/W) 5 Jan 18, 2002 09:17 PM
Brushed ESC Steve KO Electric Heli Talk 1 Dec 10, 2001 04:07 PM