MCLR_FROM_RUN

luckyluke · Joined: 18 Apr 2006 Posts: 45

hello
I am working on a robot controller card which have main 18f2520 and 12f629.

Because of robot unfit chassis in some certain maneuvers robot make drifts and this cause 18f2520 have reset MCLR_FROM_RUN (12f629 never resets).

So what exactly MCLR_FROM_RUN is and how can i suppress it ?

Thanks

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

This thread talks about the problem and has some solutions:
http://www.ccsinfo.com/forum/viewtopic.php?t=56109

luckyluke · Joined: 18 Apr 2006 Posts: 45

Is there any difference about placing the 100nf or is it best to solder directly to power pins?

temtronic · Posted: Wed Mar 28, 2018 9:51 am

OK, I'm 'old school' but I'd never,ever put components UNDER a PIC or other device as shown in your "B" or right side version.

Murphy's Law of Black Smoke says that component , under the PIC WILL fail.

Jay

luckyluke · Joined: 18 Apr 2006 Posts: 45

OK
beside your point what i want to ask is electrically which one is correct

B regulator --> bypass cap --> power pins
A regulator --> power pins --> bypass cap

is there any difference?

edit note: A and B corrected according to the image

Josep Robert · Joined: 27 Mar 2018 Posts: 25

Very well explained here:
https://electronics.stackexchange.com/questions/74509/how-to-place-decoupling-capacitor-in-four-layer-pcb

Josep Robert

temtronic · Posted: Wed Mar 28, 2018 11:14 am

Yes, though how significant depends upon a lot of things.
A does allow 'bad' energy to enter the PIC.
B allows the capacitor to absorb the 'bad' energy BEFORE it enters the PIC.
So B would be the better choice.
Usually the manufacturer of the devices will show suggested PCB layouts for bypass caps as well as xtal/cap configurations.
What is also important is in using the correct TYPE and value for the capacitor(s). Again, the mfr will inform you of the type/value, sometimes even telling you a specific device to use.

Just noticed, when I mean B, I refer to the drawing NOT your 2nd post.

So in your text power pins-cap-PIC (a).
Jay

RF_Developer · Joined: 07 Feb 2011 Posts: 839

temtronic · Posted: Thu Mar 29, 2018 5:47 am

re: That's the old way of thinking about it. Yeah, well I get to collect the guv pension 2 months from now when I turn 65, so , OK...I'm old.....
When I designed Optical Emission Spectrometer Data Acqusition PCBs the ground planes were a 'checkerboard' of traces not a 'flooded' or solid copper. Never had a problem with RF doing that.
I do like 'fat' traces whenever possible,then again I was a 'tape and donut' guy..that was 1/2 a lifetime ago, sigh....
Putting parts under PICs just ain't 'right' to me though.....

Jay
Smile

RF_Developer · Joined: 07 Feb 2011 Posts: 839

Ttelmah · Joined: 11 Mar 2010 Posts: 19535

I've been forwards, then back, then forwards again.

Did projects more than 30 years ago, which were all surface mount, padless via's, but were still being drawn with tape and pads. There was a new PCB package at the time called 'Vutrax', but it had limited board size capabilities.
Then had to start designing for stuff that could be field serviced, so switched back to using a lot of through-hole parts.
Then because fewer and fewer parts are available in through-hole, the sheer packing density possible with SMD, and costs have become so low that the engineers can carry complete boards, so went back to SMD again...

Now on the examples being posted, the big thing is to make the power tracks thicker. If power planes can't be used, thicker tracks, and critically (not shown), where they go back to, and how they route relative to other tracks carrying power. The routes could also be straightened with smoother corners.

luckyluke · Joined: 18 Apr 2006 Posts: 45

Thanks for opinions.

In case of somebody have same problem i have solved it with a diode between Vpp and 5V.

I was making hobby circuits with lcd,l ed and some sensors. It was easy work with them.
Problem started after i used some motors. I made a lot of development on pcb design with the great help of Ttelmah.
I added power planes, tvs, zener, inductor then i add LC filter, two battery. I even tried with LF series pic in order to avoid brownout. I doubled drivers, solder 1000uf direct to power pins, 100 nf to power pins but resetting was always there.
No matter what i did, resetting problem always came back. When there was nothing to change, i changed my point of view.

My first idea was pic not resetting. It was going in programming mode and with no incoming data, program start from begin.
it sounds ridiculous to me but trying a diode to icsp pins was very easy compared to other changes.

i connect a diode Vpp--->|---5v and it was fantastic!! solved the problem immediately. In my case fast diodes shows better results.

Ttelmah · Joined: 11 Mar 2010 Posts: 19535

That implies you are getting RF pickup on the Vpp pin. Question what circuit you have on the pin?. Obviously a diode won't work if you ave any form of reset circuit. If it wasn't connected, then 'asking for trouble'...
Glad you have found a solution. Smile

luckyluke · Joined: 18 Apr 2006 Posts: 45

There was only a 10K pullup resistor on MCLR now there are diode and 10K.

gaugeguy · Joined: 05 Apr 2011 Posts: 303

If the MCLR pin has an extended trace to a programming header then a small capacitance of 100pF or so can be placed to ground along with the 10k pullup to Vcc. This will aid in the rejection of noise that could cause an unintended reset and still allow proper programming.