TheBackShed.com

Forum Index : Microcontroller and PC projects : Problem with PIC32MX470F512L

Page 2 of 2

Author

Message

robert.rozee
Guru

Joined: 31/12/2012
Location: New Zealand
Posts: 2447

Posted: 01:15am 03 Mar 2015

Copy link to clipboard

Print this post

Frank N. Furter said
I have lost again my MX170 basic program after a short disconnection of my batterypack WITHOUT my MX470 connected and without a connected PC on my USB-serial converter...
Frank

frank, can you possibly expand a little on this? if i am reading things right:

1. an MX170 connected to a USB-serial converter: TxD, RxD, gnd.
2. no connection between converter and PC.
3. MX170 powered from a battery pack (via 3v3 regulator).
4. running micromite basic version 4.xx?

what chip is used on the USB-serial converter? if it is a bare-board, can you post a photo? have you been able to reproduce the effect?

cheers,
rob :-)Edited by robert.rozee 2015-03-04

Frank N. Furter
Guru

Joined: 28/05/2012
Location: Germany
Posts: 950

Posted: 08:19pm 03 Mar 2015

Copy link to clipboard

Print this post

Hi Robert:

1. Yes, USB-serial converter connected to TxD, RxD, Gnd
2. Yes, it wasn't connected with PC
3. Yes, MX170 powered from a battery pack (2x 1.2V accu with step-up converter) via 3v3 linear regulator
4. I am not shure at the moment - it's v4.6a or v4.6b - I can check it this evening...

The chip on the USB-serial converter is a Prolific in a housing (no bare-board).
It was the second time that I've lost my basic program with this constellation just after a short disconnection between battery pack and my 3v3 linear regulator...

Frank

MicroBlocks

Guru

Joined: 12/05/2012
Location: Thailand
Posts: 2209

Posted: 09:03pm 03 Mar 2015

Copy link to clipboard

Print this post

You experienced an unwanted 'reset' that clears the code from flash.
It happens when a short between the rx and tx is detected, and unfortunately it also happens sometimes when it is connected to a usb-serial converter and power is switched on/off or wires are wiggled a bit.
There is/was a discussion going on about this subject on tbs, i am in favor of removing that 'feature' as it sometimes ends up being a 'bug' instead, worst case leaving the chip in a state that it has to be reprogrammed with the firmware.

Microblocks. Build with logic.

Frank N. Furter
Guru

Joined: 28/05/2012
Location: Germany
Posts: 950

Posted: 11:59pm 03 Mar 2015

Copy link to clipboard

Print this post

Hi TZAdvantage,

you are right! I think the idea of resetting the chip with a short between rx and tx is quite good - but it is in practice to accident-sensitive!!!

Is it not possible to send a sequence from tx to rx over this short to avoid a unwanted reset???

Frank

robert.rozee
Guru

Joined: 31/12/2012
Location: New Zealand
Posts: 2447

Posted: 01:12am 04 Mar 2015

Copy link to clipboard

Print this post

Frank N. Furter said Is it not possible to send a sequence from tx to rx over this short to avoid a unwanted reset???
Frank

alas, for every clever solution or procedure that someone comes up with, a new set of issues emerge that prevent it from being effective. it is rather telling that there have been extensive discussions on the matter over many many months, without arriving at a foolproof solution (my own favourite is a 1-transistor inverter used as a 'key' between TxD and RxD).

in your own case, you might try adding a resistor from the micromite's RxD line (pin 12) to ground. try 3k3, or even as low as 1k. this may prevent the reset signal from being seen when it transitions through the unpowered prolific chip.

cheers,
rob :-)

Page 2 of 2

Print this page

To reply to this topic, you need to log in.