variable analogue output
- Thread starter Diesel62
- Start date
Reading analogue inputs, say a temp sensor, is no problem. The chips generally support 3 or more ADC inputs which are converted to 8-bit or 10-bit digital values.
None of the chips output analogue voltages. However they do support 1 or more PWM outputs (Pulse Width Modulation). Usually a PWM output will meet your needs.
If you really needed a true analogue voltage output the PWM signal could be converted by an external RC network.
-- Dan
None of the chips output analogue voltages. However they do support 1 or more PWM outputs (Pulse Width Modulation). Usually a PWM output will meet your needs.
If you really needed a true analogue voltage output the PWM signal could be converted by an external RC network.
-- Dan
premelec
Senior Member
PWMOUT can run continuously on 08M from a single pin... PWM needs updating and I use serial D/A converter chips like TI TLC520 for stable continuous D/A outputs....
oops - that's a TLC5620 4 x 8 bit D/A converter chip... basically a resistance string with 255 taps...
Edited by - premelec on 14/01/2007 16:51:17
oops - that's a TLC5620 4 x 8 bit D/A converter chip... basically a resistance string with 255 taps...
Edited by - premelec on 14/01/2007 16:51:17
demonicpicaxeguy
Senior Member
get a parallel adc and a parallel dac
all along with a multiplexer and bit of creative writing you'd be able to reproduce te signal and do the processing as well...
whats this for anyway?
all along with a multiplexer and bit of creative writing you'd be able to reproduce te signal and do the processing as well...
whats this for anyway?
At the very simplest level this is one picaxe and a low pass filter using pwmout. The correct R and C will need to be calculated.
I think it will be the peripheral stuff that will be complex. I'm still trying to get a toy car to move using pwmout (a simple problem that is actually quite complex - latest problem is RF interference from the motors upsetting the picaxe). A car has all sorts of motors plus the hash from the ignition, and you don't want the picaxe giving up the ghost when you wind up the electric windows.
I think at the very least this will need a shielded box with ferrites on all wires going in and out, plus really good power supply filtering. Then you have to consider how things might fail - eg if a picaxe fails then pwmout will go to zero, and what will the engine MCU register that temperature as?
Also programming might be a challenge. If you want to adjust something by +1 you will need to undo the box, pull the picaxe out of a socket, reprogram it and put it back. Or will it be programmed in situ?
As this is a 'mission critical' application (see Rev Eds comments on that), I'd seriously think about a picaxe sending out an AC 'I'm alive' waveform, which goes via a blocking capacitor into a diode and RC network, thence into a transistor driving 4PDT relay, and if the picaxe isn't producing that waveform the relay drops out and reconnects all the engine MCU wires as they originally were.
I'm sure this can all be done but it would be well worth getting a testbed circuit working on the bench before cutting any wires in the engine.
I think it will be the peripheral stuff that will be complex. I'm still trying to get a toy car to move using pwmout (a simple problem that is actually quite complex - latest problem is RF interference from the motors upsetting the picaxe). A car has all sorts of motors plus the hash from the ignition, and you don't want the picaxe giving up the ghost when you wind up the electric windows.
I think at the very least this will need a shielded box with ferrites on all wires going in and out, plus really good power supply filtering. Then you have to consider how things might fail - eg if a picaxe fails then pwmout will go to zero, and what will the engine MCU register that temperature as?
Also programming might be a challenge. If you want to adjust something by +1 you will need to undo the box, pull the picaxe out of a socket, reprogram it and put it back. Or will it be programmed in situ?
As this is a 'mission critical' application (see Rev Eds comments on that), I'd seriously think about a picaxe sending out an AC 'I'm alive' waveform, which goes via a blocking capacitor into a diode and RC network, thence into a transistor driving 4PDT relay, and if the picaxe isn't producing that waveform the relay drops out and reconnects all the engine MCU wires as they originally were.
I'm sure this can all be done but it would be well worth getting a testbed circuit working on the bench before cutting any wires in the engine.