There are 6 messages in this thread.
You are currently looking at messages 0 to 6.
What would probably be the likely cause of the output bouncing on the lower 4 bits, bits 0,1,2,3, but has good stability on the upper bits. I'm using a junk box Pic 16f807 with one input line to AN0 which is being fed from the output of a LM324 op amp wired as a non-inverting dc gain. The signal to the op amp is coming directly from a thermocouple. If it's do to noise why does it only bounce on the lower bits? Pj
Pj wrote: > What would probably be the likely cause of the output bouncing on the > lower 4 bits, bits 0,1,2,3, but has good stability on the upper bits. > I'm using a junk box Pic 16f807 with one input line to AN0 which is > being fed from the output of a LM324 op amp wired as a non-inverting > dc gain. The signal to the op amp is coming directly from a > thermocouple. > > If it's do to noise why does it only bounce on the lower bits? > > Pj Lower bits are the smaller voltage detection bits, right...
On Fri, 22 Feb 2008 07:21:52 -0800 (PST), Pj <f...@yahoo.com> wrote: >What would probably be the likely cause of the output bouncing on the >lower 4 bits, bits 0,1,2,3, but has good stability on the upper bits. >I'm using a junk box Pic 16f807 with one input line to AN0 which is >being fed from the output of a LM324 op amp wired as a non-inverting >dc gain. The signal to the op amp is coming directly from a >thermocouple. > >If it's do to noise why does it only bounce on the lower bits? > >Pj It's analog noise from the opamp; digitized, the lower bits wobble. An LM324 is a terrible opamp for this. But expect noise anyhow, since thermocouples make tiny voltages. Use a better opamp, and take lots of samples and average. John
John Larkin wrote: > On Fri, 22 Feb 2008 07:21:52 -0800 (PST), Pj <f...@yahoo.com> > wrote: > >> What would probably be the likely cause of the output bouncing on the >> lower 4 bits, bits 0,1,2,3, but has good stability on the upper bits. >> I'm using a junk box Pic 16f807 with one input line to AN0 which is >> being fed from the output of a LM324 op amp wired as a non-inverting >> dc gain. The signal to the op amp is coming directly from a >> thermocouple. >> >> If it's do to noise why does it only bounce on the lower bits? > It's analog noise from the opamp; digitized, the lower bits wobble. > > An LM324 is a terrible opamp for this. But expect noise anyhow, since > thermocouples make tiny voltages. > > Use a better opamp, and take lots of samples and average. Or use a better ADC. For instance, the AD7792 has an integrated amplifier, and 50/60 Hz filters. With some careful layout, it will get you 16 stable bits.
On Feb 22, 3:21=EF=BF=BDpm, Pj <frysn...@yahoo.com> wrote: > What would probably be the likely cause of the output bouncing on the > lower 4 bits, bits 0,1,2,3, but has good stability on the upper bits. > I'm using a junk box Pic 16f807 with one input line to AN0 which is > being fed from the output of a LM324 op amp wired as a non-inverting > dc gain. =EF=BF=BDThe signal to the op amp is coming directly from a > thermocouple. > > If it's do to noise why does it only bounce on the lower bits? > > Pj A low pass filter to reduce the bandwidth will reduce the noise, you only need dc to 0.5Hz at most.
On Fri, 22 Feb 2008 18:49:44 +0100, Arlet Ottens <usenet+5...@c-scape.nl> wrote: >John Larkin wrote: > >> On Fri, 22 Feb 2008 07:21:52 -0800 (PST), Pj <f...@yahoo.com> >> wrote: >> >>> What would probably be the likely cause of the output bouncing on the >>> lower 4 bits, bits 0,1,2,3, but has good stability on the upper bits. >>> I'm using a junk box Pic 16f807 with one input line to AN0 which is >>> being fed from the output of a LM324 op amp wired as a non-inverting >>> dc gain. The signal to the op amp is coming directly from a >>> thermocouple. >>> >>> If it's do to noise why does it only bounce on the lower bits? > >> It's analog noise from the opamp; digitized, the lower bits wobble. >> >> An LM324 is a terrible opamp for this. But expect noise anyhow, since >> thermocouples make tiny voltages. >> >> Use a better opamp, and take lots of samples and average. > >Or use a better ADC. For instance, the AD7792 has an integrated >amplifier, and 50/60 Hz filters. With some careful layout, it will get >you 16 stable bits. Maybe he only needs 8 or 9 bits... not everything has to be a frick'n gold plated design y'know. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" s...@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com