In this system, I'm not sure how much noise there is. Everything has=20 very short distances (inches to a foot at max), and RS-485 for the=20 digital, etc. Motors are smal Nema-23 servo drives, which are nicely=20 encased, so I expect it's all engineered to be low noise. I do average a number of samples, and filter, but here I'll need to make=20 very rapid changes to motor speed based on the sensor value so I'm not=20 yet sure how many samples I'll be able to grab before causing an lag=20 issues. There's really one way to find out at this point. Cheers, -Neil On 5/5/2016 10:53 AM, alan.b.pearce@stfc.ac.uk wrote: > I would have thought there would be enough noise in the system to just ov= ersample on the ADC, and add an appropriate number of samples together. > > There is an appnote around about doing this to get higher resolution, but= I can't remember if it is a Microchip one. > > Found it on my drive, it is an Atmel one - "AVR121: Enhancing ADC resolut= ion by Oversampling". Probably find it as Doc8003 on their website. > > > >> -----Original Message----- >> From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] On Behalf >> Of Neil >> Sent: 05 May 2016 15:03 >> To: Microcontroller discussion list - Public. >> Subject: Re: [EE] Increasing A/D resolution for low-voltage input signal >> >> Ah okay... but I'll have to investigate creating a noise source... which= I >> believe I can do with a zener. >> >> The issue with the op-amp is finding a suitable one in my junk bin. >> Cringing at the thought of paying for overnight shipping for a $0.50 op-= amp, >> but I need to get this running this weekend. >> >> Will investigate the LED as a clamp. >> >> Cheers, >> -Neil >> >> >> >> >> On 5/5/2016 3:24 AM, Richard Prosser wrote: >>> Hi Neil, >>> >>> You can add dither (noise) to the wanted signal and then average the >>> results to give intermediate steps. Generally OK for a couple of bits >>> of extra resolution. >>> >>> Low voltage zeners tend to have a fairly soft turn on point - not what >>> you want. You may even find a LED of the right type could be better, >>> or a string of std. diodes. Better again might be an opamp to clamp >>> the maximum level. (And amplifiy). >>> >>> RP >>> >>> >>> On 5 May 2016 at 17:53, Neil wrote: >>>> Hi all, >>>> >>>> I'm looking for a way to get more resolution on a PIC32 A/D input >>>> (I'll actually be using a Chipkit board) for a low-voltage input signa= l. >>>> Official spec =3D as good as I can get it (in limited time). The >>>> sensor/amp puts out a 0-10V signal for 0-to-10 Kg force, but I only >>>> need to measure about 0 to 0.5kg (up to 0.5V), so only 5% of the full >> range. >>>> Voltage-dividing the signal reduces the resolution/range, so I'll >>>> protect the A/D input with a zener to clamp the signal instead. >>>> >>>> Now, with a 2V external reference (lowest allowed by datasheet), 0 to >>>> 0.5V nets only 8-bit resolution. I'd really like to get better. >>>> Other than adding an op-amp into the circuit, is there any other >>>> trick to bump the resolution up? >>>> >>>> If I do happen to have a workable op-amp on hand (though I doubt) to >>>> amplify the signal x4, I know that the zener will affect the >>>> linearity of the signal as it gets close to the zener voltage, but >>>> would a 3-ish volt zener cause any noticeable error at around 2V? Or >>>> is there some parameter that indicates that a zener would have a >> "sharper" knee? >>>> Cheers, >>>> -Neil. >>>> >>>> >>>> -- >>>> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >>>> View/change your membership options at >>>> http://mailman.mit.edu/mailman/listinfo/piclist >> -- >> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >> View/change your membership options at >> http://mailman.mit.edu/mailman/listinfo/piclist --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .