Post by thread:12F675 Documentation for A/D Converter

While looking at the 12F675 for a totally different project, I started reading about the A/D converter. As a computer user who is blind, I read the PDF documents in Linux using either pstotext or ps2ascii, which ever one works on a given document without crashing. It reads text with no trouble, but tables and charts get so mixed up that they are hard to read at least, I don't necessarily trust what I am reading. There is a figure called FIGURE 7-2: 10-BIT A/D RESULT FORMAT This tells one how fast the A/D converter takes to do its successive approximation routine to arrive at a value based upon various clock frequencies. I get the impression that I can use the internal RC clock at 4 MHZ or use an external crystal or other clock source at 20 MHz. It looks like that the A/D conversion completes in 400 ns at 20 MHZ and 1.6 us at 4 MHZ. If this is correct, then one could sample at a little more than 2 MHZ sample rates at 20 MHZ clock or 625,000 samples per second at the 4 MHZ clock frequency. If this is true, the PIC could be possibly made to behave like a low-frequency radio receiver. Obviously, the higher the sampling frequency, the less time there is to do anything useful with the results, but I was thinking about maybe trying to decode the WWVB signal at 60 KHZ which should be well within the capability of the A/D converter if I am reading the table right. Thanks for your help. When I actually get ready to hook one up, I may have similar trouble with the pin-out, but I have also run across some of the PIC data sheets in which that table was easy enough to read with confidence. Here is what the decoding process does to this kind of information so you can see why it is a bit hard to de sypher. FIGURE 7-2: 10-BIT A/D RESULT FORMAT A/D Clock Source (TAD) Device Frequency Operation ADCS2:ADCS0 20 MHz 5 MHz 4 MHz 1.25 MHz 2 TOSC 000 100 ns (2) 400 ns(2) 500 ns(2) 1.6 us 4 TOSC 100 200 ns (2) 800 ns(2) 1.0 us(2) 3.2 us 8 TOSC 001 400 ns (2) 1.6 us 2.0 us 6.4 us 16 TOSC 101 800 ns (2) 3.2 us 4.0 us 12.8 us(3) 32 TOSC 010 1.6 us 6.4 us 8.0 us (3) 25.6 us(3) 64 TOSC 110 3.2 us 12.8 us (3) 16.0 us(3) 51.2 us(3) A/D RC x11 2 - 6 us (1,4) 2 - 6 us(1,4) 2 - 6 us(1,4) 2 - 6 us(1,4) Legend: Shaded cells are outside of recommended range. Note 1: The A/D RC source has a typical TAD time of 4 us for VDD > 3.0V. 2: These values violate the minimum required TAD time. 3: For faster conversion times, the selection of another clock source is recommended. 4: When the device frequency is greater than 1 MHz, the A/D RC clock source is only recommended if the conversion will be performed during SLEEP. Note: The GO/DONE bit should not be set in the same instruction that turns on the A/D. -- http://www.piclist.com hint: To leave the PICList spam_OUTpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

Seeing the mess the converter made of the table I can understand your confusion. I might have a different version of the datasheet, as my table is titled "TABLE 7-1:" But either way I think the more relevant piece of information is in the paragraph before the table. It says: For correct conversion, the A/D conversion clock (1/TAD) must be selected to ensure a minimum TAD of 1.6 micro seconds. So 1.6 micro seconds is the fastest it can read a sample, regardless of clock frequency. So I think that works out to 625 kHz max sample speed. Using the internal A/D for a radio receiver would be interesting, but I think fitting it into the 1024 byte flash of the 12F675 would be more impressive :-) -Denny {Original Message removed}

I think you're missing the part where it says how many Tad cycles it takes to do a conversion - ISTR it's something of the order of ten for 8 bits, probably 12 for 10 bits. On Sun, 8 Aug 2004 16:01:22 -0400, you wrote: {Quote hidden}>Seeing the mess the converter made of the table I can understand your >confusion. > >I might have a different version of the datasheet, as my table is titled >"TABLE 7-1:" But either way I think the more relevant piece of information >is in the paragraph before the table. It says: > >For correct conversion, the A/D conversion clock (1/TAD) must be selected >to ensure a minimum TAD of 1.6 micro seconds. > >So 1.6 micro seconds is the fastest it can read a sample, regardless of >clock frequency. So I think that works out to 625 kHz max sample speed. > >Using the internal A/D for a radio receiver would be interesting, but I >think fitting it into the 1024 byte flash of the 12F675 would be more >impressive :-) > >-Denny > > > >{Original Message removed}

> On 08th August 2004 Mike Harrison wrote: > >I think you're missing the part where it says how many Tad >cycles it takes to do a conversion - ISTR it's something of >the order of ten for 8 bits, probably 12 for 10 bits. > The AD converter takes 11 cycles, each of which must be 1.6us or greater. This gives a minimum conversion time of 17.6us. You then need to add any acquisition time on top of this. The PIC's ADC can not be described as "high speed"! Regards Mike RJ ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. ======================================================================= -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

My thanks to all of you. It is always good to provide important specifications in redundant form. I knew that table was trying to tell me something, but it was so mangled as to be useless. There is actually a better PDF converter in Linux, but that one choked on this particular data sheet and wouldn't get past ten or twelve pages in before "abending" as they say in the computer world. Even the better converter does enough weird things to tables that I don't trust what I am reading at times. I believe that the problem is that the PDF standard, such as it is, doesn't seem to be able to tell a PDF converter enough lay out information to always make the columns stay together. Ideally, the font and size is preserved, but the reality is that the character size on what amounts to an ASCII VT100 terminal is immutable so it is apt to virtually spill off the lines and make a total mess. Michael Rug bee-Jones writes: >The AD converter takes 11 cycles, each of which must be 1.6us or greater. >This gives a minimum conversion time of 17.6us. You then need to add any >acquisition time on top of this. The PIC's ADC can not be described as >"high speed"! Yup, that doesn't quite reach the 60 KHZ frequency of WWVB and it is even further from the frequency of the Rug bee time signal which I think is 77 KHZ although one might use a sub multiple scheme and receive it that way. Synchronization to the carrier would be another issue. The WWVB signal advances 45 degrees at the top of the hour and then retards 45 degrees at 5 or 10 minutes after as a sort of marker so one would have to stay synchronized. If the converter ran much faster, then it would be possible to over-sample and find the peaks. You'd sure want a band-pass filter or every bit of noise around would look like a peak.:-) -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads