Post by thread:[PIC]: Reading Mains Voltage From a PIC

Hi, I'm into a UPS application which requires a PIC16F877 to monitor the mains voltage (230 Volts AC here in India). The most obvious answer may be to use a step down transformer. But How do I read the AC level. PIC ADCs are supposed to take the DC voltages. And since its AC what if I read the Line during Zero Crossing ? Adding a small bridge rectifier out of 4148 diodes may help. but still, there will be some Alternating component left. Adding a capacitor (1 mfd tantalum) might increase the voltage a bit. I guess i'll have to Compensate for this increase in software. Worst of all problems is that since it's a UPS application, the program flow will be highly non-linear as TMR0 interrupt will be called very often. And It'll be very hard to synchronize with the zero crossing of the line voltage. Can someone suggest a good solution ? Thainking You, Jeethu Rao http://www.jeethurao.com -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

The simplest one you've pointed partially, step down transformer, rectifying bridge, small filtering capacitor ( which will not increase but keep constant Vmax, or Uvv/2, a resistive divider to obtain effective value and drop to pic AD ( less than 5V ). No software for calibration... Vasile On Thu, 21 Jun 2001, Jeethu Rao wrote: {Quote hidden}> Hi, > I'm into a UPS application which requires a PIC16F877 to monitor the mains > voltage (230 Volts AC here in India). The most obvious answer may be to use > a step down transformer. But How do I read the AC level. PIC ADCs are > supposed to take the DC voltages. And since its AC what if I read the Line > during Zero Crossing ? > Adding a small bridge rectifier out of 4148 diodes may help. but still, > there will be some Alternating component left. Adding a capacitor (1 mfd > tantalum) might increase the voltage a bit. I guess i'll have to Compensate > for this increase in software. > > Worst of all problems is that since it's a UPS application, the program flow > will be highly non-linear as TMR0 interrupt will be called very often. And > It'll be very hard to synchronize with the zero crossing of the line > voltage. > > Can someone suggest a good solution ? > > Thainking You, > > Jeethu Rao > http://www.jeethurao.com > > -- > http://www.piclist.com hint: PICList Posts must start with ONE topic: > [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads > > -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Jeethu, imho look into a National analog databook and build one of the 'perfect rectifier' circuits pictured there. They accept AC (you use a resistive divider) and output DC which can be fed to the A/D. They are rather good (i.e. precise) even if using cheap parts. Peter -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> The simplest one you've pointed partially, step down transformer, > rectifying bridge, small filtering capacitor ( which will not increase but > keep constant Vmax, or Uvv/2, a resistive divider to obtain effective > value and drop to pic AD ( less than 5V ). No software for calibration... > Vasile Just remember, this will read PEAK voltage. To read RMS (assuming a sine wave!) just divide by a factor of SQRT(2). Bob Ammerman RAm Systems (contract development of high performance, high function, low-level software) > On Thu, 21 Jun 2001, Jeethu Rao wrote: > > > Hi, > > I'm into a UPS application which requires a PIC16F877 to monitor the mains > > voltage (230 Volts AC here in India). The most obvious answer may be to use > > a step down transformer. But How do I read the AC level. PIC ADCs are > > supposed to take the DC voltages. And since its AC what if I read the Line > > during Zero Crossing ? > > Adding a small bridge rectifier out of 4148 diodes may help. but still, > > there will be some Alternating component left. Adding a capacitor (1 mfd > > tantalum) might increase the voltage a bit. I guess i'll have to Compensate > > for this increase in software. > > > > Worst of all problems is that since it's a UPS application, the program flow > > will be highly non-linear as TMR0 interrupt will be called very often. And {Quote hidden}> > It'll be very hard to synchronize with the zero crossing of the line > > voltage. > > > > Can someone suggest a good solution ? > > > > Thainking You, > > > > Jeethu Rao > > http://www.jeethurao.com > > > > -- > > http://www.piclist.com hint: PICList Posts must start with ONE topic: > > [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads > > > > > > -- > http://www.piclist.com hint: The list server can filter out subtopics > (like ads or off topics) for you. See http://www.piclist.com/#topics > > -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

If you rectify the 230VAC first you can use a very non-perfect rectifier. The errors will be insignificant vs. the 300VDC result. Then use a resistive divider to reduce the 300V to 5V for the PIC A/D. The will give you the average voltage level. If you are looking for single cycle dropouts you will have to do something faster, probably synchronized to the power line frequency. Do you need to isolate the PIC ground from the AC mains? Sherpa Doug > {Original Message removed}

> I'm into a UPS application which requires a PIC16F877 to monitor the mains > voltage (230 Volts AC here in India). The most obvious answer may be to use > a step down transformer. But How do I read the AC level. You could low pass filter the AC line signal first to make sure it is mostly a sine wave, then peak detect it with a diode, then low pass filter that and trade off ripple rejection with response speed. How fast do you need to know the AC line voltage changed? If you don't want the diode drop in there you can do an absolute value circuit with opamps. Fortunately you have very low bandwidth requirements. You could also read the LPF AC line signal fast enough to "follow" the sine wave. At 50Hz you have 20mS per cycle, which is quite a lot. If you want to throw hardware at it and simplify the software, you could sample and hold the line voltage at its peak once per cycle. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, spam_OUTolinTakeThisOuTembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

>using rectified AC to measure AC mains voltage I would only use a rectifierbridge/capacitor as a cheap and dirty alternative to a *real* analog measuring system. If you want a cheap and dirty solution, that's OK nothing wrong with it. I do it all the time myself, too. The signal from a transformer is distorted by the diode drop ( 2 diode drops in a full wave bridge), only the peak values are filtered by the capacitor, and the precision of the transformer turns will not be very consistent. I would not use such a circuit and expect any precision from it unless I used a specially made transformer. If you want something more accurate, I'd suggest an op-amp precision rectifier. This could be driven by a simple resistor divider, or a low-cost low-power transformer if you need mains isolation. This could produce an instantaneous value, useful in detecting fault conditions. Could also calibrate out transformer variations. If you want an RMS value, you can obtain it several ways. There are some dedicated analog chips for this purpose (Try http://www.national.com ) or you could do it in software. Take a number of readings over the AC cycle, take the square root of each, average them, and square the result (root mean square). If you want to filter out spikes and noise, you need to average the value for a long time, several cycles. Or use a software median filter http://members.socket.net/~llile/ . If you are concerned about READING spikes, you might want a sample-and-hold analog circuit. This would hold the spike level until you can get around to measuring it. At any rate, a general rule could be that processing analog signals directly from a PIC A/D port is OK, if you want a cheap design. If you desire any sort of precision, noise immunity, filtering, etc. you should consider a dedicated anlog section before the A/D input. -- Lawrence Lile -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

At 09:33 AM 6/22/01 -0500, you wrote: >If you want an RMS value, you can obtain it several ways. There are some >dedicated analog chips for this purpose (Try http://www.national.com ) or you >could do it in software. Take a number of readings over the AC cycle, take >the square root of each, average them, and square the result (root mean Just the opposite, actually. You want to integrate the instantaneous power in an assumed constant resistance load over one cycle, divide it by the duration of the cycle, then get rid of the assumed load resistance and take the sqrt to get back to the voltage. (In fact you don't use any resistance value, it's just a way of thinking of it that gives the calculations some physical meaning). There's other ways besides just summing the squares of the measurements to perform the numerical integration, but I think the simple way will work ok if the waveform is reasonably symmetrical about the middle of each half cycle. Otherwise, you'd better go to the trapezoidal rule or Simpson's rule (see any numerical analysis text or do a web search). Best regards, =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Spehro Pefhany --"it's the network..." "The Journey is the reward" .....speffKILLspam@spam@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com Contributions invited->The AVR-gcc FAQ is at: http://www.bluecollarlinux.com =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Hi, I like Mr Doug's Idea as its the cheapest and would save quite some space required by the transformer. I could just rectify the AC Voltage and then use a resistive divider to scale the voltage down to say 3 volts. But i'll have to add a couple of Mica Capacitors of say 0.47mfd/300 Volts to flatten out the DC. And after the Voltage Divider, I might also put up a small LC network based Low Pass Filter. And finally, a Kamikaze Zener Diode of say 5.2 Volts to Commit Suicide and Protect the PIC if one of the Divider resistors is to fail. So, the voltage I'll be getting will be the peak voltage. How do I get the RMS Voltage ? Any ready made algorithms on PICList would be great. And finally, anyone's got any comments on this design ? Jeethu Rao http://www.jeethurao.com {Original Message removed}

> I like Mr Doug's Idea as its the cheapest and would save quite some space > required by the transformer. I like it too as long as the PIC does not need to be isolated from the AC line. > I could just rectify the AC Voltage and then > use a > resistive divider to scale the voltage down to say 3 volts. But i'll have to > add a > couple of Mica Capacitors of say 0.47mfd/300 Volts to flatten out the DC. > And after > the Voltage Divider, I might also put up a small LC network based Low Pass > Filter. I don't see any need for LC. RC should be fine unless you're right on the edge of the tradeoff between ripple rejection and response time. > And finally, a Kamikaze Zener Diode of say 5.2 Volts to Commit Suicide and > Protect > the PIC if one of the Divider resistors is to fail. > > So, the voltage I'll be getting will be the peak voltage. How do I get the > RMS Voltage ? For a sinusoid, RMS = peak / SQRT(2). If the waveshape is not known, then you have to take lots of samples, square them, average them, and take the square root of the average. RMS = "root mean square". Please fix your reply address. The default reply goes to you instead of the list. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, olinKILLspamembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

On Sat, 23 Jun 2001, Jeethu Rao wrote: > > So, the voltage I'll be getting will be the peak voltage. How do I get the > RMS Voltage ? > Any ready made algorithms on PICList would be great. Of course you dont need any PIC algorithm to decrease a voltage with 1.41 ( sqrt 2 ) from initial value. You need just a potentiometer or two resistors. And keep your electronics isolated from the 220...240V mains. Else you'll can't talk with other devices only through optoisolated devices: RS232/485 etc. Vasile {Quote hidden}> > > If you rectify the 230VAC first you can use a very non-perfect > rectifier. The errors will be insignificant vs. the 300VDC result. > Then use a resistive divider to reduce the 300V to 5V for the PIC A/D. > > The will give you the average voltage level. If you are looking for > single cycle dropouts you will have to do something faster, probably > synchronized to the power line frequency. > > Do you need to isolate the PIC ground from the AC mains? > > Sherpa Doug > > -- > http://www.piclist.com hint: The list server can filter out subtopics > (like ads or off topics) for you. See http://www.piclist.com/#topics > > -- > http://www.piclist.com hint: The PICList is archived three different > ways. See http://www.piclist.com/#archives for details. > > -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam.....mitvma.mit.edu

> > And finally, a Kamikaze Zener Diode of say 5.2 Volts to > Commit Suicide and > > Protect > > the PIC if one of the Divider resistors is to fail. The Microchip app note on connecting a PIC to the power line (AN521) recommends using two resistors in series so that a simultanious short failure of both resistors will be extreemly unlikely. {Quote hidden}> > > > So, the voltage I'll be getting will be the peak voltage. > How do I get the > > RMS Voltage ? > > For a sinusoid, RMS = peak / SQRT(2). If the waveshape is > not known, then > you have to take lots of samples, square them, average them, > and take the > square root of the average. RMS = "root mean square". Two stages of RC filtering with 2.2meg resistors and 6800pf caps will provide a safe resistance and will filter out harmonics leaving nothing much more than a 50/60Hz sine wave. Then the RMS = Peak/sqrt(2) approximation will be pretty good. Sherpa Doug -- http://www.piclist.com hint: To leave the PICList EraseMEpiclist-unsubscribe-requestspam_OUTTakeThisOuTmitvma.mit.edu

> The Microchip app note on connecting a PIC to the power line (AN521) > recommends using two resistors in series so that a simultanious short > failure of both resistors will be extreemly unlikely. > Also, be sure you use resistors rated at sufficient voltage. you would *not* for instance, want to risk 0805 surface mount resistors in this case - insufficient creepage path! I think you'll be stuck with good ol' 1/4 watt thru hole resistors - but I can't find the reference just now. -- Lawrence Lile -- http://www.piclist.com hint: To leave the PICList piclist-unsubscribe-requestspam_OUTmitvma.mit.edu

I use Philips VR25 resistors for this purpose. Daniel... ----- Original Message ----- From: Lawrence Lile <@spam@llileKILLspamTOASTMASTER.COM> To: <KILLspamPICLISTKILLspamMITVMA.MIT.EDU> Sent: Tuesday, June 26, 2001 00:27 Subject: Re: [PIC]: Reading Mains Voltage From a PIC > > The Microchip app note on connecting a PIC to the power line (AN521) > > recommends using two resistors in series so that a simultanious short > > failure of both resistors will be extreemly unlikely. > > > > > Also, be sure you use resistors rated at sufficient voltage. you would > *not* for instance, want to risk 0805 surface mount resistors in this case - > insufficient creepage path! I think you'll be stuck with good ol' 1/4 watt {Quote hidden}> thru hole resistors - but I can't find the reference just now. > > > -- Lawrence Lile > > -- > http://www.piclist.com hint: To leave the PICList > RemoveMEpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu > > -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads