Post by thread:[PIC]: PIC16F84A + Ampire AC162B Rev E LCD (KS0066

Problem Description: ------------------------------ * I cannot get charecters to show on my LCD Screen. * Im not sure if my circuit is correct in terms of an external clock for the PIC. * I have tried nearly every circuit diagram i have found on the net and the matching software, which is for the hitachi controllers none of which seem to work. Parts List: --------------- - PIC16F84A Microcontroller. - Ampire AC162B Revision E. 16x2 Backlit LCD screen -> LCD Driver IC is a samsung KS0066U. - 4.0 Mhz Crystal - 2x 33pF Ceramic Capacitors - 470 Ohm Resistor. - LED Circuit Description: --------------------------- PIC RB(0-7) Pins are connected to LCD (DB0-7) Pins. PIC RA(0-2) Pins are connected to LCD RS, R/W, E Pins. PIC Vss grounded. PIC Vdd ===> 5v Powersource. LCD Pin1 ===> Grounded. LCD Pin2 ===> 5v Powersource. LCD Pin3 ==> 10Kpot for contrast PIC RA3 => 470 Ohm Resistor => Red LED => Grounded. PIC RA4 => 470 Ohm Resistor => Red LED => Grounded. PIC OSCI => Leg 1 of 4.0Mhz Crystal => 33pF Capacitor1 = > Ground. PIC OSCO => Leg 2 of 4.0Mhz Crystal => 33pF Capacitor2 = > Ground. Questions: ---------------- 1) Is the KS0066U LCD driver compatible with the Hitachi driver for a simmilar 16x2 LCD ? 2) Do you connect the ceramic capacitors between the OSC pins of the PIC and the 4.0Mhz crystal or between the 4.0Mhz crystal and ground. I have seen pictures on websites where the caps are connected between POC OSC pins and the Crystal, whereas the PIC datasheet shows it connected between the Crystal and Ground ? 3) I dont have access to an oscillascope and im not sure if the crystal is actually working, or if the PIC is actually running so i connected a resistor to RA4 to and LED, and added code to my program to light up the LED if the LCD is busy and it didnt come on at all. I then cut out all my LCD code and just put in code to turn PORTA HI->LOW to see if it would light up the led and it didnt. I checked the pins of the PIC with a multimeter, some were 5v, some were 0, but my port A wasnt changing as i would have hoped. 4) Is there an easy way of checking things are all working without having to spend hours wiring up a test circuit? 5) Has anyone managed to get an Apire LCD to display anything, if so do you have a circuit diagram that i could look at? 6) Is there any ticky things i have to keep in mind when writing my software and controlling the Samsung KS0066U LCD driver? Dylan. -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

----- Original Message ----- From: "Dylan" <spam_OUTdylanTakeThisOuTFAST.FUJITSU.COM.AU> To: <.....PICLISTKILLspam@spam@MITVMA.MIT.EDU> Sent: Tuesday, May 22, 2001 10:33 PM Subject: [PIC]: PIC16F84A + Ampire AC162B Rev E LCD (KS0066U Driver) Wont Work? > Problem Description: > ------------------------------ > * I cannot get charecters to show on my LCD Screen. > * Im not sure if my circuit is correct in terms of an external clock for > the PIC. > * I have tried nearly every circuit diagram i have found on the net and the {Quote hidden}> matching software, which is for the hitachi controllers none of which seem > to work. > > Parts List: > --------------- > > - PIC16F84A Microcontroller. > - Ampire AC162B Revision E. 16x2 Backlit LCD screen > -> LCD Driver IC is a samsung KS0066U. > - 4.0 Mhz Crystal > - 2x 33pF Ceramic Capacitors > - 470 Ohm Resistor. > - LED > > Circuit Description: > --------------------------- > > PIC RB(0-7) Pins are connected to LCD (DB0-7) Pins. > PIC RA(0-2) Pins are connected to LCD RS, R/W, E Pins. > PIC Vss grounded. > PIC Vdd ===> 5v Powersource. > LCD Pin1 ===> Grounded. > LCD Pin2 ===> 5v Powersource. > LCD Pin3 ==> 10Kpot for contrast How is the 10K pot hooked up? > PIC RA3 => 470 Ohm Resistor => Red LED => Grounded. > PIC RA4 => 470 Ohm Resistor => Red LED => Grounded. You can't do this. RA4 is open collector, meaning it can't source current. > PIC OSCI => Leg 1 of 4.0Mhz Crystal => 33pF Capacitor1 = > Ground. > PIC OSCO => Leg 2 of 4.0Mhz Crystal => 33pF Capacitor2 = > Ground. What about /MCLR? Make sure to tie it to Vdd. > > Questions: > ---------------- > 1) Is the KS0066U LCD driver compatible with the Hitachi driver for a > simmilar 16x2 LCD ? > > 2) Do you connect the ceramic capacitors between the OSC pins of the PIC > and the 4.0Mhz crystal or between the 4.0Mhz crystal and ground. I have > seen pictures on websites where the caps are connected between POC OSC pins > and the Crystal, whereas the PIC datasheet shows it connected between the > Crystal and Ground ? The datasheet is the correct way. It looks like you have it wired correctly. > > 3) I dont have access to an oscillascope and im not sure if the crystal is > actually working, or if the PIC is actually running so i connected a > resistor to RA4 to and LED, and added code to my program to light up the > LED if the LCD is busy and it didnt come on at all. I then cut out all my See above comment about open collector output. Turn the LED around and use Vdd instead of ground. BCF PORTA,4 will turn it on and BSF PORTA,4 will turn it off. > LCD code and just put in code to turn PORTA HI->LOW to see if it would > light up the led and it didnt. I checked the pins of the PIC with a > multimeter, some were 5v, some were 0, but my port A wasnt changing as i > would have hoped. > > 4) Is there an easy way of checking things are all working without having > to spend hours wiring up a test circuit? A frequency counter to see if the clock is running. An oscilloscope would really be handy; beg, borrow or steal one. > > 5) Has anyone managed to get an Apire LCD to display anything, if so do you {Quote hidden}> have a circuit diagram that i could look at? > > 6) Is there any ticky things i have to keep in mind when writing my > software and controlling the Samsung KS0066U LCD driver? > > > > > > Dylan. > > -- > 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: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

> >1) Is the KS0066U LCD driver compatible with the Hitachi driver for a >simmilar 16x2 LCD ? That may be a fatal problem. >2) Do you connect the ceramic capacitors between the OSC pins of the PIC >and the 4.0Mhz crystal or between the 4.0Mhz crystal and ground. I have >seen pictures on websites where the caps are connected between POC OSC pins >and the Crystal, whereas the PIC datasheet shows it connected between the >Crystal and Ground ? www.dvanhorn.org/Micros/All/Crystals.php Short version, caps go from the xtal pins (or the xtal pins on the chip) to the chip's ground pin. Not some other ground point. >6) Is there any ticky things i have to keep in mind when writing my >software and controlling the Samsung KS0066U LCD driver? The init phase of the Hitachi chips is VERY picky. Do one thing slightly wrong, and nothing happens. :( When in doubt, add delays. You can't get in trouble adding a few extra uS between operations, but get it slightly too fast, and you're in trouble. It's a variation of the "Rope Rule".. Ropes come in two lengths only. Too long, and too short. I can fix "too long". :) -- Dave's Engineering Page: http://www.dvanhorn.org I would have a link to FINDU here in my signature line, but due to the inability of sysadmins at TELOCITY to differentiate a signature line from the text of an email, I am forbidden to have it. -- 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

>How is the 10K pot hooked up? The Pots hooked up middle pin to LCD Pin for Vee (3), one pin to Vss, the other to Vdd. Contrast works fine, LCD powers up as i can see faint blocks when i tweak the Pot. > > PIC RA3 => 470 Ohm Resistor => Red LED => Grounded. > > PIC RA4 => 470 Ohm Resistor => Red LED => Grounded. > >You can't do this. RA4 is open collector, meaning it can't source current. LOL.. that would explain it. > > PIC OSCI => Leg 1 of 4.0Mhz Crystal => 33pF Capacitor1 = > Ground. > > PIC OSCO => Leg 2 of 4.0Mhz Crystal => 33pF Capacitor2 = > Ground. > >What about /MCLR? Make sure to tie it to Vdd. Yeah, i forgot to mention that pin, it is infact connected to +5V (Vdd). Could i put a small button here to reset the PIC? >The datasheet is the correct way. It looks like you have it wired >correctly. Ok, sounds like im on the right path.. > > 3) I dont have access to an oscillascope and im not sure if the crystal is > > actually working, or if the PIC is actually running so i connected a > > resistor to RA4 to and LED, and added code to my program to light up the > > LED if the LCD is busy and it didnt come on at all. I then cut out all my > >See above comment about open collector output. Turn the LED around and use >Vdd instead of ground. BCF PORTA,4 will turn it on and BSF PORTA,4 will >turn it off. Thats a mighty fine idea, i will try that tonight. So being an Open collector it can take in current but not put out. You mentioned setting and clearing bits of port A... this is ofcourse with pic PortA pins RA3->4 set to input with the using TRIS register? > > 4) Is there an easy way of checking things are all working without having > > to spend hours wiring up a test circuit? > >A frequency counter to see if the clock is running. An oscilloscope would >really be handy; beg, borrow or steal one. That could mean a day trip to my uni to use their equipment.... i wonder if you can hire them.. hrmm.. -- 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

>>1) Is the KS0066U LCD driver compatible with the Hitachi driver for a >>simmilar 16x2 LCD ? > >That may be a fatal problem. I have only seen two types of LCD Drivers on common LCD panels, and that is either the Hitachi, or what i have on mine, a samsung. What others are available. {Quote hidden}>>2) Do you connect the ceramic capacitors between the OSC pins of the PIC >>and the 4.0Mhz crystal or between the 4.0Mhz crystal and ground. I have >>seen pictures on websites where the caps are connected between POC OSC pins >>and the Crystal, whereas the PIC datasheet shows it connected between the >>Crystal and Ground ? > >www.dvanhorn.org/Micros/All/Crystals.php >Short version, caps go from the xtal pins (or the xtal pins on the chip) to >the chip's ground pin. >Not some other ground point. I have only one power supply and one ground, connect everything that needs grounding to the -side of the power supply, everything that needs +5v to the +side of the power supply. My power supply is a home brew rectifier, could it be that because there is a delay between when i actually plug the supply in (because of the electrolytic capacitors i use) and when the pic gets fully powered up that the initialisation of the LCD Driver dies? >>6) Is there any ticky things i have to keep in mind when writing my >>software and controlling the Samsung KS0066U LCD driver? > >The init phase of the Hitachi chips is VERY picky. Aparently the samsung ones that i have do all that stuff themselves with the busy flag set, then clearing it when its all ready for its first command. >Do one thing slightly wrong, and nothing happens. :( Strange that lots of people claim to have theirs work first time, especially when connecting it to LPT ports for MP3 based projects. I have been trying to fix this for weeks with no success. I started with LPT ports and writing my own C program, and thought it more reliable to use fixed timing of a PIC, and thats where i am now still with no luck. >When in doubt, add delays. I will keep this in mind when updating my software. -- 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

> >I have only one power supply and one ground, connect everything that needs >grounding to the -side of the power supply, everything that needs +5v to >the +side of the power supply. My power supply is a home brew rectifier, >could it be that because there is a delay between when i actually plug the >supply in (because of the electrolytic capacitors i use) and when the pic >gets fully powered up that the initialisation of the LCD Driver dies? You should tie the xtal caps to the pic's ground pin. Homebrew power supplies can cause trouble too, if the voltage isn't within spec ALL the time during operation. Of course all power supplies ramp up, but once you're there, you should stay there all the time. Another place where even a cheap scope will help a lot. >Strange that lots of people claim to have theirs work first time, >especially when connecting it to LPT ports for MP3 based projects. I have >been trying to fix this for weeks with no success. I started with LPT ports >and writing my own C program, and thought it more reliable to use fixed >timing of a PIC, and thats where i am now still with no luck. The delays of doing it on a PC may be a key factor there. I rather doubt you can get the I/O lines to move as fast as in a microcontroller. -- Dave's Engineering Page: http://www.dvanhorn.org I would have a link to FINDU here in my signature line, but due to the inability of sysadmins at TELOCITY to differentiate a signature line from the text of an email, I am forbidden to have it. -- 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

Mr. Dylan. First of all remove PF capictors from crystal 4 Mhz and then check the lcd work. If still lcd does'nt work tell me I can give you the complete routines for lcd "HD44780" which is compatible of Hatachi. Kashif ali Dylan wrote: {Quote hidden}> Problem Description: > ------------------------------ > * I cannot get charecters to show on my LCD Screen. > * Im not sure if my circuit is correct in terms of an external clock for > the PIC. > * I have tried nearly every circuit diagram i have found on the net and the > matching software, which is for the hitachi controllers none of which seem > to work. > > Parts List: > --------------- > > - PIC16F84A Microcontroller. > - Ampire AC162B Revision E. 16x2 Backlit LCD screen > -> LCD Driver IC is a samsung KS0066U. > - 4.0 Mhz Crystal > - 2x 33pF Ceramic Capacitors > - 470 Ohm Resistor. > - LED > > Circuit Description: > --------------------------- > > PIC RB(0-7) Pins are connected to LCD (DB0-7) Pins. > PIC RA(0-2) Pins are connected to LCD RS, R/W, E Pins. > PIC Vss grounded. > PIC Vdd ===> 5v Powersource. > LCD Pin1 ===> Grounded. > LCD Pin2 ===> 5v Powersource. > LCD Pin3 ==> 10Kpot for contrast > PIC RA3 => 470 Ohm Resistor => Red LED => Grounded. > PIC RA4 => 470 Ohm Resistor => Red LED => Grounded. > PIC OSCI => Leg 1 of 4.0Mhz Crystal => 33pF Capacitor1 = > Ground. > PIC OSCO => Leg 2 of 4.0Mhz Crystal => 33pF Capacitor2 = > Ground. > > Questions: > ---------------- > 1) Is the KS0066U LCD driver compatible with the Hitachi driver for a > simmilar 16x2 LCD ? > > 2) Do you connect the ceramic capacitors between the OSC pins of the PIC > and the 4.0Mhz crystal or between the 4.0Mhz crystal and ground. I have > seen pictures on websites where the caps are connected between POC OSC pins > and the Crystal, whereas the PIC datasheet shows it connected between the > Crystal and Ground ? > > 3) I dont have access to an oscillascope and im not sure if the crystal is > actually working, or if the PIC is actually running so i connected a > resistor to RA4 to and LED, and added code to my program to light up the > LED if the LCD is busy and it didnt come on at all. I then cut out all my > LCD code and just put in code to turn PORTA HI->LOW to see if it would > light up the led and it didnt. I checked the pins of the PIC with a > multimeter, some were 5v, some were 0, but my port A wasnt changing as i > would have hoped. > > 4) Is there an easy way of checking things are all working without having > to spend hours wiring up a test circuit? > > 5) Has anyone managed to get an Apire LCD to display anything, if so do you > have a circuit diagram that i could look at? > > 6) Is there any ticky things i have to keep in mind when writing my > software and controlling the Samsung KS0066U LCD driver? > > Dylan. > > -- > 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

----- Original Message ----- From: "Dylan" <dylanKILLspamFAST.FUJITSU.COM.AU> To: <.....PICLISTKILLspam.....MITVMA.MIT.EDU> Sent: Tuesday, May 22, 2001 11:19 PM Subject: Re: [PIC]: PIC16F84A + Ampire AC162B Rev E LCD (KS0066U Driver) Wont Work? > >How is the 10K pot hooked up? > > The Pots hooked up middle pin to LCD Pin for Vee (3), one pin to Vss, the > other to Vdd. Contrast works fine, LCD powers up as i can see faint blocks > when i tweak the Pot. That sounds pretty good. The Hitachi types power up with dark blocks on the first line. If you are only getting faint blocks then possibly this Samsung type may need to be pulled below Vss (-voltage) to obtain best contrast. But if you get faint blocks you should still be able to see faint characters when it is working. > > > > > PIC RA3 => 470 Ohm Resistor => Red LED => Grounded. > > > PIC RA4 => 470 Ohm Resistor => Red LED => Grounded. > > > >You can't do this. RA4 is open collector, meaning it can't source current. {Quote hidden}> > LOL.. that would explain it. > > > > PIC OSCI => Leg 1 of 4.0Mhz Crystal => 33pF Capacitor1 = > Ground. > > > PIC OSCO => Leg 2 of 4.0Mhz Crystal => 33pF Capacitor2 = > Ground. > > > >What about /MCLR? Make sure to tie it to Vdd. > > Yeah, i forgot to mention that pin, it is infact connected to +5V (Vdd). > Could i put a small button here to reset the PIC? You would have to pull /MCLR up with a pullup and use a button to pull it to ground to do the reset > > >The datasheet is the correct way. It looks like you have it wired > >correctly. > > Ok, sounds like im on the right path.. > > > > 3) I dont have access to an oscillascope and im not sure if the crystal is > > > actually working, or if the PIC is actually running so i connected a > > > resistor to RA4 to and LED, and added code to my program to light up the > > > LED if the LCD is busy and it didnt come on at all. I then cut out all my > > > >See above comment about open collector output. Turn the LED around and use > >Vdd instead of ground. BCF PORTA,4 will turn it on and BSF PORTA,4 will > >turn it off. > > Thats a mighty fine idea, i will try that tonight. > So being an Open collector it can take in current but not put out. > You mentioned setting and clearing bits of port A... this is ofcourse with > pic PortA pins RA3->4 > set to input with the using TRIS register? No, they should still be set to output using the TRIS. Writing a zero to an output pin pulls it to ground. Writing to an input pin does nothing usefull. (in this case) > > > > 4) Is there an easy way of checking things are all working without having > > > to spend hours wiring up a test circuit? > > > >A frequency counter to see if the clock is running. An oscilloscope would > >really be handy; beg, borrow or steal one. > > That could mean a day trip to my uni to use their equipment.... i wonder if > you can hire them.. hrmm.. > > -- > 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

> > > >I have only one power supply and one ground, connect everything that needs > >grounding to the -side of the power supply, everything that needs +5v to > >the +side of the power supply. My power supply is a home brew rectifier, > >could it be that because there is a delay between when i actually plug the > >supply in (because of the electrolytic capacitors i use) and when the pic > >gets fully powered up that the initialisation of the LCD Driver dies? I want to see the code. > > You should tie the xtal caps to the pic's ground pin. Dave, I didn't know this could be a problem. I kind of figured that at four Mhz, the wavelength is so long that unless my ground was more than about 50 feet away from the pics ground that it couldn't hurt. Why does it need to be hooked so close to the pic's ground pin? > Homebrew power supplies can cause trouble too, if the voltage isn't within > spec ALL the time during operation. Of course all power supplies ramp up, > but once you're there, you should stay there all the time. > Another place where even a cheap scope will help a lot. When I am testing, I keep the power supply on all the time and just use a clip lead to hook it to the pic circuit. Instant ramp up. I haven't hurt anything yet. ;-) -- 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

I have seen the LIGHT!!!!!!!!!!!!!!!!!!!! ! See below... > > >See above comment about open collector output. Turn the LED around and >use > > >Vdd instead of ground. BCF PORTA,4 will turn it on and BSF PORTA,4 will > > >turn it off. > > > > Thats a mighty fine idea, i will try that tonight. > > So being an Open collector it can take in current but not put out. > > You mentioned setting and clearing bits of port A... this is ofcourse with > > pic PortA pins RA3->4 > > set to input with the using TRIS register? > >No, they should still be set to output using the TRIS. Writing a zero to an >output pin pulls it to ground. Writing to an input pin does nothing >usefull. (in this case) Ok, turned the LED around, added a 470 Ohm resistor, programmed the pic with the following code: LIST P=16F84A, F=INHX8M __CONFIG _CP_OFF & _WDT_OFF & _XT_OSC #include <d:\programs\mplab\p16f84a.inc> BUSY_BYTE EQU 0x0F START CALL SELECT_BANK1_RFM <-- this just selects as it says Bank 1 of the RFM. MOVLW 0x00 ; RB0-7 MOVWF TRISA CALL SELECT_BANK0_RFM START_LP CLRF LCD_CTRL ; Port A CALL Delay CALL Delay CALL Delay MOVLW 0xFF MOVWF LCD_CTRL ; Port A CALL Delay CALL Delay CALL Delay CALL Delay GOTO START_LP Delay clrwdt movlw d'255' movwf BUSY_BYTE loop decfsz BUSY_BYTE,F goto loop return END The Result is: My Red LED is flasshing away prominantly on my prototype board. IMPORTANT THINGS I HAVE DISCOVERED ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - The LED does NOT flash as smoothly as it probably should on and off. I figure its got something to do with the fact the time delays are very short. -HOWEVER, if i move my board around a bit it seems to change the speed of the LED flashing ??????? So perhaps my wires arent soldered on correctly or there is a huge amount of interference from the crystal on the wires causin th epic to run erratically ? - My wires are probably too thick for use with IC's, i have seen some circuits wired up with nice looking thin wires apposed to my thick wire strand covered in thick plastic covering. Which is better, thick or thin for this application. - This erratic behavour of my prototype board and pic would without a doubt stuff up my attempts to correctly control the Samsung LCD driver... bummer. Can anyone else think of things that cause this type of erratic, fast flashing, slow flashing to occur or the pic to run at variable speeds? -- 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

> > > You should tie the xtal caps to the pic's ground pin. > >Dave, >I didn't know this could be a problem. I kind of figured that at four Mhz, >the wavelength is so long that unless my ground was more than about 50 feet >away from the pics ground that it couldn't hurt. Why does it need to be >hooked so close to the pic's ground pin? It's not a wavelength thing. First, it's not nice to radiate tons of EMI. Second, if you make that ground return path longer than it needs to be, then any EMI that it intercepts can affect your crystal operation. When you're prototyping, you already have enough problems, no? >When I am testing, I keep the power supply on all the time and just use a >clip lead to hook it to the pic circuit. Instant ramp up. I haven't hurt >anything yet. ;-) No, and I wouldn't expect damage. However, a non-monotonic rise can leave you with the CPU or peripherals in a non-recoverable confused state, till power goes away again. As before, you already have enough problems. -- Dave's Engineering Page: http://www.dvanhorn.org I would have a link to FINDU here in my signature line, but due to the inability of sysadmins at TELOCITY to differentiate a signature line from the text of an email, I am forbidden to have it. -- 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 11:07 PM 5/23/01 +1000, Dylan Hillier wrote: >For those of you watching this thread, there was talk about grounding pins >earlier. > >I found some good information (although not directly related to PIC's) on >grounding and decoupline capacitors at the following URL: > >http://mechsys2.me.berkeley.edu/ME235/LabFiles/index.html > >Very interesting indeed. I didn't see anything specifically on grounding, but there is a lot of good info here. -- Dave's Engineering Page: http://www.dvanhorn.org I would have a link to FINDU here in my signature line, but due to the inability of sysadmins at TELOCITY to differentiate a signature line from the text of an email, I am forbidden to have it. -- 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

----- Original Message ----- From: "David VanHorn" <EraseMEdvanhornspam_OUTTakeThisOuTCEDAR.NET> To: <PICLISTspam_OUTMITVMA.MIT.EDU> Sent: Wednesday, May 23, 2001 11:43 AM Subject: Re: [PIC]: PIC16F84A + Ampire AC162B Rev E LCD (KS0066U Driver) Wont Work? > > > > > You should tie the xtal caps to the pic's ground pin. > > > >Dave, > >I didn't know this could be a problem. I kind of figured that at four Mhz, > >the wavelength is so long that unless my ground was more than about 50 feet > >away from the pics ground that it couldn't hurt. Why does it need to be > >hooked so close to the pic's ground pin? > > It's not a wavelength thing. > First, it's not nice to radiate tons of EMI. > Second, if you make that ground return path longer than it needs to be, > then any EMI that it intercepts can affect your crystal operation. > When you're prototyping, you already have enough problems, no? OK, I see. I wonder why they put the Vss pin on the other side of the chip then. How long is too long? 1cm? 1inch? Or, just make it as short as possible? > > > >When I am testing, I keep the power supply on all the time and just use a > >clip lead to hook it to the pic circuit. Instant ramp up. I haven't hurt > >anything yet. ;-) > > No, and I wouldn't expect damage. However, a non-monotonic rise can leave > you with the CPU or peripherals in a non-recoverable confused state, till > power goes away again. > As before, you already have enough problems. I probably should have been more clear. The circuit has a regulator with a big cap on the output. {Quote hidden}> -- > Dave's Engineering Page: http://www.dvanhorn.org > > I would have a link to FINDU here in my signature line, but due to the > inability of sysadmins at TELOCITY to differentiate a signature line from > the text of an email, I am forbidden to have it. > > -- > 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

> >OK, I see. I wonder why they put the Vss pin on the other side of the chip >then. How long is too long? 1cm? 1inch? Or, just make it as short as >possible? The practice that I follow, is that this track gets #1 priority to get to the CPU ground. The components are placed as close to the chip as practical. I don't allow any other "ground" connection to the caps, or ground lead of a resonator. >I probably should have been more clear. The circuit has a regulator with a >big cap on the output. That's probably monotonic then. Another gotcha is rise time. The on-chip brownout detectors often have a spec for minimum risetime. Too much C, and the chip isn't guaranteed to wake up. Isn't it fun? :) -- Dave's Engineering Page: http://www.dvanhorn.org I would have a link to FINDU here in my signature line, but due to the inability of sysadmins at TELOCITY to differentiate a signature line from the text of an email, I am forbidden to have it. -- 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

>I probably should have been more clear. The circuit has a regulator with a >big cap on the output. How big a cap? Part of your problem may be that the cap is too big, and because the regulator goes into current limit the rise time becomes too slow. I would say that a cap of 10 to 100uF would be heaps. The other gotcha that you should be aware of is when disconnecting the power. 7800 series regulators do not like having volts on the output, and no volts on the input. Other regulators also have this problem. If you have a capacitor on the output that will hold up the voltage for a significant time after input power is removed it pays to put a diode across the output to input of the regulator (reverse biased in normal operation) so the input is never less than 0.6V below the output. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.