Post by thread:[PIC]: Free ICD issues

Hi all. I guess this question was answered before, but I can't seem to find the answer in the archive, so here it goes. I built the ICD circuit from http://stolz.de.be/. I can't tell if it works or not. I loaded the BL010101.hex file from the ICD2 folder in MPLAB, to a F877-20 (non-A), through NOPPP for windows. For testing, I bridged MCLR to 5v, but the crystal refuses to oscillate. As in that site, I'm using a 20MHz crystal with 22pF caps. If I pull out the 877 and make a test circuit (Just MCLR/VDD/VSS/XTAL), using a 4MHz Xtal it works, 4.0000000092 MHz, according to my counter. I don't have another 20MHz to test it (and i dont want to mess with the board i made because it's just a universal board and it breaks easily). Looking at the project in piclist.org, I see it uses a 3.6864MHz crystal. So what's the deal? Should I use 3.6864 (as in piclist.org) or 20 (as in stolz.de.be)? Is the firmware (Bl010101.hex) for 3,6864 or 20? Is the guy at stolz.de.be using other firmware? Regards, hjf

I tried connecting it but I don't know if my MAX232 even works because I got it from my junkbox. I have the +5V in /MCLR from the RS-232, but no connection to MPLAB. I'm pretty sure my circuit won't work the first time but I can't even get oscillation from the xtal (OK maybe my scope's probe capacitance is screwing up the xtal but I don't know. I think i should have SOME signal, right? all i get is DC. I'm using "plate" capacitors (the tiny ones, gray with a black dot on top), 22pF. hjf On Thu, 20 Jan 2005 12:24:38 -0800, Al Dynarski <spam_OUTaldTakeThisOuTsonic.net> wrote: > I built the stolz ICD2 and it works great. I used a 16F877 (non-A) with a > 20MHz crystal and 22pf caps. I burned BL010101.hex into the 877, powered it > up, set the RS232 parameters per MPLab's instructions and was able to get a > 'Connect' confirmation from MPLab. Never having used an ICD2 before, what > wasn't clear to me was that you must then manually 'Download the ICD2 > Operating System' from MPLab, at least this first time. Once I figured that > out it works as advertised. > > -Al

I've had some luck picking up oscillation by using a wire or loop of wire on my scope probe held near the crystal. That way you aren't messing up the circuit at all. It doesn't take much, just a few mV, to rule out failed oscillation as your problem, and you should see that if it's actually there. Mike H. > I tried connecting it but I don't know if my MAX232 even works because > I got it from my junkbox. I have the +5V in /MCLR from the RS-232, but > no connection to MPLAB. I'm pretty sure my circuit won't work the > first time but I can't even get oscillation from the xtal (OK maybe my > scope's probe capacitance is screwing up the xtal but I don't know. I > think i should have SOME signal, right? all i get is DC. > I'm using "plate" capacitors (the tiny ones, gray with a black dot on > top), 22pF. > > hjf

----- Original Message ----- From: "Mike Hord" <.....mike.hordKILLspam@spam@gmail.com> Subject: Re: [PIC]: Free ICD issues > I've had some luck picking up oscillation by using a wire or loop of > wire on my scope probe held near the crystal. You beat me to it. I was going to suggest the same thing. My next move would be to yank out the caps. I've never seen a PIC fail to oscillate without caps unless the PIC or crystal was toast, but the description of the caps sounds a little suspect, and it sounds like we've ruled out the PIC. The crystal is a possibility. Those things are subject to damage if you have a heavy foot on the slobbering iron. --McD

i have an excellent soldering station and I'm pretty good at soldering. but I don't know, shit happens and this may be the case. these caps, i've been using them since I first started with PICs because these have "zero i don't know what". bias or coefficient or something like that, and that makes them ideal for oscillator. i think it's temperature coefficient. Ok, time to get my hands dirty and dig deep in the junkbox,in search of another 20Mhz hjf On Thu, 20 Jan 2005 16:18:05 -0500, John J. McDonough <mcdKILLspamis-sixsigma.com> wrote: > {Original Message removed}

ok. i'm officially lost now. i replaced the 20mhz xtal and it doesn't oscillate either. a 4mhz one does, but if the code is for 20MHz then it won't work, right?. so what I dont understand is why both circuits (stolz's and piclist.org's) use different xtals but the same software hjf On Thu, 20 Jan 2005 18:34:43 -0300, Hernán Freschi <.....drgenioKILLspam.....gmail.com> wrote: {Quote hidden}> i have an excellent soldering station and I'm pretty good at > soldering. but I don't know, shit happens and this may be the case. > these caps, i've been using them since I first started with PICs > because these have "zero i don't know what". bias or coefficient or > something like that, and that makes them ideal for oscillator. i think > it's temperature coefficient. > Ok, time to get my hands dirty and dig deep in the junkbox,in search > of another 20Mhz > > hjf > > On Thu, 20 Jan 2005 16:18:05 -0500, John J. McDonough > <EraseMEmcdspam_OUTTakeThisOuTis-sixsigma.com> wrote: > > {Original Message removed}

ok now I loaded the hex with fpp and it says config word: 3F3A. that is 11111100111010. according to the data sheet, the last two bits (10 in this case) set the osc type. and 10 is HS osc. so I'm in a dead end again. the 877 refuses to oscillate even in a "test circuit" (just mclr/vdd/vss and the xtal). with 4mhz, it oscillates perfectly. this is driving me crazy! it's a -20 part. i've run -4 F84's at 20MHz with no problems... hjf On Thu, 20 Jan 2005 23:22:18 -0300, Hernán Freschi <drgeniospam_OUTgmail.com> wrote: {Quote hidden}> it may be. but i doubt microchip would make such a mistake, i'm using > their bootloader. and, i don't have the .asm to recompile, and my > programmer (noppp) won't let me mess with the config fuses. > > hjf > > > Could be that XT type oscillator is chosen in configuration word? > > For 20MHz it should be HS. > > > > Regards, > > Oliver >

Let's take it one step at a time: Has the MCLR line been released (placed at 5V)? Is there a bypass cap across VCC and GND? Something like 100N ? have you connected ALL the GND pins to GND and all the VCC pins to VCC? Have you installed a test program, so that some of the pins are driven hi then low, so you can see activity on a scope? Make PB0 and PB1 toggle, watch 'em on a scope. At 20Mhz, its VERY hard to be certain that the oscillator is running. Hold an FM radio near the PIC. Does it place a signal ("birdie") on the FM band? How about the caps across the crystal to force a good startup. They should be about 22-27pF. Are they there? Is the GND leg actually GROUNDED? There is ALWAYS a reason why this stuff doesn't work. Let us know. We want to help. --Bob Hernán Freschi wrote: {Quote hidden}>ok now I loaded the hex with fpp and it says config word: 3F3A. that >is 11111100111010. according to the data sheet, the last two bits (10 >in this case) set the osc type. and 10 is HS osc. so I'm in a dead end >again. the 877 refuses to oscillate even in a "test circuit" (just >mclr/vdd/vss and the xtal). with 4mhz, it oscillates perfectly. this >is driving me crazy! it's a -20 part. i've run -4 F84's at 20MHz with >no problems... > >hjf > >On Thu, 20 Jan 2005 23:22:18 -0300, Hernán Freschi <@spam@drgenioKILLspamgmail.com> wrote: > > >>it may be. but i doubt microchip would make such a mistake, i'm using >>their bootloader. and, i don't have the .asm to recompile, and my >>programmer (noppp) won't let me mess with the config fuses. >> >>hjf >> >> >> >>>Could be that XT type oscillator is chosen in configuration word? >>>For 20MHz it should be HS. >>> >>>Regards, >>>Oliver >>> >>> > > > -- Note: To protect our network, attachments must be sent to KILLspamattachKILLspamengineer.cotse.net . 1-866-263-5745 USA/Canada http://beam.to/azengineer

I've had similar issues before and the problem usually turned out to be related to board capacitance and RFI type problems. Are you using a breadboard? If so, try lifting the two oscillator pins and soldering the crystal and caps directly to the pins. Another alternative is perhaps the crystal is being overdriven and you need a series resister. AFAIR, the oscillator pins are not as well protected from static electricity as the other pins of the chip are, and due to the nature of the oscillator circuit are more sensitive. If you haven't been taking good static precautions perhaps you've moderately damaged one of the pins. Some limited types of damage could cause a circuit to function under some circumstances and not under others (such as 4MHz vs 20 MHz). Lastly, make sure you have adequate caps across the power pins of the PIC. Hope this helps, I certianly understand how frustrating this kind of problem can be. If nothing else works you should be able to purchase a 20MHz canned oscillator and drive the input directly instead of having the chip drive the crystal. I'm sure it would be instructive to find the real problem rather than patching it, though. Good luck! -Adam Hernán Freschi wrote: {Quote hidden}>ok now I loaded the hex with fpp and it says config word: 3F3A. that >is 11111100111010. according to the data sheet, the last two bits (10 >in this case) set the osc type. and 10 is HS osc. so I'm in a dead end >again. the 877 refuses to oscillate even in a "test circuit" (just >mclr/vdd/vss and the xtal). with 4mhz, it oscillates perfectly. this >is driving me crazy! it's a -20 part. i've run -4 F84's at 20MHz with >no problems... > >hjf > >On Thu, 20 Jan 2005 23:22:18 -0300, Hernán Freschi <RemoveMEdrgenioTakeThisOuTgmail.com> wrote: > > >>it may be. but i doubt microchip would make such a mistake, i'm using >>their bootloader. and, i don't have the .asm to recompile, and my >>programmer (noppp) won't let me mess with the config fuses. >> >>hjf >> >> >> >>>Could be that XT type oscillator is chosen in configuration word? >>>For 20MHz it should be HS. >>> >>>Regards, >>>Oliver >>> >>> > >

Have you verified that the config word really IS programmed correctly? Just because the hex file has a config value doesn't mean it was programmed. The fact that you get oscillation with a 4 Mhz crystal suggests that you could be in R/C mode. It can't hurt to pull the crystal and see if you are oscillating with the 27pF caps. Add a few hundred pf and you'll know for sure that you're really in RC mode. And don't forget that all unused pins that aren't set to output must be pulled high or preferably, low. If your config word is not set you'll be in LVP mode if RB3 floats high. And if you're using a breadboard you may have too much capacitance for the 20Mhz crystal. Try connecting the xtal directly to the pins by bending them up and soldering directly. Be careful about static damage if you do this. R Bob Axtell wrote: {Quote hidden}> Let's take it one step at a time: > > Has the MCLR line been released (placed at 5V)? > > Is there a bypass cap across VCC and GND? Something like 100N ? have you > connected ALL the GND pins to GND and all the VCC pins to VCC? > > Have you installed a test program, so that some of the pins are driven > hi then low, > so you can see activity on a scope? Make PB0 and PB1 toggle, watch 'em > on a scope. > > At 20Mhz, its VERY hard to be certain that the oscillator is running. > Hold an FM > radio near the PIC. Does it place a signal ("birdie") on the FM band? > > How about the caps across the crystal to force a good startup. They > should be about > 22-27pF. Are they there? Is the GND leg actually GROUNDED? > > There is ALWAYS a reason why this stuff doesn't work. > > Let us know. We want to help. > > --Bob > > Hernán Freschi wrote: > >> ok now I loaded the hex with fpp and it says config word: 3F3A. that >> is 11111100111010. according to the data sheet, the last two bits (10 >> in this case) set the osc type. and 10 is HS osc. so I'm in a dead end >> again. the 877 refuses to oscillate even in a "test circuit" (just >> mclr/vdd/vss and the xtal). with 4mhz, it oscillates perfectly. this >> is driving me crazy! it's a -20 part. i've run -4 F84's at 20MHz with >> no problems... >> >> hjf >> >> On Thu, 20 Jan 2005 23:22:18 -0300, Hernán Freschi <spamBeGonedrgeniospamBeGonegmail.com> >> wrote: >> >> >>> it may be. but i doubt microchip would make such a mistake, i'm using >>> their bootloader. and, i don't have the .asm to recompile, and my >>> programmer (noppp) won't let me mess with the config fuses. >>> >>> hjf >>> >>> >>> >>>> Could be that XT type oscillator is chosen in configuration word? >>>> For 20MHz it should be HS. >>>> >>>> Regards, >>>> Oliver >>>> >> >> >> >> >

Hernán Freschi wrote : > i'm using their [Microchip] bootloader... I might be missing something, but have you mentioned anything about a bootloader before in this thread ? Have you been using that bootloader for any other project ? Does it have any limitations (some bootloaders put some restrictions on the application) that might be a problem for this particular application ? > and, i don't have the .asm to recompile, The ASM for the bootloader or for the application ? > and my > programmer (noppp) won't let me mess with the config fuses. You should not "mess" with them. You should verify they are correct and, if needed, set them correctly. Have you been using noppp to program the 877A with any other application (and it worked) ? Hernán Freschi also wrote : > ok now I loaded the hex with fpp and it says config word: 3F3A. that > is 11111100111010. according to the data sheet, the last two bits (10 > in this case) set the osc type. and 10 is HS osc. so I'm in a dead end > again. And what about the other config settings ? Have you verified those ? > the 877 refuses to oscillate even in a "test circuit" (just > mclr/vdd/vss and the xtal). "Just" ??? No caps for the crystal ? No decoupling caps between vdd/vss ? There is no reason at all to describe your curcuit if you leave half of it out anyway. > i replaced the 20mhz xtal and it doesn't oscillate either. Have these 20 Mhz xtals ever been working with *any* PIC ? If so, you can probably rule out them as a problem. Regards, Jan-Erik.

OK, to recap: Browsing PICLIST.org, I found out about an interesting tool, called the ICD. So I read about it and it seemed easy to build. Then, at the bottom of the page, I found a link to a website, stolz.de.be. I followed the instructions in the site. I built the circuit in a drilled board and tested it. First, I loaded the file BL010101.HEX, from the Microchip\MPLAB IDE\ICD2 folder. Programmed it into a PIC16F877-20/P. Then I put it in the board I just made. Powered it up, connected it to the serial port. I have 5,1V in the /MCLR pin, 4,98V at VDD and 0 at vss. I also clamped LVP to ground through a 2k2 resistor. Fired up MPLAB, chose the right serial port, disabled FIFO buffers, and tried to connect. Of course it didn't work the first time. So I started debugging it. I checked all the resistors, transistors, diodes, etc. And it all seems OK. There were only 2 pins wrong in the MAX232A (they were reversed). I solved that and re tested. It didn't work either. So I went back to the lab, and put the circuit to a power supply. /MCLR has 0V because it takes the 5V from the serial port. So I tied it, through the resistor and zener to 12V (just like the serial port would do). Again I have 5,1V. OK. Checked voltages: all OK. Scope to the osc pins: nothing. Just DC. So I decided to build a little test circuit. Breadboard. MCLR to VDD (5V), VSS (0v), LVP (0V), 100nF between VDD & VSS. 20MHz xtal. 18pF caps. Power it up, and then nothing. Scope shows nothing. So I take the XTAL out. Put in a 4MHz and check it: oscillates perfectly, 4.000009 MHz. Remove the caps to see what happens. Keeps oscillating, but now at 4.000100 MHz. Put a 20MHz. Doesn't work. Put in another 20MHz. Doesn't work either. Tried with 22pF and 27pF caps. Nothing either. 100R Resistor didn't work either. So i go back to the computer. Read the PIC. Config word is OK (HS osc). Program is OK. I know the xtals work because I used them with other circuits before. any ideas? hjf

Y'know, it occured to me, perhaps the difference in xtal values (20MHz and that 3.X one) could be because you're looking at an ICD and an ICD2, which are different? Just a theory. FWIW I haven't gotten my Stolz version working yet either, but that's because I only have the A version of the chip, and at the moment I haven't had time to hook up my programmer that will handle the A (we use older programmers here because that' s all that's needed). Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams On Fri, 21 Jan 2005 10:31:43 -0300, Hernán Freschi <TakeThisOuTdrgenioEraseMEspam_OUTgmail.com> wrote: > OK, to recap: > > Browsing PICLIST.org, I found out about an interesting tool, called > the ICD. So I read about it and it seemed easy to build. Then, at the > bottom of the page, I found a link to a website, stolz.de.be. I

well so I now finally have communication with the ICD2. I can see the voltages and it passes the self tests. But I have a problem when trying to download the operating system. It goes on a little and then MPLAB just exits without previous warning. Any ideas? I'm using XP SP2 on an Athlon XP 2200. hjf On Fri, 21 Jan 2005 11:47:46 -0300, Hernán Freschi <RemoveMEdrgenioTakeThisOuTgmail.com> wrote: {Quote hidden}> OK I just downloaded the firmware again, just for the heck of it. And got this: > > MPLAB ICD 2 Ready > Connecting to MPLAB ICD 2 > ...Connected > ICD0082: Failed MPLAB ICD 2 operation > > some progress finally! > now I have to check what that error is > > hjf >

seems like I found the problem. I was using a very long cable (5m). I found a shorter cable (2m) and now it works. hjf On Fri, 21 Jan 2005 13:46:42 -0300, Hernán Freschi <drgenioEraseME.....gmail.com> wrote: {Quote hidden}> well so I now finally have communication with the ICD2. I can see the > voltages and it passes the self tests. But I have a problem when > trying to download the operating system. It goes on a little and then > MPLAB just exits without previous warning. > Any ideas? I'm using XP SP2 on an Athlon XP 2200. > > hjf > > On Fri, 21 Jan 2005 11:47:46 -0300, Hernán Freschi <EraseMEdrgeniogmail.com> wrote: > > OK I just downloaded the firmware again, just for the heck of it. And got this: > > > > MPLAB ICD 2 Ready > > Connecting to MPLAB ICD 2 > > ...Connected > > ICD0082: Failed MPLAB ICD 2 operation > > > > some progress finally! > > now I have to check what that error is > > > > hjf > > >

You mean the RJ12 cable was over 2 METERS long? I never allow mine to be longer than 18 inches. --Bob Hernán Freschi wrote: {Quote hidden}>seems like I found the problem. I was using a very long cable (5m). I >found a shorter cable (2m) and now it works. > >hjf > >On Fri, 21 Jan 2005 13:46:42 -0300, Hernán Freschi <RemoveMEdrgenioEraseMEEraseMEgmail.com> wrote: > > >>well so I now finally have communication with the ICD2. I can see the >>voltages and it passes the self tests. But I have a problem when >>trying to download the operating system. It goes on a little and then >>MPLAB just exits without previous warning. >>Any ideas? I'm using XP SP2 on an Athlon XP 2200. >> >>hjf >> >>On Fri, 21 Jan 2005 11:47:46 -0300, Hernán Freschi <RemoveMEdrgeniospam_OUTKILLspamgmail.com> wrote: >> >> >>>OK I just downloaded the firmware again, just for the heck of it. And got this: >>> >>>MPLAB ICD 2 Ready >>>Connecting to MPLAB ICD 2 >>>...Connected >>>ICD0082: Failed MPLAB ICD 2 operation >>> >>>some progress finally! >>>now I have to check what that error is >>> >>>hjf >>> >>> >>> > > > -- Note: To protect our network, attachments must be sent to RemoveMEattachTakeThisOuTspamengineer.cotse.net . 1-866-263-5745 USA/Canada http://beam.to/azengineer

no, the RS-232 cable. afaik, rsr232 allows up to 15M On Fri, 21 Jan 2005 13:07:24 -0700, Bob Axtell <EraseMEengineerspamspamBeGonecotse.net> wrote: {Quote hidden}> You mean the RJ12 cable was over 2 METERS long? > I never allow mine to be longer than 18 inches. > > --Bob > > Hernán Freschi wrote: > > >seems like I found the problem. I was using a very long cable (5m). I > >found a shorter cable (2m) and now it works. > > > >hjf > > > >On Fri, 21 Jan 2005 13:46:42 -0300, Hernán Freschi <RemoveMEdrgenioKILLspamgmail.com> wrote: > > > > > >>well so I now finally have communication with the ICD2. I can see the > >>voltages and it passes the self tests. But I have a problem when > >>trying to download the operating system. It goes on a little and then > >>MPLAB just exits without previous warning. > >>Any ideas? I'm using XP SP2 on an Athlon XP 2200. > >> > >>hjf > >> > >>On Fri, 21 Jan 2005 11:47:46 -0300, Hernán Freschi <drgenioSTOPspamspam_OUTgmail.com> wrote: > >> > >> > >>>OK I just downloaded the firmware again, just for the heck of it. And got this: > >>> > >>>MPLAB ICD 2 Ready > >>>Connecting to MPLAB ICD 2 > >>>...Connected > >>>ICD0082: Failed MPLAB ICD 2 operation > >>> > >>>some progress finally! > >>>now I have to check what that error is > >>> > >>>hjf > >>> > >>> > >>> > > > > > > > > -- > Note: To protect our network, > attachments must be sent to > spamBeGoneattachSTOPspamEraseMEengineer.cotse.net . > 1-866-263-5745 USA/Canada > http://beam.to/azengineer > >

I think an important point might be that the ICD2 is using some RS232 signals for uses other than that which they were designed for. For instance, MCLR of the ICD2 PIC is driven using one of the RS232 pins (I can't remember which). Perhaps voltage drop on this and other pins caused a problem? Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams On Fri, 21 Jan 2005 18:42:53 -0300, Hernán Freschi <KILLspamdrgeniospamBeGonegmail.com> wrote: > no, the RS-232 cable. > afaik, rsr232 allows up to 15M

at least the /MCLR had 5,1V when I measured it, even with the long cable. Maybe, if that's the problem, it can be solved just by adding a buffer (transistor, etc). But anyway, my ICD stopped working. It was working perfectly and then bam, can't connect. I'll keep trying hjf On Fri, 21 Jan 2005 17:47:12 -0500, Josh Koffman <EraseMEjoshybearEraseMEgmail.com> wrote: > I think an important point might be that the ICD2 is using some RS232 > signals for uses other than that which they were designed for. For > instance, MCLR of the ICD2 PIC is driven using one of the RS232 pins > (I can't remember which). Perhaps voltage drop on this and other pins > caused a problem? > > Josh

well it's no longer working. I don't know what happened to it. Suddenly, it's not responding anymore. damn. hjf On Fri, 21 Jan 2005 21:32:24 -0300, Hernán Freschi <spamBeGonedrgenioKILLspamgmail.com> wrote: {Quote hidden}> i'd be glad to add this info, when I get this little thing working > 100%. as of now, it's not working :(. i'll keep a log of events here. > should I relabel this topic? > > hjf > > On Fri, 21 Jan 2005 15:54:25 -0800, James Newtons Massmind > <.....jamesnewtonspam_OUTmassmind.org> wrote: > > Is there any of your experience that would be valuable to others? Could I > > ask you to add that to the page at piclist.com? Use the little form at the > > bottom of the page... > > > > --- > > James. > > >

OK so, I've been working on this since wednesday and it still doesn't work. It worked for a little while. I programmed a PIC, debugged some code, all was great and then bam, no more. No matter what I try, it just won't work. I even made a PCB for it, and no luck. I took it to the lab and with the scope in the serial line I tried to debug it and found some interesting things. The serial port does send the reset. The PIC gets the first messages and answers them and the computer hears it and tries (because I see lots of activity in the computer's TXD when the PIC is on, but not when it's off. But maybe that can be because of the serial control lines. One thing that I noticed is that the pic strobes one bit constantly at a low frequency (I forgot to measure it). Seems strange. Everything else is OK. The firmware is loaded OK. The pic is a 16F877-20/P. So unless I'm using the wrong firmware, I don't know what the problem is :(. I'm using MPLAB 7, windows 98 (on the laptop) and XP on the desktop. hjf On Sat, 22 Jan 2005 00:37:07 -0300, Hernán Freschi <TakeThisOuTdrgenio.....TakeThisOuTgmail.com> wrote: {Quote hidden}> well it's no longer working. I don't know what happened to it. > Suddenly, it's not responding anymore. damn. > > hjf > > On Fri, 21 Jan 2005 21:32:24 -0300, Hernán Freschi <TakeThisOuTdrgenioKILLspamspamgmail.com> wrote: > > i'd be glad to add this info, when I get this little thing working > > 100%. as of now, it's not working :(. i'll keep a log of events here. > > should I relabel this topic? > > > > hjf > > > > On Fri, 21 Jan 2005 15:54:25 -0800, James Newtons Massmind > > <.....jamesnewtonRemoveMEmassmind.org> wrote: > > > Is there any of your experience that would be valuable to others? Could I > > > ask you to add that to the page at piclist.com? Use the little form at the > > > bottom of the page... > > > > > > --- > > > James. > > > > > >