Thanks Jari,Just what I was looking for. Found all kinds of sites for monostable and astable circuits but not for this purpose! Great! :-)Regards,

Oh well, :-newburnSpent the last couple of days trying to find any fault on the COM board and now I've tried everything I can think of.- Measured all ground connections. 0 Ohm resistance between all points.- Changed both voltage regulators. No change in voltage.- Changed MAX232 and NE555 circuits. No change.- Detached transistors, MAX and 555. No change in voltage.I'm lost here... :-( I can't seeem to find any shortsI can't find any errors in the groundingI can't find unexpected resistancesI still have 4.5V from the 4805. Board stripped down essentially.With the NE555 attached and TR1 detached I have 13.0V on the output pin on the 78L12 in NORM mode, but as I change to PROG mode the voltage drops to 12.5V. Measured no unexpected resistances!!!I'm going INSANEEE!!!!!! :-ukliam

Ok, don't give up, yet . ;)See my experience:http://www.fsnordic.net/discussion/index.p...;threadid=10610Dirk is right, just try it with another PC.Regards,Pedro

Mats,Don't give up. A touch of insanity can make it easier to meet the world on its own terms!Was the key card connected when you made your earlier measurements?You may have several things going on at the same time.The 7805 producing only 4.5 volts is certainly one. Possible causes:1. Input has a large AC component. 7805 can't regulate when input voltage drops too low.2. Output has large current load (reversed cap perhaps, or something on connected board)3. 7805 is defective.4. R1 , R2 wrong value.5. ?Your 555 voltage pump seems to be working. The low output voltage may be a result of a low input voltage. The 78XX series of regulator typically must have 2~2.5 volts more on the input than on the output. If you expect 13.5 out you may need as much as 16 volts in. This may be at the limit of the circuit's performance on 8.5 volts in. If you are sure your 9 volt supply is producing good DC, you can get a little boost by shorting out the input protection diode D4. This should get you back to 9 volts going into the 555.Measuring the input and output voltages on the 78L12 in flash mode would be of interest.Mikewww.mikesflightdeck.com

If you'll give up finally, send the two boards for me for check up if you dare.I promise you you will get it back.

Pedro, Mike and KozakThanks for your replies. They made me realize I wasn't quite finished with this...lolPedro, I checked the link you gave me. Good stuff. I think I've come to stage: Change computer.Mike, here's some for ya; I actually tested the 7805 without anything, connecting to 9Vdc and still got 4.5V out. Changed to a new one and still the same. So either it gives 4.5 or my multimeter is way off, probably the latter as it was pretty cheap :-) Or it could be a bad DC i'm inputting into the board. I have a standard battery eliminator 9Vdc, 1A, 9VA max. Thinking of buying me one of them Mascots. Can't one build ones own power supply unit?No reversed caps or wrong valued resistors. I tried with a 12Vdc battery eliminator as well and that raised the MCLR signal to 13.5V To the PIC I still had 4.0V on pin 12 and 13 and 4.5V on pin 14. With the 12V p/s I got alot of volts on the IN pin on the 78L12, and 13.5V on OUT. I then did a test erasing and writing and I noticed it wrote two positions (words?) I got 3003 and 01B9 for the first two. I then Erased the PIC and all was back tp 3FFF. So apparently it works but not very good to put it nicely ;-) Tried this a dozen times and some of those times it managed to write those two first words.I then changed to my old 9V p/s and did the ERASE/READ/WRITE/READ test again. Same result as with the 12V p/s. SO DOES THIS MEAN IT IS WORKING? Despite the fact I only have 13.5 IN on the 78L12 and 12.5 OUT. 4V on PIC pin 12 and 13 and 4.5V on 14. Actually I read that the MCLR has to be 7.25V higher than Vdd so that means you could escape with 12.5V. Any comments on that? It seems as it's not going the whole way through though. To this I can add I use an old P2 300MHz machine, and I have tried an 800Mhz to no avail. So I guess I have to try it on the main computter as it looks as the speed has a lot of significance when writing. Isn't there something one could do about the computer problem?Kozak, thanks for your offer, I'll remember that next time. ;-) But for now I'm on the move again. Again, thanks guys for keeping my spirit high about this! :-)Regards,

Mats,I guess it is possible that there is some peculiarity about the computer causing the problem. However, the low +5 supply continues to attract my attesion. I took a look at the Microchip family guide for the mid range PICs. 4.5 volts is the minimum voltage allowed for running the internal oscillator in HS mode. This is the mode normally used with an external crystal. Perhaps this is not a problem, but in flakey operation problems it is often a good idea to eliminate the power supply as a possible source of the problem before chasing other possibilities.First, though, you have mentioned that your meter may be in error. Checking that seems a good idea. Some possibilities: a fresh single cell alkaline battery has a terminal voltage slightly higher than 1.5 volts. I just measured four brand new ones at 1.60 volts no load.Another possible check of your voltmeter is to measure the voltage drop across a silicon diode. Connect a 4700 ohm resistor to the +9 volt terminal of your supply. Connect the other end to non-banded end of a 1N4148 diode. Connect the banded end to the other terminal on your power supply. Measuring the voltage across the diode should give you very close to 0.60 volts.Are you using a digital voltmeter?If your meter is reading low, your circuitry may be operating as designed and the solution may in fact require using a different computer (or serial ifc card??).One the other hand, if your meter appears to be functioning properly, making sure the 5 volt supply is providing a solid 5 volts would IMHO be a good thing.And yes, it is quite possible to make a power supply yourself.Mikewww.mikesflightdeck.com

Hi Mats,Dont give up yet because you are very close to solving this.First of all, Mike is right. You need to figure out your +5V from the regulator. Because the chips might work with 4.5V but will be very unstable and unreliable.Try this,Just build the circuit below on a proto board or wire it up with only the components shown and connect your 9V supply. Be sure to use Polarized electrolytic capacitors for 220 & 10uf caps in the circuit. There are Non polarized electrolytic caps so make sure you are using the correct ones.Are you getting +5V out? if yes, then check the other components on your COMII board.If the answer is no, then remove the 10uf cap and measure for 5VThen measure without 220uf, and next without the 100nf cap. See if anything cahnges. If not, your power supply and or the regulator is defective. Idea here is to ensure that you have at least 5VDC out from the regulator.You wrote;"I then did a test erasing and writing and I noticed it wrote two positions (words?) I got 3003 and 01B9 for the first two. I then Erased the PIC and all was back tp 3FFF. So apparently it works but not very good to put it nicely ;-) Tried this a dozen times and some of those times it managed to write those two first words."I had the same problem. I built my com board on a pcb as a prototype. I did not use the 555 timer charge pump circuit, but instead, used a 12v (actally 13.8V), DC power supply and connected it directly to the 5v regulator and then directly connected the 13.8V (w/o a 12V regulator), for the programming voltage. I did use all filter capacitors in the circuit. Then the fun began. I tried flashing it using a PII 300MHZ IBM 390E laptop running WIN2K.Rread after Write error every single time, but it did manage to write the first two memory locations, but nothing else. Tried all possible combinations of COM port settings from the lap top but no improvements.Then, I used a DELL Dimension P4 2.66 GHZ DESKTOP running WIN2K. Viola!It worked each and every single time.Next I tried flashing using a DELL Inspiron 8000 P4 1.8 GHZ laptop running WIN XP. Same results. Worked beautifully. I even connected a 25 foot serial cable from this laptop to my home brewed COMIII board and it still flashed without any problems. So try a different PC and see what happens. In the meantime, I will try to figure out how to get it to flash with my IBM PII 300MHZ LAPTOP.Anyhow, try with another PC and let us know what happens.Good luck and hang in there!Sean W.

Good replies from others. I don't have much to add.But, one reason might be broken capacitor what drops your voltage little bit. Also that 9 volt is not pure DC. Building requlator chip to other board is good idea. Also check voltages without any other cards. Just plain COM-board.And i have noticed same about computer speed. Flashing is critical operation for timing. Some 300MHz pentium is not fast enough for flashing. Also some motherboards have slow UART's. Try faster computer. I have even found problems with my 800MHz Dell. AMD XP1800+ usually flashes everything just fine.- J

>And i have noticed same about computer speed. Flashing is>critical operation for timing. Some 300MHz pentium is not fast>enough for flashing. Also some motherboards have slow UART's.>Try faster computer. I have even found problems with my 800MHz>Dell. AMD XP1800+ usually flashes everything just fine.We also had a problem here, trying to flash on a (even moderately fast) computer with Windows98 failed with the "read after write" error, but using a laptop with Windows2000 worked nicely. I think a bunch of people have said that they had issues with flashing on Windows98.If you can try on the fast one, give it a go, and if you can try somewhere with a Windows2000 / XP, it might be worth a try also. You only need to flash the PIC once anyway, so whatever computer works is OK.Tuomas

Hi!I think there are more clever people in electronics than me so I have only one idea:Maybe this thread shold be put on www.fsbus.de for those people who have problem with their comIII.It could be a selected zip file, downloadable.Hope you gonna find the error on your board finally.

Hi all, Sorry for being away for long but I just had twin babies!!:-)I had about 5 different types of problems in flashing the PIC - maybe this info will help:I successfully flashed with both W98 and W2K and with a PII and a PIII machine. I also successfully flashed with a laptop. Also, very important to note that I also get only about 4.6 volts from the 7805, so I cant think that could be the problem... I use a 1000 mAmp 9 volt supply for my board.My basic problems in flashing were:1. The RS232 cable: I first did not use a ribbon cable, but instead soldered individual wires to the connectors. Length was short, but was not able to flash with it. When I changed to the flat cable, I flashed immediately.2. Bad PIC: Out of the 20 odd PICs I flashed, there were 2 that were plain bad. Even with the correct voltages, they did not flash.3. Dry solder: After checking the entire circuit board tracks for breaks/shorts (and there were a few even in my professionaly made board!!!), I found a dry solder problem. That meant that the solder at the back was not actually connected to the IC pin at all. So although the track was fine, one IC pin was not connected on one board.4. Bad power supply: Once when I used a different supply, I had flashing problems. I found out that the power supply wires were causing a lot of interference on the rest of the circuit. If I kept the supply wire away from the serial cable, I flashed.5. COM Port assignment: Once when I used a new PC, there were basic COM port problems, that is the port itself was not configured properly. When I went into Start->Settings->Control Panel->System->Device Manager, the COM port was in error.I have created the FS Key, FS LED, FS Display, FS AD and FS DO successfully. It was a lot of pain at first, but just hang in there with a multi-meter. I would say that I solved most of my flash problems with my mullti-meter as 90% of the problems were on the board side.All the best...Rahul SharmaNew Delhi, [email protected]

Congratulations!

I can't help you with your voltage problem, but I had a similiar occurance with it writing to the first few blocks, then giving the read/write error deal. I guess the problem I had at the time was that I was using Windows ME. I know a few of you have gotten it to work with different OS, but in my case I had to throw XP onto my FS computer in order for the router program to recognize the com board and allow me to flash the PIC (in my case one of the Display PICs). What I saw before was the blue and white fsbus icon in the router had a red x over it and the message at the bottom said com1 open in polling mode whatever thats supposed to mean. If I tried to flash it would change the beginning to 2003 then error. Once I loaded up XP and it recognized my com board it sayd Com1 connected then everything went as advertised. You probably use XP already, but maybe you don't as I didnt at the time. I had tried other computers but they didnt have XP either so the results were the same. It may have nothing to do with the solution, it just happened to work in my case. Good luck.Jason Tighe

Thanks for the input Jason. I wonder if some of the problems relate to OS differences in the system level com port drivers. Possibly not all offer the same degree of control on the RS232C control lines. Different mother boards is another possibility. Who knows?Mikewww.mikesflightdeck.com