You can't programm the chip with micro-C, use the program *icprog*
http://www.test.gompy.net/icprog.zip

This pads are for the leads to incicuit programming.
It's better not to use it for your first programming [8D]
Otherwise, if the programvoltage is to low from your rs232-port, use the leads (only the data, not the powerleads) and a extra batterie of 4,8 - 6 Volt (4 pennlites)

Maybe I'll just make a stupid comment, but...
I've never made a homemade PCB. Some friends of mine tell me is not hard, it's just getting the hang of it, but that means making several and learn to get things right. Surely, I don't want to destroy a plane that cost me many time, money and efforts to get flying because of a poorly-made PCB.

I've seen in other forums ideas like this, so it could work here also.
If the creator (Bigboat) agrees with the idea and there's enough people interested (so that PCBs come cheap), we could have the PCBs done by one of the many companies that make small series for a relatively small price. We should make a list with all modellers and PCBs they want to buy, and once a certain number is reached, send them to be made. Finding buyers should not be that much of a problem; I'm sure that by spreading the word at the flying fields we could soon reach the target number. The "difficult" part is finding someone who is willing to act as "coordinator", to contact the companies, pay, and send them to other users. Well, it's just an idea, don't throw too many stones at me!

To let a company make pcb's is very expancive, not the pcb, but the machines they use.
One pcb for this cdi will cost you about 15 euro, incl the holes, excl sending to the buyer, and you have to order most of the time 1000 in ones.
Find someone who can make pcb's in your own town, put a advertice in the local stores or newpaper.

Here in the US some of the pcb makers give us hobbiest a deal like 5 small ( 2 x 3 inches or so) double sided boards for $50. You might seach around in your part of the world for something similiar.

Steve

Hi to all,i was surfing the net and got across this site http://www.pcbcart.com/.
they manufacture pcb's and they ship around the world AND they are cheap!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Thing that you would find this interesting and i saw that that the last posts were about pcb manufacturing.
I did not buy from then yeat ,but i the future i sheel buy.
Regards ,Zoli.

Lumpylumpy, that's one of the sites I had seen. If a certain amount of PCB were ordered, perhaps about 50 (I think it wouldn't be hard to reach that number), then they would be cheap enough. Those sites have fairly nice prices.
I see no interest in the idea, so I'll just call a friend to make me a "PCB manufacture tutorial".

Hi,sorry it wasent eny help,and about PCB manufacturing tutorial here is a link that i used to http://www.hobby-elec.org/e_pwbm.htm
These are things that i use for PCB manufacturing at home:
-negative exposure printed board(this kind of pcb i can find in my country ,and its cheap ,board size 160X100 mm cost 1,5 euro)
-glass clamp for exposing the pcb, from conrad www.conrad.de
-ultaviolet bulb from conrad www.conrad.de
-mask patern(layout) printed with a laser printer on a transparent sheet, home made
-developer for pcb(after exoposure), home made
-etching, home made
Hope its helpful.
Regards Zoli.

Hi ,
I am is stage of accumulate devices. I brought 3 PICs but 16F628-20p. Is good for our application ?
Is not newer version "A" and i think is 20mhz internal clock.
Use it or trashed it ?

Another question. Mechanically the sensor position is 40 deg BTDC and max retaining is up to 15 deg. Result a 25 deg retaining by PIC program.
If i will put mechaniclly at 25 deg the sparck at low rpm will by at ~0 deg ?
Y need this for a YS 110 4T methanol.

Regards,
Dino

In terms of program code, the 628 and 628A are identical, except you will need different header and linker files. The biggest difference is in the programming of the chip. Some older programmers that worked with the 628 version will not work with the 628A version unless modified. I use PicStart Plus and MPLab. Once modified and updated for the 628A, however, the programmer is backward compatible. You will need to make adjustment for the different pin assignments in the 20-pin package. The 20-pin package has two each Vdd and Vss pins (4 total); whereas, the 18-pin package has only one each. Thus, your layout on the PCB will be different. If you have no experience in doing that, I would suggest getting the 18-pin version.

If you have little or no experience with Microchip processors, I would recommend going with what is in the original design. That is not because the versions are greatly different, but because the small differences can nevertheless keep the project from working, if you do not know how to make adjustment for them. Just the value of your time to make a new PCB layout to account for the two extra supply and ground pins may not be offset by the savings of using the older ship that you have.

Good luck. John

Hi jpanhalt,
I am more confuzed. I have the ICs in my hands and i checked twice, from L to R and from R to L . Not more than 18 pins (9+9) so your theory is not exact.
I have also few IC's with this marking 16F628-04/P. what diferent from -20/P to 04/p ?
Again ..can i use it for our application ?

Best regards,
Dino

A 20-pin version is made. In my haste, misread your original post that referred to your chip as (-20/P) and assumed you were referring to the number of pins. However, my comments about the program header and linker files I believe are still correct. In my own endeavors, I migrated from the 16F84 to the 16F628A without actually using the 16F628 version. I did need to update my programmer to program the 628A. As to the effect of oscillator frequency on the function, it should have very little effect, except timing loops will need to be adjusted if you use a different frequency. I recommend checking the datasheet for the version you have to determine how the frequency may be set to the same frequency used in the original version by Bigboat.

Sorry for the confusion I caused. John

Ok John,
Thanks for explication.

lumpylumpy, I would not suggest to use the method you outlined. Google for 'ink transfer'. It is very easy and cost next to nothing. For the etching I use the 2 part oxygenated water (I think that's what it is called) and 1 part sulfuric acid solution. Again, cheap and works great. Add some of the water if the etching slows down to much.

Yoram

I believe the chemical term is "peroxide". There are two formulations in common use, one is based of persulfate (which is the one you suggest) and the other is based on hydrochloric acid. I suggest that those who are interested visit: [email protected]

Myself, I use the old, reliable ferric chloride method. I don't know that it has any advantages, except stabliity. The peroxide solutions need to be replinished or made fresh much more frequently. Ferric chloride will sit for years and still be active. It has several disadvantages, including that you cannot see the progress of the etching; whereas, the peroxide-based etchants are relatively clear. However, removing the PCB to check the etching is not really a big problem on a small scale. John

I finished my PCB..less holes. I used Pres'n Peel method from http://www.techniks.com/how_to.htm
Ferric Chloride for etching. carrefully with this make yellowish spots ....
Little messy but not violent like acids. In the past i used a diluted solution of nitric acid (HNO3). Works fine but you must have a right dilution.
If not "eat" copper under mask.

Rob,
What is correct value for C3 abd C7 ? from yous picture is an electrolitic cap.

any progress boys ?

Both 0.1 uF (100nF) and electroliticcap.

nice to see this project is going somewhere.

i've made a few ignitions similar to the one posted by JMJ a couple of years back with good success but winding tiny toroids is no fun at all...and i doubt hacking little E-cores would be much more fun either. a while back i got some of these [link=http://www.digikey.com/scripts/DkSearch/dksus.dll?Criteria?Ref=3851&Site=US&Cat=35062915]transformers[/link] from digikey and hope to try them out soon.

spark coils for CDI don't need insulation between layers if done right. the secondary winding is 5 or 6 narrow coils stacked on the bobbin (made from plastic washers glued to a plastic tube) and the primary is 30 or so turns directly on a ferrite rod and placed inside the bobbin. this is the same as the trigger coils used in strobes, camera flashes and the tiny coils found on zenoah engines. works good, easy to do and best of all they come out small and light.



good luck,
dave

I tried the link to the transformers but it says it is expired. I bought some transformers. The were 10 for about $5 plus shipping. They fit on the board with some minor modification.

Do you have a link to a site where it is described how to do the coil you've mentioned?

Yoram

Shalom Yoram,
try this link http://www.hobbyengineering.com/H1958.html for e-core trans.

Dino

sorry about the bad link, the data sheet for the transformer is at http://www.bourns.com/pdfs/pm61300_series.pdf and the digikey number is M8101-ND. these are about $4 each and what i like about them is that they seem ready to use with the primary, secondary and feedback windings in what looks to be the proper ratios. not having to wind a transformer will make this ignition so much easier.

in the attached image (sorry about the poor quality) you can see the type of spark coils i was talking about. the cut away coil on the left is from an echo CDI, the bottom layer of the coil is the primary winding with about 27 turns and the four upper layers are the secondary. the ferrite core is 6mm diameter by 13mm long. on th right are parts from an old 3W coil. the ferrite core is 6mm x 15mm with 30 turns of 0.4mm wire in one layer as the primary. this fits into a pair of finned bobbins (only one is shown) where there are 6 layers of windings for the secondary, each about 500 turns of 0.1mm wire.

the coils i made were copies of the 3W style but on homebrew bobbins. i wound them on a slow, variable speed drill with a revolution counter cobbled onto the shaft....took a couple of tries to figure out how much tension to put on the wire so as not to break it. never really searched but there has got to be some kind of transformer or inductor out there that might work for this kind of secondary.....if all we had to do was wind the 30 turn primary these ignitions wouldn't be so scary looking as a DIY project.




dave

I just tried the Bourns PM61300-1 and PM61300-2 transformers based on the lead given above. They give plenty of voltage, as expected, but seem short on current capacity. My test setup used a 0.47 microfarad capacitor for the HV discharge, a coil from a Zenoah G62, and two different oscillator circuits. One of the circuits was from a popular commercial imported unit. Charge voltage was to around 350 V and the oscillator was fixed at 60KHz as per the datasheet. (The commercial unit used a voltage doubler output, so I had to modify that to a full bridge to keep the voltage below 500V.) Also tried higher and lower oscillator frequencies, but with no improvement. I could get good sparks, but could not go above 6000 to 8000 rpm reliably in simulation due to the low current capacity. Also tried a higher voltage, but was limited to about 400V by the components.

Does anyone have any other source of commercial transformers to recommend? I can adapt the cricuit to use transformers with either a singler primary winding or a primary with the second, feedback winding.

John

Hi can find only tantalum electrolitic cap. for this value (100nF).
Is ok ?

HV-transformers http://www.amazing1.com/transformers.htm
No, you can't use tantalum elo's