iura

Ok, Jake! Don’t huff it’s only my thought, not more.
Hi, Jr! Which scheme did you use as HV board in your CDI?

gr8flyer55

I am going out on a limb here and assuming that there is confusion on which schematic is best for this project. I will post a close up picture of a bare board, both sides, and an edited pre-existing schematic to use with the board. Planning for a busy season with fall/winter repairs and new projects, I will have some kits available shortly. First I must stock up on the parts. The inquiries are becoming an every day occurrence. I anticipate sending out quite a few kits and assembled units. As most of you know, I do repairs on ignitions. If after you build a kit and have a non-functioning unit, I will offer a troubleshoot/repair of your assembly, for a very minimum fee. Let's face it folks, I am not going to get rich doing this! I only offer this service to help out my fellow electronics enthusiasts.
My only problem in supplying the kit of parts is the sparkplug caps. There are many types and I can't have a supply of each type on hand. It gets costly to stock numerous caps that I will never sell.
That said, are any of you interested in a kit version after the final software version is released?
If so, I will start a name list for them. Jake is doing more than his share of the work in this project and I don't think he needs another responsibility right now. Jake mentioned the possibility of getting the timer boards done commercially. I am all in favor of this, since it would make my workload a lot less. As I see it now, one or two changes will be made to the timer board for extra wiring of switches etc. When that is accomplished, we can go ahead with producing a drawing in Eagle format and get them done!! I for one can't wait!

Pictures coming shortly of what we are doing with the project.

Many thanks to all who contribute to this project, past and present!!!

John

Gompy

I got the engine, broken coilholder, but thats no problem if I mounth the CDI on it.
I want to start first the engine like it is to know if he is OK.
With the flashlight I want first to see the factory settings of the original ignition.
Then I remove the flywheel and the ignition coil and will place the magnet and hallsensor.
I need also a smaller prop, this one is 20 x 8, I don't think the engine can make max rpm.

iura

As I see this is HOMELITE trimmer 28cc engine. Better chose is 16*8 prop. Max RPM as I remember is about 7400, not more.

Gompy

It's my second Homy http://modelbouw.gompy.net/homelite/index.htm
The first one I rebuild it with a waterjacket to use it into a speedboat.....no sucses, far to low rpm.
But into my Zodiac it was a great sucses http://modelbouw.gompy.net/zodiac/index.htm
The motor was cover with a hollow babydoll and lot of people say I'm a fool to put a baby (...) into the Zodiac )

jakestew

Naw, not huffing. Just explaining.

When I get serial coms going on the 1840 we'll need a program to monitor the engine communication. The way I see it though is that the best way to work with curves and graphs is with a common and power program that's geared to it, like excel.

Having a spreadsheet makes it easy to change the calculations and tweak things. If I dumped all the excel calculations into the source code for a program it would be a total mess. I'd get lost and soon wouldn't be able to make heads nor tails of it. It would also be hard for others to jump in and figure out what's going on. Excel splits up all these calculations and organizes them in a way that people are used to and can understand.

I'm glad Gompy is working on the schematics and hardware. I'm sure John will also help when he gets back. I got an assembled HV board and parts to make another one or two from John, but when I got them and tried it out I was just too impressed and excited to start on the programming, so I haven't even put the other boards together yet.

This is the great thing about open software + open hardware projects, the hardware got me going on the programming and now I'm trying to encourage the hardware guys. I also think that the demand for units will pick up and that will encourage all of us.

I'm hoping that a commercial CDI manufacturer will also see the value in supporting open source and start making their CDI units with the right processor(s) and wiring up the programming header to make it available for easy programming. That would take a lot of the pressure off of the hardware end. People who are in a hurry or don't have electronics skills can buy a similar commercial unit and people that like to tinker can have fun building theirs.


-Jake

bluejets

Futurlec do a reasonably priced production of printed circuit boards.

http://www.futurlec.com.au/PCBService.jsp

Also, if you go down the page you can see they will accept 12 different drawing files.

I am not familiar with Eagle although I have downloaded it and gave it a try.

I found it to be somewhat un-user friendly in some instances requiring more time to study the workings of it.

However, I guess it is free and I assume that's what you are looking at.

With the Protel and Circuitmaker systems I can knock stuff out in 1/4 of the time.

Sent an Eagle preview pc board but I think everyone is rather busy as I never heard anything more about it.

gr8flyer55

Hi Jeff.
I am not sure, but I think Circuitmaker can do a double sided board. If it can, we are thinking of going that route with the timer board. Just talked to Jake on the phone and some things were discussed and that was one of them. It would be then possible to kind of shrink the size of the current board by at least 1/4 inch in both dimensions. I have that program but ended up being confused with it at my age I guess. Teaching old dogs new tricks sort of thing. Anyway, new version of excel sheet on the way, no more compiler!! It adds the hex to the clipboard, copy and paste and save. Sounds great to me!!
I got more parts now for the boards, will pass them out shortly.

John

bluejets

Hi John,
Seems I have grey matter lock-up also at times. Prefer to think it is because too much going on at one time.
Things seem to be moving along here at a good pace and a lot of work done in a relatively short time.
Keep up the good work.
I'll look into the "double-sided"

Gompy

Using double sided PCB or SMD is not DIY-friendly.
Using it all for this board, I dont know its wise ??
To designe a new double sided board to use a bigger PIC and communication OK.
Ill make the pcb, even if it needed tripple or quart layer....12 layer is max, sorry.
There is one thing i cant do, making a PBC and using SMD, my eyes are to bad or the parts are to small. [&o]
With using normal components its also possible to make a small pcb, if we gone use both side of the pcb.
Not all the components have to be fit at top of the pcb !!
Use the bottom for the timerboard and the top for the HV-components.....Tatatataa, small board

bluejets

Hi Rob,
I had similar thought about whether double sided would be necessary.

Seems the board is fairly small as it is with only a couple of jumpers needed.

Maybe like you said, if it gets smaller then not very DIY-friendly.

jpanhalt

Considering the two construction options < SMD and double-sided > separately, I prefer SMD every time over through hole. I dread drilling lots of holes; although, I am set up well to do it. 1206/1210 SMD are only a little smaller and actually don't save much space, but IMHO are easier than through hole. 805-size SMD is the smallest I go. They are not much harder than 1206, once you get used to them, and they save lots of space. As for the chip, an SOIC-sized chip saves lots of room. Without a socketed chip, though, you will probably want to enable ICSP for programming. As for eyesight, I am not a youngster. I wear glasses and a head magnifier.

As for the board, for those who use the photo method, double-sided and single-sided are not that much different. Just tape the transparencies together, slip the board in, and flip the unit after exposure to get the other side. However, for those using toner transfer, I can appreciate that a double-sided board would be more difficult. However, isn't the eventual plan to provide a commercial board? That would be so much cheaper and more convenient for everyone; everyone except for the poor guy who agrees to coordinate it.

Thus, if the plan is to use a commercial board, which will include a solder mask, I think SMD with 805 or 1206 is the way to go. I would keep the processor to an SOIC like size for ease in soldering.

John

Gompy

There is also a other BUT against small PCB's..........electronic smog
Timerboard and all the other stuf even the SCR can put dirt into the hole project.
Ecven if the (normal) components will be placed right to each other, it's possible they disturb each other.
Also the small housing of components can be a problem, small housing - no EMC smog protection.
A metal BC547 welding to ground is better whyen using a plastic dip BC547.
BTW, not only the HV will / can disturb, even the cristal or PIC can do the job.
And I forgot to mention the spark or radio signals from GSM, GPS or your rc transmitter.

gr8flyer55

Hi John.
I think you were using Eagle to do schematics and boards. Do you have a list somewhere of pcb fab houses that accept Eagle file here in the U.S. ? I've seen alot offered outside the country, but not many here in the states. Also, there is Circuit Maker 2000, which I have here but could never quite get the system working easily enough to do what I need. It is a very capable program!! Bluejets is able to work with it easily, but then he is one of the most talented electronics enthusiasts I have ever met. Hi knowledge has allowed me to get many things working when I thought I'd just given up hope on a project. Rob uses Orcad and can do the boards also, so it looks like the competition has started!! This is a very good thing!
I am using what I consider to be the smallest aluminum case size for my ignition & timer board in it's current state. Both fit comfortably but I feel it can always go smaller. Ignition board is very compact, but the timer board size definitely needs to be addressed. Today I'll try getting some pictures posted that I promised. Been having a problem lately uploading pictures which made me cancel the messages. I also am writing up a parts lust for both boards, with both Digikey and Mouser part numbers. My stock of parts is now down to about 50 of everything parts-wise for building boards. Not a problem with new boards, will use same basic parts anyway.
Time to get serious and try posting my pics and parts lists.

John

iura

Guys, which one scheme you are building????

jpanhalt

Hi John,

I have actually never sent a PCB out to be manufactured. I have thought about it many times, but end up DIY. Advanced Circuits is US based and advertises US manufacturing (http://www.4pcb.com ). It has a good reputation, and I believe it will take Eagle-generated files. For really cheap boards, the manufacturer will probably be off shore, but remember, everything is online. Only the finished boards need to be shipped, and international air mail is quite reasonable.

John

Gompy

All they need are Gerber files to create PCB's and all programs generate Gerber files.
So you have to send the Gerber files to any PCB company and they make the PCB for you.

www.makepcb.com

gr8flyer55

Still having trouble posting a board pic, sent it to you in an email.

I'm wondering if we made a file of the timer board and told them to step and repeat for as many times would fit on a 24" square single sided sheet, if that's an option? Seems some companies also offer this besides etch, solder mask, drilling, and cutting apart.
If we can, I think that would give us the most boards for the money. I don't know how to go about doing an order, but maybe someone has info for us. Getting a price will decide which way we go I guess.

John

Gompy

I think they say in the US "hold your horses" and thats wat you have to do John.
Without testing _this_ hard- and software for some time, it's not handy to let make now the PCB's professional.

Make a few (10 or 20) CDI-PCB's or kits, start testing the CDI (not one person, but 10 or 20 differend persons with differend engines), listen wat they say about it, change or fix program / bugs, test again and if no one have complains you can start makinf real PCB's by some compagny.

A other point to attention befor you go to the PCB-factory.
If Jake gone make a new CDI-version, hard- and software, wat will you do with the old version ?
No one will buy a old version if there will be a new one soon.

I don't know how big / small the CDI can be, but a "eruo-cart" is 4" x 6".
A single or double sided PCB I can make it in 3 minutes, etsing and lighting, not drill and cut.
The most work will be the drawing of one single PCB, not to copy / past the rest to (example) the "euro-cart" sheet.

My advice, take some time to test and don't try to run faster then your shadow.

gr8flyer55

You are right Rob. I don't want to rush into this, only to find out there are problems with certain engines. So I guess we need more engine testers. I will supply kits at my cost, very low price. We just need more people testing. When Jake releases the next version of the software, we should almost be there. What he's told me on the phone about what he did with the excel spreadsheet is a wonderful thing. There will be a button to click to take everything in user data, curves, and the source code, and convert it to a hex listing. This will then be copied to the clipboard in Windows. From there, open Notepad, paste clipboard contents to the screen, name it and save wherever you like! Much, much easier than going thru drudgery in the XC8 compiler. All that's left to do at that point is to program your PIC chip and try it. Nothing hard. Jake has taken alot of time doing this feature but well worth the wait I think!

After switching over to the next chip, (12f1840) and a slightly different timer board with in circuit programming, a boot loader for data, and the rest of the wanted features, we will have a modeler's dream come true. There are many commercially made ignitions with far fewer features for hundreds of dollars which we will walk all over with this combination. And to think, everyone participating in this will have made it all possible.
I don't want to rush Jake and have him make mistakes. So we can work on the hardware in the interim. If you remember the schematic that Nyemi posted from rccdi.com, that is basically the ignition circuit I am using with some component values changed. What was added is a way to tune it for lowest mah current draw, and the highest voltage across the .47 uf capacitor. I will show where to add resistors and suggested values for a close tune. Each board must be tuned because of variances in component tolerances etc. Not all transformers will put out exactly 400 volts for example. I have a few that reached 550 !! Those are the exceptions rather than the rule. Also, taking into consideration that the oscillation circuit transistor heating that can occur if over driven will basically rule how you tune. Alot of variables. There is room On the ignition board for doubling up on the transistors for higher current draw handling if you feel the need. So you see, it is a true user's system. You build it, you tune it, you use it. An electronic's addictiction cured.

John

Gompy

I like the things Jake do but it haven't be a problem for the tuners to use more programs to realice there project.
They know wat they are doing.......users don't know wat they are doing and thats the problem.
All the questions I get, I get from joung guys who want to use it on there scooter / moped.

For now everybody look up to the "special" transformer, but using a flyback transformer thats the problem.
If you use a transformer in reverse, example a 2x 3 Volt to the low Voltage side, you get 250 Volt output at the other side.
Nyemi have make this kind of Voltage pump into his ZVS DC CDI.
Its also possible to make hig Voltage with a UC3845 and a small easy to make transformer.
I think this is more realible to use instead of a flyback transformer.
If I feel better and at home everything will be normal, I'll gone test with this kind of Voltage pump.
I think also the timerboard need some extra components to secure it against electronic smog.

BTW, Volvo put the sparkcoil right on the candle into the sparkplug.
The exclude with this construction from high Voltage inteverense to the timerboard.

jakestew

New Release!
v0.98b6 is now on the site!

Changes...
-Changed table calculations to load a minimum of 7 instead of 0. This should make the processor always fire a spark, even if it's late.
-Added the ability to save and load curves
-Spreadsheet will now generate the HEX file! No compiler needed!
-Too many spreadsheet changes and tweaks to list.
-All settings now stored in the table, so everything changes when you change the table switch!

Since this is still a beta version the spreadsheet has not been protected or locked in any way. Nothing has been hidden from the user, except empty rows and columns. So be careful to only edit what you need to change and to not accidentally change cells you shouldn't be changing.

This version will generate the hex, but remember that it's not compiling anything. It is just entering the table data into a precompiled hex. I highly suggest not even clicking on the "HEX_Calcs" sheet.

There is nothing for the user to edit on the "Table_Values", "Code", and "HEX_Calcs" sheets, so do not edit anything on these sheets. The "Copy Code" button is on the "Code", sheet though BTW.

I tried to make things easy with the color coding and buttons.
Green = settings/cells you can change
Red = calculated values used in the program
Yellow = intermediate or informational cells
Blue = Various information or settings that should not normally be changed by the user

The default curves that come with this sheet are based on replicating the stock ignition on my 4-stroke engine, which has a fixed 25 degree advance. This might not work well on your engine, so don't load it if you think it won't work!


Quick run through on the spreadsheet...
The first thing you need to know is that you must turn on macros for the spreadsheet to work right. When you first open the spreadsheet excel will pop up a bar across the top about a security warning. Click what you need to so that macros are turned on. You can also go into excel options and change the settings related to this or make it always trust this specific spreadsheet file.

If you don't trust me to run macros on your computer or somehow get this file from a untrustworthy source then you can still use the spreadsheet. None of the buttons will work, so you will have to manually copy the source code into a file, manually copy the table values, and paste them into the source code, then compile the source. To do this just edit everything how you want, then copy cell G3 from the "Table_Values" sheet and paste it into the source code at the proper place, repeat with cell N3, then compile.

Sheet 1, "Setup_Settings"
-Put in your hall sensor degrees BTDC. My program times from when the magnet approaches, so it will probably seem 1-2 deg. MORE compared to programs that time from when the magnet leaves the sensor. If you use a 30 deg. magnet/sensor position, it will probably be 31-33 deg. according to my system. To find this just load up a fixed advance flat curve and time the motor to see where the spark hits, then adjust the Hall Deg. setting so the spark hits where it should. Finding this and adjusting it in the program should be easier than trying to move the sensor/magnet around!

-Set your "Est. Start RPM" and "Start Advance deg." settings. 5 deg. advance is suggested for easy starting, so the main thing you have to figure out is how fast you are spinning the engine when you are starting it. Probably around 200-500 for handstarting, higher if you use a mechanical starting device or pull start.
* This doesn't actually change anything anymore! It's just for seeing the calculations*
Start RPM is now down lower in the user settings for each advance table, and the start advance degree is taken from the "0 RPM" setting in the advance curve.
There is also a deg to digi-deg calculator in the top right cells of the page.

-The rest of the settings are related to the MSD feature. With the default settings it will throw three sparks at start up, and two sparks from there until it hits 916 RPM. Otherwise it should only throw one spark. These settings should work best for most people, only change them if you want to try playing with the MSD feature.


Sheet 2, "Advance_Curves"
-Here is where you program your advance curves.
-You can change the RPM set points and the advance settings.
-Don't cross your points. That means they should all be in order from lowest to highest.
-Point 0 is for starting, Point 500 is for 120-915 RPM. You can't change these points, but you can change the advance setting for them.
-You can use the buttons to change the values. If you need a specific value you can key it into the cell.
-Hitting tab instead of enter makes excel move to the right instead of down. This makes it easy to key in the values.
-The graph shows your curve line in blue, the back-calculated setting is shown by the red line.
-Red points at the top, bottom, and in the line show where the actual control points are according to the table (CurrentRev), these can't be changed.
-Use the red marks and line to see what effect your curve is having on the actual table values.
-You can go above the red line, but the red line represents what is actually put into the table and where sparks will actually fire.

-You can use the button to copy your settings down to table two. Use this to quickly align your control points or change the curves. If you want to swap 2 -> 1 then use one of the save slots on the "Saved_Curves" sheet.


Sheet 2, "Saved_Curves"
-Here you have 6 slots to save and load curves
-The preview button loads the saved curve into the graph at the bottom so that you can see it before loading it.
-The load buttons load the saved curve into table 1 or 2 on the "Advance_Curves" sheet.
-The save buttons save curve 1 or curve 2 into the save slot.
-Loading/saving curves does NOT change the user settings, except for the start and low_rpm advances.
-Always double check your other settings and remember that they are not saved with the curves.
-The sheet will prompt you to name the curve and will automatically save the date and time.


Finishing up!
Once all your settings are right you can go to the first sheet and click "Copy HEX". This copies the hex file into your clipboard. Open a text editor, hit paste, then save the file with a .hex extension. You're now ready to program your PIC with the most advanced CDI program ever created! (and the programmer is as humble as he is skilled)


-Jake

gr8flyer55

Jake, seems something is preventing me from seeing where I should be able to see sheet 2. Almost like when it was saved, no other sheets were saved. Anyone else have this problem or am I the only one? Nevermind..went to full screen view and I can see it..Looks very good with the features..Editing a few curves and trying it out shortly!

John

jakestew

One thing I forgot to mention... the code sheet doesn't automatically put in the table values anymore, it's just the regular code. You'll need to paste in the table values when you go to compile. Now it's only two lines though since the user settings are moved into the table.


-Jake

gr8flyer55

Jake, This thing is the cat's meow!

I love it, and everything works great. Just to prove it for myself, I programmed two curves into a previously unused chip so there was no mixing of data from a preprogrammed curve. Popped it into the timer and tested. Works really nice. Now if the weather holds tomorrow morning, I'll give it a run on the engine and test both curves.
I need to make a good 4 stroke curve for my converted O.S. 60. It was setup for throttle coupled mechanical advance and always stumbled and hesitated at mid throttle settings so I want to see if it can be cured. If it works, Look out Mr. Saito!! Theres a new kid on the block...
Anyone interested in sharing a 4 stroke curve for testing so I have a starting point?

John