Notices
Engine Conversions Discuss all aspects of engine conversions in this forum

CDI gr8flyer55

Old 11-23-2012, 01:32 PM
  #876  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Uhhhh, smaller transformer.....the one I get from you and modify is 15x15mm.
Thats the one I use on the pictures above.

I just wind me own transformer 250x AWG33 and 2x 12 AWG25.
It fit on a 20x20mm E-core transformer.
360 Volt at 6 Volt, 283 Volt at 4V8.
This kind of transformers everybody can wind.
Using a normal transformer can't be, the e-core don't have a aircap !!
Old 11-23-2012, 03:52 PM
  #877  
gr8flyer55
Senior Member
 
Join Date: Nov 2005
Location: Hamburg, PA
Posts: 805
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

My smaller transformer was only a test to see if it worked and it did, but lower voltage. It was only 15x15 size, laminated core, not ferrite.
A 20x20 is an easy size to wind. It should be fine.
I'm just looking for a reliable source here in the U.S. , haven't found any yet. They make it hard to find both ecore and bobbin to fit it. You usually can find one or the other but not both.
Anyone have a source for a 15x15 or a 20x20 ecore and bobbin? We need the source and part numbers.

I tried a toroid today with the same winds, but got different results. 400 turns was fine, but the primary winds were a problem. You have to go bigger on the core size than .5" od. I tried a .625" od and got better results but still not enough voltage output, only 250 volts compared to 375 with an ecore. Must be the length of wire on the core and total resistance for 20 winds on each. My ohm meter doesn't show a difference.... Not sensitive enough.

John
Old 11-23-2012, 09:09 PM
  #878  
bluejets
 
Join Date: Mar 2009
Location: xnot applicable, AUSTRALIA
Posts: 293
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

http://australia.rs-online.com/web/c...ock=4294960100

anything like this John..????
Old 11-23-2012, 09:58 PM
  #879  
nyemi
Senior Member
 
Join Date: Feb 2010
Location: szarvas, HUNGARY
Posts: 133
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55


ORIGINAL: Gompy

Uhhhh, smaller transformer.....the one I get from you and modify is 15x15mm.
Thats the one I use on the pictures above.

I just wind me own transformer 250x AWG33 and 2x 12 AWG25.
It fit on a 20x20mm E-core transformer.
360 Volt at 6 Volt, 283 Volt at 4V8.
This kind of transformers everybody can wind.
Using a normal transformer can't be, the e-core don't have a aircap !!
I congratulate you, Rob.
What are the secondary circuit are you using?
Doubling the voltage?
You tested resistance inverter? 4.7 k/5W (~10000RPM) switched capacitor to charge. You diagnose, inverter performance.
If the capacitor voltage, 130V below, does not fall. Proper performance of the inverter.
Examples:
200V/
4,7K=0,0426A *200V=8,5W/166Hz(10000RPM)= 0,051J=51mJ Spark Energy.
160V/4,7K=0,034A*160V=5,45W/166Hz(10000RPM)= 0,033J=33mJ Spark Energy.
130V/4,7K=0,027A*130=3,6W
/166Hz(10000RPM)= 0,021J=21mJ Spark Energy.
My opinion: 20mJ sparks of energy, 10000rpm.
Appropriate model airplane engine.


Even a circuit. Very promising.
See the picture. link:www.next.gr/inside-circuits/power-mosfet-inverter-l9896.html
I will start testing this out.
In order to increase reliability.
Components is necessary.


Attached Thumbnails Click image for larger version

Name:	Us53498.jpg
Views:	269
Size:	53.0 KB
ID:	1823212   Click image for larger version

Name:	Rp44231.jpg
Views:	516
Size:	71.3 KB
ID:	1823213   Click image for larger version

Name:	Ge96503.jpg
Views:	182
Size:	52.6 KB
ID:	1823214  
Attached Files
File Type: pdf
Mh20340.pdf (287.7 KB, 198 views)
Old 11-24-2012, 02:52 AM
  #880  
jakestew
Senior Member
 
Join Date: Aug 2011
Location: Moscow, ID
Posts: 144
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

ORIGINAL: Nav-Aids
I don't think we are at a point when we can lock-down the PCB design of the 683, at least until Jake splits the development of the 2 processors. everything that has be done so far has been development work. Once the 683 is locked down a separate thread for the 1840 should be made.
Rob and I have been talking about how best to do the switches. I'd like for the kill switch to work (kill the engine) both if the switch comes loose OR if the wires short out. If anyone knows a good way to do this feel free to chime in!

Other than that the PCB design works just fine. For the next version I'll make the switch normally closed (closed = run). That will work with the same hardware. But that won't solve if the switch wires short out. If there's an easy way to fix that it will be great. But if not I figure that someone can just rip the switch out and the circuit will go open and the engine will die.

But the timer board is pretty well locked down. I have no problems with the design. Some filtering on the hall sensor would be nice, but it has always worked well enough for me so far.

I'd like to keep the 683 and 1840 together because they run essentially the same code. With the plan I have there won't really be a "split" in the software... when I've done all I can with the 683 it will pretty much just be dropped from further development. I thought that time had already come, but I figure I can add the instantaneous measuring and learning feature to the 683. Since it can't write it's own memory or do serial communication, I just simply can't think of any more features to add to it!

A couple people wanted me to do a 683 version and it has kept me writing good code that will benefit all the future versions. That's really the only reason I wrote code for the 683. There's no reason for people not to switch over to the 1840. They're pin compatible and almost the same price (the 1840 is actually cheaper from my supplier last I checked). We're only talking about $1.80 for the 1840, so I just don't really see many people using the 683 much longer. That's why I don't want to invest much time in doing anything special for it or trying to separate the two.

> I recommend that it be switched to RCGroups, this way posting / uploading of files won't be a hassle.

I'm fine with that. RCU can be a real pain sometimes, and it's lost just enough of my posts to piss me off to the point where I wouldn't mind switching. We've got a lot of good contributors here though, so we'll have to make sure we can get them to come along!

> I've been working on a program flow chart for newbie’s which makes things a lot clearer to follow.

Good instructions is key to a good project. I couldn't make heads or tails of what was going on when I first started. There are lots of VERY long threads here and it seemed impossible to figure anything out. If it wasn't for John PM'ing me and sending me out working hardware I'd have probably never gotten anywhere.

I've also gotten hardware and help from other people, but if you have to post and ask to find out what's working and how to get or build the hardware... most people will figure that it's just too hard and look elsewhere. AFAIK we have the best system and CDI project out there, so if people overlook us it's a real shame. It took me a LONG time to even find this project and figure out that it was the best one going.

I'm fine with however everyone wants to split up the threads, as long as it prevents the current situation of long and confusing threads. For me it always has seemed like there is a lot of hardware discussion that is hard to understand, and that makes things confusing for a lot of people I think. I suppose the hardware people find the software discussion equally confusing, but I can only speak from my own perspective. I like reading the hardware discussion, but it does make the threads a lot longer to have everything lumped together. Better to have two or three shorter threads than one massive and confusing one.


-Jake
Old 11-24-2012, 03:06 AM
  #881  
jpanhalt
 
Join Date: Oct 2005
Location: Parma Heights, OH
Posts: 168
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Rob and I have been talking about how best to do the switches. I'd like for the kill switch to work (kill the engine) both if the switch comes loose OR if the wires short out. If anyone knows a good way to do this feel free to chime in!
One way might be to put a signal on the switch circuit, say 10 KHz. Watch transitions. If transitions are not seen for 200 uS (or whatever period you chose), kill the ignition.

John
Old 11-24-2012, 03:18 AM
  #882  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

@Jake
With a resistant to ground and a jumper (killswitch) to power you can tackel few problems.
If the killswitch input have broken leads, the input will be allways low
The killswitch input will be low if batterie fails (low Voltage - PIC still working)
If I don't want some one start the engine I remove the jumper of the killswitch.
You can test the complete RC-model with batterie on and killswitch removed.
Most safetyswitches from store, like lost RC, are working with switches / signals to ground.
Thats why I want the killswitch input with a resistant to ground and a jumper to hold him high.

So high is no problem and low is shut off or kill PIC output.

The internal pullup resistants are to weak to hold the input stable high.
Disable them into the PIC and use the external resistants.

BTW1, disable the resistants will also open the way for other applications (see above).

BTW2, if people use two b atterie packs, much safer, and the batterie pack of the radiocontrol failed the CDI will still running !!!
The killswitch like it is now hold the killswitch input high with the internal pullup resistats.

If we all agree with this changes and the changes are made, I will draw and publish the schematic and PCB.
The timerboard get a optocoupler into the output to the HV-board protect the timerboard.
I publish at the same time the schematics and the PCB so there can't be a discussion about wat kind of soft- or hardware there will be used.

@Nyemi
I use a cascade, doubler, after the transformer.
The transformer is wind 1:40, with 4V8 the Voltage from teh transformer is far to low.
I have still test how much power the transformers can make, this is only a "is it work test"
I use both schematics you show and both work good, but I like the ZVS for the LC ocsilator (no heat)
Now I have to find out the components I want to use, they must be small but reable.

I want to know witch timerboard we gone use A or B ?
Old 11-24-2012, 05:48 AM
  #883  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Gompy want - B
Old 11-24-2012, 12:42 PM
  #884  
jakestew
Senior Member
 
Join Date: Aug 2011
Location: Moscow, ID
Posts: 144
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

B looks like the switch lines are pulled low. I think it will be better to pull them high, and that's how it's done right now so there's less change.
Old 11-24-2012, 01:17 PM
  #885  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

If you use the internal resistors, you pull the inputs high with a very high value resistor.
This is a problem if you want to kill fals pulses / dirt on the input.
With the external resistors to ground you pull them standaard low and make them hard high.

If you forget a jumper for the table, table one is allways activate.
If the wire is broken or the Voltage is low, table one is allways activate......the most important table.
If you forget a jumper for the killswitch, you never get the engine start.
If the wire is broken or low batterie, the killswitch will allways be activate by it selfs, it will be pull hard to ground.

If schematic A will be used and the broken lead connect to the engine, most likely with heat, the switch don't do anything.
If this happen in schematic B, first you get low batterie if the plus is connectect to ground but also the input will be pull low.
In both situations the CDI switch off.

With a two batteriepacks you connect the minus together and the plus of batteriepack two to the killswitch input.
If for some reason the lead will broken between batteriepack two and the (killswitch) input, the CDI stop.
Same with the batterie low or lost model, the modules are all switch to ground.
Simple low batterie controle, relais connected between + and -, if the Voltage fails, the relais switch allways off.

If some one want to use the killswitch for something else, he have first to switch off the internal resistors.......but he don't know nothing about programming and anything about electronics.
With some luck he blowup only the pullup resistor, but I think he blowup the complete PIC.
Old 11-24-2012, 03:42 PM
  #886  
bluejets
 
Join Date: Mar 2009
Location: xnot applicable, AUSTRALIA
Posts: 293
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Jake,
Is it possible to look at the kill switch input as a mid point voltage. i.e. if there were say 2 x 4K7 in series across the supply and let the program look at the mid-point. That way if someone wanted to use a low kill or a high kill or if the wires came off the switch or something shorted out, basically anything other than mid-point voltage, then ignition would be killed.
It can be done in other pic such as picaxe but not sure about in C.
Old 11-24-2012, 03:52 PM
  #887  
jpanhalt
 
Join Date: Oct 2005
Location: Parma Heights, OH
Posts: 168
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

That was one of my thoughts too, but it cannot be done with a simple comparator, absent a lot of hysteresis. If you add in an ADC, then you may end up changing a lot of things.

John
Old 11-24-2012, 04:57 PM
  #888  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

If there is in the connection not a real hard high or low, the input will see anything you like with something in the middle.
Old 11-24-2012, 10:50 PM
  #889  
jakestew
Senior Member
 
Join Date: Aug 2011
Location: Moscow, ID
Posts: 144
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Rob makes a good point... if low = run then if the switch or wires short out it will be stuck on. So design B seems like the best choice.

> Is it possible to look at the kill switch input as a mid point voltage.

Sure, we could use the AD converter to sample the voltage. That way only the proper voltage will work and either high or low will kill the engine.
The AD converter can sample from pins 3,5,6,or 7. Looks like it would be no problem to pull samples pretty much as fast as we want.

I guess it's just a question of how much work the kill switch is worth. I'm up for doing the programming if it's something that sounds worthwhile.

The only problem I see is what to do if someone doesn't want to use the kill switch at all. They'd have to jumper it with a resistor I guess.


-Jake
Old 11-25-2012, 01:13 AM
  #890  
nyemi
Senior Member
 
Join Date: Feb 2010
Location: szarvas, HUNGARY
Posts: 133
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Hi Rob
I will help you calculate secondary circuit capacitor.(C2-main Cap)
You, the main capacitor, what do you use (1uF-680nF-470nF/400V)?


I used a transformer, power.
See the picture.
(EE20-N27
material-air gap=0,09)=26W
The document is EPCOS.
FET push - pull inverter. Best choice: N87 ferrite material.
Hungary is not available.(EE20/10/6)

Attached Thumbnails Click image for larger version

Name:	Ig11821.jpg
Views:	131
Size:	79.4 KB
ID:	1823563   Click image for larger version

Name:	Hc93225.jpg
Views:	162
Size:	54.3 KB
ID:	1823564  
Old 11-25-2012, 05:41 AM
  #891  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

It's a example, but this will be board B (34x27mm).

<RC-universal server error>
500 - Internal server error.
There is a problem with the resource you are looking for, and it cannot be displayed.







Old 11-25-2012, 06:34 AM
  #892  
gr8flyer55
Senior Member
 
Join Date: Nov 2005
Location: Hamburg, PA
Posts: 805
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

I will etch both A and B versions today. When the software is finalized, I will be ready to test them.
Both boards are similar in size so that is good. When the HV board is ready, I can do an aluminum box pattern to put both inside for a small case size with everything inside it. Only external wiring and switches to be added of your choice.

John
Old 11-25-2012, 08:59 AM
  #893  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

If we choose of version B, Jack have make the following changes.

If jumper is on killswitchjumper set - killswitch is not active
If jumper is on tablejumper set - table one is active

I don't know the jumper function on pin 4 - maybe can be used for low batterie detection ?
This can be the second batterie pack detectiopn of the receiver if used.
If high no problem, if low.......LED flash at pin 7 ?

@Nyemi, this week I will buy a 4K7 / 10W resistant to test the power of the HV-board.
I use a 470nF/400V capacitor for the moment, but everything can changed.
I try to use the 4V7 diode between Gate and Source, but the HV stop working
Also if I use the IRL540 FET, maybe need small FETs and not MTPE3055V (?)

BTW1, I have to add the schematics on my own server, RCuniverse picture server is down.
If the server is up again I'll upload anything to the RCuniverse server.

BTW2, I was forgotten to draw the 78L05 on the timerboard and into the schematic.
Old 11-25-2012, 03:02 PM
  #894  
bluejets
 
Join Date: Mar 2009
Location: xnot applicable, AUSTRALIA
Posts: 293
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Rob,
Do you realise your output to the "signal output" is now reversed from the original..??

I did drawings to show what I mean but the site is having some internal error apparently.
( 2 drawings, jpg. about 160kb each...)
Old 11-25-2012, 03:21 PM
  #895  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

I know, but thats how the inputs works normaly.
I was working in the automation and if there was a broken lead the inputs will pull low.
Why, there no power anymore and with pull high you need power.
It's also for safety, you have to place the junpers / hold the switches if you want to start the engine.
In trains, planes aso, there are switches into the seats, if the pilot leave the seat you make a short circuit to ground.
We use low Voltage, but IRL the Voltage can be high enough to kill, thats also a reason why to make a short circuit to ground.
Most simple killswitch will be a relais, relais is connected between power and ground and use a NC to kill the circuit.
No power relais quit, NC contact make short circuit........all safety circuit work like this principal.



I want to know, can the left circuit work with low batterie ?
Old 11-25-2012, 03:23 PM
  #896  
jakestew
Senior Member
 
Join Date: Aug 2011
Location: Moscow, ID
Posts: 144
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Now that I think about it more, I favor design B for the switch setup.

A couple changes will make this timer board better suited for future features though. Here are some changes I'd like to see...

- Jumpers use 3-pin connector for Jmp1,2,3 (Gnd, Signal, +5v)
This will give more flexibility if we want to change things up. Just make the pattern for this. You don't have to actually drill the hole or put in all three pins, but it will be useful later if needed. Leaving this space will also make sure that people using three pin connectors will have the space to plug them in.

- I don't think we need LED2. I have planned to make more LED modes for the next release. I will make the processor switch the LED according to the hall sensor when the engine is not running. Maybe LED2 can be a power LED instead? That would have saved me a couple drained batteries already!

- I don't think we need T1 on LED1. The PIC can source or sink 25mA from any GPIO pin. There is a total limit of 90mA for GPIO. So we should be able to drive 3 LEDs at 25mA (=75mA) safely. You'll have to check the current draw on the HV output transistor and make sure the LEDs are limited to under 25mA.

- It would be nice to have a jumper/header for connection to a timing light or flash bulb. No need to worry about the actual flash circuit yet, but having the connector will make it easy in the future.

- Pin 4 is input only, and I'd like to keep the hall sensor on pin 5. Otherwise anything can be moved around. Feel free to move things around if it makes things easier to lay out.

- I'd like to put the kill switch on Pin 4.

- I'd like the table switch on Pin 2. (This is also the RX pin)

- I'd like a 3-pin connector for serial comms. on pins 2+3. (GND,RX,TX)(GND,Pin2,Pin3) There should be a ~2k resistor in line with the TX,RX lines for protection (optional). This way I figure the PIC can always send data on the TX pin and if we want to use the serial RX we can just open the table switch. If the serial is hooked up we can use that to control the table setting.

- I'd like an ICP header on the board for programming. My PIC pins are about to start breaking off from swapping them around so much.
You can find details about the ICP header connections here...
http://www.piccircuit.com/shop/pic-c...ic-series.html
And look at this schematic...



- I'd like to see some filtering and protection on the hall sensor input.
I read this page, which seems to have pretty good info on this sort of stuff.
http://www.digikey.com/us/en/techzon...ectronics.html
Signal coming from long wires showing ringing...

Filtering circuit...

Filtered input...



I think if we make these changes we will have the best board possible for our PICs. I'll be etching some boards for another project hopefully sometime this week. If we figure out the best design I'll make some of these boards for testing. Everybody needs to look over these designs and double check all the values and circuits! If we can get it right this v2.0 board design should last us a good long time.


-Jake
Old 11-25-2012, 03:42 PM
  #897  
bluejets
 
Join Date: Mar 2009
Location: xnot applicable, AUSTRALIA
Posts: 293
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Rob, you quote" I know, but thats how the inputs works normaly."

The 2 diagrams have the output to the signal output reversed. How can both be correct..??
Old 11-25-2012, 03:46 PM
  #898  
bluejets
 
Join Date: Mar 2009
Location: xnot applicable, AUSTRALIA
Posts: 293
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

Jake,
I think if you look back a couple of weeks, that filter arrangement was what I suggested back then but everyone said correct with software.
I was of the opinion that the micro should get what it is expecting to see.
Old 11-25-2012, 03:51 PM
  #899  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55

LED 1 will be the timinglight.
LED 2 will be the degree LED and check the hallsensor.

On the PCB there are pins to plug-in the LED's.
If you can see there are allready 3 LED's into the schematic, LED 1, LED 2 and OPTO 1.
Pull LED 1 or 2 or both if you don't use them, they are on pins.
Plug a rotationmeter to the LED-place and you can see how fast your engine run.

The killswitch can't be used if you want to use pin 2 and 3 for communication.
Jumper 2 and 3 are allready on the PCB, so communication is possible to use the jumperplaces.

Move the table switch to pin 4, jumper is allready be avalible.
Changing the otherside of the PIC pin 5, 6, 7 and 8 is allmost impossible without making a complete new PCB-designe.
But there are two designes feel free to use A or B.

You have also thinking about the size of the board, we use normal size components.
Using on my bike it's no problem, but putt the same size into a model airplane can be a problem.
Old 11-25-2012, 03:58 PM
  #900  
Gompy
Senior Member
 
Gompy's Avatar
 
Join Date: Jul 2010
Location: Alkmaar, NETHERLANDS
Posts: 404
Likes: 0
Received 0 Likes on 0 Posts
Default RE: CDI gr8flyer55


ORIGINAL: bluejets

Rob, you quote'' I know, but thats how the inputs works normaly.''

The 2 diagrams have the output to the signal output reversed. How can both be correct..??
I'm confused.......wat diagram you mean......the low batterie schematics ?

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service -

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.