jpanhalt

I can't read your schematic. The lousy JPG resolution makes reading any part identifiers almost impossible. On Bigboat's site, T1 is an ignition coil. Q1 is the thyristor.

As for why it is a thyristor, I suspect the high pulse-current capability, easy gate control circuit, and easy turn off contribute to the reason. John

Futterama

Check again, I changed the attachment so it's now readable and both the bobine and the thyristor is called T1.

The attachment was created from this image:

http://www.modelbouw.gompy.net/newcdi/cdi-a.jpg

Bigboat

OK, T1 can be the SCR and the coil......my mistake, and I was thinking you mean the ocsilator transistor to change into a FET (thats possible todo).
To use a FET for the SCR will cost you a fortune AND you have to make a signal to close the FET after the ignitionpuls.
A SCR will close by him self after the *ignition* becourse the *backflash* of the ignitioncoil create a negative puls.

Futterama

I bought 3 BT151-800R for 30DKR ~ $6 (USD) so I'll just use them.

I have a lot of different MOSFETs but they are low voltage (20-30V) and high power (80-150A) so because of the low voltage, they are not suitable.

Bigboat, I also bought the Velleman FLPS, and got it disassembled but I can't figure out how to wind 400 winds of 0.25mm (AWG30) copperwire onto the plastic holder.

According to my calculations and measurements on the plastic holder, there is only room for 365 windings on 3 "sections" and the last section is used for the 0.5mm (AWG24) wire, and I tried to wind 20 windings with AWG24 and it can barely fit.

How did you wind it? Did you remove the wall seperating the sections to get more room for windings?

Bigboat

Yes.

rudrich

Hi BigBoat,

sorry for I guess asking again, but did anyone ever find a coil that would fit and is buyable to use as transformer in the HV PCB? I'm no very good in winding coils...


Best,
Phil

geaditk

Hi to all i'm from thailand, i'm intresting in build a 2 strokes ignition system.
i'm not understand about pickup position degrees use in Source1.asm
please tell me the Information form my motorbike in my pictupe.
Thank you very much
Geaditk

Bigboat

If the original ignitionpoint of your engine 38 degrees BTDC, you have to place the pick also 38 degrees BTDC.
At the start of the engine, the CDI-software will change the ignitionpoint to 5 degrees BTDC (easystart position).
If the engine runs 1800 RPM, the ignitionpoint change to 38 degrees BTDC.
Now comes the most important thing you have todo..........make the ideal curve.
At high RPM, the ignitionpoint have to change a few degrees more (38 degrees+ X-degrees)
With the ideal curve, your engine makes more RPM and get more power.

Futterama

Bigboat, can you explain why the B1 high voltage transformer is separated into 3 coils?

If someone told me to generate high voltage, I would just use a square AC on a single primary winding.

Bigboat

One of this coils is the ocsilator together with the Resistor, Condensator and Transistor.
It's impossible to create with a normal AC-transformer the same.

Phatburner

Hi Bigboat;

I have been reading this forum and the related materials. Thank you for all your work and sharing. I am trying to find out what is the frequency of the free running oscillator in the inverter? Does it run fast enough to allow triggering of multiple sparks?

Bigboat

Test let us see we can run over 30.000 rpm with a good and bright spark.

aerodactyl

Hi to all people interested in CDI's.
My name is Uwe and I'm from Germany. Please excuse my none perfect English. The reading of this thread was really interesting and you did already a good job, especially bigboat from NL.

Years ago I tried to build a CDI by myself, but just with medium success. The base I use is slightly different to the version your are talking here. I work with an external oscillator for the DC intermediate circuit. The rest looks very similar. Basically my system works but I have problems to raise the voltage fast enough for high speed engines. The best result I ever reached was 6 ms for 150V in a 0.47 µF capacitor after the fire signal. My target is to reach less than 3 ms as it was reached with the poor system described in an article od Floyd E. Carter. Unfortunately I couldn't find the link any more. The way how to create the DC voltage is the same as the version of bigboat.
I already read the full thread but I couldn't find any no. of ms or a scope diagram. But if the CDI will go up to 30000 RPM the time have to be less than 3 ms.
The problem I encounter is the "transformer". I tried different version but with no success. I found out using a ferrite ring is more insensitive to frequency variations of the oscillator than other versions.
Before ordering the FLPS board just to have the e-core is somebody of you able to tell me your reached values of:
- max. DC voltage
- raise time up to 150V
- used capacitor (µ47F?)

Would be nice if you could support me with some technical details.
Thanks in advance.
Uwe

Bigboat

Hi Uwe,

If you buy the FLPS-boards, you have rewinde the core......you can't use the core like he is.
If you can buy a HF-E-core it will work too.
Our HV-coil makes about 400 till 600 Volt, depent how the coil is wind and how the ocsilator is working.
Raisetime untill 150 Volt less than 3ms, 400 till 600 Volt within 3ms

aerodactyl

Hi Rob,

thanks for the fast reply.
I made tests with a coil where 2 E was "inserted". I bought [link]http://www1.conrad.de/scripts/wgate/zcop_b2c/~flNlc3Npb249UDkwV0dBVEU6Q19BR0FURTEyOjAwMDEuMDA3N C4xODEyZWVmYiZ+aHR0cF9jb250ZW50X2NoYXJzZXQ9aXNvLTg 4NTktMSZ+U3RhdGU9MjMxODc4Mzk0Ng==?~template=PCAT_A REA_S_BROWSE&mfhelp=&p_selected_area=%24ROOT&p_selected_area_fh=&perform_special_action=&glb_user_js=Y&shop=B2C&vgl_artikel_in_index=&product_show_id=&p_page_to_display=DirektSearch&~cookies=1&zhmmh_lfo=&zhmmh_area_kz=&s_haupt_kategorie=&p_searchstring=516643&p_searchstring_artnr=516643&p_searchstring_manufac_artnr=&p_search_category=alle&fh_directcall=&r3_matn=&insert_kz=&gvlon=&area_s_url=&brand=&amount=&new_item_quantity=&area_url=&direkt_aufriss_area=&p_countdown=&p_80=&p_80_category=&p_80_article=&p_next_template_after_login=&mindestbestellwert=&login=&password=&bpemail=&bpid=&url=&show_wk=&use_search=3&p_back_template=&template=&kat_save=&updatestr=&vgl_artikel_in_vgl=&titel=&darsteller=&regisseur=&anbieter=&genre=&fsk=&jahr=&jahr2=&dvd_error=X&dvd_empty_error=X&dvd_year_error=&call_dvd=&kna_news=&p_status_scenario=&documentselector=&aktiv=&gewinnspiel=&p_load_area=$ROOT&p_artikelbilder_mode=&p_sortopt=&page=&p_catalog_max_results=20[/link]this.

Is it similar to the core of FLPS?

Uwe

Bigboat

The url doesn't work, Artikel-Nr.: 516643 - 62 at www.conrad.de.

It's hard to say if this core will work.
Normaly HF-cores will be 10 Mhz and more.
If the core can't handle the ocsilator frequentie, the ocsilator will shut-off.
But the price isn't to high to test this core.
BTW, if the freq isn't enough to charge the C to the right Voltage it can be also the transistor.......low HEF-type !
Some built the ocsilator with a FET, I don't try it yet, but maybe this can help to raise the freq and the Voltage.

aerodactyl

Hi Rob,

sorry for URL but you found the right part.
I just rewind this core with 300 to ?? windings. I assume it was 22 from the test before. More than 300 windings were not possible because of the E ferrit.
The result was better but not good enough.
I work with just 1 primary winding. The clock is generated separately. This is much more easier to vary the frequency. The switch is a MosFET with very low resistance and should be OK.
Looking on the oscilloscope the voltage raise up very rapid to about 100V, then the curve becomes very flat.
That's still my challenge

Later this week I will reduce the primary no. of windings and see what happens.
Next step will be to remove the "walls" of the core in order to get more space and to wind the primary on top of the secondary winding.

Best regards
Uwe

Futterama

Hi Uwe,

Look at this link:

http://www.sportdevices.com/ignition/inverter.htm

They claim that the circuit charges a 1uF capacitor at 250 volt in approx. 1 ms.

I couldn't find any recommended transformers for Texas Instruments UC3845 but then I found the Linear Technology LT3750. Linear claim this about the LT3750: "A typical application can charge a 100μF capacitor to 300V in less than 300ms" so I guess it can charge the 0.47µF in less than 3ms. They have several recommendations regarding the transformer: TDK, Sumida, Midcom and Coilcraft. The datasheet contains the part numbers for the recommended transformers.

I ordered a Coilcraft transformer and the LT3750 but I won't have the time to test it anytime soon.

I hope some of this can help you.

Regards,
Martin from Denmark

jpanhalt

Futterama,

Do you know if the 100 ohm resistor in Vcc is necessary (presumably it helps stabilize power to the controller)? If needed, what is the smallest wattage resistor you been able to use? The LT3750 and Coilcraft DA2032 look pretty good, or are you using the larger DA2033/2034?

John

Futterama

John, I have not been working with the schematic at SportDevices. I just got some inspiration there, to look for a flyback controller and that was how I found the LT3750. I haven't had the time to test anything, actually I haven't even received the parts as I ordered them only a few days ago. I ordered the LT3750 and Coilcraft DA2034 and I'm waiting for them to arrive, and for me to get time to work with them.

The reason why I picked the LT3750 was because they had some transformer recommendations so I knew I didn't need to wind my own transformer. A lot of people here have requested a schematic with a ready-built transformer. Perhaps this is the answer, only time will tell (we have the answer when someone have tested it).


Martin

jpanhalt

Martin,

I appreciate the update. I had the same goal of a CDI with commercial transformer, but got onto another project 6 or 8 months ago. That project is nearing completion, so I may get back back to the CDI. Please keep us updated on what you find. The components you chose certainly look promising.

John

Bigboat

I order about 10 or 12 read to use transformers, but all stop working around 10 - 15.000 rpm.
Normaly the transformers will be used for DC/DC conversion, 12 Volt to 240 Volt.
They don't need a high frequentie like we do.
The FLPS-transformer will work without conversion, but you do modify our schematic.
The Voltage from the FLPS will be about 250 or 350 Volt, but it's to slow.
You have to change the C on the FPLS to get a higher frequentie.

jpanhalt

Actually, I was going in the direction of the FLPS with a Coilcraft transformer. I recently got the FL2015-9L, but have not had time to experiment with it. I think Martin's approach with voltage feedback is also worth looking at. At least, we have a long Winter ahead of us in which to test our ideas. John

aerodactyl

Martin,

I just studied the LT3750 device. It is similar to the one I already tested but nevertheless I ordered some samples too. Same for the coils from coilcraft. Let's see what will come out.
The device I used was a MC34063. The min. time I found to reach 200V was about 6 ms with a several self-made transformers.
I expect better results from the LT3750 because of the tuned coils for the LT device.

Hope the parts will arrive within the next 2 weeks.

Best regards from Germany

Uwe

Futterama

Uwe, I hope you have more time to experiment than I do, because I would really like to know if it will work for the CDI.

Good luck :-)