oyk

hi guys, just few minutes ago i manage to start my 1994 fe 600 WITHOUT the freaking charging coil
the charging coil (black and red leadings) is always the one that gets open and causes all the problems at the sem stators.

i only used 1 component for this operation and the results were really unbelieveble !! unfortunatelly i have no further time this week for a ride report, i only test it stationery.

i used neither a battery for this and the good thing is that it is fully adjustable for the optimum result.

i just want to complete testings so as to share the circuit with you everybody.

by the way, total cost for this was about 2.5 euros.

any comments will be appreciated

oyk

hello everyone,
after playing around with some batteries and some inverters for replacing the charging coil, i concluded that my first thought of using a 220/12 volts transformer connected in reverse is the best.

all is needed is a 220/12 volts tranformer (actually i use an industrial 400/24 volts one) 4 diodes for rectifiying and one electrolytic capacitor rated as 2 or maybe 4 microfarads / 450 volts for "smoothing" the output.

i connected the secondary coil (12volts) directly between the yellow leads of the stator (that are connected in paralel) and the blue one.

this this gives the transformer constantly about 13 volts in the secondary coil in a constantly changing frequency (increased revs=increasing frequency).

the result is about 220-250 volts (it depends on the frequency) at the primary coil.

then it has to be rectified with the bridge and the capacitor to give a maximum of 300 dc volts more than enough to supply the cdi.

connect "-" to the frame or the black lead and "+" directly to the red lead of the cdi and VOILA!

i get about 40 volts at kickstarting (at least!!).

the HINT is the selection of the transformer:

if possible to have a ferrite core because high frequencies beside higher voltage
generate losses too.

must have low impendance at the primary coil (bellow 200ohms if possible).

it is better to use one of about 100VA.

some 12 volts halogen lamps use some very compact transformers of a 100w power.

Smorgasbord

Almost done that.
Not on the SEM though, it was done on a 50 cc Victoria moped engine.
A transformer rated at 4W @ 50/60 Hz was (more than) enough! There was no regulator to bring the AC out of balance. The frequency was not as high as the sem will give in a husa engine. The discharge capacitor was only at 0.5 to 1uF and 400V. So it was easier to make it work. It even overcharged the capacitior and eventually it... Umm... Didn't I describe this previously? Wait a bit...
Here:http://www.husaberg.org/index.php?na...t=2408&start=0
Diodes and especially bridges for 50/60 Hz are not optimal, the berg will produce up to 1000 Hz. The Victoria couldn't have produced much more than 100 Hz! Try fast diodes instead. Also, try to run it without the smoothing capacitors. The voltage from the original charger coil is far from smooth. Also, the regulated electric tension from the lighting coils is AC with overlayered DC which is something that transformers do NOT like, the DC part that is. DC will lead to magnetic saturation and excessive heat build up in what you use as the primary winding (which was originally intended as a secondary one). Either filter the DC out with a large bipolar capacitor or else get a symmetrical regulator.Then will perhaps your small transformer be more than enough in your case too.
Maybe could one lighting coil be used unregulated if a suitable inductor was connected in series to protect the discharge capacitor from too high voltages at high revs. But that would take some experimenting.
A 100 VA transformer for 50/60 Hz is a bit too large, if you can get an aircraft transformer, they are designed for 400 Hz and are much smaller for the same power.
Your trick with a transformer is definitely what I'll try as soon as the stator breaks. But it has been around for a number of years and it just won't quit. But when the day comes...
An inverter is an alternative. Especially if the bike has got a battery. Emco is one inverter brand that could be considered unless you want to design your own.

Good initiative with the transformer. Thats the way they should have been built from the beginning.

Regards

oyk

thanks for the good points smogasbord, core saturation from the dc current is sotmething i intent to eliminate with a capacitor as you correctly mentioned.

a for the 100va is for now what i could possibly easily found in my storage house !

the problem i came up with smaller ones is that they had high output impedance so there was excessive voltage drop in rpms more than 2000

thats why for now i went with 100va one.

inverters needs a bit of attention (i mean the ones you can buy). they need very good power supply with not lot of voltage variations and harmoniques.
might also need a 1:1 transformer at the output.

an alternative idea is to built a simple oscilator with an ic 555 which can be supplied flawessly with a voltage from 5 volts to 15, and feed a reversed transformer.

thank you very much for your experienced advice and if any one else has a point is welcome

oyk

i also tried unregulated lighting coils with a smaller transformer but the voltage drop was the same with the dc overlayered voltage

Smorgasbord

The voltage drop can also come from the diodes if they are not well chosen as well as from the smoothing capacitors. At the higher revs the regulator skews the waveform to give more DC and all these things will work against you as revs goes up.
The classical 555 is great for many things. But look at for instance IR2153, IR2155, you will be delighted!, and there are probably even newer and better stuff around nowadays. These are bipolar oscillators and drivers in the same package and are useful with ferrite transformers operating at very high frequencies which therefore can be very very small for the power they produce. I'm sure that you can find suitable application notes on the internet to inspire you to make a good design. HOw much power does the discharge capacitor really require?, it should be possible to calculate it somehow. It will be proportinal to the revs, me thinks.

I hope to read about your progress here at the forum in the near future.

Regards

oyk

i will keep my progress posted,
it works perfect for now and with no heat produced from the transformer at all.
its size is about 7 by 7 cm but no problem i have plenty of room under the seat

jmb

Hello OIK and Smorgasbord

I have electrical troubles with my FSE 650 2002. It seems that the SEM charging coils are the cause (but i'm still not shure). Please take a look at my new topic at http://www.husaberg.org/index.php?name= ... pic&t=6998.

Are you in the opinion that the method you described in this topic "sem's charging coil replacement!" could be easily applied to my 2002 berg ? Any changes should be implemented ?

One more question (which might sound stupid but please take account that i'm not an electrical engineer..) : do the lights still function with your method ?

Thank you in advance for your reply.

oyk

hi jmb, of course you can do this modification to your 650 and get rid for good the sem problems. mine is working like a clock for about 50 hours now (from february 07 until now-temperatures from -5celsius to 48 celsius).
well by reading your post i would first eliminate the possibility of a faulty stator and a faulty cdi.

to test my cdi i just connected a simple 4 wire cdi out of a honda c90. it cost me 5 euro and 2 euros for the external hv coil necessary.

if you get a nice spark with the testing cdi then indeed your cdi has gone.
be sure to find a 4 wire cdi (ground-pick up signal-charging voltage-hv coil output).

you can also use this cdi for experimenting with the transformer modification which is highly recomended.

if you decide to do this please notify me so as to do this together step by step fo as to avoid any mistakes because several measurements must be taken first.

see at my gallery at the most recently pics and locate the white 50w trans i use now.

oyk

of course the lights are functioning