Original schematic, design done by : Tony van Roon, thanks alot.
First I changed Q1, Q2, D1, R2 into one MOSFET N-FET. The current into D3 is not inrush protected so I added 10 Ohm in series with D3
I found out the 555 can go into lock-up mode ! where the PWM output is constantly HI !! this will blow up my glowplugs
The 10 ohm resistor I added in series with D3 helped alot on the lock-up, but I could not take anymore chances,
so I added a timeout circiut on the PWM output to the FET. Now I feel much more safe :-)
Since the PWM is a very low dutycycle, the glow will change alot when the input battery voltage changes,
So I am adding a voltage compensation circuit that will increase the dutycycle a little bit when the voltage drops.
The starter motor draws extensive current from the same battery so it is normal to see a voltage drop of 1 - 2 volts,
But without voltage PWM compensation the glow is much decreased if the battery goes from 12 to 10 V.
This is my modified schematic that works ! but still no voltage compensation added..
The TO-220 FET is not even cooled but still ice cold ! the input current is only 500mA at 12.0V when a normal glow plug is lit fine !
Resistor Rm is very important, I have tried without it but the glow is harder to adjust and the peak current is much more extreame,
also this resistor add the feature of over current protection.
Here is my final version ! I am proud to say that the glow level is now constant from 8 to 16 V input
Min pulse time = 50uS, Repeat time at min = 10mS (100Hz), dyty cycle = 0.5%
Max pulse time = 1000uS, Repeat time at max = 9mS (111Hz), dyty cycle = 11%
Here is my ugly breadboard test setup, looks funny, but easy and fast to find the best way to solve stuff.
Here is my PCB layout
The PCB is not made as small as possible, since I have plenty of room inside my fieldbox.
It is 72 x 50mm i have cut some of the edge away. it was fast and easy to make, not one single drill hole or via that is nice.
I have tested a Robbe 100 Powerpanel for comparison, here is the result:
Min pulse time = 55uS, Repeat time at min = 7mS (142Hz), dyty cycle = 0.78%
Max pulse time = 1200uS, Repeat time at max = 8mS (125Hz), dyty cycle = 15%
Same glowplug was used and adjusted to same glow, but the robbe uses 750mA at 12.0 V so the efficiency is alot lower
Also the Robbe 100 Powerpanel was really terrible unstable and could not lit up the glowplug,
before I mounted a 1000uF capacitor at the battery/starter plugs.
The Robbe 100 Powerpanel gives a loud zumming sound from the glow regulator, and it has NO voltage compensation !
Thomas Scherrer, July 2004