It is used as a current limiter.
- 1N4007 has 1A continuous current rating and 30A surge current rating. I don't want to burn it at the begining of the capacitor charge up.
- When the flash is fired, this resistor will slow down the charge up. That will let the voltage drop so the flash can stop. Otherwise, the flash will keep on burning until it overheat and break.
The exact value of this resistor can be varied and it should be determined by your expectation and real performance.
Also, if you check it carefully, the rating of the fuse is confusing. The initial input current should be 200v/39ohm=5A. I choose 1A because it works and it looks more safety (should choose slow blow fuse).
The resistor will not be hot unless you flash it in high rate.
It's so great that somebody is doing the same project.
Sorry, I am too busy so I don't have time to continue my posting now but I will make it ASAP.
I have checked my prototype. The rating of the fuse I posted is wrong. The actual rating is 3A. Choose slow blow type or use 5A.
Again, the initial current is controlled by the resistor at the input side. A lower rating can make the charging faster. How fast it should be? I don't know. The commercial product may be lower than 1 second. It's no problem to make it 1 second (or even 1/10 sec). It depends on the your expectation and real performance. Change the value of the resistor and fuse can get a different result (you are dealing with 350V DC so don't forget your safety).
Yes, each light tube should have it's own set of cap and they should be seperated by a resistor.
Otherwise, there will have problem exactly like you described. I will post the circuit later.
Also, I tested some slave trigger circuits. Pre-flash trigger prevention is also part of my project. It looks not difficult but I want to prevent the use of logic circuit (TTL or CMOS) but looks impossible.