Discussion in 'Electronics & Electrics' started by RyoSaeba, Feb 15, 2021.
Oh yeh sorry it's 1k not 1000k lol. Typo. Definitely Brown Black Black Brown Brown.
I just realised it was my own post that mentioned the 1000k's first
ROFL at easy typo hahaha At least the jaycar links were correct so that's fine hehehe
Well I got extra Transistors and extra set of everything. Swapped out the BD140 transistor the way I should have it, still no ramping.
I took the extra parts and build a cleaner version again and still no ramping. Whacked a bunch of resistors to see if i can make any difference and still nothing.
Latest pic isn't working for me
I'll make this and check this out...
Much appreciated mate.
I feel really terrible about this, I apologise profusely about my oversight
It was in the back of my mind there was something goofy about that simulation as the size of the capacitor and resistor seemed low. It's running about 100 times slower than realtime and I should have double checked the calculation.
I'm just divvying up a circuit to make for bloody sure.
Actually I think I'm getting a bit confused now.
Looking at this diagram:
The BD140/BD136 PNP transistor, the incoming with the arrow is labeled E, the out going is labled C. Is that the right orientation or is that mislabled? I'm labling my Transitors like this img. So according to this, the incoming should be the collector and outgoing emitter right?
Oh ok so it's working just not at the speed we're expecting! Excellent now we're getting somewhere!!!
So the changes needed are:
1k -> 47k
10uF -> 100uF
Theoretical time constant (time to rise to 63% of applied voltage for the RC section) is 4.7 seconds, calculated by T = RC.
My test looks like this (note the resistors have big shadows behind them)
The results in video
With the LED strip, it doesn't noticeably illuminate much until about 6V.
Awesome! Thx again mate. I'll head down and grab them from Jaycar tomorrow.
You should also get a pack of 10k so you can tweak the timing if needed.
It works!!! Thanks mate
But I did have to use all 8 resistors to get it at a satisfactory speed. Can I just use something like a 470k 1W resistor instead? Like this one:
Also Is there smaller option for the 47 Ohm 5W resistor? Or should I just chain a bunch of smaller ones together as you suggested before?
Probably a 1 watt carbon film type resistor could be used for the 47 ohms. Metal film resistors don't like overload as much so they're not the choice in this scenario. The impulse peak is theoretically still on the high side, but a 2 watt one would be much larger again. Another option is to use 2x 100 ohms or 3x 150 ohms to make up the overall pulse handling ability (1.5 watts rated would be comfortable and should survive the most rapid of repeats). You can see in my test I used a half watt metal film and it would be fine most of the time, but it's possible that it could die down the track given enough ramps.
Another thing that you should check is what your actual voltage across the NPN C-E and the 47R resistor is, as this information confirms our actual losses in the operating state.
What value did you end up with in your tests?
The cap is getting filled by a current coming out of the PNP transistor as well as the resistor, so the higher the value of the resistor the less the timing follows the RC time constant. At 470k the timing would be all dominated by this timing, so if the difference is that large we need to either increase the cap size or do something different.
Since you probably have another PNP transistor handy, one option is that we can buffer the pnp transistor to reduce the effect of the base current on the timing components. The layout of this is below.
Even though the values aren't being changed significantly this will lengthen the timing by itself, then you can also increase the value even further. Beware though it will take time for the capacitor to discharge (reset) as well. The ramp will restart at whatever the capacitor value is.
So 47 Ohm 1W carbon film type.
Yup checking the voltage across the 47R shows that it does indeed start at 12V and drops to 0 eventually.
Yup that works. Now it's a tad slow though. The discharge time is ok. Not that long. It's tolerable. Which item do I adjust to change the speed of the ramping? Probably about halfway between having 1 and 2 PNP transistors.
And none of the parts got warm during operation. So I can safely house them in a printed housing I think.
If the 47R is dropping to close to 0V (not like 300mV or something like that) then actually the CCFL might be drawing much less current than 1A when on. If this is actually the case we can look at downsizing the components.
After the adding the pnp transistor, both the capacitor and resistor can be changed to adjust the ramp time as before. You can try reducing the resistor value to increase the speed.
Oh that's only because I've only got 1CCFL on there for testing. I know without any in between, the CCFL I use typically draws around 0.5A each. The inverter can support 2 CCFL so that makes 1A. I'll drop the resistor tonight to see how it goes. So close now I can taste success!!
Ok finally got time to do some more testing. Got it down to a desired speed with 10k Ohm resistor. Also the 47R is showing 0.4V when I put 2x CCFL on the inverter. None of the parts are warm to the touch so that's good. Now to work on tidying it all up. With the 47R, the thing is that it won't see much rapid resets. Collector's don't turn off/on their collection in rapid succession often. Usually they turn it on and then watch it for a while and turn it off. So it's not really going to see much rapid reset. With that in mind, what resistor would you recommend for this?
I had a look around, unfortunately I can't say for certain for the Jaycar sourced resistors, but it looks like 1/2W would be right on the edge of being within spec for another general purpose carbon film type resistor I could find for this kind of loading. There are ones that are rated for more in a small size but it becomes a lot more niche.
Being that you're putting it in resin, a thermo fuse of maybe around the 120deg.C could be an advisable addition. Look it's the fact that you're intending to also give it to other people, I'd tend to say it's better to be conservative. If the ramp portion becomes stuck at a mid voltage due to part failure down the track the module can excessively heat up.
Other things that can help in downsizing is that the PNP transistor can both be replaced with 2N2907's. They're not really that much smaller than the ones you have in practice but it could help. In SMD this could be made truly tiny - like a thumbprint size - but the board would have to be designed and assembly is a complication etc.
Oh I won't be buying it off Jaycar. I'll be buying either off Aliexpress or Ebay in bulk. So yeh anything is fine.
It won't be actually encased in resin. I'll just be using my resin printer to make a case for it that's all. I agree with over engineering it with conservative options though. It won't cost that much more anyway.
Size wise it's ok as it is. Only the 47R that's a bit awkward. But yeh with a change of the resistor I should be good to go. I could do it SMD, but yeh like you say that's a bit more complication with SMD. One day perhaps I'll get that done. Right now I got it down to this: