I have a 48V 1000W controller I want to run 14S (51.8V) on. It will be about 58V full charge. I am wondering if the large resistor at the bottom of the board is used to step down the main voltage before going into the regulator then to processing circuit? This is a little different design then the older stuff I have worked on. I would always replace the large resistor when overvolting controllers, to put less voltage load on the regulator. I can't remember the math I used to calculate the replacement resistor, I will have to look into that. All the other components seem to be able to handle the voltage. I did purchase some rubycon caps incase I should replace the cheap ones on the board. Any advice would be appreciated! Thanks
48V YK33F controller circuit question for 52V
- Thread starter TMaster
- Start date
That's pretty likely, as that is the most common design for >24v controllers (brushed like this one, or brushless) that need an LVPS to provide 5v, 12v, etc within the controller. It drops the voltage down to something the linear regulators (LM317 etc) can handle, by wasting the excess as heat. The same thing is often done on SMPS type LVPS as well, though those are at least able to have a higher-voltage / wider input range if properly designed.TMaster said:
Those little SOT-223 parts near one end of the resistor are probably an LM317 and a 7805, just like in the typical brushless controllers.
That's usually a good idea, it will also reduce heat from the resistor itself.
I'm terrible at coming up with formulas, but you should be able to divide the new voltage by the old one, and multiply the existing resistor by the result to get the new higher value.
If you can then install it so it has more lead length and can be pushed up and away from both the PCB and the capacitor, it will help it not heat those up during use. The closer you can get it to the controlelr casing the better, without actually touching it. (a layer or two of Kapton tape between it and the case can help prevent touching and eventual abrasion of the ceramic coating allowing the resistor to electrically short to the metal casing...if both ends did that it would short across the resistor).
I'd use the rubycon anyway, since they're probably MUCH better caps and will have a longer lifespan while doing a better filtering job.
Thanks for the quick reply Amberwolf! The 2 regluators in the picture are a 78L05 5V and the one next to the end of the big resistor is a 78L15, 15V. Calculating 0.035A x 1180 ohms (1.18k) Im getting a voltage drop of about 41V. That seems like a lot, considering Normal nominal voltage of a 48V pack would drop it to only 7V for the regulator input. Now, I'm not totally sure what the current or amp draw is on the circuit, I'm just guessing on the .035, from previous experience. Both the 05 and 15 regulators can handle 35V max.
Thoughts?
Thoughts?
After thinking about this some more, I think i need to power it up and probe it with different voltage to see what's going into the regulators. There is a big difference between .02 amps and .035 amps on the voltage drop. Then I can get a real value on what the resistor needs to be changed to, or it may be fine like it is!
Interesting. Most of the stuff I've seen uses the LM317 instead of the 7815 to create the 15 (or 12) v output, that then feeds to the 7805, to create the 5v output.TMaster said:
I wonder if they are also running their inputs in parallel from the resistor, rather than the above series usage?
I also wonder if they are running the LM339 (the quad comparator that makes up the controller "brain") off the regulators, or off the resistor input? (could be done either way, as it usually handles the same range as the 78xx series)
Either way, the resistor is still being used the same way--drop the input voltage down just enough to keep it within the limits of the regulators--it just depends on the load being relatively constant and sufficient to cause enough current to create enough voltage drop.
For 48v, that's a max charge typically of 52v for lithium (13s "liion"), but lots of brushed controllers are used iwth lead systems, which is four 12v (13.6v-14.4v "full" or "charging") bricks, so 4x 14.4v worst case usually, or 57.6v. Max charge of another common "48v" battery in these, LiFePO4, would be 16s x 3.65v = 58.4v.
The current itself doesn't really "matter" for your purposes.
If you need the current for something (such as power dissipation required by the resistor) you can calculate it as volts / ohms. Just measure the actual voltage across the resistor, which you already know the resistance value of (it's marked by the stripes as 1.2kohm 5% tolerance per https://www.calculator.net/resistor-calculator.html). .
If the voltage drop on the existing resistor means it's output is well below the max for the devices it feeds, then you can probably use the existing one as-is. If it's marginal, you could change it for a higher value.
If you need to calculate a new resistor value, then if I did it right then the "divide the new voltage (58v) by the old one, and multiply the existing resistor by the result to get the new higher value." would (assuming 13s lithium, or 54v full previously) be 58v / 54v = 1.074 x 1200ohms = 1288ohms.
