$2 for 5PCBs - order with coupon code: JLCPCBcom Any questions about JLCPCB or ordering, pls email support@jlcpcb.com Fitting opto-isolators to the super capacitor protections boards.
can you share the updated PCB stl (whatever PCB formats use), is there a video on building the super caps and what is the tool your using to pick up the SMD components?
I was just wondering where that useful little soldering iron stand came from? I just typically use a regular stand doesn't necessarily fit that great but it works that one seems very simple useful I've printed a few different designs on my 3D printer to hold the irons but none that I fell in love with not one's not anything special I'm sure I could probably get one from China but was curious where you got that one? Also I was wondering if you were to change the values on those load resistors would that change I'm assuming those are for discharge but I don't know exactly. I'm asking because I'm looking at changing them on my liitokala li500 charger. Would like to try to change them to get it to discharge more It says that it discharges at 1000mA But that is not the case would actually like to be able to discharge it at one amp just curious if I could make that happen by changing the load resistors on the charger? Look forward to hearing from you thanks! Also nice content keep it up!
you're going to put the optos in series so it only signals when all are charged or you will use some multiplexers to pass to the controller which ones are ready ?
Would be interested in seeing the inside of some of your broken supercaps. Find out if they were faked in some way or if the shelf life really is as low as 2 years.
For mounting, I would have suggested drill the holes 1 size too small (over here you can get saws in 1/4 to 1/8 inch steps depending on needs) and then cut the wood lengthwise and put screws thru to firmly clamp the caps between them.
I'm also in the process of doing a complete overhaul of the protection boards for my 3000F 2.85V supercaps (12 caps, 1000F 17V max in series-parallel). Because I'm using my supercap modules as a car battery replacement, I want charging to be aggressive (200A+ from the alternator). *-Transistor Swap:* replace existing 2SD1804 NPN transistor with NCE2060K N-channel MOSFET with 1V gate threshold voltage 6mΩ RdsOn, 60A drain current and 60W(Tc) heatsinked power dissipation rating . I found that the old NPN setup dropped as much as 1.3V across it and maintained an average temperature of 135-160 degrees C while balancing using 0.5Ohm resistors. The poor performance was likely due to insufficient base current. There aren't a lot of cost effective high current (>10A) NPN transistors on the market, so switching to FETs was an obvious choice. *-Logic Transistor Swap :* replace existing S8550/2TY PNP transistor with HX2301A P-channel MOSFET with 1V gate threshold voltage. PNP signal transistors have a fairly high collector saturation voltage: They're not ideal for driving MOSFETs with only 2.5V of headroom when they themselves drop 0.6-0.8V. Also the previous PNP transistors maximum collector current would be exceeded if I wanted to drive the 2SD1804 load transistor at any current greater than 5A (gain drops severely with current). A P-channel MOSFET will allow the main N-channel MOSFET to be driven with the highest possible gate voltage and lowest RdsOn. *-Balancing Resistor Swap:* replaced pair of 6.8 Ohm 2W surface mount resistors with 50W power resistors (0.135 Ohms) to upgrade balancing current to 20A per cell from a paltry 400mA . I could have purchased 0.015 Ohm resistors to max out the transistor rating, but at 20A, the FETs stay cool, and the total balancing current will be 240A, which should be more than enough for a car alternator. I was skeptical that the SMD heat sink for the 2SD1804 I also purchased would be able to dissipate 20W of heat at 60A and 6mOhms of resistance anyways. Here's a photo album showing the caps and balancing boards. My MOSFETs and heat sinks are currently shipped and on the way, but the system is currently functional with an acceptable balance current around 3-4A with the existing NPN system. The only component I'm concerned about now is the 18 AWG wire going to the power resistors. Two AWGs in parallel drop the effective AWG by 3 sizes, so two 18 AWG wires in parallel would be enough for even 30A continuous. That being said, balancing will never be continuous, since alternator voltage will never exceed 14.8-15.2V which is 2.5V per cell (balancing is set to kick in at 2.67V with an onboard voltage divider). In a solar setup like yours, however, I would have rated my balance current to bleed the full power at MPP. photos.app.goo.gl/NPHGyuRrHg5sK2Wa8 edit: formatting
@@atskijvinny Certainly, First, here is Julian's unmodified balancer circuit, credit to wictor on the eevblog forums: www.eevblog.com/forum/projects/chinese-tl431-super-capacitor-balancer/ Next, here is my modified balancer circuit made for 20A balancing per cell (more if R5 resistance is reduced) R7 is fairly flexible, it serves as a soft pull-down resistor photos.app.goo.gl/yV3n1sAmudKa6ZR98 You could also place an LED + resistor in parallel with R7 for balancing indication.
@@atskijvinny Hey atskjvinny, upon futher investigation, the original BJT setup could be upgraded to 5.5A (3A in Julian's circuit) by replacing the TL431 with TL432 upside-down and using two identical resistors > 10k in R2 and R3 to balance at 2.5V. If R3 is 10% higher than R2, the balance circuit will kick in before the supercap hits its maximum rated voltage, and prevent it from being overstressed. The problem with TL431 was that it only pulled its output down by 0.5V to 2V because it shares its VCC supply pin with the output, and the minimum voltage for TL431 to operate is 2V. This is why the balance circuit was performing so poorly. MOSFETs are still a superior choice if you want a balance circuit that can keep up with rapid charging, however.
weird and probably stupid question but...can I rearrange the circuit on a 60S, 160V BMS PCB without disassembling it? As in by drilling out or otherwise severing and and then reverse wiring connections turning it into a 6S 10P board? It's a high end bank and I really don't want to completely destroy the board...nor desolder 60 capacitors - ugh...or, barring that, how do I step it down to a more common voltage?
im not sure i understand the thinking on that lay out . i would imagine keeping the shunt in place too lower any over voltage would be usefull and then use the opto's on the points you see the led's on would work best so you still have a safty feature as well as your power shut down .
I see your reasoning that the red LED lit, so should be enough to light the infrared LED in the opto, and it is switching, I guess it depends on the characteristics of the opto, I still can only think its not fully on, so we shall see what happens when you 'gang' them all up...cheers!
In theory those optos work great in this setup but.. The combined way the transistor sides are in parallel they come on too early without a load. I tried this before. You really need Schmitt trigger opto isolators is what I found.
I actually quite enjoy the dead tree carcasses and lack of 3D printing, I'm actually tired of seeing people here on youtube using 3D printing for everything, even stuff that would be cheaper, better, stronger and faster just using off the shelf stuff.
The red LED is so dim, what's the voltage drop across it ? that opto is probably not fully turned on because clearly very little current is flowing through the LED as there is no current limiting resistor...so my guess is that the opto isn't switched on...cheers.
I've found a different way to check the optos - DMM set to ohms across the opto-transistor (which are now all bussed together). When the protection reaches maximum current, the opto 'resistance' is about 1k.
Does look like the opto is not being turned on enough. Need to know were Julian connected it to, by just cutting the track you are limited to whatever current was flowing and it would seem not to be enough.
Yes they work great and they do not absorb heat like other tweezers so it's easier to hold things without sucking heat out of them. I would get the I don't remember exactly what they're called but I would get the ones that are reversed. Kind of like the forcep tweezers where they lock and hold except these hold under their own reverse tension They work great I use them all the time to hold small stuff in place every time I use normal tweezers I seem to either not squeeze it hard enough or squeeze it too hard and those little fuckers go flying and I can never find them lol
@@luckylarz4269 They aren't AS thermally conductive, but ceramic is still pretty thermally conductive. You can tell by how ceramic feels cool to the touch. However, it's like night and day versus metal tweezers, but there are better materials for low thermal conductivity. Just not in tweezers :)
A current limiting power supply (lab PS) would work *_much better_* than a coin cell for testing. That small battery just can't provide much current and frankly is a waste of money to use here. Not to mention even if you recycle it it isn't good for the environment whereas your PS running off your solar power costs nothing and there is nothing to recycle. (Until the solar panel dies in 25 years I guess!)
Using a pen is a *bad idea* to hold down the SMD LED while soldering! Unless you want to colour the whole PCB blue. ;-P Try a toothpick if you don't have the right tweezers handy.
