• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

Any electronics hobbiests want a project?

MakeMePower said:
based on what I am seeing, you need at least 10v to get to 60v for a 48v pack. thats 4 li-ions in series to get there. Box would actually be kind of big...

the other issue I am having, is how exactly would you use this? you need a charger, and how long does this thing need to run to wake a bms up? and what amperage? considering this would need to be a universal unit, i think we need to know more about how long at the voltage and what amperage. you might need IMR cells, this could get expensive fast.
Click to expand...

So a rough thoughts because I can't sleep it seems right now.

1. You don't need 4 cells to get to up to the output voltage of 16s LFP packs from a single NCM cell. There are some new dc-dc controller chips that support a 100% duty cycle for what is definitely not the best idea. Would definitely be easier with a higher input voltage

2. My working design idea was 1a at 52v max output, though I may reduce that, worst case that's less than 30A. Completly supportable by single units of cells like a Samsung 40t, Moilcel p45b or EVE 40PL. (Shill time, data sheets in the resources section)

3. The bom cost for doing this on a single cell pack is going to outweigh the the cost of the cells listed above. It's not a cheap converter.

4. I am not trying to design this as a useful product, I want to do it for fun as a learning experience. Through most of the converters I seen to be doing at the office are similarly crazy.
 
Last edited:
MakeMePower said:
chargers are cheap for this, may take hours, I am more concerned about what kind amperage you need to kickstart them, and why exactly the charger your using can't do it.
Click to expand...
Because a lot of times the only charger you have on the system won't turn on because the battery has gone into safety mode. So you need to charge the battery to turn it on to turn on the charger to charge the battery. If you can get just enough juice to wake up the battery, the rest of the charging system will come online and you're good to go.

And if anyone says "Just buy a $500 chargeverter and a $400 gas generator and wake it up that way", you can suck my dirty, hairy... toe!


SparkyJJO said:
Feeding those lions could be problematic too...
Click to expand...
Friskies? MeowMix? 😜
 
zcskywire2 said:
So a rough thoughts because I can't sleep it seems right now.

1. You don't need 4 cells to get to up to the output voltage of 16s LFP packs from a single NCM cell. There are some new dc-dc controller chips that support a 100% duty cycle for what is definitely not the best idea. Would definitely be easier with a higher input voltage

2. My working design idea was 1a at 52v max output, though I may reduce that, worst case that's less than 30A. Completly supportable by single units of cells like a Samsung 40t, Moilcel p45b or EVE 40PL. (Shill time, data sheets in the resources section)

3. The bom cost for doing this on a single cell pack is going to outweigh the the cost of the cells listed above. It's not a cheap converter.

4. I am not trying to design this as a useful product, I want to do it for fun as a learning experience. Through most of the converters I seen to be doing at the office are similarly crazy.
Click to expand...
im just looking at cheap converters on ALI, what converter are you using? single cell capacity could be a issue no?
 
MakeMePower said:
im just looking at cheap converters on ALI, what converter are you using? single cell capacity could be a issue no?
Click to expand...

Ground up custom converter built on a TI LM5171

LM5171 data sheet, product information and support | TI.com

TI’s LM5171 is a 80-V dual-channel bidirectional buck-boost controller. Find parameters, ordering and quality information
www.ti.com

Runtime will suck, mabye 5-6min at best, but that's plenty to get an actual charger plugged in and running.
 
Bit more thoughts as a follow up,

Power stage would be a TI LM5171 2 switch buck boost in multiphase mode. It's a dual phase chip. TI doesn't make a boost only version yet otherwise I'd have used that.

General design design would be rated 15Ah/phase with 80V rated parts. Estimated duty cycle is 96% at worst , lm5171 supports up to a 99% duty cycle so we're good there?

USB charging, likely 3A, without USB-C PD support. Simple TI BQ25611D. No point in charging those cells faster which is what the PD support would be good for

MCU host is needed to control the LM5171 over i2c probably drop a GD32E230, unless I decide on a screen (never did anything with screens yet). Probably go with a faster one then if I did
 
zcskywire2 said:
Bit more thoughts as a follow up,

Power stage would be a TI LM5171 2 switch buck boost in multiphase mode. It's a dual phase chip.

General design design would be rated 15Ah/phase with 80V rated parts. Estimated duty cycle is 96% at worst , lm5171 supports up to a 99% duty cycle so we're good there?

USB charging, likely 3A, without USB-C PD support. Simple TI BQ25611D. No point in charging those cells faster which is what the PD support would be good for

MCU host is needed to control the LM5171 over i2c probably drop a GD32E230, unless I decide on a screen (never did anything with screens yet). Probably go with a faster one then if I did
Click to expand...
yea... I'm not at that level, how you going to handle the 4 levels of output? fixed resistor values?

If your not doing PD, most people won't have a 3A capable USB charger, most chargers will barely break 2A on a good day if your at 5v. Unless your talking QC capable, which most people prolly don't use. or even know what we are talking about. I love my PD chargers, but people won't spend that coin, cheaper to buy a wall wart and just plug it in, at least you can control the charge design
 
MakeMePower said:
yea... I'm not at that level, how you going to handle the 4 levels of output fixed resistor values?
Click to expand...
Programed over I2C via a host processor. Adjustable on the fly

Most idiot proof design is to have the code start at a low output voltage and slowly ramp the output voltage up until we see a solid load on the output. At that point stop raising the output and hold until shutoff or out of battery


Edit on wall warts:
Most USB-C PD chargers will give the full 2.2A on a simple USB 2.0 implementation on USB 3.0 type c cabling. I used the same charger IC for a suspiciously similar, though much more reasonable PCB.
 
Last edited:
zcskywire2 said:
Programed over I2C via a host processor. Adjustable on the fly

Most idiot proof design is to have the code start at a low output voltage and slowly ramp the output voltage up until we see a solid load on the output. At that point stop raising the output and hold until shutoff or out of battery
Click to expand...
that's impressive, your well beyond my ability, never thought of using a mcu to make the voltage move, thats neat. I have limited AT and ESP experience.
 
MakeMePower said:
that's impressive, your well beyond my ability, never thought of using a mcu to make the voltage move, thats neat. I have limited AT and ESP experience.
Click to expand...
If you have a "dumb" fixed ratio dc-dc IC that uses a resistor divider to set output voltage you can use spare I/O pins as a way to change that divider. Put a bunch of resistors in series an use the spare I/O pins as connections in between them that can be switched to ground. Simple way to have a few fixed output points. In a more advanced version you can use digital potentiometers to have a wide range of voltage settings.

Granted I'm cheating here by using a smart dc-dc IC, I tell it to do an output a voltage through software and it does.
 
Last edited:
zcskywire2 said:
Edit on wall warts:
Most USB-C PD chargers will give the full 2.2A on a simple USB 2.0 implementation on USB 3.0 type c cabling. I used the same charger IC for a suspiciously similar, though much more reasonable PCB.
Click to expand...
As i state most people won't have PD chargers. most will have dollar store usb and cable.
 
zcskywire2 said:
If you have a "dumb" fixed ratio dc-dc IC that uses a resistor divider to set output voltage you can use spare I/O pins as a way to change that divider. Put a bunch of resistors in series an use the spare I/O pins as connections in between them that can be switched to ground. Simple way to have a few fixed output points. In a more advanced version you can use digital potentiometers to have a wide range of voltage settings.

Granted I'm cheating here by using a smart dc-dc IC, I tell it to output a voltage through software and it does.
Click to expand...
yea i was just thinking a rotary switch to be honest, don't see the need for the mcu, turn a knob push a button etc.... but i get your example to control it, I am cheating with gpt to learn your i2c code for that chip referenced above.
 
Rednecktek said:
You looking to build or just design?
Click to expand...
Probably get to at least layout, which is when the PCB becomes orderable. An actual mechanical casing is beyond me. Code is also beyond me on these, may move it away from the GD32 just so things are more programmable for the average person.
 
When the PCB layout is finished and all x-y-z dimensions are established. Maybe brainstorm about already existing boxes that would fit the project? Instead of designing and manufacturing from scratch.
 
Rednecktek said:
Are you implying that someone on here DOESN'T need an excuse to use their new 3d printer? 😜
Click to expand...
using and designing are completely separate entities. going to need a fully functioning item to design it. sounds like we don't have that
 
Rednecktek said:
Are you implying that someone on here DOESN'T need an excuse to use their new 3d printer? 😜
Click to expand...
🙃
Well, I think the one does not have to exclude the other and two options are better than one and better than none.

For prototype, sure, might be 3D print custom. But is 3D printing on bigger scale economically viable compared to finding already existing that fits?
 
You must log in or register to reply here.

Similar threads

M
Recovering a usb-c battery pack, suspect electronics issues
Replies
4
Views
234
mrfrozenpotatohead
M
liftbikerun
Ryobi RY48ZTR100 Aftermarket Blade Motor ESCs Replacement Tutorial
Replies
14
Views
3K
Ryobi480
R
O
Finally going off grid, need a bunch of help
2 3
Replies
51
Views
6K
michaeli
M
J
Building A New 48 Volt System, Planning To Execution...
2
Replies
25
Views
5K
JeepHammer
J

diy solar

diy solar
Back
Top