0
\$\begingroup\$

Per a members request I am making a post compressing multiple possible posts into one and providing additional info.

I am making a 18650 powered portable electronics workstation that will have many features. I would like the following features:

  • upgradable at any time for more capacity and still be able to use the same components
  • have integrated charging capability
  • be as cheap as possible
  • automatically switch to charging mode when cable is inserted
  • display an approximate battery level
  • be fan cooled while charging and when a power output goes past 50 degrees celsius.
  • have adjustable voltage and current output , a 24v output port for a ts-100 soldering iron, a 12v soldering fume extractor port, a 20v protected laptop charger port, and a 5.2v protected phone charger port.

Pseudo circuit schematic is below. I'm sorry it's messy; I was drawing fast. schematic

Can anyone recommend a better system, safer laptop/phone charging, and/or check over my "schematic"? I will be putting the tp4056's, relays, and cooling fan regulator on one perfboard sheet.

Every battery has its own protection circuit. I forgot to put the battery meter on the drawing, and the switches. Every power output will have a switch on the positive wire BEFORE it goes to the regulator.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Don't. Use. The TP4056. Unless you need exactly one cell to be charged. It's a single cell charger. And a less than perfect one, too. And the fact that you want to actively cool those things is another indication you shouldn't be using it, but a proper switch mode multi-cell battery controller. \$\endgroup\$ – Marcus Müller Jul 30 '18 at 22:48
  • 1
    \$\begingroup\$ and, seriously, where does your 5V 8A come from? Sounds like an extremely strange choice of power source. \$\endgroup\$ – Marcus Müller Jul 30 '18 at 22:49
  • \$\begingroup\$ Converting this to an answer. \$\endgroup\$ – Marcus Müller Jul 30 '18 at 22:58
4
\$\begingroup\$

Don't do this. It's a bad design!

  • You have more than one cell. Don't use the TP4056. It's a single cell-charger, and a less-than-perfect one at that. You don't want a lot of current going through linear regulators. A good design wouldn't need a fan for cooling.
  • A proper battery management IC would make the relais unnecessary. And thus, save a lot of money in your design.
  • You want outputs that are all above your battery voltage. That requires stepping up. Stepping up takes place, cost, and efficiency. Avoid when possible. And it's easily possible here: just put your batteries in series to give you e.g. 6·3.7 V = 22.7 V and charge with a multi-cell charger IC.
  • 5V 8A: That's an insufficient power supply. With linear regulators, charging your 8× 3.4 Ah will take eons. With more efficient switching regulators, still ages, if, and only if, your power supply doesn't give in: it's severly underdimenstioned for the job.
  • since you should be using cells in series, you'd normally just go for a cheaper, higher-voltage supply. What's wrong with 48 V? You can get such supplies cheaply; and then, feed switch-mode regulation from it.
  • Perfboard: nope. This is a high-current, explosion hazard if short-circuited design. It gets so much easier if you just layed out a PCB. Easier to design, since you can lay arbitrary thick traces. Easier to build, since you don't have to solder anything to connect things. Easier to test, becasue you can trust your traces to work. And safer. And saner. And prettier. Probably way more compact, too.

All in all, I think your design process was more like "I have this 5V supply, and I've heard of the TP4056, this must be the way to go" than "OK, I have these requirements, how do I fulfill them?".

That's a typical beginner's mistake – don't worry, we've all been there. So, take a step back. File your schematic under "learned a lot while drawing it, but not worth following"!

Start by making a proper list of what you want to do:

  • Which output voltages
  • Which output currents
  • Maximum instantaneous overall output power
  • Duration for how long you want to supply that

Based on that, decide on a multi-cell configuration that suits your needs.

Based on that, decide on output regulators that fulfill your needs (hint: drop the laptop charging idea. It's inefficient to discharge one battery to charge another, and laptops take a lot of current, so this is a very hard and expensive one to implement!) .

Based on the batteries, decide on a charger. There's multiple companies that build good charger ICs for multi-cell things. Maxim Integrated and Texas Instruments spring to mind; especially the MAX1640 might be the right choice for you. It can deal with 6 LiIon cells in series. At the powers you're dealing with, going with a linear regulator is really no option, so you'll need to go with the datasheet of the IC you select, and design a switch mode regulator. It's not that hard, but not trivial, either. Good news is that switching regulators are way more efficient than linear ones, so you might even get away without a single fan.

That also sets your charging voltage and current (hint: voltage will be a lot higher than 5V, and current potentially higher than 8A, too, depending on your design). Based on that, and thus, pretty much as last design step, choose a power source. It's not going to be your 5V 8A thing, whatever that was.

All in all, I think you might be underestimating the complexity of building the things you have planned. I think you even might have skipped the step where you roughly estimate how much power your primary power source can offer, and how much it'll take, even if you had 100% efficiency, to charge all your batteries (hint: energy = time · power= time · (voltage·current); energy in battery [Wh] = voltage [V] · capacity [Ah]; your 5V 8A supply only offers 5 V · 8A max = 40 W).

\$\endgroup\$
5
  • \$\begingroup\$ Thank you for a very thorough answer. I was planning on using perfboard because I have never made an actual pcb before, and because if I make the distance between the pins of components the wrong size I can always just use a different hole and slightly adjust the design. Is there a software option to make a pcb without to steep of a learning curve. I wanted to use at least 8 batteries for around 100 watts of power @ 3.4 ah each. I harvested a few 18650s from some old power tool battery packs but they have different capacities and different amount of use so I will have to buy new cells. \$\endgroup\$ – 5hutd0wn R35tart Jul 31 '18 at 18:27
  • \$\begingroup\$ I don't know – I use kiCAD these days, and am pretty happy with it. Many beginners start with Eagle. Sites like oshpark.com simply take the files these produce directly and send you boards. \$\endgroup\$ – Marcus Müller Jul 31 '18 at 18:29
  • \$\begingroup\$ I would think that an 8s charger would be even better because that would give 29.6v average which is above all of the outputs even when drained to 2.75v per battery. Alternatively if this system saved significantly enough money I could probably spring for 12 18650s and have a 6s 2c pack if you think that would be better. \$\endgroup\$ – 5hutd0wn R35tart Jul 31 '18 at 18:32
  • \$\begingroup\$ I believe that a TI BQ24610 could be what I need. I cannot find the maxim chips current rating but it looks like it could output 1A max. The schematic for both of the chips only has one battery output so how do these chips balance charge? I do not believe that you can charge li-ions in series without balance. \$\endgroup\$ – 5hutd0wn R35tart Jul 31 '18 at 20:36
  • \$\begingroup\$ The BQ24610 you linked does not balance, it's just a charger. For cell balancing you need a BMS chip like the BQ76930: ti.com/product/bq76930/description \$\endgroup\$ – Nate S. Dec 6 '18 at 18:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.