Good morning, all new printed circuit boards from jlc pcb uh the box kindly reshaped by dhl. Let's take a look inside here and these are oh very nicely wrapped. My new cell connect pcbs, which connect uh. My new lithium ion phosphate cells together and have a little chippy up at the top.

Let's take a closer look: yeah, let's get in close on that chip and use the magnifying glass and it is an acs-712 o five b. So it's a five amp uh current sensor, which will measure, i hope, the balance current and i just hesitated there hesitated there for a second, because i thought wait a minute. Surely the current measurements not on pins one, two, three and four, but actually it is so those are the high current pins. Why isn't that? Focusing yeah, so um, coming from the cell, where the uh silk screen text? Is there that's the cell interconnect that goes into pins one and two and then pins? Three and four go to those three bolt holes which will be for the balance lead so that chip just simply measures any balancing currents.

That's going to those three holes. Of course it's not measuring the main cell current going between these two large holes, because this is only a five amp sensor. There are a couple of uh 100 nano farad capacitors there. One is across uh vcc, which will be on that three pin.

Header um that'll be an arduino dupont, style header, the other one is the filter capacitor. So i went for 100 nanofarads again, i seem to remember um in a lot of the data sheet stuff. It's got 0.1 nanofarads and it was absolutely hopeless so for slow. Moving.

Current values, you want a bigger capacitor. I think i used 470 n, but that's a 100 n, so that should give a reasonable compromise. Um filtering function, so i went for 10 of these now. I think i'm going to need 9 because we've got 8 cells and i could potentially measure certainly all of the points where the two cells connect together.

But the extremes you might want to measure currents flowing in and out of the end points of my eight cell pack, so let's bring in my pack and see if these fit so here's my uh pack of uh lithium-ion, phosphate cells. It's quite heavy is this 32 cells, it's a 8 in series 4 in parallel, and i noticed that um, the two boxes of these cells actually had different, measured values. Some of them were around 6 000 milliamp hours somewhere around 6 200 milliamp hours. So what i've done is, i think, i've put them in alternate rows now, they're all in parallel, so the amp hour values will all just add up so hopefully all of these eight series cells should be roughly the same, but that's not what i wanted to check.

I wanted to check whether my boards, these fit over these studs hmm, so here we go. That's going to fit kind of like that. Yes, that fits rather well, let's get another one here we are and that's going to go there and then, on the other side, on the back of this uh, we'll put one across there, of course, but on the other side, so that looks pretty good. Let's get another one on there, yeah they fit rather well, and i did notice that um there are these flats if you can see them on the plastic moldings the frame, that's holding all this together and if i start putting bolt nuts on these, it will pull The cell into the plastic molding so it'll help to hold this thing all together in a sort of coherent, homogeneous thing now notice that the braised on copper thread has a slight ridge on it, which sits slightly off the surface of the cell here and that's good Because it looks like it's at about the same height as the plastic molding there.

So when i tighten the nuts onto here, let's put a nut on here um. This should all tighten down and all be reasonably flat. There won't be too much flexing of the pcb, so that looks pretty good now. The other thing was that i noticed that these studs weren't necessarily exactly centered um on the cells, so i've made these holes a bit bigger they're.

I think five mil and these studs are four mil thread, so there is a bit of looseness there, but i think i've got that about right because those fit pretty fantastically. Let's get some nuts on here and so the balance points are sitting up above uh, the top of the frame here, so you can put up to three bolts in there now those are m3s, so i can use the smaller ring terminals to connect uh well, they'll. Be the bms there'll be um, a set of balance leads for my little um monitor, display, of which this is one, but i've also got the little um oled version as well, and then the third one will be for the the active balancer. So you can have up to three devices on each of the cell interconnect points, but let's get these four on now i made these um not full height, because the problem is the distance between these cell centers is something really horrible.

It's not on a nice grid. Point it's not quite 34 millimeters. I think something like 34.25, so i figured. What i might do is make a second pcb, which is just literally a square to go down here and then bridge across with a couple of my wire links that i made up these.

I mean i could bridge all of this bottom section with these wire links, but i worked out. I'd need 32 of them, which is quite a lot so i'll. Just put the fourth one of these on there and i'll stick some nuts on here. I'm going to need a lot of these m4 nuts.

Aren't i that'll be the next thing i'll run out of m4 nuts um and then i'll put some of these on the other side. Now um, these are going to go in the middle positions. On the other side, so three of them will fit neatly the two others. If i use them, will sit out of the battery pack, so they won't be quite as neat, but i could make another board with an acs-712 which is just a sort of single width.

Just linking two cells together, but there's no real need for that, and i do have 10 of these pcbs. So let's just see how we go with these now quite clearly. I don't want to put this board on directly opposite that board, because then i'll have a complete loop and very large amounts of current will flow. This um copper area here, unlike my distribution, xt90 board, oh, which i have here this one - where the top was 24 volts and the bottom ground on this board top and bottom copper areas are both the same net they're, both a net, which i called cell, because It's the cell interconnect and i've actually peppered this thing with a few wires.

I'm not sure that was absolutely essential, so we just got two copper planes carrying the same current um. So yes, i very much definitely have to put that here. So that should be fine to put it there in the gap opposite these ones. So let's get some of those on um.

I was going to say something else about this. Oh, yes, this business, about current flowing uh left to right here could create a magnetic field. So i've got the current flowing in the current sensor, chip in the perpendicular direction and i'm hoping that currents flowing in this main copper area won't affect the current reading of this chip, because it is a hall effect. Magnetic chip won't affect this too much, but only time will tell on that yeah.

So this one will go there. Won't it! Yes, no explosion! That's good um! I'm i've found i've got washers for these, so i'm gon na put washers on and then, like i say on the end, it's going to be like that, so this will stick out over the side, but i may want to measure the balance currents coming on to This end point and also this end, point um the way this is laid out because it's eight cells, fortunately, both end points are on one side, so i can put this on the back of my shelf. This is an end point. This is an end point.

In fact, if i get all these pcbs on, i should be able to put my 24 volt bulb across there and it should light up nice and bright. Now. Naturally, i used the jlc pcb smt surface mount uh technology assembly service for these boards, because i didn't want to have to solder this chip and those two capacitors on. So that was all done for me.

Um these chips actually were surprisingly expensive. I think they were three dollars each so of course, 10 of those uh 30. I think the total order value and, of course, that's including dhl postage, was something around 75. So just go around these with a seven millimeter.

It is a spanner tightening these four millimeter. M4, nuts just tighten these up, i'm not gon na massively overdo it. I just want enough tension that we get a good connection to the pcb. Someone was talking about the crush force of these fiberglass pcbs um.

I think i looked this up and it was something similar to aluminium, so they take a fair amount of crushing force uh. I don't think you're going to collapse, that's the through plated hole by applying a huge force. I mean i'm not going to talk these up with a massive force, i'm missing a nut there, let's find a nut and then i'm going to see whether that across there, yes, 24 volts is safe. The phone's ringing right - let's flip this over and that's the other side.

I haven't put washers on this, so i'm going to take all these nuts off the few that there are put washers on put nuts on and that's my pack now, because i'm only using the top two rows and, like i say i'll, perhaps make some more pcvs For the bottom rows and then make up link wires for that, i think that's what i'll do that uh phone call was uh. Thank you for placing a 400 pound order for something you don't remember, placing a 400 pound order for now that you're suitably shocked and appalled that uh you've placed this 400 pound order phone phone, this number and be scammed out of all your money. Yes, so we don't respond to those. Do we good so gradually populating this thing with pcbs? Why isn't that one going down? I don't know um washers and nuts.

Let's get this whole thing linked up and i should have a 24 volt lithium ion phosphate battery pack and that's it all. The nuts are done up and tight. Now you can see what's coming next, can't you um, i must admit i did hope. I was hoping that this three pin array might be clear of the top of this plastic, but it's not so the pins will have to come out this side.

But, yes, you can see where we're going. I'm going to put a board on the top of here, which is going to be a homebrew arduino board, a couple of oleds and all the wiring connection points for these acs-712s and this board will show at all times on these oleds um all the balance, currents That are coming into and out of all these cell uh interconnects, so that could be quite exciting right, quick test for the 24 volt bulb. Let's go most negative over there. Um i'll hold that onto the side of my 24 volt bolt.

This is going to be bright, isn't it put it on there 24 volts going into my bulb very nice. Of course that's only using the top row of cells, so i don't have a link down there, but let's try putting two more of these boards on that. Stick out the size and just see how stupid it looks. So that's the five interconnects - and these are the extremities of course, that's most positive and that's most negative.

So i'll do my bulb test again, which will go on there and that lights up nicely. Um, yes, and that course means that we can measure all nine of the balance currents from battery minus the most negative point to battery positive. I mean there will be a current flowing uh into and or out of all nine of these connect points. Five on this side and four on the other side - and that of course leaves me with one of these left over - i ordered 10, but that's okay, that can act as a spare.

Now this issue about using uh pcbs as bus bars. I reckoned, i think it was something like um. This would be good for about 20 amps per side, so i'm thinking 40 amps um for this entire pcb. Now you wouldn't normally.

I think if you were going to pull an arbitrary amount of current from this pack, say you're going to hook it up to an inverter and i know, run a toaster or something you probably wouldn't choose to use printed circuit boards as your cell interconnects. But of course, i'm only pulling a known current, which is going to go into my ant miner um. Now, if i'm saying that i can pull about 40 amps through this board, then that's at 25 volts at 12 volts once it's gone through the butt regulator. Of course, it can't handle 80 amps, but we should be looking at about 80 amps um into the crypto miner, before i'm likely to have any uh temperature significant temperature rise on these boards.

I suppose thinking about it. These uh vias i put in here if this started to get warm, you could perhaps think about it, delaminating slightly away from the fiberglass substrate, and these wires could help to keep the two sides held together. I think if you were getting to the point where this was getting significantly warm you'd be in trouble anyway, but uh yeah. I certainly think these are going to be fine in my application, but i probably wouldn't recommend pcbs as bus bars if you're going to pull an arbitrarily large current from your battery, and most people probably would want to do that.

I don't yeah. So the next couple of pcbs, i've kind of spoiled the surprise a bit really um. As i say, i'm gon na need a square one for the bottom and then i can just link the top and bottom current connect, pcbs, um and actually, if, if i'm saying, i can put 40 amps through there, then of course with a second board down here And some thick interconnect cables. I should be looking at 80 amps.

I don't think it's going to get to that point. Um, but yeah pcbs can take a fair amount of current um yeah. So these bottom pcbs will i'll have to do and, like i say, the whole point of making these boards with the acs-712 current sensor and of course i chose this because it's galvanically isolated, there's, no electrical connection between the current measuring thing. Here all that's inside this chip is a piece of wire running across there and then a hall effect device on this side, which you can reference, and i can reference the five volt um supply supply power to these chips, all from the same five volts, because, as I say: there's a galvanic isolation down the middle of the chip here, um.

Yes, an arduino board here with some displays telling us um the in and out currents on every single one of these nine pcbs. So that is my latest uh printed circuit board. It is a cell interconnect pcb with integrated current measurement chip. I think that's quite neat.

What do you think cheerio.