Project PITS update

RE: Project PITS update

three-phase — Mon, 10 Dec 2018 21:04:14 GMT

Thanks for your help Gene, I will hook up this circuit over the next few days and see if I have any better success with it.

Kind regards.

RE: Project PITS update

genebren — Sun, 09 Dec 2018 23:55:53 GMT

Donald,

OK, the modified circuit does make a bit more sense now. I am not sure that the gate voltage is a part of your problem, as you did initially turn on the MOSFET. The extremely low collector current (~7.5uA) might be keeping the Vce of Q3 from keeping the voltage on C5 low enough to turn off the Q2 (which in turn shuts off Q1 MOSFET). You can verify this by measuring the voltage on C5 while the MOSFET is on (the voltage should be less than 0.65V or Vbe).

Looking around, I found a similar circuit might work a bit better. Here the timing cap (C1) is low when the circuit is off and high when on. This better much insures that you should be able to turn off. (I am a little concerned that it might take a bit of time for the cap to discharge enough to properly turn the circuit back on)

[View:/resized-image/__size/954x899/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e2b70ded-484b-44bb-a4d3-ff92aa79641e/contentimage_5F00_189366.png:954:899]

Good luck!

Gene

RE: Project PITS update

three-phase — Sun, 09 Dec 2018 21:47:12 GMT

I have tried testing the soft latch circuit over this weekend and haven't had much success.

[View:/resized-image/__size/620x414/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e2b70ded-484b-44bb-a4d3-ff92aa79641e/2604.contentimage_5F00_189364.jpg:620:414]

Initially, I found an error in my original circuit. The connection to the 650k resistor (R6) needed to come from the live battery feed and not the switched feed of the mosfet. With that corrected, I had minor success of being able to switch it on but, I could not get it to switch off.

[View:/resized-image/__size/620x428/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e2b70ded-484b-44bb-a4d3-ff92aa79641e/8551.contentimage_5F00_189365.jpg:620:428]

I tried a number of variations of the circuit available from the EEVblog and the comments surrounding it, but did not have total success with any of them. Some options I couldn't try due to my minimal collection of components to test with.

There were some comments about the circuit, that high capacitance loads would prevent the power from being switched off, not sure if that applies to my circuit though.

I will now start to look at the MAX16054 controller, that can do the soft latch, but will need to order a DIP adapter for it before I can use it on the breadboard. I think I will also look for a mosfet with a lower gate voltage to get a better switching action.

I will then have to go back and redesign the PCB, so it looks like the PCB won't be ordered until after Christmas now.

RE: Project PITS update

three-phase — Thu, 06 Dec 2018 20:57:52 GMT

I have now added a single 1A poly fuse to the battery positive connection. This pretty much completes the initial circuit. Over the weekend, I hope to set up the soft latch circuit and try it with the prototype as all the components have arrived.

If that goes well, I will look to send out the PCB for manufacture and build up this version and box it.

I will then look at the options provided by [mention:b0bc65b9ecdc4307bd967592f00e340a:e9ed411860ed4f2ba0265705b8793d05] and [mention:d214a0a0f5594ee19515b2a3782e7070:e9ed411860ed4f2ba0265705b8793d05] over the next few weeks to see if I can build up a more professional version.

A second unit will be useful for me, as I could then leave a unit with each of my injection test sets.

[View:/resized-image/__size/620x407/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e2b70ded-484b-44bb-a4d3-ff92aa79641e/0878.contentimage_5F00_189363.png:620:407]

RE: Project PITS update

genebren — Mon, 03 Dec 2018 23:10:05 GMT

Donald,

The updates to your project look great. The MCP73831 is a great choice for the Li Ion charging circuit. On the boost side, you could greatly simplify your circuit by using a boost convertor with an enable input. I have used the PAM2401 part in some similar designs. By using the enable pin, the circuit uses < 1uA of current when shut down. Your switch could control the enable line and you would then not need the inline MOSFET (and associated circuitry) to turn off power to the boost.

Gene

RE: Project PITS update

shabaz — Mon, 03 Dec 2018 01:14:34 GMT

Hi Donald,

Nice-looking design!

Regarding the clips, I've used the ones that you've specified, and with hindsight I'd definitely remove all ground plane on the top surface, in the U-shape that the clip occupies.. something like this basically:

[View:/resized-image/__size/620x431/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e2b70ded-484b-44bb-a4d3-ff92aa79641e/3036.contentimage_5F00_189359.png:620:431]

The yellow area would be the minimum area where there should be no ground plane I think. Otherwise, the metal clip is only separated by thin solder resist, because it sits very flush on the PCB.

As you say, it could also be good to have some extra protection. A nice bit of protection would be a polyfuse, they are super-cheap.

Even without that, it can be handy to have space for a series resistor, that could be useful for desoldering to make current measurements during development for instance, and in normal use is a zero-ohm resistor.

But if you use a polyfuse, you don't need space for an additional zero-ohm resistor; the polyfuse can be desoldered for the current measurement.

EDIT: Now I think about it, an even better way would be to have a small U-shaped copper area on top, which is then separated by an air gap to the ground plane. That way, when the clip is soldered down, the solder from the underside of the board can fill up the via too. But, this is overkill, because the metal connector is actually very high quality, and doesn't have much force applied in a direction that would cause it to weaken the joint, because it sits very tightly square against the PCB. Plus, if it really was a concern, epoxy glue could be used.