I need more current but I do not want to change my pcb board so can I piggy back two L293D pin for pin? I am driving dc motors and I am not seeing any gross failures but I am very close to the max ratting and so I wanted just a bit more headroom.
2
-
\$\begingroup\$ I would ask the vendor's FAE. \$\endgroup\$– Brian CarltonOct 16, 2012 at 16:08
-
\$\begingroup\$ I am just a hobbiest so I am not sure how I would get a hold of one or if they would even bother to answer my questions. \$\endgroup\$– stephenmmOct 16, 2012 at 16:15
Add a comment
|
3 Answers
\$\begingroup\$
\$\endgroup\$
6
That's not generally recommended. If you know the devices are from the same batch (i.e. are more closely matched) and you can ensure that they are thermally coupled (hard to do since they are in packages) then it's not recommended.
It would safer to get another similar device that has lower Rds(on).
answered Oct 16, 2012 at 15:52
-
\$\begingroup\$ So are you saying that one could turn on before the other one? And even if it did would it really do damage? \$\endgroup\$ Oct 16, 2012 at 15:55
-
\$\begingroup\$ Turning on is the least of your worries, it's about current hogging. You don't say what the device is, so I'll assume a MOSFET of IGFET. One device will take a majority of the load and as it heats up it will pull more. \$\endgroup\$ Oct 16, 2012 at 15:57
-
\$\begingroup\$ Oh, okay. I see where you are going. So where will it end up (stable state), will it be that one is taking 75% of the total current or are you saying that it would end up that one would end up taking >95% of the current. I have not tried to run the numbers in a long time but from a gut feel it seems like they might differ slightly but that one would not end up taking a large majority of the current. \$\endgroup\$ Oct 16, 2012 at 16:05
-
\$\begingroup\$ Obviously you're at the limit of one driver otherwise you'd not be thinking of doing this. BUt you've not said how badly or marginally you're presently failing. Maybe all you need is a heat sink. \$\endgroup\$ Oct 16, 2012 at 16:08
-
1\$\begingroup\$ Looked at the datasheet and it's a bipolar device, which is not good news as they can have thermal run away. However, since it's marginal now, any current that is diverted into a parallel circuit should increase your margin as LONG as the mounting of the 2nd package doesn't compromise the the total thermal system. If this is a one off thing then experiment. If a production unit, I'd not recommend it. \$\endgroup\$ Oct 16, 2012 at 16:21
\$\begingroup\$
\$\endgroup\$
I've definitely seen this done with exactly this chip, so it can work. As rawbrawb indicates, though, it's a hack - it might work, but it might cause problems due to mismatched components. Your worst case scenario: you burn out both of them from excessive current. Since they're cheap parts, it might be worth a try anyway, depending on your constraints.
answered Oct 17, 2012 at 9:07
\$\begingroup\$
\$\endgroup\$
If you have through hole chips, it could be feasible to increase the amount of power on your board in a safer way than soldering two chips on top of each others.
As explained in the datasheet, the safe way to double the power of l293 chips is to use 1 chip per motor. To do that, you have to connect channel 1 to channel 4 and channel 2 to channel 3.
In other word, if you want to connect two chips, you'll have to solder them above the board next to each others. One chip for the left side and one chip for the right side.
The tricky part is to have both pin enable, vcc connected correctly as they are not parallel to each others.
I'd flip the chip on the side like this:
simulate this circuit – Schematic created using CircuitLab
The advantage of this method against piggy backing two chips is that if one chip fails, then the second motor will still work. And you only have to replace one chip.
One way that would look less like a hack would be to have a new PCB that you can stack on top of the l293d socket.
answered Aug 29, 2017 at 9:22