Hi all, I have been using a 2n2222 as a switch to turn on and off a very small servo, for battery saving purposes (save the 3mA drawn from it being ON). So it draws only 100mA up to 200, so a 2n2222 is ok for the job at 5v. But when I use batteries (18650 or AAs) the voltage is lower, and barely enough for the servo to operate. Therefore,. The 0.6v drop across the BJT (2n2222) makes the voltage too low for the servo. I am looking for a mosfet that can easily be used as a switch from a logic gate with 5v source drain. The voltage is 5v for the gate too. Which device can I use? Im a bit lost.
Or instead of a mosfet, would it be a good idea to keep the 2n2222 with a little 5v boost (a 34063 or something) between the servo and transistor?
What package style is acceptable, SMD, TO220, TO92?
larryd:
What package style is acceptable, SMD, TO220, TO92?
Any, really, but I'd prefer if the package was available in any Trough Hole package for prototyping plus available in SMD for my final version of my project(s), because I like to have a polished finish with SMD circuits!
Hi,
How are you switching the power to the servo, high side or low side switching?
Thanks.. Tom... 
I often use an IRL3705 TO220
A SOT23 IRLML2502 should work too.
jackthom41:
I would recommend BC547 or IRF540, both work great for me.
Guess you didn't read the OP if you suggest a BC547. An IRF540 isn't a great choice either at 3.7 V.
As a result of another thread, I have come across the SI2302.
Of course you have to remember to set the servo control pin as INPUT if you use low-side switching.
Paul__B:
Guess you didn't read the OP if you suggest a BC547. An IRF540 isn't a great choice either at 3.7 V.As a result of another thread, I have come across the SI2302.
Of course you have to remember to set the servo control pin as INPUT if you use low-side switching.
Interestingly, a SI2302 a N025P and a IRLML2502 are equivalent.
TomGeorge:
Hi,
How are you switching the power to the servo, high side or low side switching?Thanks.. Tom...
High side, however that's what I tend to do naturally and I am not really familiar with the ups and downs of each setup. I have seen a few mosfet circuits placing the load between the fet and the ground. Maybe this is what I should do? thanks!
larryd:
Interestingly, a SI2302 a N025P and a IRLML2502 are equivalent.Alright, so that gives me 4 choices:
IRL3705 TO220
SOT23 IRLML2502
SI2302
N025P (which actually gives results for IRLML2502 on ebay)Just to make sure I understand how to pick the right one, I'm looking at Vgs for IRLML2502 (http://skory.z-net.hu/alkatresz/irlml2502.pdf) and it says: max Vgs at 4.5V is 4.2. Does that mean that I need at least 4.2V to turn the switch On (if VCC is 4.5), otherwise it will be off? There also is a Vgs = +/- 12V, does this one mean the max voltage that can come across gate-source?
I was thinking I could only buy smd and use some sort of DIP mounting pcb like I would do for 8pin soic packages for prototyping (ex: 5PCS SOP8 SO8 SOIC8 SOT to DIP8 adapter PCB convertor(gold-plated) | eBay)
but I cant find any of those for 3 pin transistor. Anyone had any luck? Thanks again
You have been using an NPN transistor as a high-side switch?
No wonder it doesn't work, you'll have been losing volts and risking burning out the 2222.
PNP for high-side, NPN for low-side, that's the rule. Use the 2N2222 low side and it's likely to work.
Ditto with a PNP switching transistor on the high-side.
Use a 150 ohm base resistor.
Despite what you think, that servo will draw a lot more than 200mA peak, it'll max out the '2222
but that's probably going to be OK and average is a lot less than peak (unless you jam the servo
against the end-stops)
MarkT:
You have been using an NPN transistor as a high-side switch?
No wonder it doesn't work, you'll have been losing volts and risking burning out the 2222.PNP for high-side, NPN for low-side, that's the rule. Use the 2N2222 low side and it's likely to work.
Ditto with a PNP switching transistor on the high-side.
Use a 150 ohm base resistor.Despite what you think, that servo will draw a lot more than 200mA peak, it'll max out the '2222
but that's probably going to be OK and average is a lot less than peak (unless you jam the servo
against the end-stops)
Thanks it does work for the intended voltages!! thanks for the rule I wrote it down
Its worth studying the difference between using a transistor as a saturated switch in the common emitter
configuration, and what you had, an emitter-follower circuit (sometimes called common collector).
The latter cannot saturate as to saturate the base has to get to a higher voltage than the collector
(for NPN). Without saturation the transistor is dropping 0.7V or more and getting hot.
Darlington pairs have the same problem of being unable to saturate, so run hot.
In a properly saturated state the output voltage drop for a common-emitter switch can be as low as 0.1V.
Some more modern switching transistors can do even better, such as the ZTX851 which is the best
performing device I've encountered, 50mV forward voltage is achievable, and very high pulse current
rating.
Of course this is pretty moot, MOSFETs are pretty universal these days...