Printable View
Does anyone know of a transistor I can use that has enough gain to sink the ground connection on a relay that's drawing about 180ma? (Open collector configuration). Using a 16C65B...can't remember how much current a PIC can source.
Tried a 4401 and some others but they won't latch on.
THANKS
2n3904 200 mA
or you can go wild with a 2n2222 1.0 A
Yeah but the PIC won't supply enough current to keep them latched on. I tried a 4401...same animal as a 2222.
Will you provide a part of scheme ... ???
Alain
Hi Michael,Quote:
Originally Posted by Michael
The PICs supply 25 mA per pin. That should be more than enough to hold those small signal transistor on. You may be looking at a different type of problem.
-Adam-
Do you have a ~1K-ish resistor on the base?Quote:
Originally Posted by Pic_User
use this circuit found towards the bottom of the page
http://rentron.com/PICX6.htm
I'm driving a 2N4401 with a 1K....even did a 560 ohm. An open collector config to turn on/off the ground circuit to the relay.
I'm actually using the input to the LM317 on my power supply (around 8 volts) for positive relay voltage. (5v relay)
Using pin D4 on the PIC, which works fine (goes high for 15 seconds) but using the transistor to the relay will just give a momentary pulse and won't engage.
However, if I tie the ground circuit directly and bypass the 4401, the relay works perfectly....so I know the power supply is fine.
Maybe the transistor on resistance is too high but like I say, surely there's a part someone has used to sink 180ma without a problem.
Hi, Michael
Could ever your Pic have a RESET when power applied to your relay ... mmmmh ???
That could be a very good explanation ...
do try a LED instead of the coil ... just to see what's happening ...
sooo ...
1) do you have a BIG tank capacitor on your pic supply ( say 1000µF + )???
2) do you have the .1µF capacitor between vdd and vss ( the closest to the Pic )
3) did you think to the reverse diode ( 1n400x) // on the relay coil ???
Alain
PS ... once more: a scheme IS necessary to the troubleshooting !!!
Well....I changed the DC+ from the LM317 raw input to the 5v regulated output and all is fine.
Drawing 120ma as compared to 180.
??
I didn't want to draw current through the regulator, hence my using the input side.....it should have worked that way....are you guys sure the 2N2222 can do 180ma with a pic port?
Works fine now....another one of those "twilight zone" moments in electronics.
I could have been keeping the 317 from doing it's job and therefore affecting the 5v supply to the PIC....the transformer was only rated for 300ma.
Probably the case.
Hi Michael,Quote:
Originally Posted by Acetronics
I agree that a schematic can say 1000 words (in all languages too!).
Eight volts input for five volts out, is cutting it close. The relay is probably pulling the (regulator input) voltage below “drop-out” level, during the valleys, in the ripple, of your supply. Causing the PIC to reset.
Yes, the transistor should hold the relay on.
Just take the base resistor off the PIC pin and connect it to the 5 volt supply. This will tell you if the transistor can do the job...
-Adam-
The other thing you could so is use a N-Channel FET to low-side drive the Relay. In otherwords, keep your driving + to the relay and let the FET close ground on the Relay... My circuits generally use this method where we are driving a relay.
NDS7002 is one N Channel MOS you could use, and put a pull down between it and the micro for insurance. this particular FET is rated at 280mA on the Source to Drain. The gate drive is well below the 25mA.
As for a flyback diode, any general purpose Diode will also work...
and the Rds is interesting too... interesting current limiting/protection device in some case.
2N4401
40 Volts collector voltage
600 mA collector current
DC current gain with 180 mA collector current is about 100 minimum
With a minimum gain of 100 it should be able to switch 180 mA collector current with 1.8 mA from the PIC.
Less than 2 mA.
You may be looking at a different type of problem.
-Adam-
MMBD4148SE would be a good general purpose diode, you could also put a 5.5v varistor between the IO on the pic and the gate on the FET if you wanted maximum protection, although I think it's generally not needed. In my opinion if you were to put an over-voltage device on the circuit the Varistor would work better because it acts faster than a zener. Any reasonable package (assuming surface mount)...like 0805 would work.
Incidentally, I was looking at your requirement to not draw more current thru the regulator and in my mind this low-side drive would ensure you are not sourcing thru your power supply for the pic. I am assuming this is an external relay or something you don't want mixing in with your signals on the board.
This particular circuit i have suggested has been tested to well beyond 1Million cycles on a bench. we finally decided we'd beat it to death and took it off the supply during testing. This circuit would also appear to reduce or at least simplify your drive circuit design. BJTs are cool, but the FET is a bit less mysterious for driving a circuit like this.
Thanks.
I need to stock up on a good general purpose FET so I'll grab some NDS7002.
I would like to keep my relay from drawing through the regulator, so I'll breadboard something.
As they say in Indiana....Cheers.
I have always used bd681 with a 1k on the base, supports upto 3 or 4 amps with a heatsink.
Emitter to ground and load between + and collector. if driving a relay, then put a 1n4004 across the transistor
I just poped-in the thread and I bet the problem in the first place was the reseting of the PIC. So as very good stated by Alain the circuit should have good decoupling on the 317 and the PIC. The coil of 180mA sure spikes hard!
Every regulator need large capacitors on the in and out pins! Do not forget it!
The transistors like the ones refered are enough to drive the relay. Also add to the list the BC337 (0.5A/45V)
Ioannis
My supply circuits are generally as shown in the dodgey mspaint attachment.
This only powers the pic, any other devices ie relays etc are powered from the source or seperate regulators. I use this circuit in automotive projects without issue.
100 uF at the output might be high and loose some fast spikes, i would add 10uF tantalum on input and output as well. At the input you can use higher than 100uF.
Without having the whole thing here, it's hard to find the best suggestion. Many nice tips have already been said.
Hi Michael,
How did the new transistors work out?
-Adam-