Newbie transistor question

Status
Not open for further replies.

slopagafud

New Member
Ive recently built a timer for a 9v pulser using a picaxe 08m.I am using a 2N3906 to turn the power to the pulser on and off but I am not getting a full "turn off" (pulser still putting out 1-1.25v during "off" time.) I have a 2N4403(found it in the package of 3906) which has all the same characteristics as the 3906 but a higher ma rating on the collector current.(3906=200ma, 4403=600ma) If I was to use this instead, would I get a better "turn off"? for the pulser?
I have included the schematic for the timer.


https://www.electro-tech-online.com/attachments/picaxepowerswitch9v-jpg.16538/
 
The problem is in how you are biasing the transistor so a different PNP will not change things. The emitter has 9 volts applied. The uP output cannot go higher than 5V so it is incapable of pulling the transistor base to a voltage higher than 8.3 volts which is what is necessary to turn off the transistor. There are a few different ways to fix this. One way would be to add a 5.6 volt zener diode in series with R4 (anode towards the uP). Another way would be to place 7 silicon diodes in series with R4 (cathodes towards uP). Or you might consider changing the output transistor configuration to something else instead of a PNP type. For example, you could put an additional NPN in between the uP and the PNP but then of course you would have to invert the logic of your output to compensate.
 
I tried the 5.1 zener in series w/ R4 and still didnt get a full shut down from the pulser,though it did reduce the output voltage by about .2-.3v. Just for kicks and giggles I added a 2nd 5.1 to the series and got another drop in voltage but still not shutting the pulser down completely.My stupid question of the day is: If I was to use a 12v zener diode would I get better results? I dont have enough room on this board to try adding an NPN to the circuit.
I forgot to mention before,and Im sure this makes a difference,I am no longer using a 9v alkaline batt,Ive switched to a 10v NiMH pack to increase the pulser output Voltage.
 
Well, there could be one of two problems happening with the zener diode idea. One is that the leakage current of the zener below its zener voltage is too high for this application. The second is that perhaps you put the zener in backwards? Please confirm which way you put it in.

As for the leakage problem I'm not sure if there is an easy fix so perhaps this idea is not so great after all. In this case, using a 12V zener won't help.

What amount of current is flowing into the pulser when it is normally on?

The idea of adding an NPN between the uP and the 3906 is bulletproof, so its too bad you can't find a way to add a 2N3904 and one 10K base resistor. The circuit would be uP -- 10K ---2N3904base --emitter grounded --- collector to existing 10K base resistor etc.
 
With the pics pin at 5V and the supply at 9V the voltage across the 1k resistor will be less than 0.4V and so won't the transistor be off? Or, am I misunderstanding something here? Could the solution be to put a resistor across the output?

Mike.
 
Pommie said:
With the pics pin at 5V and the supply at 9V the voltage across the 1k resistor will be less than 0.4V and so won't the transistor be off? Or, am I misunderstanding something here? Could the solution be to put a resistor across the output?

Mike.
Click to expand...
Well, yes, but he'll only get about 130uA of base drive, so if his load is more than a few milliamps, or capacitive, his output voltage waveform will suck.
What does a PIC output do when you try to pull it up above Vcc? Does it have an internal diode to Vcc? The datasheet says the maximum voltage on an I/O pin is Vcc+0.3.
In any case, his scheme is not a good way to get output voltages higher than Vcc.
 
Last edited:
He could use an NPN common base stage (no phase inversion) in front of the PNP, but this requires 2 base resistors instead of one, and he's already said he doesn't have room for the NPN.
 
He could replace the PNP with an NPN and reverse his logic. Or, a 2N7000 and loose the base resistor as well.

Mike.
Roff, the pic pins have internal diodes to the supply rails. There are Microchip application notes where they feed mains into a pin via a large resistor.
 
Oh yeah, now I remember.
 
I am pretty certain I installed it correct.I tried installing it both ways and when reversed,the pulser never shut down at all. Thats when I decided to try the 2nd one in series, and got a better result.

I haven't yet measured the current into the pulser.

Im gonna try to squeeze the NPN in somewhere, somehow. Just hold my hand here for a second, the 10k & 3904 would be installed before R4? What about R5? I assume that stays in place?
 
2n7000? speak to me! but be gentile, remember Im a newbie! I dont know poop from pudding about this but I can read a mean schematic!lol
 
RadioRon asked,
What amount of current is flowing into the pulser when it is normally on?
Click to expand...
Before you can get a definitive answer to your questions, you need to answer that question.
 
The 2N7000 is a MOSFET (TO92 package) designed for switching small loads from a logic level. Connect Source to ground. Gate to Pic pin and drain to pulser - other side of pulser to 9V. You will have to reverse the logic in your program as logic high will now be on.

Mike.
 
unable to measure current right now,its a little too late,that will have to wait till tomorrow.

I thought I had an extra 2N700, but I cant locate it right now,I guess Ill have to order more.
 
If his load must be grounded, it could also be used as the level translator, either inverting (common source) or noninverting (common gate), with no additional resistors required.
 
That's true, more efficient than the NPN.

I avoided recommending low side switching with a FET just because it tends to freak people out, although I use it often myself.

I think I miscalculated the effect of that 1 K resistor across the PNP's emitter base junction. I'm so used to people putting very large values there that I just ignored it, but yes, it would tend to limit the eb voltage wouldn't it? But if so, then this thing would work by now. Perhaps there is some other issue besides the bias. A circuit fault?
 
I am noticing a lot of "if/then" in these posts. you can find details about the pulser Im using here: **broken link removed**
there is also a schematic for the pulser on the page.I built mine from the book.
I attached the device to a harness, and use it on my dog,shes got a rare heart disease. I added the timer to make my life easier. Long story short, she was supposed to be dead 6 months ago,today, you'd never know shes sick.
Sorry to get off topic, I just wanted to provide as much info as possible.
 
Status
Not open for further replies.

Similar threads

H
  • Solved
Newbie: Help finding replacement diode and transistor
2
Replies
20
Views
5K
shokjok
S
5
Newbie
Replies
5
Views
1K
Nigel Goodwin
A
  • Question
A couple resistor and transistor questions
2
Replies
39
Views
6K
ZipZapOuch
F
  • Question
Need some help with a simple project (Newbie here)
Replies
2
Views
2K
rjenkinsgb
I have another question.
Replies
6
Views
1K
aliarifat794
A
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…