voltage selector

[erkanyigiter]

hello,
i am looking for a circuit which switches according to input=12V or 24V. i mean, if the input is 12V, it will forward the input to A, if the input is 24V, it will forward the input to B. i know, i can do it using relay but is there a way to achieve it without of using relay?<p>(look like a multiplexer)<p>[ November 03, 2004: Message edited by: yigiter ]<p>[ November 03, 2004: Message edited by: yigiter ]</p>

[Engineer1138]

Q: What are the voltages for? i.e., is there an AC signal riding on the DC voltage? does a certain amount of current have to be transferred?<p>If you just need to set an output to 12 or 24V at low current, you don't even need a mux; you just need to replicate the incoming signal at the right output. This can be done with an LM339 quad comparator. Set one channel up to switch when the input is between 9-15V (will require two channels to build a window comparator, or if 0-12V is fine, just switch if it's below 15V) and the other when the input is >20V (assuming these levels are close enough for your needs) and pull up the 9-15V output to +12 and the >20V output to +24 and you're all set.<p>If there *is* a signal riding on that voltage and you need to mux that, then you can use the LM339 to switch a DG508 (or similar) multiplexer. The 40xx series of CMOS multiplexers won't take 24volts.<p>If this won't work, give us more information on what you're trying to do and I'm sure someone can find a solution.

[erkanyigiter]

hello again,
i have posted a block diagram to make more clear. i've painted the needed circuit in blue. also the the needed circuit must be avaliable for high current(5Amp for 24VDC and 10Amp for 12VDC)
tahanks for your helps..<p>

[Enzo]

OK, so you have a thing that runs on 24VDC and you want it to run with either 24 or 12 volts applied, right? And you don't want to have to select it, you want it to deal with it automatically. Yes?<p>Will the DC/DC converter handle a 24V input? If so, then run the thing on the converter all the time. or is it important to keep the 24VDC off the converter input when the input is 24V?

[erkanyigiter]

as you said i have a printer that runs on 24VDC and i want it to run with either 24 or 12 volts applied. And i don't want to have to select it, i want it to deal with it automatically. also my dc-dc converter is only for 12VDC +/-%5 input. so i mıst keep 24VDC off my converter. i know i can do it using relay. but there may be another way to do it. any idea?

[dyarker]

Put power diodes where the block diagram says "IF THE INPUT IS 24VDC" and "24V OUTPUT". They keep the 24V converter output from getting back to the input side (smoke), and the 24V input getting to the converter output (maybe smoke).<p>Power NPN transistor (or FET) collector to "12VDC OR 24VDC" input, emitter to converter input.<p>About 15V to 18V Volt zenner and resistor in series from "12VDC OR 24VDC" input to base of a small NPN transistor. Collector of small NPN to resistor to "12VDC OR 24VDC" input, and to base of power NPN. Emitter of small NPN to common.<p>When input is 12V small NPN is off, and power NPN is on. When input is 24V, current through zener turns on small NPN which pulls base of power NPN near to common, so power NPN is off and not passing current to the converter. NOTE - check reverse Voltage ratings of power transistor!<p>Cheers,

[erkanyigiter]

hello Dale,
i have tried to draw schematic as you explained. i want to learn if the drawing is true. also are the resistor values important?
thanks,<p><p>[ November 04, 2004: Message edited by: yigiter ]<p>[ November 04, 2004: Message edited by: yigiter ]</p>

[dyarker]

Except for R1 the drawing is correct. R1 goes from D5 anode to Q2 base. R1 limits the base current.<p>Looking at it again, I think you'll need an R3 where R1 is currently drawn also. It's purpose is to pass zener leakage current when input is less than 18V. It should be 10 to 20 times R1. To make sure Q2 is completely off with 12V in.<p>The resistor values are important.<p>If Q1 has a minimum gain of 150, the Q1 base current needs to be 34mA for 5A to the converter.<p>... oh-oh, just spotted a problem ...<p>The circuit is logically correct, but Q2's collector current is wastefully large with 24V in.<p>I'll get back to you. Q2 driving a relay would be easier, but a no relay solution must be possible. I'll work on it.<p>C U L -

[erkanyigiter]

Dale,
i am waiting for your post and i want to remind something. when the input is 24V we don't need 34mA base current for 5A because when the input is 24V it will be forwarded directly to output. i mean Q1 is off. but when the input is 12V, Q1 will be on and the collector current will be 10A max. so the needed base current for minimum gain of 150 is about 68mA. when you work on circuit you could bethink it. thanks for your helps. i am waiting for your reply.<p>.........<p>Dale, may be i could use a thyristor instead of power transistor. what do you think?<p>[ November 05, 2004: Message edited by: yigiter ]</p>

[dyarker]

The problem with version 1 is R2. For Q1 to saturate at up to 10A when input is 12V, R1 turns out to be about 10 Ohms. That means when input is 24V, Q2 must saturate at about 2.5A to keep Q1 off.<p>For version 2, I changed the power transistor to PNP darlington (didn't show two transistors in the symbol), and renumbered parts left to right.

If 5A at 24V you mentioned is happens not too often and only for short peaks, then Q3 can be a 10A transistor like 2N6667. If 5A at 24V is more of a continuous load, the Q3 should be a 15A (or more) transistor like BDW42. The parts are the same except for R4.
D1, D2 = 8A, 35V schottky diode, 80SQ035 okay
Q1, Q2 = 2N3904
Q3 = PNP darlington, see notes
Z1 = 18V, 500mW zener diode
R1 = 680 Ohms, 1/8W
R2 = 50K Ohms, 1/8W
R3 = 8.2K Ohms, 1/8W
If Q3 = 2N6667, then R4 = 620 Ohms, 1/4W
If Q3 = BDW42, then R4 = 120 Ohms, 1W<p>With 11V to 15V in, the 10uA (max) reverse current through Z1 causes only 0.0068V across R1. So, Q1 is OFF. Current through R3 turns Q2 ON, causing current through R4 turning Q3 ON, allowing current to the converter.<p>With 22V to 26V in, Z1 conducts and has an 18V drop. R1 limits the current to around 8mA, so voltage at junction is 5V to 8V. Q1 turns ON, turning Q2 OFF, turning Q3 OFF.<p>You could use a thyristor, if input voltage is completely disconnected when switching between 12v and 24V.

[erkanyigiter]

Dale,
Thanks for your helps,
i did a simulation and i found the values as drawing below. for 12V, the output voltage is 9.791V. it could be harmful for my converter. because it was designed for 12V +/-%5. 9.7901V is very small for it. what is the reason of voltage dropping? is it possible to increase it about 11V?<p>

[rshayes]

The problem is in the PNP darlington transistor. The output transistor in a darlington does not saturate. The total voltage drop is at least the base-emitter voltage of the output transistor plus the saturation voltage of the input transistor. There are also some series resistances involved, such as the base resistance in the output device.<p>If you can get it, a large P-channel MOSFET is a simple solution. Unfortunately, P-channel devices do not have on resistances as low as N-channel devices. Look for a low voltage (about 30 volts), low on resistance (.1 ohm or less) P-channel MOSFET. This may be rare and expensive. Most MOSFETs are N-channel. A .1 ohm on resistance will drop about .5 volt at 5 amps.]<p>The second choice is a high current PNP bipolar transistor. Choose a high current device, and give it a base current about 10% of the collector current. This will be about .5 amp for a 5 amp collector current. Look for a low collector voltage rating, probably in the 30 to 40 volt range. High voltage transistors usually have higher saturation resistance due to the lightly doped collector region. A large area transistor should give a collector-emitter drop of about .3 to .5 volts. Look at packages such as the TO-3 or its plastic equivalent. This should be capable of the .5 volt voltage drop requirement.<p>Use a PNP emitter follower to provide the .5 amp base drive. The collector will connect to ground instead of the output. Some series resistance can be used to limit the current to .5 amp. Power dissipation will be about 6 watts, so a TO-220 device with a heatsink may be adequate. This transistor will probably need about about 20 milliamps of base drive.<p>Another alternative would be a darlington made up of a PNP transistor for the output transistor and a P-channel MOSFET as the input transistor. The voltage drop will be a little higher than the saturated PNP transistor since the output transistor in a darlington does not really saturate.