7v low, 24v high to cmos logic?

[schrodingerscat]

Hey,

Im currently trying to come up with a simple circuit that will interface a pulse output of a wattmeter to a chip. The levels of the output of the wattmeter are 24 volts for the high part of the pulse, 7 volts for the low. I wish to convert this to 3.3 volts for either the high or low, and 0 volts for the other. Does anybody know of a simple circuit that could do this?

Cheers in advance! :wink:

 

[Jay.slovak]

Hey,

Im currently trying to come up with a simple circuit that will interface a pulse output of a wattmeter to a chip. The levels of the output of the wattmeter are 24 volts for the high part of the pulse, 7 volts for the low. I wish to convert this to 3.3 volts for either the high or low, and 0 volts for the other. Does anybody know of a simple circuit that could do this?

Cheers in advance! :wink:

Hi, I would use Comparator with ~10V reference and voltage divider on output (to reduce 24V output to 3.3), this should be very easy to do. Or you can put divider to input, so that you don't have to have a high voltage reference and Comparator won't have to be powered from 24V...

When I re-read your post, I think that regular NPN would do the job.

 

[audioguru]

If your Cmos logic uses a 3.3V supply, anything higher than 2.2V is considered to be a valid high. Anything lower than 1.1V is considered to be a valid low. Therefore use a two resistors voltage divider to divide the input voltage.
By dividing by eight, 24V becomes 3.0V and 7V becomes 0.88V.
By dividing by nine, 24V becomes 2.7V and 7V becomes 0.78V.
A divide by nine voltage divider is a 12k resistor feeding a 1.5k resistor to ground. :lol:

 

[eblc1388]

You want 3.3V and 0V when the input is 24V & 7V.

Both Audio and Jay offers a workable solution. I hope mine works too.

A 21V zener in series with a 1K resistor to common will do. When input is 7V, zener diode is OFF and so no voltage across 1K resistor. When input is 24V, zener diode drops 21V and leaving 3V on resistor.

 

[schrodingerscat]

cheers for all the replies, i was thinking about a few of them there. I quite like the simplicity and cheapness of your idea,audio. But i think i will try Jays or eblc's idea as it will give me alot more flexability with the power supply voltages (i think i will be connecting the wattmeter up to different voltage supplies and so it will be giving different level voltages out). Thank you again for the replies, you have been a great help.

 

[audioguru]

Hi L. Chung,
I also thought of using a zener diode, but it has a high capacitance across it which will result in a 17V overshoot to the Cmos input when the signal goes high. :lol:

 

[Roff]

Here's another way. By changing resistor values, you can make this work for almost any common value of VCC. Using simultaneous equations, you can solve for the resistor values in terms of VCC and required base current (which depends on the value of the collector load resistor). I chose base current=30uA in this example.
The 5pF cap can be omitted if you don't need fast switching.

 

[eblc1388]

Hi L. Chung,
I also thought of using a zener diode, but it has a high capacitance across it which will result in a 17V overshoot to the Cmos input when the signal goes high. :lol:

Hi Audio, good point, but mostly taken care of by the manufacturer.

A quote from Don Lancaster's CMOS Cookbook, P.15:

"...all CMOS ICs have input gate protection built in, along with other forms of static protection."

Many projects I seen use a 1-10M resistor connecting high value AC voltage directly to a CMOS input gate although some would duplicate the two-diode input protection for extra peace of mind(I'm among one of them as I feel more at ease doing so).

All PIC inputs also have these two diode built in. Guess what, I would still fit the diodes & please don't ask why.

 

[audioguru]

Hi L. Chung,
Ron's PNP level shifter takes care of converting +24V to 0V and +7V to about +3.2V very well.

I also have and like the Cmos Cookbook. As it says, you can use the input's protection diodes to limit voltage only if you also limit the current though the diodes with a series resistor from the high voltage source.

BTW, in addition to an input voltage spike of +17V when the signal goes from +7V to +24V, there will also be a voltage spike of -14V when the signal goes from +24V to +7V, due to the capacitance across your suggested 21V zener diode. :lol:

 

[eblc1388]

BTW, in addition to an input voltage spike of +17V when the signal goes from +7V to +24V, there will also be a voltage spike of -14V when the signal goes from +24V to +7V, due to the capacitance across your suggested 21V zener diode. :lol:

Are you worrying that this might cause damages to the CMOS input or creates unwanted outputs?

 

[audioguru]

Hi L, Chung,
I don't think a Cmos will create an unwanted output due to overvoltage on its input unless it gets fried because the input current wasn't limited by a series resistor. :lol:

 

[Styx]

just put claming diodes to 0V and the top rail to stop the spikes being seen by the chip

 

[Styx]

does it have to be 24V, if not can it be 15? if sa you could use 4000-series logic

 

[audioguru]

Hi Styx,
Cmos inputs already have clamping diodes. L. Chung likes to add his own in addition to them. I just use a series resistor to limit the current and let the input diodes do their job.

4000 series from TI and a few other manufacturers have an absolute max supply voltage of 22V! They are happy at 18V but get hot if they feed a load or operate at a very high frequency. :lol: