6 input analog comparator

[Isenbergdoug]

First off, I'm having trouble with TI's LM339N quad comparator. My design is based on three of these, none of which seem to work. The problem is that they will not switch from low to high. They will go low alright, but when the input voltage is changed so that the output should swing high, I measure around 250 mV and rising on the ouput. The data sheet indicated that this was a cmos compatible device, so I'm trusting it, or is it otherwise? Has anyone had this problem, or know what might be wrong?
Also, the circuit monitors 6 input voltages, and right now I have all combinations compared, for example lines a,b,c,d,e,f I would compare based on this table
a b c d e f
b c d e f
c d e f
d e f
e f
where the left most column is the Vin+ input and the elements to the right are the ones being compared to it, so that when the element in the right most column has a high voltage, the comparator goes high. Then I AND the outputs of each row, the idea being that the element with the highest voltage will swing all outputs high and thus AND to a 1. Finally, but not important to the question, I encode these 6 lines into a 3-bit code. Does anyone have a simpler solution to this. Basically I am looking at 6 voltages and I want to know which one is the largest, or possibly the smallest If it makes the circuit smaller and cheaper. <p>Douglas Isenberg

[Isenbergdoug]

I got the LM339's working, thanks to National Semi, but I am still open to suggestions on a better way of doing this in analog. <p>Douglas Isenberg

[L. Daniel Rosa]

I don't know how much impaact this will have on your circuit, but a 339 comparator can't output positive without help. The "typical applications" on the spec sheet show output pull-up resistors ranging from 2k to 1M. It may be simpler to use opamps instead, without feedback they'll act like comparators.<p>As for the rest, I'm sorry I'm no help here.

[MrAl]

Hi there 'Is',<p>I take it you found the pull up resistors?
The LM339's have open collector outputs.<p>There might be easier ways to do this.
What kind of accuracy do you need here?
If your circuit was off by maybe 50mv or so,
would that be a problem?<p>Also, what kind of range in the levels are
you working with? (like all around 5 volts,
or from -5 to 5 volts, or very low levels like
100mv, or whatever).<p>--Al

[Isenbergdoug]

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by MrAl:
Hi there 'Is',<p>I take it you found the pull up resistors?
The LM339's have open collector outputs.<p>There might be easier ways to do this.
What kind of accuracy do you need here?
If your circuit was off by maybe 50mv or so,
would that be a problem?<p>Also, what kind of range in the levels are
you working with? (like all around 5 volts,
or from -5 to 5 volts, or very low levels like
100mv, or whatever).<p>--Al<hr></blockquote><p>Yeah...I didn't have the pull up resistors. I was figuring that I could use the 339 like an op-amp, but that's not the case. As for accuracy, I don't have a specific answer, I simply want a digital output that corresponds to the highest (or lowest) voltage of the inputs, so the only accuracy I want is whatever it will take to find this. I can see where I may have some problems in this, having invalid results on the comparators, but they are being clocked into a D flip-flop, so I can increase the sampling rate to smooth it out. I got off subject, no, being off by 50 mV should not be a problem.
And the voltages I am working with, 0-5 Volts. I want to stay away from negative voltages and something other than +5 and +12. The circuit will be used with other circuits at 5 volts, so that is where I would like to keep it. <p>Thanks
Douglas Isenberg

[bobsRAC]

Would a microcontroller be fast enough? Atmel, Microchip, TI, Motorola, and others all makret inexpensive ($2 - $10) microcontrollers that can give you multi-channel analog sample rates up to 2MHz in some cases. Also, Lattice Semiconductor offers programable analog devices that could easily handle your application.<p>http://www.atmel.com
http://www.microchip.com
http://www.ti.com
http://www.mot-sps.com
http://www.latticesemi.com