# Trimmer Potentiometer with Unexpected Voltage Range

I designed a voltage divider circuit using a trimmer potentiometer 3224J-1-103E for the first time. But, upon receiving the board, I found out that I was unable to tune the potentiometer across its entire voltage range, in this case ~0V-5V. Instead, my range was only ~0V to ~3.5V. From my understanding, I should be getting close to 0V-5V.

I've been looking for hours as to why this is happening, but can't find anything. I'm hoping someone can explain to me if this is expected and how I can predict it in the future from something like the datasheet. Also, potentially why it is only clipped in the 5V direction, but not the 0V direction.

As a side note, there was several of these on the board, all of which had this effect. I suspect it is unlikely a manufacturing issue.

Second note, the circuit behaves exactly as expected outside of this, I am just unable to set the range as high as I need.

As requested, this is a simplified representation of the scenario within the circuit, where it is being used to set the low/high voltages of a valid window for a signal.

simulate this circuit – Schematic created using CircuitLab

– vir
Commented Jun 26 at 0:45
• @vir Provided a schematic showing the gist of the setup. Can't provide the entire thing, but the details outside of what I've shown shouldn't matter at all. Commented Jun 26 at 0:57
• Is the 3.5 V measured at the input to the LM324, or at its output? Because the LM324 isn't a rail-to-rail op amp (and really shouldn't be used as a comparator either, dedicated comparators are much better for this use case). Commented Jun 26 at 1:02
• @Hearth input of the potentiometer (i.e., directly at the output of the potentiometer) and I understand but the two other op amps in the quad chip were used for amplification purposes so I settled for it. Perhaps the design can be improved, but it is not in the scope of this question. Commented Jun 26 at 1:08
• @imthegman55 Can you link to the datasheet iin question then? Because the one I found says it needs about 1.5 to 2 V headroom on the high side, which I doubt anyone would call rail-to-rail. Commented Jun 26 at 1:17

I'm pretty sure this is your problem:

You are exceeding an absolute maximum rating, specifically the max diff voltage on the inputs. If you exceed the limits, performance, operation, and reliability is not guaranteed.

Only rare exceptions can be made to use an op-amp as a comparator. And this isn't one of those times, especially if your sensor changes more than almost 200mV; you won't be able to find a pot setting that will allow for proper operation as a comparator.

If you're jammed up with your poor design decision and it's too late to corrected it, then one trick just to get the circuit to work the way you have it is to limit the operating range of the sensor output:

simulate this circuit – Schematic created using CircuitLab

I'll leave it up to you to do the resistor math. Hint: you want the input signal going into your "comparators" to span about 150mV over the full sensor operating range.

Also, it will be difficult to "trim" the pots with this arrangement. The trick to get more control over the (presumably) single-turn pots is to add additional resistance with the pot legs (not the wiper lead). And you are going to want to put a cap to ground across this adjustable reference. Especially due to the increased impedance.

This is not a great solution, but it will be good enough just to show your boss you got something "working" until you get a real solution. And the real solution is to use a comparator when you need to compare 2 input voltages.

• I think that spec should be read as "the difference of the supply voltages, plus 0.2V", which the potentiometer output is within.
– vir
Commented Jun 26 at 1:50
• Where did you get this snippet? LMV324 is okay with +/-5.5V differential input. Commented Jun 26 at 1:51
• @SpehroPefhany page 5 of the datasheet from the comment link. Commented Jun 26 at 1:53
• @vir I read it as the differential signal inputs. But you could be right, let me check. If so, my answer is worthless. Commented Jun 26 at 1:55
• @MOSFET it seems it applies to the LMV324A vs. LMV324. Those rotten S.O.B.s actually radically changed the front end to a CMOS type and just added an 'A'. Hence the comment about continuous differential voltages causing Vos shifts. Nasty. Marketing wanted to get sales easily, I guess. Commented Jun 26 at 1:59

There is no reason for the behavior you describe, based on the information given.

The LMV324 is not R-R input (it is R-R output) so it won't actually function with both inputs outside of that range.

If you are inferring the range from behaviour, rather than measuring it as your question states, then this is likely the reason.

You can fix this by replacing the op-amps with CMOS-input RRIO types.