LM2902N

Bob,

The lowest output voltage is based on how much current flows into the output pin. (sinking current).

The data sheet table setup has zero sink current.

There is a 30uA sink driver in the LM2902 that has a low VOL. 
For higher current, the PNP emitter follower takes over but has a higher VOL value.

Figure 1 in typical characteristics section of data sheet has more detail.

Hello Ron,

Thanks for the reply.  I understand what you sent.  However, my application is sourcing current, not syncing current.  We have a Vcc of 11V.  Our output is a 10K ohm load to ground.  Based on your chart, we should be operating pretty close to the lower left side of the curve, pretty close to zero.  But, we are seeing 500 mV above ground. 

FYI, our Pspice analysis is showing us approximately 30mV above ground.

Any suggestions?

Bob,

The current to the feedback path is also a load on the output. Output to inverting input in most cases.
What is the voltage on the inverting and non-inverting inputs?

Ron,

I have a circuit diagram attached.  The input voltage is approx. 8v peak-to-peak, centered around 5.5v reference.  The output is 4v peak.  I'd like it to start at zero but, it's around 500mV about zero. Here is the circuit diagram.

Bob,

The highest input common mode will be (5.5V + 4V) / 2 = 4.25 V
The guaranteed sink current is 12 uA.
Therefore the resistors need to be increased to 4.25V / 12uA = 354 k ohms.

Hello Ron,

Thank you for the info.  That really helped.  We did not have any 354K ohm resistors handy.  We used 453K ohm resistors instead.  The voltage came down to 80-90 mV.  Any suggestions as to how we can get it closer to the 20mV max listed on the data sheet?

Here is a picture of our output waveform.

Bob,

Leveraging a negative supply would help. Otherwise changing to another op-amp that has a drain or collector negative rail output drive. look for rail to rail output check box in search.

Good morning Ron,

All we have is a single 11volt supply.  Other than what you just suggested to us, is there anything else that we can try?

Thanks!  Happy New Year!

Bob

Hello Bob,

My apologies for the delay in response.
Ron is out right now and I am covering for him.

I don't know of any additional way to decrease the output voltage, but Ron may be able to comment further on Monday.
Otherwise, we have a very large selection of amplifiers that have output voltages that can get closer to the rail.

Using a quick filter, I found the TL972 or TL974, but understandably so its more pricey.

Please let us know if there is any other way we could help.

Best,
Michael

Bob,

As Micheal suggested a better op-amp will work.

Otherwise three NPN transistors plus a single resistor could be used to make a current mirror sink to aid the 30uA sink inside the LM2902.

Ron,

Thank you very much for the info. It is much appreciated! You've been very helpful.

Bob