PPM Difficulty

JT,

As you stated, as long as you reference the MAX6325 and all additional components that would normally be grounded to the negative rail you should be good. Curious,

what is the current source/bias schematic? Could the drift be in the current source?

Regards,

Jeff

Jeff,

Thanks for the quick answer and confirmation. The drift absolutely could be another point in the circuit, as I've been measuring the TC of the whole system, not just the voltage reference, but the previous component I tested as a reference (zener diode) matched the TC curve exactly. This could be coincidence, or I could have gotten the one bad chip in the bag, damaged it soldering, or some other TC function of the circuit, etc.

I'll poke around on my end some more and confirm everything before messaging again.

Thank you,

JT

How exactly are you measuring TC? At what temperatures? What is the soak time? Thanks!

Brendan,

Thanks for the reply, the chip I was testing ran for about 3 days before I started any TC measurements. I am measuring the voltage on the output side of a divided differential amplifier, which is just taking the output of the reference as it's difference. The method of measurement certainly could be the culprit, but I do not have a 7.5 digit voltmeter, so I've need to divide it to get some resolution. And the measurement temperatures are that of a hot room, generally between 21-27 (C).

Any advice is greatly appreciated.

Thanks for the details!

It is very difficult to assess TC on a small temperature scale. We generally look at 0C to 50C at a minimum. This accounts for both measurement error and curvature of the reference over temperature. Because I would expect MAX6325 to be basically flat around room temp, I believe you are seeing something other than TC. It could be a supply or ground issue, load regulation, even oscillation.

Have you looked at the output with an oscilloscope to confirm it is stable (not oscillating)? It would be helpful to see scope shots of both the supply and Vout of the MAX6325.

Can you look over a wider temp range to see what the actual shape of the curvature vs. temp is? Even 25C to 40C would give some useful information.

Brendan,

Thanks for all your feedback, I was able to confirm that the TC issues were unrelated to the MAX6325, but because of my power limitations, I won't be able to use it. I have a split-rail +/-5 V available to me for power, and the MAX6325 worked perfectly fine using the 10V difference (~9 with some low resistors). The issue I had is that with this set-up, I can't reference the reference voltage from 0V GROUND later in the circuit, and I really need to. So unless I upgrade my power rails, I can't really use this chip.

Thanks again for all the help.

Thanks for the follow-up! We have other products that are low TC/high stability and low-dropout, including MAX6126 and MAX6226 (in ceramic package), as well as ADR4525 and LTC6655-2.5.