Help me choose buffer circuit for LM399/ADR1399 7V to 1V

[qat]

I have come up with two different circuits for buffering and trimming the output of the ADR1399 reference from 7V to 1V, see image attached (5.1ohm / 1uf not shown).

Is there anything to watch out for when using the top circuit, which appears to outperform in various aspects based on my initial assessment?

Edit: Initial assessment complete, and I now realise why the first circuit will not work here...

[iMo]

The optimal strategy is to have the adjustment trimmers as small as possible (in value), such their contribution is the smallest possible (they have got high TC, low stability, etc.).
The output buffer - you have to consider the planned load - the output currents and capacitancies..
Also a protection of the output is important (it depends what you want to do).

[qat]

Thanks for the input
In my case the load will consist of inputs to other opamps and the reference's buffered output will only be accessible for calibration purposes.

[CurtisSeizert]

Presuming it needs to not significantly degrade the stability of the reference, the approach I generally take is to have extra footprints for discrete resistors in parallel with each leg of the divider. At the ovenized reference level of stability, I would say that potentiometers are generally frowned upon, though if you are using one with a trim range of perhaps 100 ppm, you could probably make a reasonable case for using it. It seems like your circuit is probably part of an autocalibration block, so you could probably just deal with deviations from nominal in software. I think the conventional wisdom on this board would also be to use something like a NOMC 8-resistor array as the divider. I don't know offhand the best way to do 6:1 with that, but I think that for statistical averaging to be most effective, something other than the obvious 6 in series : 1 would be best. Perhaps three in series for the upper leg and two in parallel for the lower leg with the resistors interleaved would be good. You probably don't want to go higher than 500 uA divider current because the Joule heating tends to worsen the stability of the divider and make it more susceptible to thermal EMFs.

Many of the op amps one would normally use for buffering a small voltage with minimal error have current noise that could be problematic, especially if you have the inputs of multiple op amps connected to that node without a buffer. If you are using an autozero amplifier, you will probably want to bypass the lower leg of the divider with maybe 10 to 100 nF (C0G) as the switched inputs have lower current noise when there is some bypass capacitance. Depending on your target bandwidth, you could add some more bypass capacitance, but at <10 Hz you usually just deal with it.

Less obviously, you need a good way of measuring that voltage if you're going to trim it without needing to hold the test leads, preferably with things as close to fully assembled as possible. Through-hole test points where you can solder magnet wires (or similar) are good. Keep them next to each other to make sure the junctions experience similar thermal gradients or just use the footprint for a 2-pin 50 mil pitch header.

[Andreas]

Hello,

look at the LM399 datasheet.
"Standard Cell Replacement" has a good trimming scheme.

with best regards

Andreas

[qat]

Presuming it needs to not significantly degrade the stability of the reference, the approach I generally take is to have extra footprints for discrete resistors in parallel with each leg of the divider. At the ovenized reference level of stability, I would say that potentiometers are generally frowned upon, though if you are using one with a trim range of perhaps 100 ppm, you could probably make a reasonable case for using it. It seems like your circuit is probably part of an autocalibration block, so you could probably just deal with deviations from nominal in software. I think the conventional wisdom on this board would also be to use something like a NOMC 8-resistor array as the divider. I don't know offhand the best way to do 6:1 with that, but I think that for statistical averaging to be most effective, something other than the obvious 6 in series : 1 would be best. Perhaps three in series for the upper leg and two in parallel for the lower leg with the resistors interleaved would be good. You probably don't want to go higher than 500 uA divider current because the Joule heating tends to worsen the stability of the divider and make it more susceptible to thermal EMFs.

Many of the op amps one would normally use for buffering a small voltage with minimal error have current noise that could be problematic, especially if you have the inputs of multiple op amps connected to that node without a buffer. If you are using an autozero amplifier, you will probably want to bypass the lower leg of the divider with maybe 10 to 100 nF (C0G) as the switched inputs have lower current noise when there is some bypass capacitance. Depending on your target bandwidth, you could add some more bypass capacitance, but at <10 Hz you usually just deal with it.

Less obviously, you need a good way of measuring that voltage if you're going to trim it without needing to hold the test leads, preferably with things as close to fully assembled as possible. Through-hole test points where you can solder magnet wires (or similar) are good. Keep them next to each other to make sure the junctions experience similar thermal gradients or just use the footprint for a 2-pin 50 mil pitch header.

In my scenario, the reference voltage will regulate the current flowing through a laser diode and manage its temperature. It's practical to adjust the reference voltage using a potentiometer to fine-tune the current, thereby controlling the laser's wavelength. Since laser's characteristics can vary significantly, I'll probably need to adjust the reference voltage output by approximately 5% (equivalent to 50000 ppm ). I'll incorporate your suggestions to optimize my design and conduct simulations and calculations to see what performance I'll actually get.

For calibration I'll use SMT test points or pins for attaching leads for calibration using a 6.5-digit multimeter to achieve a target voltage of roughly 1V (or whichever reference voltage I ultimately select). The final calibration will utilize an optical spectrum analyzer for precise adjustments.

[Kleinstein]

For a larger adjustment range it could make sens to use the trim resistor not so much as a variable resistor but as a potentiometer. So the pot creates an auxiliary voltage (e.g. full 0 - 7 range)  that is than mixed to the main voltage with a higher value series resistor. This way the pot TC is only relevant as relative TC. One also get a reasonable linear scale and if needed gets a test voltage that is easy to measure the pot position.
For more resolution one could use a multiturn pot (not just the cheap trimmers that are made for only few operation cycles). If need one can extend the range similar to a KV divider with a 2 level step switch. The simple one are 6x2 and thus only 6 or 5 steps, but could still be enough with 1 switch.

[CurtisSeizert]

If calibration isn't too much of an issue, you could just put the wiper of a good trim pot (maybe a Vishay 1280G or similar) as the center tap to the divider and have resistors on the upper and lower legs to constrain the trim range to 5%. It is kind of a matter of your error budget, but I would think that any design that uses a pot probably would not need an ADR1399 unless it is needed for some other part of the system. A key question is also what the calibration interval is and how much the diode is anticipated to drift. From what you have said, I would tend to think you can keep it simple here and make calibration easy.