Is this not clear by inspection? It is a LED drop plus a BJT drop. What are likely values?
"All LEDs exhibit forward voltage variation as junction temperature changes. ........ .........InGaN LEDs (Blue, Verde Green, and White) have a coefficient of between –3.6 mV/K and -5.2 mV/K."
https://www.onsemi.com/pub/Collateral/TND328-D.PDF
"vBE ... ... for constant Ic the well-known value −2mV/K applies."
transistors - Why does the base-emitter voltage of a BJT decrease with temperature?
also Vbe temperature coefficient of transistors
Assume 4.4mV/K for green LED, 2mV/K for BJT. We have 6.4mV/K (5.6mV-7.2mV) total.
In 50 degrees, 320mV.
In 2.5V nominal, this is like +/-6%.
In Maine we call that 50F to 122F. That is a w-i-d-e range for a music system. My home audio rarely plays outside 18C-24C. 6K range, 39mV drift.
Attachments
Thx Mark I lot of reading stuff and info.
What I meant is if someone could simulate or test the whole circuit from post 8 just with a 2V7 zener instead of a green lead to get a fine and stable 3.3V reference?
What I meant is if someone could simulate or test the whole circuit from post 8 just with a 2V7 zener instead of a green lead to get a fine and stable 3.3V reference?
Use a green 4231 led in series with 2 x 1N4149 ( yes 49 For low capacitance) for about a 3.15 V reference or 2 of the 4231 LEDS for 3.7 volt reference at 1 ma current. You can modify the LED voltage by varing the current . Mark Johnston has done a graph showing the single LED voltage Versus current. This varies as the LED warms up and is temperature Sensitive for for drift, as Walt suggested potting The assembly will reduce the drift once its settled after a few minutes. I use Walt’s 2N5486 cascoded jfet currents sources with differing number of the LT 4231 leds as voltage references in dacs, power amps, preamps, and regulators -
Measured data and an LTSPICE model fitted from those data, are attached to post #38 of this thread.