If you care about 8.321 - 8.28 =3D 0.041 Volt in 8.28 V =3D 0.041/8.28 x 10= 0% =3D 0.5% change, then a zener is probably not your bests solution. As Steve says and as you note, they are PTC and the effect is temperature dependent so it seems likely without doing the calculations that that is the main cause. (Current is ~~=3D (19-8.2)/10k =3D~ 1.1 mA so heating energy is about 10 = mW which isn't much. I'd imagine that variations in air flow from other sources could vary the result noticeably. Consider using two zerners in series with one a PTC part and one an NTC part. One value has about temperature neutral result (5V1?) and above and below this results are PTC and NTC so judicious combination may provide a useful result. (That's from memory with no checking). As Steve notes, a TL431 is an excellent means of voltage control and the price is usually acceptable. (10's of cents in US in small quantity, a few cents in volume in Asia). These can be had down to 0.5% "accuracy" . Overall it depends on whether you care about stability or absolute accuracy. If the latter and if 0.5% absolute is too much then even a TL431 will not suffice. If you want good stability at low component cost then consider a zener plus resistor feeding a second zener and resistor. Relative voltages and drives can be arranged for best result. Russell On 31 January 2012 12:08, alan smith wrote: > what might cause a zener voltage to slowly rise? Load isnt changing, its > constant, driving into an opamp to distribute the voltage. Its a 8.2V > part, MMSZ5273B with a 10K series. It starts at 8.28V, over 5 min rises = to > 8.321V. Cooling it down..drops it so its temp related....current related= ? > Too much current so it heats up? Source is 19V as well. > > > -- > http://www.piclist.com PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .