Off-Label LTZ1000

[zhtoor]

-

[cellularmitosis]

A naive reply from a newbie:

Increasing the Zener current in an LTZ1000 would be a trade-off of worsening the long-term drift performance in order to gain lower noise performance.

If low noise is the important bit (and not long-term drift), a 2DW232 would be much cheaper and lower noise.

If you need good long term drift and lower noise, my guess would be that multiple LTZ's averaged with resistors is the only solution.

Not only is this super expensive due to the cost of multiple LTZ's, but it is worse than that: the importance of the averaging resistors (their contribution to tempco errors) is proportional to how far apart the LTZ's are in terms of voltage.  So ideally, you'd have a number of LTZ's to choose from, so that you could select several which were already bunched close together in terms of Vz voltage.

[kj7e]

Increasing the Zener current in an LTZ1000 would be a trade-off of worsening the long-term drift performance in order to gain lower noise performance.

is that a proven fact?

-zia

Bob Dobkin and Jim Williams say it is.  If they say it is, I'm sure they tested this and know for a fact.

[JS]

  There is also the tempco thing everybody talks about, there is a current in the zenner for which this is lowest... That's the middle ground.

  I think the bigger contributor to the long term drift is the Pz as affects the zener junction directly (the difference is not in the temperature, but the current) If the heaters gets worn out, let's say it's resistance raises, the temperature compensation should get it back up. I don't even think this should be the case with a reasonable resistive material and the temperatures we are talking about. (in the resistor the temp does mean much more as it's the heating element, the temp will be directly correlated with the current) In really high temps every resistive thing drifts significantly.

JS

[BravoV]

Increasing the Zener current in an LTZ1000 would be a trade-off of worsening the long-term drift performance in order to gain lower noise performance.

is that a proven fact?

-zia

Bob Dobkin and Jim Williams say it is.  If they say it is, I'm sure they tested this and know for a fact.

did they make that statement while assuming that the heater was un-used?

best regards.

-zia

Quite surprising you don't know well who these two guys are.

All things you've learned, heard or read here that you respect here in this forum, when it comes to LTZ1000 are nothing compared to what these two people said.

[Kleinstein]

There is some evidence that a lower heater set-point can reduce long term drift a little. However this is more like comparing very high temperature (some HP refs) to more normal temperature. The downside with a lower temperature is that it also takes longer to finish the initial settling. So the advantage is more in years 5 and later, at the cost of possible not so good in the first few years.

For the LTZ1000 the unheated TC is positive for all reasonable currents. So the TC of the zener itself is not a good argument against a higher current. However with higher current the resistance of the pins/bond wires gets more important - especially in relation to the usually 120 Ohms resistor to set the current. So there is a trade of besides classical aging. So one might be able to go a little higher than the usual 4 mA, but not so much to get really much lower noise.

The internal heat is kind of required to get the very stable temperature to get low low-frequency noise.


For just low noise the 2DW232 Chinese zeners (if you get the right ones) are supposedly very good. They tend to run at a rather high current anyway to get zero TC. Due to scattering some selection might be needed and the best current is different for every unit. The main unknown about these diodes is there long time drift. Chances are it can be a bit higher.

[borghese]

If I remember correctly the bond wires are aluminium (Jim or Bob dixit, somewhere in "Ultra Precision Reference LTZ1000")
Best regards
Borghese

[chuckb]

Attached are some measurement I gathered from a new LTZ1000 and an aged LTZ1000A. The chips were in an environmental chamber from -50 deg C to +150 deg. I have not seen the data altogether in one spot like this so I decided to post it. Testing details are in the last attachment.

The heater resistance and the Zener resistance both seem to have a 4000 ppm / deg temp co. So maybe the resistance is all from the aluminum traces on the die. The kelvin voltage sensing of the Zener voltage by the base of Q1 (at the center of the die) may remove the resistance of one of the Zener traces. ...

Reference the above post. I measured the heater temp co at around 4000 ppm / deg C. It is probably aluminum.

[Edwin G. Pettis]

The TCR of aluminum is +4500 PPM/°C, ±5% (25°C - 125°C) and +4300 PPM/°C, ±5% (-55°C - +25°C) so you are in the neighborhood.  The TCR of the actual heater is likely different as it is not aluminum.