Hello fellow forum users,
I just finished with a telephone conversation with Bob Dobkin, and he was short on time, so I was not able to get all of the answers to every question I [and we] had, but here are the ones I was able to get:
1) [FYI]: The applications circuits in the LTZ1000 data-sheet and other ap-notes were designed by Bob Dobkin, not Jim Williams.
2) Regulation in the LTZ1000 is more about controlling the power rather than controlling the voltage. If you accurately control the power, voltage stability will be achieved as a by-product.
3) The LTZ1000A has a different die attach than the LTZ1000, but they are both isolated quite a bit from external lead stress. Both of them will respond a small amount to lead stress, but the 'A' version if far less susceptible to this. Bob said that this can show up as a drift in output voltage with changes in barometric pressure.
4) It's been 20 years, so he didn't know the exact numbers, but Q1 is only rated to a few hundred micro-amps of collector current. [This means my single-resistor idea will *not* work! -- back to the original circuit from the ap-note!] Q2 will have a max rating of around 1mA, and the Zener can take quite a bit more.
5) They *could* make a model for LTspice, but the temperature effects are very difficult to simulate, and the market need is not large enough to support the development effort [my words]. He said the best way is to build the circuit and test it out in real life.
6) After telling him that I wanted a voltage transfer device, he said a better way might be to use [at least 6] LM399's in parallel [like the Bob Pease idea]. He said that the burn-in procedure would be to operate these in an oven set to 125C for 2 weeks, which would be equivalent to 1000's of hours of normal operation. Any LM399's that are drifting too much after that can be replaced [i.e., you burn-in more than you need, and select the best units for the array]. The LM399 is much more sensitive to board stress than the LTZ [because the LTZ has a special mechanical arrangement in the die mount]-- so the LM399 should be mounted off of the PCB a little bit to allow for this. The long term stability of the array of LM399's will be directly related to the power required to run the heater-- and this can be minimized with insulation-- the more the better! The LM399's do not age when they are turned off, and have almost no hysteresis-- so keeping the reference *off* until a few hours before you need to use it [and/or calibrate it] is the best way to keep the long-term drift minimized.
7) Note that for LM399-based designs, the slots in the PC board [plus a lot of insulation top and bottom] make sense-- Bob said that the less power the heater requires, then the more stable the output voltage will be. So, in this case, the slots [plus insulation] are helping with this. Oh-- and he also said that the LM399 should be run at about 1mA of Zener current for best stability. The more stable you can make the Zener current, the more stable will be the output voltage. He said that there is about 1uV of voltage change for 1uA of current change.
8: All voltage references *do* age faster at higher temperatures. So, since the LTZ1000 can be run about 10-deg-C lower than the 'A' version, it will have lower [about 1/2] long-term drift than the LTZ1000A under otherwise identical conditions. [But, the 'A' version has *other* differences that are positive-- my words. There are other sources of drift that the 'A' version is less susceptible to-- barometric pressure changes is one of them].
9) For the LM399 and both of the LTZ parts, he said a good burn-in routine would be to operate them at 125C [in an oven] for 2 weeks. After that, you can cycle the power on and off 10 to 15 times at normal operating temperature, and this will get them to settle down. This process should remove most of the initial drift that these devices exhibit. [So, my initial guess for a burn-in cycle was pretty good-- there is nothing wrong with a burn-in 10 times longer than this.]
10) If properly insulated, an LM399 can have a ppm/K figure of between 0.1ppm/K to 0.2ppm/K-- the data sheet is *very* conservative. [Note that the DMM manufacturers are not insulating these other than the plastic insulator that they come in-- they would benefit by better insulating the top and bottom of the LM399 to keep the heat in, and the heater power down.]
11) They are working on releasing more parts in the LS8 [8-pin ceramic LCC] package-- they should be coming out later this year. Some of the parts will not fit in the LS8, and they will have to use a larger ceramic LCC for those. He said the 1021 series will also be on the list of parts that will be released in LS8.
12) The ceramic LCC's have Kovar in them to get out of the package-- so, you don't get away from the thermal EMF problem, and these parts should be insulated from air currents [just like the TO-style packages] after they are mounted on the board.
13) The LTZ parts do not need to be spaced above the PC board like the LM399 does due to the unique mechanical de-stress design inside the package.
14) They will not be building a reference design for the LTZ part-- this is a niche product that does not warrant the development time. [That does not mean that *we* cannot come up with our own reference design! Let me think about this-- perhaps if there is enough interest, I can take the risk to do this...]
15) The TO packages are filled with *DRY AIR*-- not even nitrogen-only [so you would expect some degradation from this-- my words]-- but, Bob says that they don't have any parts that are degraded by this.
16) The LTZ1000 will use a lot more power than the LTZ1000A, so for battery operated circuits, it's best to go with the LTZ1000A, and insulate it to save power.
17) There is probably not a scenario where you can get zero-TC out of the LTZ-- he said you probably would have to operate Q1 at around 1uA, and that is not practical.
18) He said that it is not necessary to use the LTC2057, as the drift and noise of the Zener contribute far more than the LT1013. Bob said that you could use almost any precision op-amp, and the drift would not be affected by much even by the less precise amps. [I'm still going to use the LTC2057 anyway-- they are as cheap as popcorn, and have very little noise. Since I *do* need a zero-drift amp in the follow-on boost circuitry, and it's more economical to keep the number of unique parts on the BOM as low as possible.]
19) The Zener in the LTFLU-1 is the exact same one as is in the LTZ1000. He said that even though the exclusivity portion of the contract with Fluke is no longer in effect, they are not going sell these to the public because they don't want to anger a customer that buys millions of dollars in other parts from them. Even if they did want to sell it to others, it has not been characterized and there is no data sheet-- so there would be a lot of work just to build that. It's just not going to happen...
20) Bob said that he would join this forum, but he doesn't have enough time in the day to deal with what he already has to do-- so, it is not likely that he will participate. He said that I can email him with any questions I [we] have in the future, and he will do his best to answer them.
That's about all we could fit in. I hope that helps everyone-- it sure helped me a lot, as I was able to see that my "single resistor" design would not work. Bob said the that the LTZ would probably not be destroyed by my circuit [because "the LTZ can take a lot of abuse"], but it still would not ever be reliable even if it did work in the laboratory.
I was furiously taking notes during the phone call, and so I may have missed a few things-- if I remember anything I missed I will post it.