Ultra Precision Reference LTZ1000

[grx]

(Analog devices refusing direct PDF datasheet access these days is a complete annoyance, btw)

All french suppliers I know (farnell,rs,mouser,digikey) only carry the pdip. Only Rochester sells cdip lt1013 them at digikey marketplace, and I must get 67 of them. Not gonna happen.

In fact I dont think you can buy ANY cerdip at these suppliers.

[Kleinstein]

With the extra gain from the transistor stages the LT1013 at the LTZ1000 should not be that critical. No need for something higher performance and the plastic case should be OK.

An extra output buffer would be something different. Here one has plenty of additional choices as no more single supply is needed and the signal is lower impedance. For an extra buffer the LT1013 never was a good choice.

[tatel]

I fear I will have to change to ADA4522 in future.

I guess I'm missing something. I was under the impression ADA4522 was indicated for the 7 to 10 volts buffer, which wouldn't be LT1013 case.

And I can't find ADA4522 in cerdip package either. but since it's a chopper, probably that doesn't matter.

Am I wrong in assuming you are thinking about substituting ADA4522 plastic packaged chopper op-amp for the general purpose LT1013 op-amp, not for the external buffer but to work directly associated to LTZ1000?

[grx]

I have made a PCB that will allow mounting two LTC2057 or ADA4522-1 on the footprint of a LT1013, but that was for fun, it is easy to design a board with those. I will probably do it soon.

In other news: GRX ref item 1 is working! I get 7.1v after a few seconds of heating! wohoo!
The missing R13 is to ensure startup, but it worked without it.
Vref is powered via a J112, and the heater transistor is a BC108.

Now is time to work on insulation, boxing, UPS and monitoring before I power it again on the long term.

[Andreas]

I guess I'm missing something. I was under the impression ADA4522 was indicated for the 7 to 10 volts buffer, which wouldn't be LT1013 case.

And I can't find ADA4522 in cerdip package either. but since it's a chopper, probably that doesn't matter.

Am I wrong in assuming you are thinking about substituting ADA4522 plastic packaged chopper op-amp for the general purpose LT1013 op-amp, not for the external buffer but to work directly associated to LTZ1000?

Yes I have alternatively populated 2 samples with LT1013AM and 2 samples with LT2057 to see long term differences.
Result was: long term stability mainly depends on date code of the LTZ1000A.

nowadays I would use ADA4522

with best regards

Andreas

[grx]

Information duly noted, thank you.

edit: Would you adapt some compensation components when using an ADA4522 instead of LT1013 or LTC2057?

edit2: Also just noticed that for the ADA4522, in the absolute maximum ratings, the differential input voltage is specified as max +-5v

This specification is temporarily exceeded when the circuit starts. I have no idea if this is a problem or not.

edit3: they also suggest to avoid large source resistances so maybe some series resistors at op amp inputs should be removed in our designs.

[iMo]

..as I wrote here in different threads - doublecheck the leakage current of that Wima red bricks foil capacitors (if those are populated on your pcb). The samples here have quite large leakage (which increases exponentially with the temperature).

[Andreas]

Hello,

description of my cirquit designed for both OP-Amps is somewhere in this thread (~2013)
some hints are also here:

https://www.eevblog.com/forum/metrology/emi-measurements-of-a-volt-nut/msg2684103/#msg2684103
https://www.eevblog.com/forum/metrology/emi-measurements-of-a-volt-nut/msg2758840/#msg2758840 and following.

And no since most precision OP-Amps have some kind of diode protection between both input pins you cannot remove series resistors which limit the current to valid levels.

with best regards

Andreas

[Andreas]

..as I wrote here in different threads - doublecheck the leakage current of that Wima red bricks foil capacitors (if those are populated on your pcb). The samples here have quite large leakage (which increases exponentially with the temperature).

Hello Imo,
are these measured values or only simulated (perhaps with datasheet values?)
Do you have any measurements from WIMA MKP2 (poly propylene) as comparison?

with best regards

Andreas

[MiDi]

film-and-ceramic-capacitor-leakage-current ff

[grx]

the only wima I used are 10 and 100nF.

[iMo]

..as I wrote here in different threads - doublecheck the leakage current of that Wima red bricks foil capacitors (if those are populated on your pcb). The samples here have quite large leakage (which increases exponentially with the temperature).

Hello Imo,
are these measured values or only simulated (perhaps with datasheet values?)
Do you have any measurements from WIMA MKP2 (poly propylene) as comparison?

with best regards

Andreas

@Andreas - yes the simulation is based on my measurement. In the sim you simply vary the cap's Gohm at 25C and the -0.xx coef in the IB equation such the Ileakage graph fits your observation and it sims pretty good. The Eq. comes from a DS I found somewhere in past..
I made measurements with many smallish sized caps MKP (150n-470n) and MKT (330n-4u7) off my junkbox, and the MKP (all used from older PSUs) show 10-20pA at ambient and 100-250pA at say 50C (+/-10C). The MKT Wima vanilla smallish red bricks 4u7 (3pcs, 50V one) showed 2000-4000pA at higher temp.
Mind the measurements were pretty simple one with cap in ser w/ 10Meg_internal at 200.00mV range (100pA=1.00mV) and hair dryer (the sample in a small glass jar while blowing the hot air into it for 2-3minutes), so nothing sophisticated.

PS: and the big Q is how that ILeakage evolves with time, of course..