[volt-nuts] any unheated Vref better than LT1021-7 (at 7 ppm/khr typ.) ?

[Andreas Jahn]

Hello beale,

the VRE305 and VRE3050 references from APEX (formerly THALER) are specced to 
have typically 6ppm/kHr.
They are in hermetically ceramic packages.

But in real life applications the long term drift at 25 degrees is only one 
of many specs that determine accuracy.
And mostly the long term drift is only valid only for the hermetically tight 
package (which is in most cases obsolete).
Additionally  its only a typical value and if you observe the parts there 
are large differences from device to device.

Generally you should look for devices with a buried zener and a hermetically 
tight package.
(AD586LQ, AD587UQ, LT1031CCH, LT1236AILS8-5, LT1021CCH, MAX6350MJA ...)
And you will have to select out the bad devices if you have high demands.

Hysteresis is another problem which arises if you do not operate the device 
at constant temperature.

> 2) I have seen mention of dipping a PCB in molten paraffin as a conformal 
> coat.
I cannot believe that this will really help. (But I have not tried this up 
to now). I guess you will get a longer time constant for humidity effects.
The problem with humidity is that it takes long time (several days) until 
the plastic case or the PCB swells or shrinks.
So even if you mount a hermetically tight package on a PCB you will have a 
humidity effect from the PCB shrinking.
Generally you can expect that a DIP device will be more stable than a SMD 
device.
And a hermetically device will be more stable than a plastic device.
Additionally you will have to take away mechanically stress of PCB 
(humidity) from the PINs of the reference.

On 2 LT1027CCN8-5 devices (only the ground pin soldered to the PCB the 
others soldered with VERO-wire)
 I measure a typical humidity drift of around 0.5 ppm/% humidity change.

with best regards

Andreas

----- Original Message ----- 
From: "beale" <beale at bealecorner.com>
To: <volt-nuts at febo.com>
Sent: Tuesday, January 01, 2013 6:50 PM
Subject: [volt-nuts] any unheated Vref better than LT1021-7 (at 7 ppm/khr 
typ.) ?


> Happy 2013 to all volt-nuts!
>
> I have a question about unheated voltage reference chips. I am interested 
> in long-term drift, for example ppm in the first 1000 hours.  The 
> manufacturers seem coy about this spec, as I have not found any parts 
> listing allowing you to sort on that parameter.
>
> The lowest drift spec I have found so far is the Linear Tech LT1021-7 
> which claims 7 ppm for the first 1000 hours (typical), although the data 
> sheet does not provide a graph. The -5 and -10 volt versions of the same 
> part are twice as bad, attributed to the internal output scaling 
> resistors. I have a few questions:
>
> 1) Does anyone know of an unheated reference with a lower 1khr drift? Not 
> counting the "deliberate lie" that 
> http://cds.linear.com/docs/Design%20Note/dn229f.pdf complains about.
>
> 2) I have seen mention of dipping a PCB in molten paraffin as a conformal 
> coat. Can anyone say if that is hermetic enough to remove the "typically 
> 100 ppm" drift from relative humidity changes as mentioned on p.13 of the 
> LT6655 datasheet http://cds.linear.com/docs/Datasheet/6655fb.pdf ?  I 
> gather the board does have to be scrupulously clean before the coating.
>
> 3) If the humidity effect is strictly from mechanical stress on the die as 
> the plastic encapsulant swells or shrinks, would a part in a metal can 
> then be immune from this effect, even if it was non-hermetic?  (Assuming 
> that it was also mechanically decoupled from dimensional changes in the 
> PCB.)
>
> Thanks for any insight on these issues!
>
> _______________________________________________
> volt-nuts mailing list -- volt-nuts at febo.com
> To unsubscribe, go to 
> https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
> and follow the instructions there.