The LTFLU (aka SZA263) reference zener diode circuit

[dr.diesel]

zlymex, is it possible to take higher res die photo?

[zlymex]

zlymex, is it possible to take higher res die photo?

I found that photo on the web

[plesa]

zlymex, is it possible to take higher res die photo?

I found that photo on the web

BU508A offered one LTFLU which he purchased for teardown, It it will be genuine we can have hires photo soon.

[zlymex]

zlymex, is it possible to take higher res die photo?

I found that photo on the web

BU508A offered one LTFLU which he purchased for teardown, It it will be genuine we can have hires photo soon.

Good news. I have several pieces of SZA263/LTFLU-1 but very reluctant to tear them apart.

[plesa]

zlymex, is it possible to take higher res die photo?

I found that photo on the web

BU508A offered one LTFLU which he purchased for teardown, It it will be genuine we can have hires photo soon.

Good news. I have several pieces of SZA263/LTFLU-1 but very reluctant to tear them apart.

We will see. There are some Fluke DMM with LTFLU, but I did not find it cheap enough on ebay for this experiment.

[dr.diesel]

If we don't end up with a high res photo from BU508A I've seen the Fluke DMMs sell for ~130 on ebay. 

I'm willing to de-lid for the cause. 

[Kleinstein]

I think there are die photos of the real LTFLU around somewhere.

The question would be if the cheap Chinese ones are real - if they are not and not working very well, than not much is lost to tear one apart. This might even if give a hint on what chip is inside and maybe even make use of them at all, even not as the original.

[manganin]

I received today the 5 "LTFLU-1ACH", here are some pictures of them.

Any suggestions / ideas?

1. Fake. Another chip which has the same 4 pin package. Easy to measure if the behaviour is totally different from the LTFLU. Any candidate for a cheap 4 pin metal can IC?

2. Fake. Built from a zener and a transistor. Requires tools for welding the metal can. Correct function but poor stability.

3. Fake. Another chip with the same function. The original Motorola version was freely available and manufactured a long time and in large quantities. But the housing was a little different? Maybe some Chinese or East European copy exists?

4. Real. Fluke and Keithley multimeters are manufactured in China nowadays. Even if real, certainly not the highest grade and possibly even rejected.

[zlymex]

I once bought five LTFLU-1ACH with date code 9515, they looks the same as BU508A's. Functionally are all good by test and two of them were being used in DIY of voltage references with very nice performance. By adjusting the resistors at the collector, a near zero tempco can be achieved. There is this DH80417B in one of my 731A transfer standard, was that made by Motorola?

[Dr. Frank]

I assume, that it's from T.I.:
https://www.eevblog.com/forum/testgear/fluke-332baf-in-the-slaughterhouse/msg393627/#msg393627

Frank

[zlymex]

Thanks Frank for the photo. Indeed the same reference IC inside TI's plastic container. I have two Fluke 335Ds but never dare to open those TI modules to see what inside. However, I find photos of other TI modules with Motorola SZA263 inside.

[fitzfish]

Hello everyone,

My first post here and work is not terribly permitting of time so please be patient If I don't get back to questions in a timely manner...

Perhaps I can contribute some additional pictures of the die. The first is the Motorola SZA236 removed from a Fluke 732A. The second is an example of the LTFLU removed from a calibrator board. You'll have to forgive the picture quality. I don't have a great tool to do this.

Enjoy,
Dave

[zlymex]

Hello everyone,

My first post here and work is not terribly permitting of time so please be patient If I don't get back to questions in a timely manner...

Perhaps I can contribute some additional pictures of the die. The first is the Motorola SZA236 removed from a Fluke 732A. The second is an example of the LTFLU removed from a calibrator board. You'll have to forgive the picture quality. I don't have a great tool to do this.

Enjoy,
Dave

Great photo, Welcome here!
Interesting to see the first photo, the zener and the transistor were separated.

[branadic]

Could someone please explain where the zener can be found on LTFLU and what this multiple taps are for? What is the transistor here?

[zlymex]

Could someone please explain where the zener can be found on LTFLU and what this multiple taps are for? What is the transistor here?

My guess, red area is two zeners in parallel, green area transistors, blue protection diodes.

[plesa]

The blue and green areas are pad for bonding wire and probing ( measurement). Similar are on almost all IC including LTZ1000.
I agree that in center area is the buried zener itself and it looks like two mirrored structures.
Transistors are between the blue/green areas and center red area.

[saturnin]

Since the photo of LTFLU die is not very clear, we can only speculate about its layout.

I attached a picture with my guess (everything is described in the picture, green rectangles are possible locations of transistor(s)). It seems to me the symbol used for LTFLU-1 is rather simplified schematic while its internal structure is more complex.

[fitzfish]

A better picture. Not much more I can do with a 30x microscope and a phone camera...

[fitzfish]

Great photo, Welcome here!
Interesting to see the first photo, the zener and the transistor were separated.

Thanks Zlymex.

I've enjoyed reading through the zener related threads.

[plesa]

It looks like LTFLU is not glued directly to package. I can imagine some thermal insulation material or something to relax stress to silicon.

[saturnin]

@fitzfish:

That's much better. Very interesting, thanks!

I think stripes at the top and on the bottom of your photo are presumably diffused resistors. They seem to be connected to pads only, so they may be used as a heater during trimming process?
I can also see two structures under the top metal in the center of the die.

It is definitely far more complex layout than the one in LTZ1000.

[Gyro]

Definitely right about the trimming fuses, some are blown and others not. There are a couple more in the top left area - in fact there's one on the thick track just above left of the first zener, a 'fuse after trim'?

P.S. Any chance that the long dark structure running the full length of the top of the die is a heater?

[Gyro]

I was told that the LTFLU has a heater, but Fluke decided not to use it.

Actually, there is another identical 'possibly heater' structure running along the bottom of the die (less clear because of the metalization above). The two appear to be wired in series and are terminated on the second and third pads from top right (1st one is bonded out).

[zlymex]

Thanks @fitzfish for the clear photo, not bad at all for a phone camera.
Some of the pads are for binary fuses.
Only "1" left un-blown, that seems to me not a trim at all, rather, it's a straight forward go(other devices could be the same).

[fitzfish]

One more... This one uses highly oblique illumination to try to bring out the features obscured by the washout in the previous pictures that used diffuse direct illumination.

The fusible links are different on this sample indicating that they are trimmed for some purpose. Could the two arrays either side of the center band be a large transistor array that are trimmed to match it's TC to more closely cancel the TC of the paralleled zeners?

The fusible links that have been blown shine brightly with this illumination method.

Enjoy!