LH0070 / LM169 - die pictures

[Noopy]

Hi all!

I got a LH0070, a 10V-reference built by National Semiconductor:





  A ceramic-@*#+%? ...I don´t know what is the english name for "Keramikträger". 
There is a 56Ohm-resistor and a 3,6nF-capacitor in series at the output. It seems that the reference was unstable at higher frequencies and they had to damp it.




Hey, it´s a LM169! A reference that is a bit more accurate.




Does anybody know what that is? There is a fusible link but what is the big thing consisting out of two parts connected unter a metal block?


More pictures here:

https://richis-lab.de/REF09.htm


 

[iMo]

"keramikträger" == "ceramic substrate" ?

[Noopy]

"keramikträger" == "ceramic substrate" ?

Yes! Thanks! 
In fact in german also "Keramiksubstrat" is possible... Well too late for today... 

[edavid]

A ceramic-@*#+%? ...I don´t know what is the english name for "Keramikträger". 

English speaking electronics engineers would call that a "thick film hybrid circuit".

[Noopy]

A ceramic-@*#+%? ...I don´t know what is the english name for "Keramikträger". 

English speaking electronics engineers would call that a "thick film hybrid circuit".

Yes, also right!  Perhaps more right. 

[iMo]

"Thick film hybrid integrated circuits" are usually built on a "ceramic substrate" 

[Theboel]

Hi,

stupid question Is this a good reference ?
I have some I pull out from old board

[Noopy]

Good morning! 

It is a ceramic substrate with a thick film resistor but there are also ceramic Substrates with thin film resistors!  Because of that I usually use "ceramic substrate" that´s alway correct. 

Hi,

stupid question Is this a good reference ?
I have some I pull out from old board

Well, good, better, good enough… In my view it´s good. A voltnut would prefer a LTZ1000.

[Kleinstein]

It is a while I looked at it. AFAIR it is a relatively good one.  Roughly comparable to LT1031 or AD581 in a metal case.

[Noopy]

A moment ago I learned that the ceramic substrate is no surprise: LH stands for Linear Hybrid! 

[Lemonizer]

Noopy, those are really nice pictures ! Thanks for that, makes me want to get a microscope sometimes

Ceramic caps/components got a bad characteristic for such voltage reference ? I don't know much about them, but I'm quite interested.

I do have a LM399 board I bought online, that I'm using as a sanity check for my various DMM's. I'd love to build a few myself in the future !

[Noopy]

Thank you! 

There are a lot of different types of ceramic caps. You can also buy very good ceramic caps.
But I don´t think that is a problem here. The RC doesn´t stabilize the voltage it musst only short the output if the frequency gets to high. That´s a very easy task.

[edavid]

A moment ago I learned that the ceramic substrate is no surprise: LH stands for Linear Hybrid! 

I suspect there was an earlier version with more than one IC on the hybrid, before they integrated it into the LM169.

[schmitt trigger]

Noopy;
thanks again for taking us to the wonderful world of micro electronics.

Back when National Semi was a major player, their prefixes in their part numbers made a lot of sense:
LM= Linear Monolithic
LH= Linear Hybrid
DM= Digital Monolithic
LF=  Linear Fet
LMC= Linear Monolithic CMOS

And so on.

I read somewher that this was a carryover from Fairchild with their  uA = micro analog prefixes.

[Noopy]

A moment ago I learned that the ceramic substrate is no surprise: LH stands for Linear Hybrid! 

I suspect there was an earlier version with more than one IC on the hybrid, before they integrated it into the LM169.

Sounds reasonable! 

Noopy;
thanks again for taking us to the wonderful world of micro electronics.

Back when National Semi was a major player, their prefixes in their part numbers made a lot of sense:
LM= Linear Monolithic
LH= Linear Hybrid
DM= Digital Monolithic
LF=  Linear Fet
LMC= Linear Monolithic CMOS

And so on.

I read somewher that this was a carryover from Fairchild with their  uA = micro analog prefixes.

Thanks! 

[kleiner Rainer]

If you look at the datasheet of most regulators/references you will find that the output impedance is rising with frequency. This means the output impedance is inductive. Depending on load and output capacitor size, you get a nice noise voltage bump at the resonance frequency of the resulting series LC circuit. One way to mitigate this is to put a resistor in series with the cap and destroy the Q of the LC circuit. I suspect this is the case here.

Greetings,

Rainer

[edavid]

If you look at the datasheet of most regulators/references you will find that the output impedance is rising with frequency. This means the output impedance is inductive. Depending on load and output capacitor size, you get a nice noise voltage bump at the resonance frequency of the resulting series LC circuit. One way to mitigate this is to put a resistor in series with the cap and destroy the Q of the LC circuit. I suspect this is the case here.

Not exactly - the explanation is on the LM169 datasheet:

The output can sink current as well as source it, but the output impedance is much better for sourcing current. Also, the LM169/369 requires a 0.1μF tantalum capacitor (or, 0.1μF in series with 10Ω) bypass from the output to ground, for stable operation in shunt mode (output sinking current).

[kleiner Rainer]

I had this text in mind:

https://electrooptical.net/static/oldsite/www/sed/ErrolDietzRegulatorNoisePeaks.pdf

Could be useful if you need a low noise reference (or low noise in a certain frequency range).


Greetings,

Rainer