LT3094 -ve LDO equiv to LT3045

[alanambrose]

Hi, FYI I just noticed this comment:

https://ez.analog.com/thread/109212-can-the-lt3045-be-used-to-regulate-both-a-positive-and-negative-voltage-rail#comment-361010

... but can't see other info atm.

Alan

{edit - I meant 3045 rather than 3095}

[TiN]

I've used LT3092, it's nearly same as 3042/45.

[coppercone2]

Can you link a picture so we don't need to register to AD website (ridiclous)

I assume you are asking about making some kind of negative LDO that has a ridiculously low noise floor and great RF PSRR ?


I don't see a LT3094 chip?
I noticed in general there are no good manufactured equivalents of chips in the same ball park, maybe because of PNP stage?

You said 94 now 95 and 45 can you fix everything I am very confused.

[Echo88]

Seems they changed the Analog.com-Forum to log in only this day. 
Anyway: https://www.mouser.com/pdfdocs/ADI_LDO_General_Presentation_2018Apr1.pdf The LT3094 seems to have the same specs as its positive LT3045-brother.

[coppercone2]

Seems they changed the Analog.com-Forum to log in only this day. 
Anyway: https://www.mouser.com/pdfdocs/ADI_LDO_General_Presentation_2018Apr1.pdf The LT3094 seems to have the same specs as its positive LT3045-brother.

I am excited for that part. The order of magnitude difference was very irritating IMO. ~0.8 vs 18 !!!

[alanambrose]

Apologies for the slightly cryptic and erroneous initial title.

>>> I assume you are asking about making some kind of negative LDO that has a ridiculously low noise floor and great RF PSRR ?

Yes I think this is of general interest here.

>>> Anyway: https://www.mouser.com/pdfdocs/ADI_LDO_General_Presentation_2018Apr1.pdf The LT3094 seems to have the same specs as its positive LT3045-brother.

Wow, that's a nice presentation.

Amazing to see the favourable comparison with Li-ion noise; the absolutely critical layout; and the magnetic field analysis.

Alan

[Echo88]

https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/LT3094_Rev0_9-5-18.pdf 

[splin]

The 3042 and 3045 are pretty expensive, so how much might an external pass device, to increase output, compromise the noise performance?

What is likely to be the biggest factor behind the big improvement over previous regulators? Reference noise perhaps?

[Echo88]

Update: Digikey has the LT3094 in stock, while Analog.com still doesnt list it as a part.

[branadic]

The 3042 and 3045 are pretty expensive, so how much might an external pass device, to increase output, compromise the noise performance?

What is likely to be the biggest factor behind the big improvement over previous regulators? Reference noise perhaps?

Designed as a precision current reference followed by a high performance rail-to-rail voltage buffer...

instead of a voltage reference.

-branadic-

[MiDi]

Seems they changed the Analog.com-Forum to log in only this day. 
Anyway: https://www.mouser.com/pdfdocs/ADI_LDO_General_Presentation_2018Apr1.pdf The LT3094 seems to have the same specs as its positive LT3045-brother.

Luckily this seems not to be an April's fool 

[MiDi]

https://youtu.be/_U-KD5tHUuE

[TiN]

I got few some weeks ago from Digikey, but they still sit in ESD bag

[MiDi]

If you are curious whats inside the christmas cookie box: take a look at AN159

[Gerhard_dk4xp]

The 3042 and 3045 are pretty expensive, so how much might an external pass device, to increase output, compromise the noise performance?

What is likely to be the biggest factor behind the big improvement over previous regulators? Reference noise perhaps?

It is not much worse.

<    https://www.flickr.com/photos/137684711@N07/29452163806/in/album-72157662535945536/       >
Circuit is like the one in the data sheet. Measurement was quick & dirty without the
second alu coffin.

<   https://www.flickr.com/photos/137684711@N07/29197476530/in/album-72157662535945536/    >

To the left / right are measurements of other devices that might be interesting.


The stronger than 1/f increase of the noise at low frequencies is my preamplfier, the input capacitor
was too small for an effective bias network short.  -3dB is not everything. :-)
That has been healed with a wet slug tantalum; I intend to repeat the measurement campaign with
the new version of the amplifier.
0 dB == 1 nV/rt Hz.

The reference is a current source; that is much easier to decouple than a low impedance
voltage source. Don't overdo it with the output capacitor or you'll get noise peaks.

regards,
Gerhard