Hi Lorien,
It seems that these two ICs are a good choice. I use LT1763 in DAC digital section, compared to standard regulators, digital power supply steel free of noise.
Hello Lorien,
No, I never use reference as PSU. If you want to use reference, perhaps the LT1019 might be a better choice.
Attachments
Hi Lorien,
Just brain strom.
The discreate regulator probrably a good choice for your low noise application.
I do beleive that IC Regulator all of them has an error amp and and Vreference, are integrated inside. The noise is dominate from i/p noise of error amp and o/p noise of Vreference. Unfortunately, IC regulator can't put the big cap inside IC package so ..... you have option to put the cap at i/p and o/p of IC regulator and some IC regulator has bypass cap option. Don't for get the 2 resistors that divide the output voltage are the source of noise also (for regulator ADJ). You need to select the appropriate value and type of resistor also. For me i love SMD so i use thin film. Conventional resistor the noise voltage propotional with resistive value.
I have no specific test set up to test the noise of regulator. I just tried many the regulators with PLL to see the phase noise and also fed the ripple to DC offset at input of regulator to see how much of rejection. Most of IC regulators are not good when compare with discrete regulator (using the high performance part). I used to try to measure the noise at low frequency it is very slow.
First, you need to understand the source of noise and you will know how to eliminate. The ripple rejection is also. I think the selection of high quality caps low ESL+ESR. Generally, just parallel many capacitive values together (bypass all frequency) and PCB layout are very important also.
I always see many regulators (********) that they claim their design is a low noise on many websides. The ******** it mean the price is not reasonable for me estimate from the BOM cost. Although they do not share the schematic I just take a look the parts that they use
too expensive.I'm not expert but I beleive in audio world is no magic. It seems all complicated things and hard to prove become to Voodoo
.How do you think?
I'm also a believer in discrete noise regulator but my application was build around a cost-efective ideea thus I can not insert, for example, Paul Hynes' regs out there (even if I have a strong wish to do it) because I'll ruin my first goal. PHs are expensive but reading many reviews on web, they act as expected. Until now, LT1763 is too noisy for my requirements, instead I would rather use a reference IC like LT1019 - following Eric Juaneda' kind suggest... though there's no 3,3V option and every atempt to force the output to 3,3V will require additional parts => perhaps more noise than is desired ?
In any case, I've ordered few samples from Linear Technology for each IC for testing purposes...
Kind regards,
L
P.S. Does someone have an ideea on how can I adapt LT1019 for 3,3V output?
In any case, I've ordered few samples from Linear Technology for each IC for testing purposes...
Kind regards,
L
P.S. Does someone have an ideea on how can I adapt LT1019 for 3,3V output?
Last edited:
PH = Paul Hynes
No, it's not sensor In fact it will be an asynchronous I2S application with two on-board oscillators. I just want that PSU(s) for these oscillators to have the best specs I can afford within reasonable prices. I know that LT1019 have a high price but it's (still) duable.
BTW I was thinking to use the 2,5V option with a resistive divisor at output (3.3v -> 2.5V). It was the fastest ideea who cross my mind. I do NOT say it's the best one...
Kind wishes,
L
No, it's not sensor
In fact it will be an asynchronous I2S application with two on-board oscillators. I just want that PSU(s) for these oscillators to have the best specs I can afford within reasonable prices. I know that LT1019 have a high price but it's (still) duable.BTW I was thinking to use the 2,5V option with a resistive divisor at output (3.3v -> 2.5V). It was the fastest ideea who cross my mind. I do NOT say it's the best one...
Kind wishes,
L
If you will use reference voltage to supply the current to your oscillator pleae keep in mind that the referrence voltage is not intent to use as PSU. It can't provide high current to the load please see the spec of your referrence voltage. And if you divide voltage to 2.5V, at that node, the voltage is not constant due to the voltage drop between referrence output to divider node, your load (oscillator) is not take the constant current.
previously you said you need 3.3V but LT1019 has version of 2.5, 4.5, 5, and 10V.
don't worry about noise too much.
previously you said you need 3.3V but LT1019 has version of 2.5, 4.5, 5, and 10V.don't worry about noise too much.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.