NAS/MEDIA PLAYER線性電源pcb

royccw

johnme 發表於 2014-4-7 18:43
Thank you for the reminder.

The choice of capacitor size depends on what frequency you are workin ...

Thanks for your info.
I don't know much about LT regulator but use NS LM1085 before, I tried to use Cadj=47uF but seems doesn't work which mean there is no output at all until I change down to 22uF or less. I guess it may related to combination of Radj & Cadj in order for internal operation.
Just guess and may difference design between NS & LT.

johnme

royccw 發表於 2014-4-7 20:33
Thanks for your info.
I don't know much about LT regulator but use NS LM1085 before, I tried to us ...

I am glad that you share you experience of working with other regulator ICs.

You are welcomed.

johnme

royccw,

is LM1085 the TI version of LT1085?

How does it compare with LT1085?

billymud

johnme 發表於 2014-4-8 21:42
royccw,

is LM1085 the TI version of LT1085?

LM by Texas Instrument, LT by Linear Technology.
Go read their LT1085 data sheets for the good, the bad and the evil on them.

johnme

billymud 發表於 2014-4-9 01:45
LM by Texas Instrument, LT by Linear Technology.
Go read their LT1085 data sheets for the good, t ...

I always read datasheets before I ask questions.

BTW,

1. LM1084 is much cheaper than LT1084, I ordered some for trial.

2. The LM1084 datasheet is just a direct copy of LT1084.

See the figure (from LM1084) below:

It seems that the ripple rejection at low frequency is better than LT1085.

So, the insertion of large electrolyte Cadj is not need for LM1085. right?

johnme

billymud 發表於 2014-4-9 01:45
LM by Texas Instrument, LT by Linear Technology.
Go read their LT1085 data sheets for the good, t ...

When it says protection diode is optional, I take it seriously.

Believe me, it is ok. I tested a prototype without any protection diodes.

IN fact, most LT regulator ICs have built-in protection.

royccw

johnme 發表於 2014-4-8 21:42
royccw,

is LM1085 the TI version of LT1085?

My understanding is that they are almost same but don't know who first made, however, for LM317 regulator, LT should be better than LM.

royccw

johnme 發表於 2014-4-9 08:08
I always read datasheets before I ask questions.

BTW,

don't be just concentrate on ripple rejection, as this figure may advantage if Vin is large ripple or noisy but in your case it would not much change.
One of the figure need to concern is RMS noise output in 10~10kHz range, in depend on IC itself, so far I found LT317A should be the second best one.
The very best one should be LT3080, although it is only 1A output.

royccw

royccw 發表於 2014-4-9 11:13
My understanding is that they are almost same but don't know who first made, however, for LM317 re ...

Orginial purpose of protection diode use is to prevent ouptut voltage is larger than input.
If your output cap is very large means it may happen o/p > i/p when switch power off, IC will easily broken.

johnme

royccw 發表於 2014-4-9 11:26
don't be just concentrate on ripple rejection, as this figure may advantage if Vin is large ripple ...

Agree with you that ripple rejection is only one aspect of the regulator performance.

As I am working with an IC (LT1084), I cannot address all performance parameters mentioned in the link below
as I would have done with a all discrete design like Sigma TMC posted in the first page #2:

http://sound.westhost.com/articles/power-supplies2.htm

Yet, in the current design, I focus on the following things commonly encountered by other c-hings here:

1. ripple rejection
2. noise
3. transient response
4. heat problem