jaygunn

Title says it all. I am planning on building a lab DC power source with very clean output, using a linear power supply terminating in an LT3080 LDO linear regulator. I want really low ripple and noise.

It would make nice eye candy to display the voltage (especially if I make it adjustable) with a very affordable LED voltmeter like the ones on eBay for under $10. But will a part like that add a lot of noise to my clean DC? Most of them say 100K input impedance FWIW
John Gunn
Chapel Hill, NC


scott2000

Pretty sure I saw somewhere you should use a separate power supply for those displays....

Can't remember where I saw it though.....

jaygunn

I could use the post-filter-caps, pre-regulator voltage, about 20.5VDC instead of the post-reg 18VDC as the power. Might help?
John Gunn
Chapel Hill, NC

scott2000

Sorry....He was talking about if you use the one with the ammeter... not even sure it's relevant....

Found the video........

https://www.youtube.com/watch?v=F460UG1vfNs&list=UUaaLFE5H7583e_znOwch1uw&index=20

jaygunn

Thanks Scott. Wordy but rational youtube(!) I'll try a pair of these voltmeters and if they are noisy how much loss will there be?

Case closed I guess.
John Gunn
Chapel Hill, NC

abbey road d enfer

It would make nice eye candy to display the voltage (especially if I make it adjustable) with a very affordable LED voltmeter like the ones on eBay for under $10. But will a part like that add a lot of noise to my clean DC? Most of them say 100K input impedance FWIW
It depends on the particular circitry of the voltmeter, but the amount of noise that can be reinjected in a decently designed PSU is much less than the PSU's self-noise.
The LT3080 has an output impedance of about 1milliohm, the voltmeter is 100 kiloohms.
Even if itwas capable of sending 5V spurious (assuming 5v internal rail, actually probably less), and it would be a serious malfunction, that would result in 50nV. Utterly negligible.
Actually, why do you want to make it a two-tier design? I believe the LT3080 itself is capable of more than adequate performance.
The real challenge here is the layout, with a strict hierarchical arrangement all the way.
The arrangement of the safety ground is not trivial.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

jaygunn

This will be my first experience with the LT3080. I have a bunch of 1000U filter caps lying around and I thought like this: "If the LT3080 gives 60-70dB of ripple reduction, and if the ripple is pretty low to start with (2mV or less anyway), then those 60-70dB will effectively make it vanish."

I noted in the datasheet that LT wants the user to put a guard rail around the very high impedance adjustment pin, so I will do that, on both sides of the board. I will use wider than usual traces--why not? I'll run a shielded cable from the pin to the adjustment pot. I'll also put a metallic vertical partition between the transformer/diode bridge/coarse filter cap section and the regulator/display/banana jacks section of the case. Sometimes overbuilding is the way to go. Am I missing something important?
John Gunn
Chapel Hill, NC

abbey road d enfer

Am I missing something important?
PCB layout. You must isolate the transformer/rectifier/smoothing cap currents from the rest of the circuit, by having only one point of junction, at the foot of the smoothing cap.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

jaygunn

Will do. Thanks.
John Gunn
Chapel Hill, NC

 

Related Topics

  Subject / Started by Replies Last post
7 Replies
6482 Views
Last post May 13, 2005, 07:19:34 PM
by
1 Replies
1469 Views
Last post May 16, 2005, 03:01:49 AM
by pstamler
0 Replies
300 Views
Last post April 06, 2019, 06:17:48 PM
by acidblue
3 Replies
457 Views
Last post September 04, 2019, 04:41:47 PM
by PRR