Sonsivri
 
*
Welcome, Guest. Please login or register.
Did you miss your activation email?
February 19, 2019, 01:47:22 13:47


Login with username, password and session length


Pages: 1 [2]  All
Print
Author Topic: About switching power supply> First Contact to Load  (Read 2551 times)
0 Members and 1 Guest are viewing this topic.
Signal
Active Member
***
Offline Offline

Posts: 181

Thank You
-Given: 96
-Receive: 74



« Reply #25 on: January 14, 2019, 03:43:47 03:43 »

I have some thoughts about linear regulator to 3 V powered by 12V.
1) Linear regulator (MCP1703) from 12V to 3.3V at 250mA radiates more that 2W. It is too much for sot23. (No requirements was stated for current limit so I took the maximum for MCP1703)
2) Serial resistor 30 Ohm before such regulator introduces additional 7.5V drop and radiates 1.875W at 250mA. Then linear regulator has only 0.3W to radiate that is already acceptable.
3) 1uF at MCP1703 input as noted in datasheet is placed near to regulator and needed for stability. (For some regulators the requirement of having input capacitance not less or bigger then output is explicitly noted.) It is not enough to effectively compensate big load transients though if there is no additional bulk capacitor near on PCB and device is powered by long/unspecified cable. Of course it depends on dropout reserve too.
4) I'd use at least 100uF right after serial input resistor. Then tau=3000us. So with 30 Ohm load there will be no such "first connection" drop and overshot or it will be effectively filtered out. Rapid change of load current will cause "line/input" level change slow enough to effectively regulate it.
5) If manufacturer states 18V as absolute maximum without specifying time conditions (or avalanche rating) then I consider it dies at 18V immediately without any optimistic assumptions. Although I know that the one that kills electronic devices is energy, not voltage.
6) if an input is exposed and power supply polarity is not guaranteed then reverse power protection is needed - at lease a diode.

My choice however for low power device is to use 2954 or 9036 (both with reverse polarity protection) as first step to get 5V followed by 1703. As a bonus - 3.3V will be cleaner. And of course I'd consider the use of 220VAC/5VDC power supply instead of 12V. But I do not know full requirements for given device.
« Last Edit: January 14, 2019, 04:35:08 04:35 by Signal » Logged

Give a right name to a right game and play it right
Sideshow Bob
Cracking Team
Hero Member
****
Offline Offline

Posts: 611

Thank You
-Given: 197
-Receive: 634



« Reply #26 on: January 14, 2019, 04:57:38 04:57 »

This is just a suggestion. If you have some old cell phone chargers that you do not use any more why not at least try some of them and see if they can do the job. They have an output of about 5 volt. Those chargers that are orginal with the phones tend to be better than the no branded aftermarket ones. This could solve your voltage and power power dissipation issues

Posted on: January 14, 2019, 04:44:26 04:44 - Automerged

Exactly !

The question is this: 12V wall wart is not a good thing as they have no regulation most jive out 11 to 19 volts.

Thank you pickit2!

take care,
Xo!
I would say it would depend. As many switch mode power supplies are actally specified not to only have maximum load current. But also a minimum load to be in spec, regarding output voltage and ripple. In general connecting a 2A switcher to a very minimal load would be a bad desing in most cases
Logged

I have come here to chew bubblegum and kick ass... and I'm all out of bubblegum
Manuel
Active Member
***
Offline Offline

Posts: 137

Thank You
-Given: 188
-Receive: 72


« Reply #27 on: January 14, 2019, 04:57:47 04:57 »

I had chosen 1703 since it's SOT-23.
Obviously light load will allow to work properly even with such small case, so serial resistor should not be necessary I would say.
Agree with point #4 of SIGNAL.
Fortunately there are other producer than MCHIP that realize same component with higher rated voltage up to 40V, for light loads.

Take care,
Xo!
Logged

-> An Apple a Day does not Let U become a Macintosh!
kreutz
Active Member
***
Offline Offline

Posts: 122

Thank You
-Given: 361
-Receive: 108


« Reply #28 on: January 14, 2019, 11:26:26 11:26 »

IMHO you don't need to mess with the 3.3V regulator, just get another 12V power supply from a reputable manufacturer, or better yet, feed the 3.3V regulator from a 5-6 Volts pre-regulator. Cheap power supplies are cheap for a reason, in your case, the "cheapness" could be also a byproduct of a substandard design, or maybe its intended use doesn't require a good load transient response. Capacitive loading a switching power supply could induce control loop instabilities.

See: https://www.edn.com/design/power-management/4427185/DC-DC-converters-and-highly-capacitive-loads
« Last Edit: January 14, 2019, 11:32:46 11:32 by kreutz » Logged
Manuel
Active Member
***
Offline Offline

Posts: 137

Thank You
-Given: 188
-Receive: 72


« Reply #29 on: January 14, 2019, 02:32:38 14:32 »

Agree with you 120%,
apart for the fact that I need a working full functional cost effective solution with a complete introspection about issues.

 Grin Grin Grin Grin

Anyhow, just to complete the question MCP1703 is really a great product thinking about cost, dimension, quality and noise.
Talking about noise some sensors I can confirm takes advantage about this question.
This is the reason I was digging around it...

Take care,
Xo!

« Last Edit: January 14, 2019, 02:36:01 14:36 by Manuel » Logged

-> An Apple a Day does not Let U become a Macintosh!
Sideshow Bob
Cracking Team
Hero Member
****
Offline Offline

Posts: 611

Thank You
-Given: 197
-Receive: 634



« Reply #30 on: January 15, 2019, 03:14:25 03:14 »

Agree with you 120%,
apart for the fact that I need a working full functional cost effective solution with a complete introspection about issues.

Anyhow, just to complete the question MCP1703 is really a great product thinking about cost, dimension, quality and noise.
Talking about noise some sensors I can confirm takes advantage about this question.
If your power demands are quite low. But noise characteristics are more important. Why do keep flogging yourself by using a switch mode power. Google "linear power supply" and you will find that all major on line electronics stores have such wall warts. And for a low power unit the price is not that bad
Logged

I have come here to chew bubblegum and kick ass... and I'm all out of bubblegum
Manuel
Active Member
***
Offline Offline

Posts: 137

Thank You
-Given: 188
-Receive: 72


« Reply #31 on: January 15, 2019, 04:03:17 04:03 »

Dear Bob,
I agree with you also !

The noise I am talking about is related to the Voltage regulator itself, not to the power supply.
The best power supply with the wrong voltage regulator will anyway create problems for some products.

Family like 78L03 (and all related equivalent design) are really poor vs  1702/3 in terms of noise.

take care,
Xo1
Logged

-> An Apple a Day does not Let U become a Macintosh!
Sideshow Bob
Cracking Team
Hero Member
****
Offline Offline

Posts: 611

Thank You
-Given: 197
-Receive: 634



« Reply #32 on: January 16, 2019, 07:59:46 07:59 »

Dear Bob,
I agree with you also !

The noise I am talking about is related to the Voltage regulator itself, not to the power supply.
The best power supply with the wrong voltage regulator will anyway create problems for some products.

Family like 78L03 (and all related equivalent design) are really poor vs  1702/3 in terms of noise.

take care,
Xo1
But well switch mode power supply spiky noise have a nasty habit of be able to penetrate voltage regulators, allthoug in a damped form. So I would be much more consered of that kind of noise in a noise sensetive application. As the residual switing noise may very well end up in the mV range, there the internal regulator often ends up in the uV range.  Of course the internal noise generated by a voltage regulator is also part of a system design component selection
Logged

I have come here to chew bubblegum and kick ass... and I'm all out of bubblegum
Manuel
Active Member
***
Offline Offline

Posts: 137

Thank You
-Given: 188
-Receive: 72


« Reply #33 on: January 16, 2019, 03:36:06 15:36 »

YES BOB, agree.

Vreg can have an OUTnoise from INnoise with attenuation from 30db to 70db in general....frequency related...

Vin noise is important too obviously.

From my personal experience I would say that a 150mV ripple IN(put) vreg .... can be easily dumped....but vreg noise (expecially if you need pulsed current and minimum noise during pulsed current) is a different question. Always I refer to Measure field.

take care,
Xo!
Logged

-> An Apple a Day does not Let U become a Macintosh!
optikon
Cracking Team
Hero Member
****
Offline Offline

Posts: 723

Thank You
-Given: 745
-Receive: 1885


« Reply #34 on: January 16, 2019, 04:12:46 16:12 »

....able to penetrate voltage regulators, allthoug in a damped form. ....

I would say NOT even in a damped form. Most the of the ringing noise tends to be so high frequency, the regulator has no gain there and it just couples right through as if it was a wire.

TI makes some good HIGH PSRR regulators that are known to be used in RF design with noisy switchers.. so while not perfect, this is the kind of thing you want. a normal jellybean linear reg wont help your noise issue.
Logged

I can explain this to you. I can't comprehend it for you.
Pages: 1 [2]  All
Print
Jump to:  


DISCLAIMER
WE DONT HOST ANY ILLEGAL FILES ON THE SERVER
USE CONTACT US TO REPORT ILLEGAL FILES
ADMINISTRATORS CANNOT BE HELD RESPONSIBLE FOR USERS POSTS AND LINKS

... Copyright 2003-2999 Sonsivri.to ...
Powered by SMF 1.1.18 | SMF © 2006-2009, Simple Machines LLC | HarzeM Dilber MC