Problem with TL431 in a SMPS - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 3rd August 2005, 08:53 PM   #1
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
Default Problem with TL431 in a SMPS

Hi,

I have started a small SMPS design with Power Intergrations circuits using aTinySwitch II Series IC.

My design is similar to "Engineering Prototype Report 14" except for the output voltage being 4,1 V/1,2 A instead of 9 V/0,3 mA and it also uses a TL431 shuntregulator instead of the zener used in their schematics to regulate output voltage seen in EPR 14 document on page 5.

The prototype started up fine and with 4,09 V at output with a 1 kohm bleader resistor on the output not leaving the output unloaded.
My problem is that the output voltage changes too much with load not expected by me with a shunregulator, max output current is around 1,2 A and at full load the output voltage is only 3,7 V, decreased with 0,4 V which is unacceptable for me, 0,1 V would be the target.

The configuration with the TL431 is the same as in Powerintergrations EPR 29 seen on page 5.

One problem is of course the HF ringing occuring at switch-off which becomes repeated much more often with increasing load.
Unfortunately I have at my hand only a 20 MHz scope and another problem is the continuosly modulating switcing frequency which makes it hard to inspect the curve traces.

BTW, I made up quickly with MSpaint copy-paste from both schemativs in above mentioned PDF's which looks like my PSU which I will try to attach here below.

Suggestions?
Thoughts regarding TL431 and high frequency disturbans?

Michael
Attached Images
File Type: jpg tny_4v1_reg_with_tl431.jpg (31.4 KB, 719 views)
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 3rd August 2005, 09:46 PM   #2
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Are you ensuring proper TL431 biasing by placing a 470 ohm resistor in paralell with the led of the optocoupler?

Note that the feedback current the phototransistor is required to draw from the tinyswitch is only approx. 250uA while the TL431 requires more than 1mA for optimum biasing.

Also, be aware of placing any capacitance between A and K pins of TL431 because it may oscillate. The datasheet includes a graph showing valid capacitance values for different 'gain settings'.
  Reply With Quote
Old 3rd August 2005, 11:03 PM   #3
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
Hi Eva,

who else then you would appear to help a guy with SMPS design issues!

I should say it again that the schematic is an approximate copy of my SMPS, I didn't take time to draw it up in a schematic editor so I just copy pasted from PI's two diffrent paper just for this post here on DIYaudio, as explained in my first post, lazy me.
The values differ just slightly so it's quite accurate.

The PC817A LED has an 1 kohm in paralell, I meassured the series resistor R9 to have a voltagefall corresponding to 1,5 mA and should be sufficient.
But maybe I will try to lower the value a bit.

The Capacitor in paralell with TL431 is from Power Intergrations EPR 29 paper, it's 10 uF electrolytic capacitor and PI explains this with giving the SMPS a softer start as it draws more current through the LED at charge up.
But I insist with out, it amazed me too a bit before I saw their explanation, actually the SMPS don't oscillate.

I have to say it again, these all-in-one el-cheapo circuits is a h*ll to work with because of their modulated switching frequency if using a modest oscilloscope, and scopes are necessary when developing SMPS's...

My theory is that it has too much HF ripple which the shunt reg cant handle every time the primary switches off, we are talking about MHz ringing.
So, when the load is increasing, the Tiny will increase the train of switching pulses (or how I should explain that as it has some type of "switch skipping" scheme) or should i say decreasing the skipping rate by load so the HF ringing will be more frequent and thereby the shunt reg can't notice the actual voltage as it contains more DC-overstored HF energy exceeding the stable charge voltage in the electrolytes seen between the switching moments...?
Anyhow, I suspect strongly the HF energy is affecting on the shunt reg in some way and have also considered a redesign of the PCB layout even if in current design I have already located the feedback circuitry couple of cm from the transformer because where they are now I could have located these components even much closer because of the free space.
Actually I even made a "dead" strip on the PCB layout between transformer and feedback circuitry connected to the transformers return/GND in an attempt to lower the stray pickup a bit.

Michael
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 3rd August 2005, 11:49 PM   #4
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
There are several things I would measure in order to understand such a problem:

- Voltage waveform at TL431 gate pin.
- Voltage waveform before and after the smoothing inductor.
- Voltage at TL431 cathode, is it saturating? it should have 2.5V or more between K and A.

Ummh... Now that I've mentioned it...

Your photodiode is dropping approx. 1.5V (1.5mA@1Kohm) and the current limiting resistor placed in series with it is also dropping some hundreds of milivolts, so Vka of your TL431 is probably lower than 2.5V. With such a low voltage it's not likely to work properly, namely the Vref may be no longer 2.5V but lower.

I would try to temporarily increase output voltage to 5V. Don't panic if this is the cause of the problem, I've already figured out some posible solutions.

PD: I neither like these compact ICs, I developed my own self oscillating flyback for low power applications:
http://eva.eslamejor.com/aux10w_0.jpg
http://eva.eslamejor.com/aux10w_1.jpg
http://eva.eslamejor.com/aux10w_2.jpg
  Reply With Quote
Old 4th August 2005, 06:03 PM   #5
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
Michael -- what are the values you are using for the transformer?
  Reply With Quote
Old 4th August 2005, 06:29 PM   #6
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
Jackinnj,

I ordered a ready made transformer from a manufacturer who claimed they make transformers for TNY circuits normaly.
The spcification I have at my hand says
* 2300 uH for primary
* leakage 13 uH
(they meassured it)
I have already EFD transformer kit but when i got this ready made transformer i didn't start to calculate the transformer.

One annoying thing seem to be that they have used cores without airgap, so they added some thin paper between or so, too bad if so because the stray magnetic field is worser for surrounding components...


Eva,

I have meassured and made some changes and tests now for couple of hours, I even made a voltage doubler just for feeding the optocoupler and I got between 10-20 V depending on load an input voltage, unfortunately it didn't help much.

I noticed that the output voltge is very stable with changing input voltage from 85 to 250 VAC(Actually I should go to 265 VAC but the variac didn't do it), at no load to almost full load the output is changing only some 10 mV with changing mains voltage, very nice.
This gives me a clue that it might be the HF energy that interfer somehow with the TL431 as the Tiny switches more frequent by load, so I have started to think if I can "rectify" that HF energy and "backfeeding" it to the shunregulator circuitry at some node and fool it so it will compensate.
Well, right now I'm pretty tired so I have to rest a bit.

BTW funny that your SMPS in your picture attached looked a lot like mine, I will post later some pictures.


A note: I should mention and make it clear that 4,1 V/1,2 A is not really full load but the mean load and should momentarily supply upto 2 A as the circuit the SMPS going to feed draws short pulses up to 2 A, but mean load is about 1,2 A. So I have to chose a balance between bulky el-lyte's or capable transformer.


Michael
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 4th August 2005, 11:06 PM   #7
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
And now some pictures as promised before!

This picture is the PCB upper side.
Attached Images
File Type: jpg tny_smps_upper_side.jpg (40.3 KB, 552 views)
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 4th August 2005, 11:07 PM   #8
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
..and PCB lower side...

As you can probably see theres 3 voltage outputs, but the others are quite neglible as theres no currents to talk about, and the voltages are 3 and 5 Volts.
Attached Images
File Type: jpg tny_smps_lower_side.jpg (43.0 KB, 530 views)
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 4th August 2005, 11:12 PM   #9
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
...and finally a oscilloscope picture.

The above wave trace is outputvoltage, settings are 20uS/div resp. 200mV/div.
Below is the voltage seen at the shunt regs ref input, 10mV/div

Michael
Attached Images
File Type: jpg tny_smps_scope_pic001.jpg (39.3 KB, 498 views)
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 5th August 2005, 01:15 AM   #10
cerrem is offline cerrem  United States
diyAudio Member
 
Join Date: Jun 2005
Location: San Diego, CA
I would HIGHLY suggest to get rid of the PI part and get a ST micro part .....
I have had it out many times with PI, they are arogant and think they are experts in switching supplies...
My job in one company was to redesign and improve the PI designs...
What are the output specs for the supply you would like to have?????
Majority of the problems like you describe are usually a result of the loop compensation.... The TL431 with an Opto is SLOW and low frequency POLE usually down about 8K to 10K at best....
Did you account for the worst case CTR of the OPTO????
The inductor size seems a bit odd..... Not to worry about the gap...Your main concern is adjusting the VOR and peak current to where you want them figuring the inductance last....the GAP will take care of itself if you design the transformer properly...
The Y cap is also a major issue here and is a HUGE current leakage.... It is there for EMI reduction, since returning it to the BUSS... But if you do not need to comply to standards , do not use it....

Chris
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
TL431 and SMPS Question pftrvlr Power Supplies 1 23rd July 2008 10:38 PM
SMPS problem? choncba Car Audio 11 8th May 2008 08:54 AM
Problem with TL431 regulation, plse help me. commstech Digital Source 27 31st March 2008 01:53 PM
What is the problem with the K6 SMPS? ghemink Power Supplies 2 16th March 2005 03:30 AM


New To Site? Need Help?

All times are GMT. The time now is 01:02 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2