Sonsivri
 
*
Welcome, Guest. Please login or register.
Did you miss your activation email?
December 08, 2016, 03:01:53 03:01


Login with username, password and session length


Pages: [1]
Print
Author Topic: Tranformer for CFL inverter ?  (Read 11546 times)
0 Members and 1 Guest are viewing this topic.
ronk
Guest
« on: September 01, 2008, 05:34:54 17:34 »

Hi, I want to build CFL Inverter from siliconchip artikel
http://www.siliconchip.com.au/cms/A_102473/article.html

Have problem with the transformer, how to construct the transformer windings (primary/secondary),
or can I use standart EI transformer 12V CT / 220V 3A ?

Thanks.
Logged
Kwaigo
Newbie
*
Offline Offline

Posts: 25

Thank You
-Given: 5
-Receive: 6


« Reply #1 on: September 02, 2008, 09:22:37 21:22 »

You can use transformer, if it works at inverter frequency.

Here is theory:
http://focus.ti.com/lit/ml/slup126/slup126.pdf

But there is simple method.
1) Choose the core, against frequency and power rate. (look for formula up there)
2) Make the primary winding with large turns reserve.
3) Use oscilloscope and low ohmic resistor to measure current in winding.
4) Step by step decrease turns count, until current rises to 3-4Amp, then increase turns count in 1.5 times.
5) Make the secondary winding, voltage is proportional with turns count.
« Last Edit: September 03, 2008, 09:54:43 21:54 by Kwaigo » Logged
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #2 on: January 03, 2009, 09:12:42 21:12 »

Hi,
Do not bother for Silicon Chip circuit. You can try the following two very easy circuits. These are reliable and tested.





Logged
Jehan
Junior Member
**
Offline Offline

Posts: 52

Thank You
-Given: 46
-Receive: 18


« Reply #3 on: January 05, 2009, 12:27:38 12:27 »

From the silicon chip Photograph, i can say that it is an ETD29 type ferrite core.
for an input of 12 volt and output of 370 volt o/p the required turns will be 7 turns for the Primary and 500 turns for the secondary. but you should observe the polarity at the primary circuit and the centre tap should be tapped in the proper polarity. if you have access to a LCR Bridge it will be very easy. Cool
Logged
sohel
Senior Member
****
Offline Offline

Posts: 437

Thank You
-Given: 140
-Receive: 124



WWW
« Reply #4 on: January 05, 2009, 01:02:01 13:02 »

he was asked for that transformer. please figureout tranformer core size, turn ,wire no etc.

if this is for his requerment why brother sending him to another world to learn smps topology. Huh Huh

take a look
http://www.pavouk.org/hw/lamp/en_index.html
« Last Edit: January 05, 2009, 04:00:59 16:00 by sohel » Logged

A Thousand Miles Journey Start With a Single Step<br />http://www.youtube.com/user/masud58
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #5 on: January 14, 2009, 03:03:01 15:03 »

Hi Jehan,
From Sohel's attachment, it could be clearly found that, the Transformer is ETD39,the max voltage in the circuit is 370 and the Topology is Push-Pull. For Primary winding, you have used 7 Turns, that means you are using 900 gauss of Flux density in the core for the circuit and with that Flux Density, the Frequency should be 40KHz. So, with that Parameter, the Secondary winding will be 212 Turns. If frequency is changed to 50KHz, with same Flux density, the Turns will be, 6 Turns Primary and 170 Turns secondary. All the calculations are based on 80% efficiency. However, let us consider the first parameter and then, our Turn ratio and other calculation will be:

7 Turns Primary
212 Turns Secondary
Flux density= 900 gauss
Frequency = 40kHz
Efficiency= 80%
Wire Size= Primary #22 and Secondary #28
Core = ETD 39, but smaller size like ETD 34 also can be used.
          EI 33 Transformer also can be used.

For Ronk,
Hope, above mentioned calculation will fulfill your requirement. For anything else,you can inform and I will try to help.
If you don't mind, I will advice you one thing. For smps circuit lot of theoretical and practical knowledge is required. Just following a circuit and try to construct one may be dangerous in some cases. Follow the safety rule and before put your hand in the circuit, discharge the capacitors first. Good Luck.
« Last Edit: January 20, 2009, 08:22:32 08:22 by tAhm1D » Logged
rajudp
Newbie
*
Offline Offline

Posts: 28

Thank You
-Given: 1
-Receive: 2


« Reply #6 on: January 20, 2009, 02:18:39 02:18 »

what is the formula to calculate code windings?. any airgap needed between cores/
Logged
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #7 on: January 20, 2009, 08:21:32 08:21 »

Hi rajudp,
For Push-pull, Half Bridge and Full Bridge Topology, you don't require any air gap. For Fly back Topology, You require air gap.
Certainly, there are lot of calculations for core winding and setting other parameters and in smps, these calculations should be followed properly to get optimum result/efficiency. You may find some calculations here:

http://www.sonsivri.com/forum/index.php?topic=19587.0
Logged
ajmain
Guest
« Reply #8 on: January 20, 2009, 04:03:16 16:03 »

what is the formula to calculate code windings?. any airgap needed between cores/

It is not necessary, but an airgap of about 10-20 mils will stabilize the transformer, ie, it will help reduce chances of core saturation to a certain degree.
Logged
rajudp
Newbie
*
Offline Offline

Posts: 28

Thank You
-Given: 1
-Receive: 2


« Reply #9 on: April 04, 2009, 08:59:42 20:59 »

Hi Tham 1D,
 i did the circuit but the mosfet is getting very hot when load is connected. if i connect about 40w afetr some time the rectifier diodes are getting damaged.

at about what frequency the BA169 can be Used. i dis the circuit "12V Flourescent Lamp Inverter  September, 2002  silicon chip "by replacing driver stage with 3525 pwm and insted of 1N4936 output rectifier i used ba159. when load is connecting the mosfet is getting very hot. draws about 5A current with 15W CFL.
Logged
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #10 on: April 05, 2009, 03:07:04 15:07 »

Hi Tham 1D,
 i did the circuit but the mosfet is getting very hot when load is connected. if i connect about 40w afetr some time the rectifier diodes are getting damaged.

at about what frequency the BA169 can be Used. i dis the circuit "12V Flourescent Lamp Inverter  September, 2002  silicon chip "by replacing driver stage with 3525 pwm and insted of 1N4936 output rectifier i used ba159. when load is connecting the mosfet is getting very hot. draws about 5A current with 15W CFL.


Hi,

There can be some reasons for your problem:

1. You could add an input inductor to prevent input current rush as power is applied.

2. You could add a deadtime by adding a resistor between pins 5 and 7 of SG3525. A deadtime of 1 to 1.2 uS or so will suffice and for this you can use a 220 ohm 1/4 watt resistor.

3. You could increase the softstart time by increasing the 100nF capacitor to 10uF.

4. You could add a snubber across the transformer primaries to suppress transients.

5. Normally in push pull circuits, you do not really need a snubber, but maybe because of your winding defect, the transformer has higher leakage inductance which is causing spikes and heating up the MOSFETs, so that's why you require the snubber here.

6. You do not need 370V output. Around 320V will be good enough. More voltage will create more problems.

7. For this circuit feedback is not required as this is a low power design. So, keep the high voltage ground and battery voltage ground isolated.

8. Add an output bulk high voltage (min 400V) electrolytic capacitor (10 - 22 uF) to work as filter. But if you do use 370V output, then the capacitor must be of minimum 500V.

9. BA159 is way too slow for 100kHz. 500nS delay is way too slow. Use something like a UF4007 or BYV26E which have only 75nS reverse recovery time, or use BYT43J (50nS). Ultrafast rectifiers should be used not only fast rectifiers.

10. 100kHz frequency is too high with the existing circuit. You should use something around 40kHz (I use 35kHz) because at 100kHz many parameters have to be taken into consideration, such as accurate transformer winding, snubber setting, deadtime setting, reducing leakage inductance, etc which is applicable for only expert SMPS designer.

In SMPS, only following circuit will not do, a lot of parameters have to be set for which a lot of practical knowledge is needed.

Honestly speaking, the circuit given in the Silicon Chip article is poorly designed.
« Last Edit: April 05, 2009, 03:09:20 15:09 by tAhm1D » Logged
rajudp
Newbie
*
Offline Offline

Posts: 28

Thank You
-Given: 1
-Receive: 2


« Reply #11 on: April 05, 2009, 03:56:03 15:56 »

Hi,
for the new design i am using toride core T 36x23x15 with 4t+4t primary and
"12V Flourescent Lamp Inverter  September, 2002  silicon chip " witn 0.75mm2 wire and sec is 134t with .4mm wire for 340V . the circuit
http://www.siliconchip.com.au/cms/A_30504/article.html
i replaced the pwm part with CFL inverter 3525 ic with rt of 5.57k Ohm and CT is 1nf.(around 256Khz oscillation and 128Khz Pwm ) but with frequency meter oscillation is showing around 186 Khz and Pwm is showing 100Khz (?)
But the drain of the IRFz44 is comming with around 40V to ground with out connecting the output rectifier. when connecting load this is again hign with high frequency oscillations. even to the gate of mosfet. i added a ferite bead to gate and a 1n4148 accross gate resister 22 Ohm

Posted on: April 05, 2009, 03:39:41 15:39 - Automerged

i used 40khz  the same design cfl inverter but when connecting  around 40W load after some time 5-10 min the rectifier getting dammaged.

for the 12V Flourescent Lamp Inverter  September, 2002  silicon chip when connecting 8W cfl thecurrent is around 1.2A

i purchased a local cfl inverter board they are not usein any rectifier at output and  using 20Khz Pwm. i disassembled the transformer and they are using flux density as 1000 ( i am using 2000 till 35Khz) so insted of 6T 6T for primary they are using 12T 12T. and 215T sec.for EI30 Core
Logged
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #12 on: April 09, 2009, 01:12:56 13:12 »

Hi Rajudp,

you can try with the following tested circuit or you can study it for further clarification. All required data are there and the parts required are also mentioned clearly and the circuit can be tested with EI 33 or ETD 34. If you study carefully, you can find many answers for your questions.


                                                         Detailed calculation part.


                                                            The circuit.
« Last Edit: April 09, 2009, 09:57:16 21:57 by tAhm1D » Logged
rajudp
Newbie
*
Offline Offline

Posts: 28

Thank You
-Given: 1
-Receive: 2


« Reply #13 on: April 09, 2009, 05:12:28 17:12 »

Hi tAhm1D,
  I will check the circuit and give feedback. i have one more doubt the given RT and  CT will give about 48khz (F= 1/(0.7RT + 3RD) CT) so each driver will have half of the frequency i.e. 24Khz. As per my previous post about local CFL inverter board the core is EI33 and primary is 12+12T and sec is 215T and not using any feedback. In that the given RT is 28.8Kohm (22k and 6.8k) CT is 1nf and RD is 1Kohm. It is giving a clean waveform (only a small peak).   It is used p55nf06 Mosfet  and the Mosfet is heating only very less at 40W load( a small heat sink is used). Please tell me the primary 4+4t will work fine. According to 40Khz it is correct but what about 24Khz ( 20Khz).

« Last Edit: April 09, 2009, 05:27:45 17:27 by rajudp » Logged
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #14 on: April 09, 2009, 10:05:37 22:05 »

Hi rajudp,
I corrected the parameter of the software and posted the corrected one. Yes, frequency should be 80khz, not 40khz as the frequency will be divided and will be 40khz in each phase. With 40khz, also it will work but efficiency will be slightly less and there will be slight heat generated. In this circuit, since it is small powered circuit, you can omit the feedback portion.
Hope, it will be useful for you.
« Last Edit: April 09, 2009, 10:19:02 22:19 by tAhm1D » Logged
HackerNo.1
Newbie
*
Offline Offline

Posts: 21

Thank You
-Given: 25
-Receive: 27


« Reply #15 on: April 25, 2009, 06:49:49 18:49 »

Hi Tahmid,
Your observations regarding Rajudps circuit is appropriate and I find it helpful for one of my circuit. The Software is also beautiful. Calculations are mostly correct. I am searching for this type of software for a long time. Hope to get help from you in this regard in future.
Logged
mitsos
Senior Member
****
Offline Offline

Posts: 449

Thank You
-Given: 1570
-Receive: 683


« Reply #16 on: May 13, 2009, 07:03:58 19:03 »

hi

I have many CFLs with open-circuited filaments.
The above circuits give life again to these lamps.
The question is: how long ago a CFL lamp can operate by this way?

regards
mitsos

PS if the operation is for example some years, it is a smart idea to buy the most cheap CFLs and to exchange the half bridge converter by a solution like here.
Logged
Prisim
Junior Member
**
 Muted
Offline Offline

Posts: 86

Thank You
-Given: 28
-Receive: 57


« Reply #17 on: May 14, 2009, 04:45:09 16:45 »

Hi Rajudp,

you can try with the following tested circuit or you can study it for further clarification. All required data are there and the parts required are also mentioned clearly and the circuit can be tested with EI 33 or ETD 34. If you study carefully, you can find many answers for your questions.


                                                         Detailed calculation part.


                                                            The circuit.
can u upload the software  showing in picture
Logged
BRANCA
Junior Member
**
 Muted
Offline Offline

Posts: 84

Thank You
-Given: 127
-Receive: 68


« Reply #18 on: August 08, 2009, 11:43:59 23:43 »

Same question-can you upload software smps topo-master?
Till now I find only a very good soft Design tools pulse transformers 4.0.0.0,but the interface is in russian and english XP cannot see fonts!The author is promising an english version,but...is a long way.
Logged
tAhm1D
Senior Member
****
Offline Offline

Posts: 336

Thank You
-Given: 103
-Receive: 180



« Reply #19 on: August 14, 2009, 04:31:10 16:31 »

Hi,
I cant upload SMPS Topo-Master right now as I still have to make quite a few modifications to it and also perfect it.
Hope I can finish in the near future.
Logged
BRANCA
Junior Member
**
 Muted
Offline Offline

Posts: 84

Thank You
-Given: 127
-Receive: 68


« Reply #20 on: August 16, 2009, 04:09:50 16:09 »

We are waiting.Thanks for your hard work!
Logged
download_crux
Junior Member
**
Offline Offline

Posts: 70

Thank You
-Given: 48
-Receive: 63



« Reply #21 on: February 10, 2010, 04:46:48 04:46 »

Hi,
I cant upload SMPS Topo-Master right now as I still have to make quite a few modifications to it and also perfect it.
Hope I can finish in the near future.

hi
tAhm1D

r u complete the modifications in SMPS Topo-Master?HuhHuh??

please upload SMPS Topo-Master.........

Regards
download_crux
Logged
Capitaine Haddock
Newbie
*
 Muted
Offline Offline

Posts: 14

Thank You
-Given: 11
-Receive: 14

Radio Actif


« Reply #22 on: March 05, 2010, 07:54:05 19:54 »

Hi Friends
i am also interrested in SMPS TOPO-MASTER download.
i saw a V2.0 somewhere but was unable to download free.
perhaps available soon here ??
archibald
Logged
Pages: [1]
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