Etd 49 & 3c90 Help!!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,

Do not try to build offline smps if that is your first project and you do not have access to isolation transformer and oscilloscope. It can have lethal consequences. Use this link page 21 to calculate power capacity of your core. Do not forget to factor in proper isolation between primary and secondary (minimum three layers of insulation and 8mm creepage distance).

Best regards,

Jaka Racman
 
This thread may contain useful tips :

http://www.diyaudio.com/forums/showthread.php?s=&threadid=30183

I recommend low voltage SMPS experimentation to grow some experience and knowledge before trying to handle high voltages. Oscilloscope >=20Mhz with >1KV internally attenuated probes, mains isolation transformers and step-down transformers are almost mandatory. Also, it's not a great idea to test prototypes with 300V DC the first time you run it, I use 50V-100V in series with a light-bulb or even 24V regulated and current-limited for that purpose, and when everything appears to work right I increase the supply voltage. Never test anything with 320V [230V rectified] before succesful testing with lower voltages or without a reasonable size fuse in series
 
The ETD49 with a 3c90 core is quite a big core. the 3c90 material should be run between 20kHz and 200kHz. I'm unsure as to the maximum flux density (or what ever it is called), but you should be able to get >300W thru it at 20kHz. I've currently got some effiency issues with one I'm bulding at the moment (same former and material), so I can't really tell you what you can do. I'm getting about 150W out for 250W in :xeye:. 60% effiency - Not good.
 
Altough english is not my mother language, I think the term 'offline SMPS' designates any SMPS circuit whose primary side is directly connected to mains line [probably from 'power taken off the line']

I don't know of any references to 'online SMPS'
 
Online/Offline

...UHPS!! I thought the other way round...?
I was thinking that offline SMPS are not directly connected to the mains, because the "off" is written with "ff" not "f".
I interpreted it the same way as I am used from the WEB, and thought offline would mean not connected.
:xeye: :xeye:
i.e. car smps or sm-regulators behind a regular isolated transformer...

??????

@George:
No matter how it is called....
The thing which you are planning seems to be the more difficult and more dangerous sort of smps. If I understood your first post correct,
you are planning t as you are going to deal with the full non isolated mains voltage...
And also your output voltage is quite high!
Do you need this for tubes, or are you going to supply a high power class D halfbridge amp?? Just curious.

Also the IR2153 is not directly implicating a certain topology of your smps. But it does not lead my thoughts to a standard single ended forward converter or a push pull topology....
Also curious in this regard ..;)

Bye and by careful
Markus
 
Hi!

I've just finished a 230AC/2*100VDC SMPS with ETD44. For a short time (some minutes) it delivers 2000W. Time is not limited by transformator, but the heat sink on semiconductors. But: the proper winding of a SMPS transformator is not easy! If the coupling is not perfect, it will produce very much heat! Current is limited by series inductance, and a fuse. Operating frequency is 50 kHz. At higher freq the core can handle more power (again: with proper winding), but switching losses increasing on transistors.

This is strange. IR didn't specify output current on IR2153!

Here is my SMPS:
http://www.hszk.bme.hu/~sp215/elektro/SMPS_top.jpg
http://www.hszk.bme.hu/~sp215/elektro/SMPS_bottom.jpg
 
Hi Pafi,
yeah, your SMPS looks simialr to something that I will need for my subwoofer project.
Does it provide self equivalent line isolation?
And power factor correction / input line harmonics acc. EN 61000-3-2 ?
...first point is a must? Second point would be nice to have...

Bye
Markus
 
Hi Markus!

I don't really know what does "self equivalent line isolation" means, but the answer is I haven't test it yet. Power Factor is a very weak point of this circuit. I'm absolutely sure it doesn't meet any standard. :ashamed: It can be easily modified to have better PF, but then load regulation and maximal power output would be lower. PFC is a future plan.
 
HI EVERY ONE

I just fiinish schmatic,but all the value is not calculated.so any one give me any advice and suggestion:)
 

Attachments

  • power s.jpg
    power s.jpg
    61.4 KB · Views: 1,112
HI Pafi!
Sorry I had syntax error.
It should comply to be SELV-equivalent saftey standards.
SELV means Safety Extra Low Voltage.
In order to achieve equivalent safety as in such low voltage circuits there have been defined several requirements for creepages, clearances, minimum thicknesses and properties of the used isolating materials, test volatges etc...
Not always easy to meet such requirements. They tend to bring down all the functional performance of the smps transformer and increase the required size....:xeye:

Please be aware that the attached links do not enable us to design according the relevant standards.
And testing if the design is matching or not requires not only a simple voltage test, but also a very detailed analysis of the construction (creepages!!!!).
All these standards are fairly complicated and you always have to consider that depending on your topolgy the occuring voltages are much higher than expected just from the mains voltage. Ie. small flybacks for mains application deal with up to 600V and their creepages must be accordingly.

http://www.i-spec.com/IEC_60950/glossary.html
http://www.evaluationengineering.com/archive/articles/1203prodsafe.htm
http://www.conformity.com/0207safety.pdf

I know that we often use rail voltages for our amps above the SELV-level, but at least for the isolation against the mains power plug I always want to make sure the right creepages etc....!

OK. I will stop now. Otherwise the moderator may close this thread....

Bye
Markus
 
@George:
THX for adding the schematic...
360V: Does this mean that you have a PFC boost converter in front of that circuit?
Simply rectified 230V mains with 4 diode bridge and cap would result in 320V. Or do you have mains around 250V?


Take care for safe isolation!!!

Cheers Markus
 
HI ChocoHolic
Thanks for you to reply,the circuit is with out PFC,is only my home electricity is unstable 220v~250v+ ,so I get the maximum voltage ,this circuit idea I get from yamaha EMX-5000 power mixer,that have 2cmx3cmx1cm x2 heat sink with out any fan and can push up to 500Wx2 with analog amplifierand very cool on heatsink I cannot get the original schmatic .on original circuit is with out lm311,all using transistor,
 
Dear George,
...great, your circuit might give me some inspiration....
I would like to have a power factor correction.
The simplest single stage designs with mains separation, output voltage regulation and PFC would be flybacks.
But I have doubts that I will manage a flyback for 1kW ;) :hot: :mad:

So I am thinking about a two stage design.
Boost converter for PFC and behind that "xxxx" (?) for mains separation and voltage regulation.....


Bye
Markus
 
Hi George!

I recently built a very similar PS:
http://www.hszk.bme.hu/~sp215/elektro/SMPS_IR2155stab.gif

I think your voltage controll works inversely! At your design with increasing voltage the freq gets lower, and this increase output voltage further!

If you don't put an inductor in series with primer, then you can count on leakage inductance, but high leakage transformators produces very much heat. I had to learn this on experimenting with this circuit.

Markus!

Thanks for the links! I will check it out soon!

I'm affraid that my design would not satisfy you. Minimal crepage distance is only 100mil on PCB, and about 1mm in transformator. The isolation between pri and sec is 2 layers of 3M isolation tape, but et the edges I couldn't apply it properly. I trust the isolation of wires, and I pray for dry environment! :cannotbe:

I agree: some kW at fly-back is not a nice thing! A 450W boost PFC is under testing by my friend with MC33262. I hope it will work well, and then maybe I'll make a bigger version.
 
More SMPS stuff :

This is an offline prototype that produces 5 to 15V at 72A max. The primary is driven from a full bridge of bipolar transistors [classic proportional base drive self-starting topology] and synchronous rectification is applied to the secondaries. It's intended to power car-audio exhibitions

I have succesfully field-tested it for 2 days working 8 hours/day at 60A average output and 25 to 30ºC ambient temperature. Efficiency is around 95% I think. The bipolar full bridge heatsink [4x MJE13009] and the buck inductor hardly require forced cooling while the transformer requires some forced air and the MOSFETs heatsinks [2x IRFZ48V] require a lot of air flow. It uses only a TL494 IC and some CMOS gates, without SMD or expensive or hard to find ICs. Oscillator frequency is 72Khz [36Khz as seen by the transformer]. Well, actually I use 4x 10uF 10V ceramic SMD capacitors in the base drive circuits of the bipolars since they are not available in leaded format and common electrolytics have high ESR and age very quickly due to the high ripple currents they are required to handle on that circuit

These are the most recent images I have, but I have changed things since then, like adding a heat sink to the mains diode bridge or increasing Ib of turn-off transistors to improve reliability and increasing the delay between syncronous rectification turn-off and full bridge turn-on [my bipolars turn on faster than my MOSFETs turn off :bigeyes: ] :

http://eva.eslamejor.com/72a_0.jpg
http://eva.eslamejor.com/72a_1.jpg
http://eva.eslamejor.com/72a_2.jpg
http://eva.eslamejor.com/72a_3.jpg
http://eva.eslamejor.com/72a_4.jpg
http://eva.eslamejor.com/72a_5.jpg
http://eva.eslamejor.com/72a_6.jpg
http://eva.eslamejor.com/72a_7.jpg
http://eva.eslamejor.com/72a_8.jpg
http://eva.eslamejor.com/72a_9.jpg
http://eva.eslamejor.com/72a_10.jpg
http://eva.eslamejor.com/72a_11.jpg
http://eva.eslamejor.com/72a_12.jpg
http://eva.eslamejor.com/72a_13.jpg
http://eva.eslamejor.com/72a_14.jpg
http://eva.eslamejor.com/72a_15.jpg


Some older images :

http://eva.eslamejor.com/1kw_osc.jpg
http://eva.eslamejor.com/1kw_trf.jpg
http://eva.eslamejor.com/72a_ctrl.jpg
http://eva.eslamejor.com/BIP_CTRL.gif
http://eva.eslamejor.com/syn_rect.jpg


These are images from a very early prototype, it all started from an old AT power supply... [Just for the curious : This big wooden box used as a table is actually a folded bass horn part of a 3 way all-horn system]

http://eva.eslamejor.com/proto60a.jpg
http://eva.eslamejor.com/proto60b.jpg


Some old switching waveforms [I have speed it up a little since then, these have been already published some time ago and explained in the forum, search for MJE13009...] :

http://eva.eslamejor.com/MJE13009.gif
http://eva.eslamejor.com/MJE_ON.gif
http://eva.eslamejor.com/MJE_OFF.gif

The full bridge and the control circuit are also useful for audio amplifier SMPS but until now I've been too lazy to draw full detailed schematics. The MJE13009s are rated at 12A but only used up to 6A in that design, so power output from the full bridge may be increased up to 1.5KW or even 1.8KW

Note also that bipolar transistors turn off faster at higher loads and conduction losses tend to be proportional to I instead of I^2 so this means that switching and conduction losses increase only slightly with the load [MOSFETS suffer for I^2*R conduction losses and slower switching for higher loads]

Mains is filtered using two 470uF 400V capacitors and it appears to be much less than enough since when testing at full power [1KW], average rectified DC drops from 320V to about 260V. I'm planning to add one or two more 470uF capacitors
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.