Leco smps some ugly tests...

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

DjLeco

Member
Joined 2006
I decided to see, how many watts can carry IGBT's...

Same setup, dropping dummy load from 3 ohms till 1,5 ohms.

Choose 720p (HD) at youtube.


Leco SMPS passing 5 Kw barrier.MOV - YouTube

I saw peak of 27Amps on clampmeter , 25-26Amps sustained after 2 secs of max load , possible not seen in videoclip...

At start to load with 1,5 ohms on output I measured 109,23mV on shunt, corresponding to 54,61Amperes through load, output tension (on load) was 95,3V.

In that case, power was 54,61Amps x 95,3Volts = 5204,33Watts.


Some interesant pictures!

30Amps main fuse before use (new one):

http://postimage.org/image/ywmgvv0qj

30Amps main fuse before after test, compaired with another new one:


http://postimage.org/image/eqsm7u0sl/

Front shunt:

http://postimage.org/image/mv09t769n/

Back shunt, observing 2 x 3mm resistive material.

http://postimage.org/image/hh62ze5wd/
 
Last edited:
D

DjLeco

Member
Joined 2006
Trafo was winded just for tests, also litz used (600x0,071) cannot carry that amount of amperes (54 amps).

Next week, I will wind trafo for desidered use (+/- 95Vdc, Class AB, regulated), or +/- 180Vdc, class D, I or TD, unregulated, and will see.

So 5200Watts obtained from a half bridge, at 62,5 Khz, also for a small ETD 59 core, I consider to be big amount of power, also alot of headroom, for 3500 watts demanded power, no?:D

I tried also to put 1 ohm on load :eek: (for 5200 watts was around 1,5 ohm load, excluding thermal derive of resistors), but IGBT, cannot resist :flame: ...

What's happened, you can watch here, also U will understand why I have uset that song before testing, knowing myself about final act... :p

Before watch mine movie, please watch fisrt 30 seconds from next clip

http://www.youtube.com/watch?v=hx4DdoSgWJ0


Mine movie, when I lowered dummy load till 1 ohm...

http://www.youtube.com/watch?v=d4H1Zi_CWeM&feature=youtu.be


IGBT, fallen on duty:

http://postimage.org/image/astwplz01/

Who liked that song, grab from here:

http://www.2shared.com/file/EWHQ0HYC/Savage_-_Radio__12_inch_versio.html
 
Last edited:
C

ChocoHolic

Member
Joined 2003
Bad boy!
DjLeco said:
IGBT, fallen on duty:
Click to expand...
Definitely the nicest description, available :p

Which average or continuous output power do you target in relation to max power? ==> Thermal design.
Or do you design for certain load/no load duty cycle?
Which is the worst condition that you expect - i.e. for dubstep?

Example:
Let's assume a PSU for a class D amp with 1kW.
Further assuming the class D amp may run with an efficacy of 90%.
If we would like to reproduce a continuous sine wave, then the PSU would need to have the following capabilities:
Average continuous power: 1.11kW
Instant power: 2.22kW (factor two from the ideal ratio of max instant power vs average power of a sine wave, assuming 20Hz the required times would be already in the tens of ms.)
There is no reasonable way to avoid the short term demand of factor two, except we would place really unsaint output caps. So we have to do an electrical design that can handle 2.22kW. But for the thermal design, I am convinced that massively lower requirements can be applied.
A PSU for an 1kW amp does not need to be able to deliver continuous 1.11kW, because music is not a continuous sine wave. So the thermal design does not need to allow continuous operation at 1.11kW, or even worse 2.22kW ON/OFF 25ms/25ms.
My question is for which average load or may be which load duty cycle do we need to do the thermal design?
 
D

DjLeco

Member
Joined 2006
Man,the heatsink will be much bigger, the one I used on test, measures 10 cm, final heatsink will have 30 Cm, exact one that I have used at mine modules.

No problem of heat, because smps will be progressive controlled fan, cooled, depending on temperature of trafo and main heatsink.

I tested alot of amplifiers with SMPS, wich loaded 1/2 like I loaded mine, go into protection mode...

Others dropp alot, the only one SMPS that stays ROCK SOLID, was the one of I-tech series Crown Amplifiers.

In rest, many dissapointings...

I'm very satisfied with mine results.

I'm intending to use one smps per one 3000Watts TD or D class amplifier, wich I will develop soon, or for 2 or 4 AB class modules of 800Watts/4 ohm.

Must take in consideration, the music signal, of any type, including dubstep or reggae, will not load smps that hard , as I have tested.

Tomorrow I will replace faulty IGBT's, to adjust protection (overcurent) and thermal protection too.

Keep close
 
C

ChocoHolic

Member
Joined 2003
DjLeco said:
No problem of heat, because smps will be progressive controlled fan, cooled, depending on temperature of trafo and main heatsink.
Click to expand...
That's likely to be pretty helpful.
On the other hand it sounds to me that you simply make it tougher than required in order to stay trouble free.
I also tend to settle my private designs more on the rugged side, but nevertheless I was curious about numbers to hit the point which is really needed.
Of course there are the standard rules like for PA (1/3) or HiFi (1/8),
but I was interested in your view.


DjLeco said:
In rest, many dissapointings....
DjLeco said:
No wonder. In commercial projects the engineers are forced to cut costs by design until the customers start to complain.
As long as nobody complains, the engineer will be blamed for expensive overdesign.
The art of business R&D is to know, where you can cut - and where you better do not cut...

...I am a bloody Idiot, starting to apply business considerations in my hobby, while I feel that one of the main relaxing factors of a hobby is: No business considerations.
Click to expand...
Click to expand...
 
darkfenriz

darkfenriz

Member
Joined 2004
DjLeco said:
Man, it was just a simple test; final one, I think will be full bridge for 1/2 curent on primary, and less Ic on output switching devices.
In that test, I was limmited from small heatsink of rectifier diodes, and resonant capacitor curent cappability...

I have very low loss on switching devices, gratefully almost perfect drive of it, very low rise and fall times.

Look here:

Proper IGBT driving, on mine SMPS (100Khz half bridhe resonant soft switching mode)


Observe ULTRAFAST switching, very low rise and fall timmings.

View image: DSCN5197

Tomorrob (today better said) I will try full bridge tipology, at 62,5 , 100 , and 125 Khz.

Mine IGBT;s are WARP ones (150Khz hard switching, 300 Khz soft switching).

Not interested to use mosfets, due internal body diode, also I don;t want to complicate output topology with schottky and antiparralel diodes too.

All modern SMPS, uses IGBT, or combination of mixed IGBT with mosfet too, also not interested...

Simple one, is most time, the efficient one!

If I lose under 100 Watts per total SMPS at 4000W total load, randament is HUGE, also, musical power is not more than 1/2 from a pure sin wave, or dummy load as I tested.

Next time, I will use, maybe bigger power resistors for dummy load, for longer time testing, If I keept that test, mine 4 ohms resistor will melt itself...
Click to expand...

So you switch more than 300V at 10ns rise time (30kV/us) and 15ns fall time (20kV/us) ?
If really so I suggest you slow it down, efficiency is important, but so is EMI.
 
D

DjLeco

Member
Joined 2006
Finally PCB's arived!

I have mounted some components...

View image: DSCN6166

View image: DSCN6167

View image: DSCN6172

View image: Aux Supply OK

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


Output tensions:

View image: Outputs configuration

Testing signal on IGBT's gates+delay+softstart:

Leco SMPS SoftStart+Delay+Thermal Protection - YouTube

And some final tests, loading till 4300Watts:


Leco SMPS Final Part 1 - YouTube

Leco SMPS Final Part 2 - YouTube
 
Last edited:
D

DjLeco

Member
Joined 2006
First SMPS made for one 1000W AB class amplifier, that I've sold last week to a friend, he asked me to make SMPS for that module.

Made at minimum costs, just +/- 100Vdc per power amp, one +12 volts for fan and one +10 volts for one small Mp3 module.

Calibrated for 1800Watts continuous, protection tripping at 2,1Kw

http://postimage.org/image/q0lo5ovtb/

http://postimage.org/image/vq1wq01zj/

http://postimage.org/image/xj4tebn67/


Uncalibrated protection test (without protection)


http://www.youtube.com/watch?v=Cf1VPQ2a_us&feature=youtu.be

Testing with protection enabled and calibrated:


http://www.youtube.com/watch?v=g7I2ws_x-KM&feature=youtu.be

And final test, a big short circuit


http://postimage.org/image/8uz0ncbgv/




Here is testing that amp loaded with 4 ohms, and power raised till around clipping level (almost 960 watts audio), put multimeter directly to + and - , to see voltage dropping at max power out:

Test Final Boxa - YouTube
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom