UCD 25 watts to 1200 watts using 2 mosfets

[manojtm]

Hi Dimonis

increase gate resistor with correct value , decrease temp.
normally we are using osc=300 KHz for good THD Better for 500 KHz

I use osc=125 KHz it is not BAD and filter 4th order one

RGDS
Manoj

 

[don2007ksa]

Hi manojtm about your UCD 625,your oscillator there is about 102 khz square wave,what if i'm going to use around 500 khz triangular wave? are there any part's to be changed? is it possible for me to use this amp for full range @ 4 ohm's load,thank's.

 

[stewin]

nice work cool video .BTW the fets in your video are they irfp250 2pairs?

 

[norazmi]

Yep Stewin, irfp250n, they run happily with it

 

[manojtm]

hi don2007ksa

don't go for 500Kz , thd+n will increase better for 250 or 300khz

 

[manojtm]

don2007ksa

use cd4060 with 8MHz crystal

 

[Dimonis]

increase gate resistor with correct value , decrease temp.

-increases the switch times of the mosfets(rise & fall) - increases the deadtime - increases the THD .
For the subamp application it'll do , but for the full range......

I use osc=125 KHz it is not BAD and filter 4th order one

Yea , at that freq you can bravely use 2-3 pairs of mosfets (4227)

normally we are using osc=300 KHz for good THD Better for 500 KHz

What driver & mosfet?
And about THD - how much better ? maybe you'll post some ?

 

[Dimonis]

Hi CPX

I am using RightMark Audio Analyzer 6.2.3(free version) and Acoustic Analyzing System 5E(THD only), they give similar results.

My soundcard (built-in on motherboard Intel DP35) has such results (measured line-in-line-out) . I think they are not good enough for measurement?
And my netbook has awful noise ~56dB and IMD ~0.5%
Can't find Acoustic Analyzing System 5E , can you share it with me?

 

[Dimonis]

Hi don2007ksa

Dimonis i have 60 hertz,+/- 90 VOLTS @ 800 watts power transformer,i'm going to used a 4 ohm's load around 700watt's full range speaker,thank's.

I think the best solution is the amp from this thread
http://www.diyaudio.com/forums/atta...tts-1200-watts-using-2-mosfets-ir2110_ucd.zip

 

[manojtm]

I think the best solution is the amp from this thread
http://www.diyaudio.com/forums/attac...ir2110_ucd.zip

Hi dimonis

Self osc. UCD is not good for Higher WATT (above 200W).It will create lot of problems
ie. Heat,Noise & high Start up Current

 

[CPX]

Hi CPX
My soundcard (built-in on motherboard Intel DP35) has such results (measured line-in-line-out) . I think they are not good enough for measurement?
And my netbook has awful noise ~56dB and IMD ~0.5%
Can't find Acoustic Analyzing System 5E , can you share it with me?

Yes..your results are pretty bad.Try decreasing output level a little(2,3 db) even if the software says volume to low in calibration windows .If the results are the same you can still test your amplifier by using substraction.

 

[CPX]

Hi dimonis
Self osc. UCD is not good for Higher WATT (above 200W).It will create lot of problems
ie. Heat,Noise & high Start up Current

Why do you say that?A bad design will generate these problems not self osc ucd topology.

 

[Ouroboros]

Why? The only reason you will get all those things is because of bad schematic or pcb design. Using an IRS20957 instead of a 2110 will get rid of those nasty level shifters, and allow a proper high-speed comparator running off +/-5v supplies. I have designed bridged self-oscillating amps, with fully-balanced input and modulator stages at significantly higher outputs than 200W.

 

[Ouroboros]

BTW, my response was to the rather foolish statement by 'manojtm', not the sensible comment by 'CPX'!

 

[manojtm]

increasing the switching frequency will reduce the THD but also decreases the efficiency
and maximum output power level before clipping. Increasing
the switching frequency increases the amount of loss because switching currents lower the efficiency across the output power range. A higher switching frequency also lowers
the maximum output power before clipping or the 1% THD
point occur.

from LM4651 & 4652 Notes .their test App. notes fsw=125KHz

 

[manojtm]

BTW, my response was to the rather foolish statement by 'manojtm', not the sensible comment by 'CPX'!

This is not a foolish!!!!!!!!!!!!!!!!!!
you just tell me the Efficiency of your AMP
output imp. change your fsw

 

[manojtm]

companies CROWN & Crest Audio they are not supporting self osc UCD for PA application

RGDS
Manoj

 

[CPX]

Output impedance doesn't change sw frequency much in mydesign(max 20 khz form no load to short).
My amp has between 10% and 90% efficiency

500 hz test:
1 ohm load: 9A 74%; 14A 81%;17A 85%
3 ohm load: 1A 41 %; 4A 77%; 8A 90%

 

[Ouroboros]

Including the fully-isolated dc-dc conversion stage to allow operation of the amplifer from a 24V dc supply (converts 24V to +/-85V together with the associated low-voltage supplies), I get over 82% efficiency at 120W out. Tested this morning at half-power on a 240W amplifier. (This is a low power amp, but it is one I'm working on at the moment).
So perhaps 94% eff for the amp stage and 87% for the isolated dc-dc stage. I'm sure that someone like Eva can do even better.
I design class-D amplifiers as my main work for a very well known international engineering corporation, so I don't expect to be lectured at by someone who thinks an LM311 is a good modulator comparator!

 

[Ouroboros]

Just in case anyone wonders why a 240W bridged amp requires +/-85V rails, that's because it has a directly-driven 100V output, so doesn't require the normal step-up transformer normally used on 100V PA amplifiers.