ucd 25 watts to 1200 watts using 2 mosfets

[lorylaci]

Originally Posted by Dzony988

TL074 is quad op amp low noise (4 x TL071 or 2x TL072) I dont have problems with noise with TL074.

Do you know any quad op amp like TL074 with pins compatibile to it, low noise ofcourse?
Thanks

TL084 has 20% less noise than TL074. But the differences are totally inmeasureable with normal instruments. High-tech stuff (and even high-tech room) is needed to measure that.
So don't blame opamps for noise, there are lots of other stuff.

CPX could you make a photo of your test setup? That could help, maybe some of us could spot a bug. (everytime there something doesnt work, I always wish to find something that I screwed up...)
If you page back, you can see my layout, which I must confess not the best, but I had low noise and low distortion.

[CPX]

Hm..strage..increasing gain at the buffer increases distorsions at the same load in amps!!!!!!!
In my head that means noise from switching stage gets somehow into buffer input....,but since i have 1 nF on positive input i thought that the negative feedback is where the noise comes in..so i put a 1 nF capacitor between buffer output and in- and it increased distorsions!!!!!!!!...i do not understand what is heapening...there is something wrong at the input.feedback stage...but what???



no snubbers at 4A load..huge distorsions...


100pF,4.7 ohm subber at 4A load


1nF,4.7 ohm snubber at 4A load


setup:

[lorylaci]

Originally Posted by CPX

Hm..strage..increasing gain at the buffer increases distorsions at the same load in ampers!!!!!!!
In my head that means noise from switching stage gets somehow into buffer input....,but since i have 1 nF on positive input i thought that the negative feedback is where the noise comes in..so i put a 1 nF capacitor between buffer output and in- and it increased distorsions!!!!!!!!...i do not understand what is heapening...there is something wrong at the input.feedback stage...but what???



no snubbers at 4A load..huge distorsions...


100pF,4.7 ohm subber at 4A load


1nF,4.7 ohm snubber at 4A load


setup:

What type of inductor are you using? It can hardly be seen.

The green thing is DIY photo-resist resin? Some of those are measureably conduct at high frequency. (and some them even arc at high voltage transients)

I got back to page 144, to see your layout, and from that the feedback path, and the comparator connenction seems totally different from the original schematic. Could you draw the exact schematic, of just that part?

[CPX]

I am using micrometals t106-2 core wound up to 30Mhenry multilayer..i wound one 20uH one layer but i dint change distorsions..the lower inductance increased 100 hz noise..
The green thing is permanet marker,the board is made with PNP.
The board was made for for ionut_gaga simetric feedback scheme..but i am using this now:

[lorylaci]

Originally Posted by CPX

I am using micrometals t106-2 core wound up to 30Mhenry multilayer..i wound one 20uH one layer but i dint change distorsions..the lower inductance increased 100 hz noise..
The green thing is permanet marker,the board is made with PNP.
The board was made for for ionut_gaga simetric feedback scheme..but i am using this now:

So the feedback seems fine.
It is interesting, because I never had 100Hz (50Hz) noise with so this is strange. The opamp before the comparator works as a buffer and for protections?
Have you tried feeding input, after this buffer stage? So as in original schematic. (into the opamp through with DC blocking capacitor, input resistor, input RF stoping resistor and capacitor)

[CPX]

Buffer op amp was also for under voltage lockout but i made a mistake..that will be corrected in new board..

I connected input after buffer and changed feedback shunt resistor to 330 ohm,hiss is huge now(60x gain) but 50 hz distorsions have decreased alot... still not perfect but a big improvement...thanks for the ideea...i do not know why the buffer distorts...

I am going to bed now..to tired...beautifull dreams for you too!!!

[nigelwright7557]

Originally Posted by CPX

Ok..i connected input after buffer and changed feedback shunt resistor to 330 ohm,hiss is huge now(60x gain) but 50 hz distorsions have decreased alot... still not perfect but a big improvement...thanks for the ideea...i do not know why the buffer distorts...
I am going to bed now..to tired...beautifull dreams for you too!!!

Are you setting the gain of the input buffer to match 0db input ? If not you are going to get a lot of distortion. You dont need a gain of 60. A gain of 10 is more than sufficient. I designed a similar amp.

[CPX]

No..no..that gain was at the comparator only for testing without buffer...

[nigelwright7557]

Originally Posted by CPX

No..no..that gain was at the comparator only for testing without buffer...

I didnt use any gain at the comparator on my amp. I didnt use positive feedback (hysteresis) either, it worked fine as a simple comparator.

[CPX]

Of course there is gain at the comparator...there is 28x in lorylaci's schematic and around 40x in the original schematic..to have no gain you need R10(my schematic) to be infinity..and in that case you are limited with output peak voltage to op voltage-1 ..so around 2 volts in the original schematic..