UCD 25 watts to 1200 watts using 2 mosfets

[grizlimedo]

People who made UCD here please post your THD measurement using PC sound card and programs like SpectraLAB, or SpectraPLUS which you can find on rapidshare

 

[grizlimedo]

People who made UCD here please post your THD measurement using PC sound card and programs like SpectraLAB, or SpectraPLUS which you can find on rapidshare

I did measure using Audio Presision, but all class D manufacturers suggest that AP requires separate low pass at least 5-tj order filter to reduce AP smapling frequency to mix with pwm frequency.
I measured without that filter and got about 0.1-0,2 % THD from 20kHz to 20kHz for 10 and 50W using +/-50V power supply and 8 ohm resistnace...

I wonder why hypex modules hav much lower THD at such powers ? maybe that low pass filter will make THD value go down ? Also CAKEAMP has about 0,05% thd at the same powers, but author measured using PC sound card, hmm

ALSO what is most weird and important, Theory about class D half bridge claims that ODD harmonics levesa are HIGHER than even. I got opposite resulty using AP and ucd, even are higher than odd which is quite good, cause human ear is much more sentitive to odd harmonics

 

[NMOS]

I short input to ground noise is less but still noticeable. And I use shielded input cable.So cable is not a problem.

Have 1 NF parallel with 22K but there is noise in my amp....

Is it better solution to replace 1 NF with 4.7 parallel with 100pf

whats best and effective solution to remove noise ?

 

[grizlimedo]

Have 1 NF parallel with 22K but there is noise in my amp....

Is it better solution to replace 1 NF with 4.7 parallel with 100pf

whats best and effective solution to remove noise ?

post schematic of feedback

 

[NMOS]

post schematic of feedback

Here is the schematic

amp is noisy like old cassette deck without Dolby B, does somebody have solution to solve this problem ?

and the good things: amp working fine now with amazing sound, have add 2 m copper wire on toroid ferrite, after this coil is cool and problems gone

 

[grizlimedo]

Here is the schematic

amp is noisy like old cassette deck without Dolby B, does somebody have solution to solve this problem ?

and the good things: amp working fine now with amazing sound, have add 2 m copper wire on toroid ferrite, after this coil is cool and problems gone

higher the amplification, higher the noise (hum old cassete) cause comparator has lower voltage difference

i use opa 134 4 times * UCD 9 times amplification

use spectraPlus and measure thd

 

[NMOS]

higher the amplification, higher the noise (hum old cassete) cause comparator has lower voltage difference

i use opa 134 4 times * UCD 9 times amplification

use spectraPlus and measure thd

ok thanks for your info, can you share schematic ?

but my question is: its better solution to replace 1 NF with 4.7 parallel with 100pf

whats best and effective solution to remove noise ?

or can it comes from the output filter... to much copper wire ???

 

[Dzony988]

On this basic schematic, on input is 47K to gnd,not 22K. I use 47K, 100p and 4.7nF with same feedback and i dont have noice...

 

[NMOS]

On this basic schematic, on input is 47K to gnd,not 22K. I use 47K, 100p and 4.7nF with same feedback and i dont have noice...

have changed 47K on input and 100p + 4.7 nf


ok have changed devices, i have no noice now !

but all clear highs and mids are gone,...

after change devices it sounds like mid and highs on equalizer -12 db

sound is poor is this normally ???

For fullrange cant use amp now ....but no noise

what happened ???

 

[NMOS]

have removed 4,7 nf and amplifier sounds great like before...sweet clear mid and highs

only 100 pf in input

but very noise....what can we do to remove noise and get sweet clear mid and highs ?

 

[Pafi]

What is the signal source?

Where exactly is the 4.7 nF?

What if you don't connect input source?

Do you have oscilloscope?

 

[Dzony988]

Try 100p and 1nF in paralell and try to reduce your voltage gain (now is 39X), reduce to 28X (R3 10k change with 1K).

Yes, like pafi says, what happens when input connect short to gnd without signal source (with 100p and 1nF)

 

[grizlimedo]

have removed 4,7 nf and amplifier sounds great like before...sweet clear mid and highs

only 100 pf in input

but very noise....what can we do to remove noise and get sweet clear mid and highs ?

try to reducee amplification

Also youcant "pick" value, use LTspice AC analysis and see what transfer function you get with certain values

Again I repeat reduce voltage gain to 10 maximum,

 

[NMOS]

Try 100p and 1nF in paralell and try to reduce your voltage gain (now is 39X), reduce to 28X (R3 10k change with 1K).

Yes, like pafi says, what happens when input connect short to gnd without signal source (with 100p and 1nF)

when short to gnd = no noise

 

[NMOS]

ok have reduce gain with 1 k

noise is a little less but not acceptable

have no 1 nf here will test again after add 1 nf on wednesday

 

[grizlimedo]

ok have reduce gain with 1 k

noise is a little less but not acceptable

have no 1 nf here will test again after add 1 nf on wednesday

what is your new gain now when you reduced "with 1k" what is new ratio?

 

[NMOS]

what is your new gain now when you reduced "with 1k" what is new ratio?

Dont know, but the difference between 10k and 1 k in noise is small.

I think it isnt worth to add 1 NF, ok with 4,7 nf noise is really goone but the good sound too...

any solutions to kill noise and save the good sound

 

[Pafi]

any solutions to kill noise and save the good sound

I have only one solution, but it's general: learn electronics and use instruments! (And your brain.)

I think the first question should be: "where does the noise come from", and NOT "how can I kill it".

I had some questions.

Yes, like pafi says, what happens when input connect short to gnd without signal source

Unfortunately I didn't say shorting input.

 

[Dzony988]

@ nmos
I dont know what heppens with your desing but with my with 4.7n on input music is great.

@pafi

This question is full theory, but if he can understand some problematic and source of problem, can you directly help to nmos and tell him how to reduce noise in this exampe, for this amp. That example + that theory that you said can help to him in his problem and in the future examples when he has same problem.

 

[Pafi]

Dzony988!

I have very limited free time, so I can't promise anything. And I absolutely don't want to tell him how to reduce noise in this particular case, partly because I don't know this particular case. I have solved many similar problem, but almost all of them were different, so searching for a simple universal solution is stupid thing. But there are more or less systematic methods to investigate this kind of problems.

f he can understand some problematic and source of problem,

If he understood the source of the problem, he wouldn't need advice any more.

PS:

I dont know what heppens with your desing but with my with 4.7n on input music is great.

Reduced highs implies high source impedance. Partly this because I asked what the signal source is.