Guys..i've build new diy amplifier.Just make a simple diy used 2x100watt class D 4 ohm...the original amplifier sound good for me..but,i want improve the amplifier with make clarity and sweet vocal...so,what should i do for the next step? what function if put small uF capacitor at the amplifier like 2.2uF or else? where to put it?
Uploaded with ImageShack.us
An externally hosted image should be here but it was not working when we last tested it.
Uploaded with ImageShack.us
Do you think putting a 2.2uf capacitor will improve the sound?
and where will you put this "magic" 2.2uf cap?
if the design is not so good, it will never sound good!!! No matter what cap you put!
Also did you try using another speakers?
and where will you put this "magic" 2.2uf cap?
if the design is not so good, it will never sound good!!! No matter what cap you put!
Also did you try using another speakers?
These amps are pretty mediocre at best. If you want anything with a bit of audio quality, look at the Hypex UcD and NCore stuff, or the Abletec modules at profusionplc.com
Technically you get better value for money with the Abletec modules as you get two channels + a psu for basically the same money as a single channel UcD/NCore module with no psu.
Technically you get better value for money with the Abletec modules as you get two channels + a psu for basically the same money as a single channel UcD/NCore module with no psu.
Just follow Trevor Marshall's advice: Trevor Marshall - Class D Audio Amplifier Design - TDA7498 Output filters
Last edited:
Tekko - What's the major disadvantage of pre-filter feedback over post-filter feedback? Link? Is this a subjective claim or can it be measured and why should I care? 🙂 What are some class D amp modules beside NCore that offer a post filter feedback design. Abletec?
major disadvantage is that the frequency cutoff varies with load. this, is taken from a behringer nu6000, which is pre-filter feedback class d amp:
the cutoff and slope change with load impedence.
some design even have a bump in the frequency response at high frequency, making the amp sounding harsh, bright, ear piercing, noisy, sharp and everything else associated with a non flat high frequency response.
post filter feedback keep the freq response flat across the range in all (most) load impedance. there's measured response over in avsforum.
to avoid this problem, the cutoff could be made high enough so that it doesn't affect the 20-20khz range, but that only leaves the amp susceptible to carrier frequency.
most post-filter feedback design that i've seen have variable switching frequency, and most pre-filter feedback have a fixed frequency. no problem for most people, but variable switching frequency bodes ill if you intend to bridge them for high power application. at least in my experience.
pre filter feedback is all good if you have a fixed, single application for them and don't load them with vastly varying impedance.
Pre filter feedback self oscillating varies as much in switching frequency as post filter feedback self oscillating, if not more so.
Only clocked/synced(frequency stabilized) class d has a constant switching frequency.
Only clocked/synced(frequency stabilized) class d has a constant switching frequency.
i believe judging from the designs i see posted and discussed in this section, those that have pre-filter feedback are generally non-self oscillating and those that have post-filter feedback are self-oscillating.
do you have any design of a post-filter fixed frequency to share?
Is that why many of these class D modules are specifically designed for a certain speaker's impedance?major disadvantage is that the frequency cutoff varies with load. this, is taken from a behringer nu6000, which is pre-filter feedback class d amp:
the cutoff and slope change with load impedence.
some design even have a bump in the frequency response at high frequency, making the amp sounding harsh, bright, ear piercing, noisy, sharp and everything else associated with a non flat high frequency response.
post filter feedback keep the freq response flat across the range in all (most) load impedance. there's measured response over in avsforum.
to avoid this problem, the cutoff could be made high enough so that it doesn't affect the 20-20khz range, but that only leaves the amp susceptible to carrier frequency.
most post-filter feedback design that i've seen have variable switching frequency, and most pre-filter feedback have a fixed frequency. no problem for most people, but variable switching frequency bodes ill if you intend to bridge them for high power application. at least in my experience.
pre filter feedback is all good if you have a fixed, single application for them and don't load them with vastly varying impedance.
Yes! Pre filter feedback can only be used with one specific speaker impedance without redesigning the output filter.
Post filter feedback is not a question of self oscillating or clocked.
One can easily design any of the permutations, even all permutations with nested/staggered feedback structures.
Self oscillating typically needs less components ==> simpler to build (not necessarily simpler to design) ==> lower BOM costs, lower manufacturing cost ==> better selling margin in mass market.
That's the key driver for popularity.
A thread with a clocked design and nested feedback loops:
http://www.diyaudio.com/forums/class-d/166961-new-dxa-400-class-d-amplifier-info.html
My thread of a clocked design with pure postfilter feedback.
http://www.diyaudio.com/forums/class-d/221498-system_d_md-class-d-like-chocolate.html
AP2's MXD is actually Pre Filter Feedback self oscillating but with a post filter feedback for Audio Frequency, and yes it suffers from the same (have to redesign the output filter going from 4 ohms to 8 ohms and vice versa) issue as all other class d not using post filter feedback alone for the self oscillation.
The trick to making a class d amp frequency response immune to load variations is to go UcD ala Bruno Putzeys, ie post filter feedback self oscillating.
The trick to making a class d amp frequency response immune to load variations is to go UcD ala Bruno Putzeys, ie post filter feedback self oscillating.
What if I calculated the Z of my speakers and tuned my class-d output filter accordingly?
That's what I did to my Sure tda7498 and 3-way TL speakers.
And now the little amp makes the music more beautifully than Dyna Mk 3s used to do.
That's what I did to my Sure tda7498 and 3-way TL speakers.
And now the little amp makes the music more beautifully than Dyna Mk 3s used to do.
Hi Tekko,
I think it is clocked. AP2 advertised the modulator board here:
http://www.diyaudio.com/forums/class-d/168411-high-res-pwm-module-hi-fi-amplifiers-diy.html
Or do you think that this is not the DXA modulator?
About the speed of the outer feedback we can only guess.
In the modulator thread it appears as a pure DC feedback loop.
But the scheme in the DXA thread it has the potential of a fast outer loop... AP2 never discloses the what he really does... 🙄
Do you have a DXA board on hand?
I think it is clocked. AP2 advertised the modulator board here:
http://www.diyaudio.com/forums/class-d/168411-high-res-pwm-module-hi-fi-amplifiers-diy.html
Or do you think that this is not the DXA modulator?
About the speed of the outer feedback we can only guess.
In the modulator thread it appears as a pure DC feedback loop.
But the scheme in the DXA thread it has the potential of a fast outer loop... AP2 never discloses the what he really does... 🙄
Do you have a DXA board on hand?
The trick? UcD is just one possibility of many (but definitely a clever & good one).
And it will do the trick over the entire audio bandwidth only if parametrized right. If someone just puts something together somehow in UcD structure - it does not necessarily provide the correct behavior.
The trick is the control theory behind, which can be applied to clocked and selfoscillating and it is not patented. ...but hides itself by its ugly nastyness, so nobody wants to look at. 😀
Last edited:
Member
Joined 2009
Paid Member
- Status
- Not open for further replies.