thanks alot guys . but tekko i was asking about the etd cores in pc power supply can they be used and how do you know the core number? by looking at the colour? or how so that you can feed the no,s to the soft ware for calculations .
iountzxpo . just add 5.6k resistor before the input signal cap (470n 63vlts) . it worked for me
iountzxpo . just add 5.6k resistor before the input signal cap (470n 63vlts) . it worked for me
recycling old E cores
hi Andrew Eckhardt i manage to get the epoxy apart by
1 removing the yellow tape
2 placing a 40 watt soldering gun on the e core ferrite for around 10 minutes and
3 using a piece of cloth to pull the ferrites appart from the bobbin while they are still hot
4 then unwinding the wires from the bobbin
(recycling for a greener earth
)
hi Andrew Eckhardt i manage to get the epoxy apart by
1 removing the yellow tape
2 placing a 40 watt soldering gun on the e core ferrite for around 10 minutes and
3 using a piece of cloth to pull the ferrites appart from the bobbin while they are still hot
4 then unwinding the wires from the bobbin
(recycling for a greener earth
)Don't heat them with a soldering iron, the thermal chock can crack them. Boil them in water instead.
The E cores in a pc psu is typically 3F3, 3C11 or similar. They are usable if the center leg have an air gap. Do NOT gap the outer legs or your asking for EMI galore.
However these cores are quite oversize for a 100 watt amp, more suitable for amps in the kilowatt range. For a 100W amp, the core from say a laptop psu would be more suitable.
The E cores in a pc psu is typically 3F3, 3C11 or similar. They are usable if the center leg have an air gap. Do NOT gap the outer legs or your asking for EMI galore.
However these cores are quite oversize for a 100 watt amp, more suitable for amps in the kilowatt range. For a 100W amp, the core from say a laptop psu would be more suitable.
hi tekko many thanks and the almighty GOD bless you very much 
the links and the softwarez u showed . have answered my many inductance questions .thanks even for pcb link, but not sure if they can ship to east Africa . i've seen the t106 26 are not usable in class d output filter cause of heat in the simulator.
can they be used as trasfo gate drivers 14 : 12:12:12:12: in full bridge smps?
and which software calculates inductance for E ferrite cores like the laptop one you are talking about
300µH, eww huge phase shift, not to mention DC resistance and resonance frequency in the audio band. You want the phase to barely start shifting at 15-20kHz and smallest possible DCR, ie the thickest wire that will fit with the required number of turns for the chosen unductance.
With 300µH, the phase will have turned 180deg way before 20kHz which will affect sound negatively, and the resonance peak will end up within the audio band.
I generally go by an inductance no higher than 30µH, preferably less than 20µH.
I think a volt of switching resudial on the output is ok. As long as its not common mode which is what radiates off speaker cables.
300µH may be ok for a subwoofer amplifier, i have a subamp that have 80µH and a few µF in the output filter.
With 300µH, the phase will have turned 180deg way before 20kHz which will affect sound negatively, and the resonance peak will end up within the audio band.
I generally go by an inductance no higher than 30µH, preferably less than 20µH.
I think a volt of switching resudial on the output is ok. As long as its not common mode which is what radiates off speaker cables.
300µH may be ok for a subwoofer amplifier, i have a subamp that have 80µH and a few µF in the output filter.
Sorry, I have incorrectly marked it as 100p. use the original value 330p. it is the feed back capacitor which filters high frequencies coming from the output. So it is true that reducing the value might increase the noise.
and your amp looks better than mine.. well assembled !
good luck with it! let us know how good is once you plug.
regards,
Lycanlk
thanks tekko you,ve saved me a lot in terms of blowing up fets i,ll have to stick with the ir2110 or the small e cores for smps gate drivers
BTW i took your advice and coiled a small E core like etd 29 (size is 33mm x 29mm x 13mm ,square at the center ,flat top not e shaped ) which i found in pc smps and used one for both channels .
i used 8 strands of 0.5mm wire and length 24" for which came to 8 turns . the other channel the same, i coiled in the same core but opposite direction for the other channel .
i played the amp in bridge mode for about an hour an the e core did not heat or warm .thanks again
the amp is working fine in bridge mode (one channel being fed from inverse hin & lin of the other channel, sound is clear, loud with small ignorable noise at very low volumes) just wondering should i continue to use it with one channel not having feedback network, or where should i place my feedback of the other channel ? stewin
BTW i took your advice and coiled a small E core like etd 29 (size is 33mm x 29mm x 13mm ,square at the center ,flat top not e shaped ) which i found in pc smps and used one for both channels .
i used 8 strands of 0.5mm wire and length 24" for which came to 8 turns . the other channel the same, i coiled in the same core but opposite direction for the other channel .
i played the amp in bridge mode for about an hour an the e core did not heat or warm .thanks again
the amp is working fine in bridge mode (one channel being fed from inverse hin & lin of the other channel, sound is clear, loud with small ignorable noise at very low volumes) just wondering should i continue to use it with one channel not having feedback network, or where should i place my feedback of the other channel ?
stewinYou can make feedback by using an opamp configured as a differential amplifier like baldin did in his soda class d fullbridge amp: http://www.sensibleaudio.dk/HIP4080_SODA_Class-D/HIP4080 SODA Class-D.pdf
Hope this clears some things up. If you include the output filter in the feedback you csan get rid of most of the distortion it adds, and the amp gets alittle more linear too.
Hope this clears some things up. If you include the output filter in the feedback you csan get rid of most of the distortion it adds, and the amp gets alittle more linear too.
Good job on the ferrite salvage. I'll have to keep some sort of direct heating plan in mind next time I wish to do that. All I thought of is an oven set very high which would have just made a mess.
Pulse transformers capable of driving a gate directly in a PWM amp need high winding inductance with a small size/ minimum of turns/ wire length. Iron powder cores basically wont work. Higher permeability ferrite like Magnetics W or J material works okay. You have to keep the DC and LF out of it. Overall a transformer gate drive circuit with no problems is not so simple.
Pulse transformers capable of driving a gate directly in a PWM amp need high winding inductance with a small size/ minimum of turns/ wire length. Iron powder cores basically wont work. Higher permeability ferrite like Magnetics W or J material works okay. You have to keep the DC and LF out of it. Overall a transformer gate drive circuit with no problems is not so simple.