On fast beats, it could EMI induced, could be coil induced. I wouldn't expect it to be perfect.
Did you try driving a speaker with it? Why do you not drive it from a soundcard, then you have some volume control at least?
">>>btw, series gate resistor (this is the 10R) right??
-yes
>>>and turn on bias resistor (is this the resistor in series with the diode, Hi side supply for mosfet drive)??? or the deadtime resistor??
-hmmm, it'd be the 470 I guess
>>>balance resistor in current mirror (resistors in the emitter legs 50R ??)
-nope, not the 50 ohms, the one tied to the bases.
>>>Input stage, i asked you before if TL072 can be used
-Don't know, check the data sheet? I never used it or looked it up
>>> about values for gate drivers, 470R,330R.. are they ok at higher voltages? I mean the mosfet drivers are at 12V only.. maybe they are sufficient??
-You can try it and see what happens, I can't say if it will hold together, this is all part of the fun!
====== need your reply on this/pls point w/c part/s=====
regarding the baker clamp diodes, i have some of those schemes with dual diode.. ill go look for them.."
-Cool
I hope you ordered more than one pair of IRF540's anyway. May consider trying some samples of other mosfets, whichever you think will work good.
Regards
Chris
Did you try driving a speaker with it? Why do you not drive it from a soundcard, then you have some volume control at least?
">>>btw, series gate resistor (this is the 10R) right??
-yes
>>>and turn on bias resistor (is this the resistor in series with the diode, Hi side supply for mosfet drive)??? or the deadtime resistor??
-hmmm, it'd be the 470 I guess
>>>balance resistor in current mirror (resistors in the emitter legs 50R ??)
-nope, not the 50 ohms, the one tied to the bases.
>>>Input stage, i asked you before if TL072 can be used
-Don't know, check the data sheet? I never used it or looked it up
>>> about values for gate drivers, 470R,330R.. are they ok at higher voltages? I mean the mosfet drivers are at 12V only.. maybe they are sufficient??
-You can try it and see what happens, I can't say if it will hold together, this is all part of the fun!
====== need your reply on this/pls point w/c part/s=====
regarding the baker clamp diodes, i have some of those schemes with dual diode.. ill go look for them.."
-Cool
I hope you ordered more than one pair of IRF540's anyway. May consider trying some samples of other mosfets, whichever you think will work good.
Regards
Chris
RX5 said:ok ... now i listened very carefully.. i hear a tiny bit of distortion.. on fast "beats".. i think this is harmonics....
ill rewind a new core... ferrite....
uhmmmmm not harmonics... I swapped 300R for the deadtime... IT(distortion) disapeared...almost...BUT mosfet/s heats up a bit more.... looks like i REALLY have to wait for that IRF540....looks like a balancing act.... I lowered AMP gain (33k) feedback...
is IRFZ44/46 "slow" switchers??
what are other possible candidates for N-Ch MOsfEts I could use???
I have datasheets available for some of the MOSFETS im using... mind telling me WHAT/HOW do I "see" or WHAT will I be looking at datasheets so i could compare "switching speed" etc..etc...
@chris,
On fast beats, it could EMI induced, could be coil induced. I wouldn't expect it to be perfect.
>>> i dont know too... but It seem fixxed after changing/lowering the deadtime resisitor BUT to the expense of MOSFET heating up.. balance..
as you said
about values for gate drivers, 470R,330R.. are they ok at higher voltages? I mean the mosfet drivers are at 12V only.. maybe they are sufficient??
-You can try it and see what happens, I can't say if it will hold together, this is all part of the fun!
>>> your expecting me to =kill= another MOSFET?? hehehe yeah that SURE IS FUN NOT!
Did you try driving a speaker with it? Why do you not drive it from a soundcard, then you have some volume control at least?
>>> what do you mean?? how would I hear the output if I wont connect it to a speaker? my sound source IS a CD player with headphone output(with output volume).. that is HOW I control the sound level...
>>>BTW, I found the circuit you where mentioning a while ago.. double diode + 10K... whats with the 10K?
I hope you ordered more than one pair of IRF540's anyway. May consider trying some samples of other mosfets, whichever you think will work good.
>>>I did not order them directly(for me) .. the store ran out of it.. hehe so I just told them to order more... actually, some stores here DO carry them.. but they aint cheap... costs about P60 (1USD = P56)... thats the second cheapest i could get...I could get mine at P15<at this certain store>.. hehehe must have been an error in there pricing... BUT its an original IR.. not just some imitation....
up till now, im gonna rest for a while... tommorow, ill add circuit with the double diode + resistor in place of baker clamp diodes...
BTW, HOW do i elimintate power-on "thump" ?? any idea? just an idea....
still contented (Even if STILL not perfect)
Raff
On fast beats, it could EMI induced, could be coil induced. I wouldn't expect it to be perfect.
>>> i dont know too... but It seem fixxed after changing/lowering the deadtime resisitor BUT to the expense of MOSFET heating up.. balance..
as you saidabout values for gate drivers, 470R,330R.. are they ok at higher voltages? I mean the mosfet drivers are at 12V only.. maybe they are sufficient??
-You can try it and see what happens, I can't say if it will hold together, this is all part of the fun!
>>> your expecting me to =kill= another MOSFET?? hehehe yeah that SURE IS FUN
NOT!Did you try driving a speaker with it? Why do you not drive it from a soundcard, then you have some volume control at least?
>>> what do you mean?? how would I hear the output if I wont connect it to a speaker?
my sound source IS a CD player with headphone output(with output volume).. that is HOW I control the sound level...>>>BTW, I found the circuit you where mentioning a while ago.. double diode + 10K... whats with the 10K?
I hope you ordered more than one pair of IRF540's anyway. May consider trying some samples of other mosfets, whichever you think will work good.
>>>I did not order them directly(for me) .. the store ran out of it.. hehe so I just told them to order more... actually, some stores here DO carry them.. but they aint cheap... costs about P60 (1USD = P56)... thats the second cheapest i could get...I could get mine at P15<at this certain store>.. hehehe must have been an error in there pricing...
BUT its an original IR.. not just some imitation....up till now, im gonna rest for a while... tommorow, ill add circuit with the double diode + resistor in place of baker clamp diodes...
BTW, HOW do i elimintate power-on "thump" ?? any idea? just an idea....
still contented (Even if STILL not perfect)
Raff
can I use this coil?
hello all,
can anyone Guestimate the value of this coil?? its ferrite material.. found it on PC PS....
thanks for reply
-Raff
hello all,
can anyone Guestimate the value of this coil?? its ferrite material.. found it on PC PS....
An externally hosted image should be here but it was not working when we last tested it.
thanks for reply
-Raff
While this is an approximation, it is correct. See
http://hyperphysics.phyastr.gsu.edu/hbase/electric/indsol.html
http://hyperphysics.phyastr.gsu.edu/hbase/electric/indsol.html
Here is the correct link. (I think.)
http://hyperphysics.phy-astr.gsu.edu/ hbase/electric/indsol.html
http://hyperphysics.phy-astr.gsu.edu/ hbase/electric/indsol.html
The approximation is as correct, as the links. There is only a small problem with all of them, but it makes them completely unusable.
This is the correct link: http://hyperphysics.phy-astr.gsu.edu/hbase/electric/indsol.html
Unfortunately the equation cannot be repaired this simple. As I said, it is for homogenous material. If all core is air, then OK. If part of the core is another material, then the magnetic resistance is sum of two parts. (Al=1/magn. resist.) Now the magnetic field goes through iron and air! The mentioned equation takes account only the resistance of inner field, but there is outer field too! Magnetic resistance is so low for iron, what is insignificant compared to air. This because inductance of this kind of coil is practically independent of permeability of iron (if it is over ~50)! (This is true for all heavily gapped cores!)
There is only a small problem with all of them, but it makes them completely unusable. This is the correct link: http://hyperphysics.phy-astr.gsu.edu/hbase/electric/indsol.html
Unfortunately the equation cannot be repaired this simple. As I said, it is for homogenous material. If all core is air, then OK. If part of the core is another material, then the magnetic resistance is sum of two parts. (Al=1/magn. resist.) Now the magnetic field goes through iron and air! The mentioned equation takes account only the resistance of inner field, but there is outer field too! Magnetic resistance is so low for iron, what is insignificant compared to air. This because inductance of this kind of coil is practically independent of permeability of iron (if it is over ~50)! (This is true for all heavily gapped cores!)
RX5 So, you tweak deadtime by R67 (on l8st_cheme.gif) ? What value you currently use? Bigger R67 means bigger deadtime?
Btw, your series gate resistors seems very small. Use 27-33 Ohm with lighter MOSFET`s (see st.com production)
PS mine implementation, based on Kartino`s pcb (slightly tweaked too) was discussed here:
http://www.diyaudio.com/forums/showthread.php?threadid=68539
Btw, your series gate resistors seems very small. Use 27-33 Ohm with lighter MOSFET`s (see st.com production)
PS mine implementation, based on Kartino`s pcb (slightly tweaked too) was discussed here:
http://www.diyaudio.com/forums/showthread.php?threadid=68539
RX5! Nice job! Don't wait for us, if you can do it!
OFF
But "huge/large" is not infinite in this case. It's a certain value which should be taken account. The magnetic resistance of air "gap" can't be 0. If that equation was right, it would be 0.
The outer magnetic field is similar to the electric field of a dipole. As capacitance of 2 given diameter sphere is a finite value, the magnetic resistance of outer space is definitely bigger then 0. I know from practice that it is bigger then iron core's, but I can calculate it if you want.
On that page ampere's law is applicable only for homogenous meterials, since they treat u as a constant. In correct form of law is integral(H)=sum(I), where H=B/u, and path should surround (once) the conductors. u depends on the place.
OFF
But "huge/large" is not infinite in this case. It's a certain value which should be taken account. The magnetic resistance of air "gap" can't be 0. If that equation was right, it would be 0.
The outer magnetic field is similar to the electric field of a dipole. As capacitance of 2 given diameter sphere is a finite value, the magnetic resistance of outer space is definitely bigger then 0. I know from practice that it is bigger then iron core's, but I can calculate it if you want.
On that page ampere's law is applicable only for homogenous meterials, since they treat u as a constant. In correct form of law is integral(H)=sum(I), where H=B/u, and path should surround (once) the conductors. u depends on the place.
@Kuzmenko,
R67, yes... im currently using 330.. i dont know about higher rail voltage.. im using +18/-18 now..... and yes, higher value, higher deadtime.. to the point that that when you go higher, NO output will result.. try 560 and see(actually hear).. if you can hear anything at all.... BTW, changing output filter coil, can also alter value of R67.. deatime resistor.... adjust accordingly... im using 26uH coil( gapped E-E ferrite core)
TIP on deadtime: experiment value, turn on amp.. see if mosfet/s heat up (even w/o signal input).. if it does heat, change NEXT higher resistor value...
actually there are 3 values to tweak... namely deadtime, Feedback, and input(current mirror).... some distortion can be heard if not balanced properly... "tweak" enjoy.. as what Chris would say...
BTW, thanks for the series gate resistor tip....
@Pafi,
right on!!
anyways, got It all in the wraps... ill post pics tommorow.. got a date tonight..
cheers,
Raff
R67, yes... im currently using 330.. i dont know about higher rail voltage.. im using +18/-18 now..... and yes, higher value, higher deadtime.. to the point that that when you go higher, NO output will result.. try 560 and see(actually hear).. if you can hear anything at all.... BTW, changing output filter coil, can also alter value of R67.. deatime resistor.... adjust accordingly... im using 26uH coil( gapped E-E ferrite core)
TIP on deadtime: experiment value, turn on amp.. see if mosfet/s heat up (even w/o signal input).. if it does heat, change NEXT higher resistor value...
actually there are 3 values to tweak... namely deadtime, Feedback, and input(current mirror).... some distortion can be heard if not balanced properly... "tweak"
enjoy.. as what Chris would say...BTW, thanks for the series gate resistor tip....
@Pafi,
right on!!
anyways, got It all in the wraps... ill post pics tommorow.. got a date tonight..
cheers,
Raff
as promised
direct link/s:
===================
http://i6.photobucket.com/albums/y209/beamrx5/top1.jpg
http://i6.photobucket.com/albums/y209/beamrx5/front.jpg
http://i6.photobucket.com/albums/y209/beamrx5/back1.jpg
http://i6.photobucket.com/albums/y209/beamrx5/underside.jpg
enjoy.....
Raff
direct link/s:
===================
http://i6.photobucket.com/albums/y209/beamrx5/top1.jpg
http://i6.photobucket.com/albums/y209/beamrx5/front.jpg
http://i6.photobucket.com/albums/y209/beamrx5/back1.jpg
http://i6.photobucket.com/albums/y209/beamrx5/underside.jpg
enjoy.....
Raff
Nice job!
While other peoples talking about manufactured module, while other peoples using SMD, multi layer PCB and latest component, here you prove that even using old fashioned style still produce a good sound.
Carlos, (destroyer X) do you interested? Where are you papa?
Thanks to UcD and Bruno!
regards,
kartino
While other peoples talking about manufactured module, while other peoples using SMD, multi layer PCB and latest component, here you prove that even using old fashioned style still produce a good sound.
Carlos, (destroyer X) do you interested? Where are you papa?
Thanks to UcD and Bruno!
regards,
kartino
kartino said:
Eeaaaasy now! Merely re-affirmed if anything, remember where you got your schematics from. But that aside the proof we don't see. More of a proud Papa syndrom, and the current is still being limited with series power resistors.Such a project can be worthwhile, but it will never compare to a proper amp, I can assure you the noise floor is high, and so is EMI.
I know, it will not comparable to original one and not for that reason we built that amps.
I just happy, while in other place you can built D-amps with fast generator, fast comparator, several choice mosfet driver, good fets, multilayer PCB. And here for some region we only find general fets, bjts and single layer. Of course also about money. These are the reason.
No worry about EMI/RMI. I ever got EMI problem when my monitor got disturbance. I know some country have regulation. Better to check againts local regulation if any.
I just happy, while in other place you can built D-amps with fast generator, fast comparator, several choice mosfet driver, good fets, multilayer PCB. And here for some region we only find general fets, bjts and single layer. Of course also about money. These are the reason.
No worry about EMI/RMI. I ever got EMI problem when my monitor got disturbance. I know some country have regulation. Better to check againts local regulation if any.
I agree, it's the perfect poor man's class D It gets difficult when they expect to get many hundreds of watts from it though. You don't even need a PCB! This say's alot for the robustness of self oscillating designs with post filter feedback.
I've encouraged Raff to try removing the current limiting resistors and see how it works.
It gets difficult when they expect to get many hundreds of watts from it though. You don't even need a PCB! This say's alot for the robustness of self oscillating designs with post filter feedback.I've encouraged Raff to try removing the current limiting resistors and see how it works.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.