Chris,
been reading the "miller effect" issue.... so whats affecting the miller effect here?? the WHOLE Hi side mosfet driver? or the 100uF capacitor on the High side?
might I change that cap to something lower?? say 50uF?
just a thought....
😀 
been reading the "miller effect" issue.... so whats affecting the miller effect here?? the WHOLE Hi side mosfet driver? or the 100uF capacitor on the High side?
might I change that cap to something lower?? say 50uF?
just a thought....
😀 Nope, it's not the output of the amp that's of concern, if you figure out what the precharge circuit is actually there for, you'll see that you can't enable the modulator to oscillate until the bootstrap cap has full charge, otherwise nobody will be home when it tries to switch high. So what you need is an appropriate delay after the power comes up before turning the amp on at all. No need for any aditional relay unless it falls into your plans of added DC protection or something.
"wikipedia.. miller effect.. gotta read that one later... CMRR ( Common Mode Rejection Ratio) ok!! "
Right on, wikipedia I've found to be an excellent site sometimes to get the basic info down, but then, you'll really want to dig much deeper with google, look at application notes and white papers, anything you can get your hands on.
I know I'm guilty of the same, alot of "must read that later" never gets read later. If you find something good, give it a real quick once over so you at least know what's in there, then when the time of need comes you'll have a few ideas immediatly worth following up on.
"been doing THAT .. changed it to 1k(470/330) its works on SIM BUT heats the MOSFETs in real life.so i got to "ratio" it out at higher values.. is that correct? "
Yes, the magic ratio is to have your gate to source waveform falling twice during turn off twice as fast as it rises during turn on, with signals intersecting as close to Vth as possible, this produces minimal deadtime which is best for THD and cool running. This is also where mosfet selection plays a large roll, miller effect.... not easy!
"nop.. it aint a lab power supply.... its a 24-0-24 @ 2A transformer... once again, half wave 60Hz... I cant find 48-0-48 here"
Yes I thought so, how were going to get 46Vdc out of that? Either way about the current... your output stage and load will actually see far greater than 2amps ... transformer will just run alot hotter.
"so a "common" value for the High an Low?? hmmm ok... been using different values for the Hi side and Lo side.... but changing supply to a lower voltage meant lowering AGAIN value of said resistor/s.... ok ill try the resisitor in collector leg.. in SIM..."
Exactly, but at least this way, bootstrap cap can "pump" up fast when it switches low, and you only have one resistor to adjust when you change supply voltage instead of two. I would size it not so that it stays cold to touch, just to ensure it doesn't see too much current, otherwise they can run pretty hot safely.
About the flux, "solder paste" with embedded solder particles with most certainly cause shorts if you don't clean it up nicely 🙂
"wikipedia.. miller effect.. gotta read that one later... CMRR ( Common Mode Rejection Ratio) ok!! "
Right on, wikipedia I've found to be an excellent site sometimes to get the basic info down, but then, you'll really want to dig much deeper with google, look at application notes and white papers, anything you can get your hands on.
I know I'm guilty of the same, alot of "must read that later" never gets read later. If you find something good, give it a real quick once over so you at least know what's in there, then when the time of need comes you'll have a few ideas immediatly worth following up on.
"been doing THAT .. changed it to 1k(470/330) its works on SIM BUT heats the MOSFETs in real life.so i got to "ratio" it out at higher values.. is that correct? "
Yes, the magic ratio is to have your gate to source waveform falling twice during turn off twice as fast as it rises during turn on, with signals intersecting as close to Vth as possible, this produces minimal deadtime which is best for THD and cool running. This is also where mosfet selection plays a large roll, miller effect.... not easy!
"nop.. it aint a lab power supply.... its a 24-0-24 @ 2A transformer... once again, half wave 60Hz... I cant find 48-0-48 here"
Yes I thought so, how were going to get 46Vdc out of that? Either way about the current... your output stage and load will actually see far greater than 2amps ... transformer will just run alot hotter.
"so a "common" value for the High an Low?? hmmm ok... been using different values for the Hi side and Lo side.... but changing supply to a lower voltage meant lowering AGAIN value of said resistor/s.... ok ill try the resisitor in collector leg.. in SIM..."
Exactly, but at least this way, bootstrap cap can "pump" up fast when it switches low, and you only have one resistor to adjust when you change supply voltage instead of two. I would size it not so that it stays cold to touch, just to ensure it doesn't see too much current, otherwise they can run pretty hot safely.
About the flux, "solder paste" with embedded solder particles with most certainly cause shorts if you don't clean it up nicely 🙂
RX5 said:Chris,
been reading the "miller effect" issue.... so whats affecting the miller effect here?? the WHOLE Hi side mosfet driver? or the 100uF capacitor on the High side?
might I change that cap to something lower?? say 50uF?
just a thought....
You "could" use a smaller cap there.... I sure wouldn't 😉
Keeeeeep reading!
The Miller effect applies to both the upper and lower devices and is unrelated to the 100uF bulk capacitance. It is due to the gate to drain cap of the MOSFET.
This has nothing to do with the upper MOSFET driver drawing more current. Cleaning the board to fix it sounds strange, but I can't see anything wrong with the circuit that would cause the diode to overheat. So maybe it did.
This has nothing to do with the upper MOSFET driver drawing more current. Cleaning the board to fix it sounds strange, but I can't see anything wrong with the circuit that would cause the diode to overheat. So maybe it did.
sawreyrw said:
I'd argue that point somewhat, but then Raff would be less inclined to keep reading.
a simple words... effect that cause need more current for gate charging by mosfets operation, such as high side mosfets, looked like that raising gate capacitance over than on spec.
GUYS!! 😀
thanks for those REPYLIES 🙂 I needed that... 🙂 heheheheheh
anyways, I have changed the "feedback" and "input" resistors.... both same value... aND WOW!! cleaner signal output...
I donnu if "crosstalk" is right term?? BUT it is much cleaner than BEFORE... and AMP GAIN.. is GOOD!! way much better THAN Before... I guess something WAS wrong with my circuit before I cleaned up the FLUX residue.... hhhmmmmm
in short: it sounds much CLEANER ==AND== LOUDER....really!!
@Chris,
about the power supply 24-0-24, i can make it +/-48V BUT halfwave only, meaning 60Hz.... as opposed to 120Hz full wave...(60Hz, its the line freq of Philippines) still cant imagine how? heheheheh ill post diagram if you want....
regarding bootstrap, output DC offset settles in 240mV in about 1/4 sec 🙂.... BUT i DO need it to sound "quiet" during turn on/off of amp... everytime I power it up, I remove 1st speaker, power up, then I connect it.... 🙂
100uF Hi side changed to 50uF.. hmmmmm can I remove the "gate cap of MOSFET" ??? hahahha 😀
and uhm ratio?? is 470/330 >>>> 1.42:1 ?? rough estimate.. was thinking 680/470.... 10R is to stay... I like it/him just sitting there
nywayz, been canvassing around.... power transformer is around P245.00 24-0-24 @2A (gonna buy 2 if I fullwave+/-48V) ...... 10,000uF 63V around P360.00 ... diode?? many of them here in my junk .. 🙂 1USD = P56.00
-OR-
I could just whip up a DC-DC converter 12V to +/-48VDC @ 3A.... I dunno w/c is better though.. cap is not CHEAP!!
If I opt for the latter, I will still need a 12VDC of at least 6A to supply the DC-DC.. but I have PC AT PS here.. no use.. might as well use it? 🙂
Be back in a few short hours/days.... need some work done!!
UcD amp ROCKS!!!!!!!
thanks for those REPYLIES 🙂 I needed that... 🙂 heheheheheh
anyways, I have changed the "feedback" and "input" resistors.... both same value... aND WOW!! cleaner signal output...
I donnu if "crosstalk" is right term?? BUT it is much cleaner than BEFORE... and AMP GAIN.. is GOOD!! way much better THAN Before... I guess something WAS wrong with my circuit before I cleaned up the FLUX residue.... hhhmmmmm
in short: it sounds much CLEANER ==AND== LOUDER....really!!
@Chris,
about the power supply 24-0-24, i can make it +/-48V BUT halfwave only, meaning 60Hz.... as opposed to 120Hz full wave...(60Hz, its the line freq of Philippines) still cant imagine how? heheheheh ill post diagram if you want....
regarding bootstrap, output DC offset settles in 240mV in about 1/4 sec 🙂.... BUT i DO need it to sound "quiet" during turn on/off of amp... everytime I power it up, I remove 1st speaker, power up, then I connect it.... 🙂
100uF Hi side changed to 50uF.. hmmmmm can I remove the "gate cap of MOSFET" ??? hahahha 😀
and uhm ratio?? is 470/330 >>>> 1.42:1 ?? rough estimate.. was thinking 680/470.... 10R is to stay... I like it/him just sitting there
nywayz, been canvassing around.... power transformer is around P245.00 24-0-24 @2A (gonna buy 2 if I fullwave+/-48V) ...... 10,000uF 63V around P360.00 ... diode?? many of them here in my junk .. 🙂 1USD = P56.00
-OR-
I could just whip up a DC-DC converter 12V to +/-48VDC @ 3A.... I dunno w/c is better though.. cap is not CHEAP!!
If I opt for the latter, I will still need a 12VDC of at least 6A to supply the DC-DC.. but I have PC AT PS here.. no use.. might as well use it? 🙂
Be back in a few short hours/days.... need some work done!!
UcD amp ROCKS!!!!!!!
Tsk Tsk.. should be more like 220uF, but it does have nothing to do with miller I'm afraid. Those little "tips" they gave you, dont' excuse you from reading up on it. In essence, they've told you nothing at all. You may want to search for example, power switching at high speed, with any kind of transistor. "anyways, I have changed the "feedback".."
It's about time, I told you to keep them = in each feedback loop at least six times now 😉
"regarding bootstrap, output DC offset settles in 240mV in about 1/4 sec .... BUT i DO need it to sound "quiet" during turn on/off of amp... everytime I power it up, I remove 1st speaker, power up, then I connect it.... "
offset is controlled by a few other factors, and I leave it to you to find them! However, it's more than capable of a silent turn on/off, and at most with highly efficient speakers, you "might" hear a little "thick" when you plug it in. If you read the threads which you brought up yourself, you'll have all these answers!!
Good luck with higher rails!
Hi Raff,
I use DC protection for speaker, cheapest one, with delay ON to connect amps and speaker. Priced 1usd in jakarta. Perfect to delay connect the speaker avoid "dug" at on and disconnect quickly without "zzzzztt". Beside as standard protection too for me.
I though you can get cheap one also at your place. Don't you afraid your speaker burned?
About 50uF cap at high side driver, Chris is right. The 'big' cap is used to storage energy during clip. You know if your amp going clipped, then it has not enough time for bootstrap circuit? That's the reason.
And did you talk about 470/330 ohm for the resistor at bjt gate drives? I had bad experience with loosing dozen mosfets to matching that value.
I use DC protection for speaker, cheapest one, with delay ON to connect amps and speaker. Priced 1usd in jakarta. Perfect to delay connect the speaker avoid "dug" at on and disconnect quickly without "zzzzztt". Beside as standard protection too for me.
I though you can get cheap one also at your place. Don't you afraid your speaker burned?
About 50uF cap at high side driver, Chris is right. The 'big' cap is used to storage energy during clip. You know if your amp going clipped, then it has not enough time for bootstrap circuit? That's the reason.
And did you talk about 470/330 ohm for the resistor at bjt gate drives? I had bad experience with loosing dozen mosfets to matching that value.
Hello again.... 😀
oops.. but I thought it would take a shorter time to charge cap... hhhmmmmm too fast reading with "miller article" I guess.. hehe ok more uF then....
😀
there WAS something wrong with the Hi side driver at that time, raising alone "input" side to 4K resulted in -mini distortions- ... well , its all OK now .....
REALLY????
am I right in presuming its got something to do with "trimmer in the differential input and middle leg connected to current source?" 😛
Chris, , you where right about the 470/330... ratio.. well, changed mine with 680/470 (close to 1.42:1).. seems ok BUT MOSFET warmer than before, so I again adjust deadtime... its OK now... offset is still below +1V... GOOD!
so i bet power consumption is much lower now, right?? I dont know if it is STILL neccesary to add limiting resistor for the TIP31C??? design overkill? hehehe
on a sidenote, changing THOSE resistor (470/330) does change the amp switching freq??
in the future, i will... but im intruiged by Chris' "However, it's more than capable of a silent turn on/off, " ... never really thought of it though....its capable of such...
too bad... Never killed mine with that prob.... 🙂
oops.. but I thought it would take a shorter time to charge cap... hhhmmmmm too fast reading with "miller article" I guess.. hehe ok more uF then....
😀
there WAS something wrong with the Hi side driver at that time, raising alone "input" side to 4K resulted in -mini distortions- ... well , its all OK now .....
REALLY????
am I right in presuming its got something to do with "trimmer in the differential input and middle leg connected to current source?" 😛
Chris, , you where right about the 470/330... ratio.. well, changed mine with 680/470 (close to 1.42:1).. seems ok BUT MOSFET warmer than before, so I again adjust deadtime... its OK now... offset is still below +1V... GOOD!
so i bet power consumption is much lower now, right?? I dont know if it is STILL neccesary to add limiting resistor for the TIP31C??? design overkill? hehehe
on a sidenote, changing THOSE resistor (470/330) does change the amp switching freq??
in the future, i will... but im intruiged by Chris' "However, it's more than capable of a silent turn on/off, " ... never really thought of it though....its capable of such...
too bad... Never killed mine with that prob.... 🙂
classd4sure said:
Chris,
Could you point out exactly what post# or thread perhaps?? I think its not in this thread I started.....
Chris go home, he no hea!!
... way too m uch coffee 🙄
You must have found the thread /posts in question since you mentioned the use of the pot. You should "try" and have the current source roughly equal to Irb (balance/driver setter resistorer..current)
Ummm... yeh, and provide a proper delay before enabling it to oscillate.
... way too m uch coffee 🙄
You must have found the thread /posts in question since you mentioned the use of the pot. You should "try" and have the current source roughly equal to Irb (balance/driver setter resistorer..current)
Ummm... yeh, and provide a proper delay before enabling it to oscillate.
😀 😀 😀
Right ON!! heheheh
anyway, I think I found the "sweet spot" of my amp..... even at 0V of power supply filter capacitor(total diacharge).. even WITH speaker connected, and with no signal present, When I turn on power, I ONLY hear small "thup".... NOT THUMP!! hehehe turn-off of aMP?? same thing, "thup"... 🙂 not loud... pretty small.... so DEADTIME setting is really very important, dont you say 🙂 😀
and regarding 220uF, it heated.... I forgot , it was only 16V.. lol... Miller effect doubles up the charge/voltage on cap... so I used 25V... now, all is good/cool 🙂
Must have found? hehehe awww c'mon 😀 current source ,balance, you mean current flowing through the emmiter legs of comparator/s?
I am lost here.... does this point out again to bootstap charging Hi side? 🙂
whew.. need coffee too... been de-BUGGing whole day 🙂
Cheers,
Raff
Right ON!! heheheh
anyway, I think I found the "sweet spot" of my amp..... even at 0V of power supply filter capacitor(total diacharge).. even WITH speaker connected, and with no signal present, When I turn on power, I ONLY hear small "thup".... NOT THUMP!! hehehe turn-off of aMP?? same thing, "thup"... 🙂 not loud... pretty small.... so DEADTIME setting is really very important, dont you say 🙂 😀
and regarding 220uF, it heated.... I forgot , it was only 16V.. lol... Miller effect doubles up the charge/voltage on cap... so I used 25V... now, all is good/cool 🙂
Must have found? hehehe awww c'mon 😀 current source ,balance, you mean current flowing through the emmiter legs of comparator/s?
I am lost here.... does this point out again to bootstap charging Hi side? 🙂
whew.. need coffee too... been de-BUGGing whole day 🙂
Cheers,
Raff
Hello GUYS!😀 good morning... its about 7:39am....just woke up!!
.................................
well I thought , mighht as well start my day with a bit of "sounds" 😀
speaker connected to amp... and power supply caps fully discharged.. been sitting there all night.... When I pluged the power, I listened carefully... and.. I ask, IS IT ON YET?? never heard any =pops= 😀
WOW!!!
well, anyways.... only sound "thup" ..but that "thup" is so small...I hear is when I turn it off... i hear sound I think after 5 o 6 secs from power off... quiet nice.... 🙂
anyway......
Chris, any more "updates/addtl refinement" ??? 🙂
good morning all!! have a nice day...
Cheers,
Raff
.................................
well I thought , mighht as well start my day with a bit of "sounds" 😀
speaker connected to amp... and power supply caps fully discharged.. been sitting there all night.... When I pluged the power, I listened carefully... and.. I ask, IS IT ON YET?? never heard any =pops=
😀 WOW!!!
well, anyways.... only sound "thup" ..but that "thup" is so small...I hear is when I turn it off... i hear sound I think after 5 o 6 secs from power off... quiet nice.... 🙂
anyway......
Chris, any more "updates/addtl refinement" ??? 🙂
good morning all!! have a nice day...
Cheers,
Raff
None that I'm able to share without violating non disclosure agreement.
You've come far enough I think if you want to go further with it, it's on you now!! Hit the old app notes, try and find weak areas that maybe you can improve on, maybe try a different layout or all SMT, maybe add some overcurrent protection, maybe get a beefier transformer (more current capability)... I dont' know 🙂
Cheers,
Chris
hahahah😀 ok no problem with that 🙂
oohhh.. hadnt noticed that ... 😀 i thought I was still 1/20 from the truth.... lol....
anyways , I was thinking about diff input pair.. how bout a matched transistor... like MD8003.. 60V..... and 1% toleance resistors on emmiter legs..???
.. maybe... dunno....this is HOW I describe UcD, see link below:
http://users.on.net/~richard.norrish/Autograph/hypex.htm
I feel you , dude!!! 😀
anyway, I think ill just? be content with +/-36V for now... too loud anyway.. hehehe (but should I NOT try +/-48V?)
maybe in the near future, I will add an op-amp input buffer.... IF I could find said chip here....
Thanks a lot Chris.. never had gone this far w/o you....
Cheers,
Raff
You can use any good op amp you like, just watch the offset if you go DC coupled.
Odds are it's not going to be half as good with them though from a neutrality point of view... really can't be driving the comp. directly.
Those were a few good ideas to try, and maybe some low noise audiopheliac type transistors too. Makes a little bit of a difference.
Glad it's working for ya... maybe I should build one 😀
Odds are it's not going to be half as good with them though from a neutrality point of view... really can't be driving the comp. directly.
Those were a few good ideas to try, and maybe some low noise audiopheliac type transistors too. Makes a little bit of a difference.
Glad it's working for ya... maybe I should build one 😀
classd4sure said:
offset... got it..... hhhmmmm you mean slight degration with an OP amp OR better WITH an op-amp input?? whew... getting slow on =english= here... tsk tsk...
Chris, dont tell me you had not DONE this before...? 🙂
RX5 said:
I mean there's no op amp so neutral sounding as no op amp at all, but it has its advantages.
Ok.. I won't tell you 😀
ohhh.. so WITHOUT an OP-amp input is much "transparent" sounding than with one??? hmmmmm I really need to find a GOOD op-amp here... I can see that op-amp is NOT part of the osc loop.... 🙂
you sly ol' dog... hahahah😀 😀 😀
- Status
- Not open for further replies.