quitar joe .....
i am not really familiar with the term u used "period" but i presume that you will explain to me ....
then on the other hand if its a perfectly made amplifier that doesnt need any of these discussed above please let us know ...i will be really interested in such a circuit .....
to an ocilating or litle ocilating amplifier that actually started this thread ...this makes a hell of audible diference .....
it could be the amps fault and not the inductor that made the magic .... but all is about the result isnt it ?????
also there is another thing that makes me very woried about .....
the name quitar joe ....are you a quitar player or constructor of amplifier for quitars ????? cause if you do there nothing for you to talk about here since your bandwidth stopes arround 4.5 khz ....tops
please dont missunderstand me ...i grubbed the oportunity to make a joke
i am not really familiar with the term u used "period" but i presume that you will explain to me ....
then on the other hand if its a perfectly made amplifier that doesnt need any of these discussed above please let us know ...i will be really interested in such a circuit .....
to an ocilating or litle ocilating amplifier that actually started this thread ...this makes a hell of audible diference .....
it could be the amps fault and not the inductor that made the magic .... but all is about the result isnt it ?????
also there is another thing that makes me very woried about .....
the name quitar joe ....are you a quitar player or constructor of amplifier for quitars ????? cause if you do there nothing for you to talk about here since your bandwidth stopes arround 4.5 khz ....tops
please dont missunderstand me ...i grubbed the oportunity to make a joke
i made a test .....
i added some variations of inductor to a P3 (by rod Elliot) amplifier just for fun.....
coils of 1.0 mm wire ,made in diameter of 12 mm and up to 12 turns tied with 2.2R 3W make no audible diference ...at least to my ears
if you exceed 12 turns and go to 14 or 16 it starts to have some audible diference and seem to loose some of the previously existing brilliance ...regarding mostly the tops .....
also if you go to diameter less than 12 and try 10 or 8 then things start to get worst .........
no scope tests ....at this point ...only listening.....
i will do so arround the weekend after i find out how i am going to give to my dummy load some capacitive behaviour ...and long cheap leads ....just to see
i added some variations of inductor to a P3 (by rod Elliot) amplifier just for fun.....
coils of 1.0 mm wire ,made in diameter of 12 mm and up to 12 turns tied with 2.2R 3W make no audible diference ...at least to my ears
if you exceed 12 turns and go to 14 or 16 it starts to have some audible diference and seem to loose some of the previously existing brilliance ...regarding mostly the tops .....
also if you go to diameter less than 12 and try 10 or 8 then things start to get worst .........
no scope tests ....at this point ...only listening.....
i will do so arround the weekend after i find out how i am going to give to my dummy load some capacitive behaviour ...and long cheap leads ....just to see
Re: lets keep ....
In post 54 I explained the implementation of a load coupling network in an actual design with superior RFI shunting characteristics that actually performs the dual function of an RC Zobel and capacitive load isolation.
In a thread devoted to amplifier stability, I guess that can be ignored.
sakis said:
In post 54 I explained the implementation of a load coupling network in an actual design with superior RFI shunting characteristics that actually performs the dual function of an RC Zobel and capacitive load isolation.
In a thread devoted to amplifier stability, I guess that can be ignored.
I think Bob summarized it nicely in his penultimate response, and then further in his last post. If th e laod is well defined, the Zobel and th e output inductor can in some cases be left off. If th e load is not defined . . . . I think we are actually all in agreement here guys . . . .
Main thing is we ar e not claiming that 2uH output inductors inherently compromise th e sound of an amp . . .
Main thing is we ar e not claiming that 2uH output inductors inherently compromise th e sound of an amp . . .
I think Bob summarized it nicely in his penultimate response, and then further in his last post. If th e laod is well defined, the Zobel and th e output inductor can in some cases be left off. If th e load is not defined . . . . I think we are actually all in agreement here guys . . . .
Main thing is we ar e not claiming that 2uH output inductors inherently compromise th e sound of an amp . . .
Main thing is we ar e not claiming that 2uH output inductors inherently compromise th e sound of an amp . . .
jagwap said:
I have 3 home designed amps with neither.
All are used for disco and guitar use.
I do understand that as the speaker inductor feeds back current back into the output stage that it needs soaking up.
But if the MOSFET is doing this then a zobel network is not required.
I reduced the bandwidth of the amps at higher frequencies by using HF feedback so this could explain why my amps dont need zobel or an inductor. At the VAS stage I use a 220pf cap from the collector to the base.
then i will
rest my case and keep in minde that your testing gear or knowledge about this is limited ....
no probs my friend if sounds good to your ears is perfect for me also ........
but on the other hand i have never seen a vas stage created with darligton any way .... so all of these is totally academic .....plus if we keep in mind that the miller cap is actually conected from the base of the driver transistor to the collector of the output transistor of the all darligton .....
so i am not familiar with this topology also .....
rest my case and keep in minde that your testing gear or knowledge about this is limited ....
no probs my friend if sounds good to your ears is perfect for me also ........
but on the other hand i have never seen a vas stage created with darligton any way .... so all of these is totally academic .....plus if we keep in mind that the miller cap is actually conected from the base of the driver transistor to the collector of the output transistor of the all darligton .....
so i am not familiar with this topology also .....
I was thinking about the question of which should come first the cap or the resistor....
My first response was, it should not matter... then it started haunting me...
Imagine the resistor is connected to ground side... the side the capacitor sees of this resisor is not at 0V DC, it is at whatever the voltage over that resistor would be at x frequency...
With the cap to ground (its internal resistance hopefully much smaller than the resitor) has a closer to 0 potential at the junction point....
I doubt it is audiable, and I certainly don't have the gear to test the validity of this conclusion, I am likely missing something very obvious....
A question for you sakis....
I think it was on one of your boards I once saw a screw on bracket for attatching boards to heatsink... I think you even had it in eagle... if it was you.. care to give me the details...
My first response was, it should not matter... then it started haunting me...
Imagine the resistor is connected to ground side... the side the capacitor sees of this resisor is not at 0V DC, it is at whatever the voltage over that resistor would be at x frequency...
With the cap to ground (its internal resistance hopefully much smaller than the resitor) has a closer to 0 potential at the junction point....
I doubt it is audiable, and I certainly don't have the gear to test the validity of this conclusion, I am likely missing something very obvious....
A question for you sakis....
I think it was on one of your boards I once saw a screw on bracket for attatching boards to heatsink... I think you even had it in eagle... if it was you.. care to give me the details...
fourth -output capacitor (for cheap speaker protection)
third (or second) -output switching relay (headphones out/speaker out)
first -Zobel 10R / 0.1u
second (or third) -Zobel L/R
fifth - add a resistor to charge and discharge the DC blocking capacitor
The combination of 1st & 2nd is the output Thiele Network.
Last edited:
Bonsai
http://www.diyaudio.com/forums/solid-state/120748-zobel-no-zobel.html#post1475450
Is your opinion stated in 2008 still the same some 6 years on?
http://www.diyaudio.com/forums/solid-state/120748-zobel-no-zobel.html#post1475450
Is your opinion stated in 2008 still the same some 6 years on?
GK is not here to reply.
Can anyone else explain GK's statement
"misguided belief that a high frequency load is generally necessary to keep an amplifiers open loop gain and phase margin under control."
I thought that the R+C part was specifically there to provide a resistive load at high frequency.
Can anyone else explain GK's statement
particularly this part:
"misguided belief that a high frequency load is generally necessary to keep an amplifiers open loop gain and phase margin under control."
I thought that the R+C part was specifically there to provide a resistive load at high frequency.
I think it would depend on what GK was driving at; the habitual use of a Zobel rather than experimental proof in each example, or that he believes that an HF load has nothing to do with the control of OLG and PM.
Douglas Self concludes with more certainty, based on his own experimental evidence from the CFP blameless design which is in all APAD handbook editions I've read:
Quote:
"There seems no doubt that a Zobel network is required with any load that is even mildly inductive."
The principle reason given being that amplifier HF instability is induced by interaction with the loudspeaker's inductance.
The load of the RC network on the amplifier is not great, being around 12 mA current @ 20kHz with 1V output. For comparison, an 8R resistive load draws 125 mA at all frequencies.
Douglas Self concludes with more certainty, based on his own experimental evidence from the CFP blameless design which is in all APAD handbook editions I've read:
Quote:
"There seems no doubt that a Zobel network is required with any load that is even mildly inductive."
The principle reason given being that amplifier HF instability is induced by interaction with the loudspeaker's inductance.
The load of the RC network on the amplifier is not great, being around 12 mA current @ 20kHz with 1V output. For comparison, an 8R resistive load draws 125 mA at all frequencies.
For interest, here's a reply by Bob Cordell, addressed generally to that comment and others in this thread by GK:
"I don't think I have ever built a solid state amplifier without a Zobel, and in my opinion only a fool would do so. Here is why. Even if the amp is stable on the bench with no load or a resistive load, you don't know what the complex nature of the load might be when speaker cables and a speaker are connected. The presence of the Zobel network largely guarantees the absence of a high-Q resonance on the output node of the amplifier.
Cheers,
Bob "
"I don't think I have ever built a solid state amplifier without a Zobel, and in my opinion only a fool would do so. Here is why. Even if the amp is stable on the bench with no load or a resistive load, you don't know what the complex nature of the load might be when speaker cables and a speaker are connected. The presence of the Zobel network largely guarantees the absence of a high-Q resonance on the output node of the amplifier.
Cheers,
Bob "
While the consensus seems to be; 'if in doubt, include a zobel + RC' as the loading effects of different speakers and cables can't be predicted, I'd be interested to know if people think this consensus still applies to a more predictable environment.
For example an active powered speaker with no passive crossover and power amps mounted inside the speaker cab with much shorter, enclosed speaker cables. Would both a zobel and the RC still be advised?
For example an active powered speaker with no passive crossover and power amps mounted inside the speaker cab with much shorter, enclosed speaker cables. Would both a zobel and the RC still be advised?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.