Usually no line signal output should carry dc. So there is actually no input cap necessary. After my experiments with caps with the audio sector amp I left out the cap - as Peter suggested - and the amp works fine with no output offset. That is great, because I do not use a feedback cap either, which sounds much better than with a cap (the three-resistor-amp!).
If this is your advice then you can explain to beginners why their amp just blew up their speakers that cost 100times what the chip cost.
In fact, you explain to us why you elect to give this advice and how you established the risks and balanced them against potential costs.
This bit I do not believe:
Under what operating/test conditions did you obtain this result. Can you offer that as a warranty to others to whom you give your advice.
wow, wow, wow!
This is my experience - no more, no less. Take it as an advice, or not.
Plenty of advices for beginners are availlable here - in this very thread or elsewhere. You have also posted a lot.
Actually I followed an advice Peter - who started AND is owning this thread - gave in reducing parts as much as possible. I worked also with his boards - which do not have options for input and feedback caps. So what problem do you have here??
Strangely my latest 3-resistor amps do not have any significant output offset - do not ask me why. Believe it or not. An no warranties.
This is my experience - no more, no less. Take it as an advice, or not.
Plenty of advices for beginners are availlable here - in this very thread or elsewhere. You have also posted a lot.
Actually I followed an advice Peter - who started AND is owning this thread - gave in reducing parts as much as possible. I worked also with his boards - which do not have options for input and feedback caps. So what problem do you have here??
Strangely my latest 3-resistor amps do not have any significant output offset - do not ask me why. Believe it or not. An no warranties.
AndrewT, lohk,
Regarding DC offset, I am using a DC B1 buffer and DC Audiosector Gainclone amp together without a cap between them or after the Gainclone, and have -20mV & -18mV DC at my speaker terminals. Although this is a minus figure, it does not suggest that there is no DC there at all. Correct me if I am wrong, but I would have thought that to have absolutely zero DC in such an event would be most unlikely.
Regarding DC offset, I am using a DC B1 buffer and DC Audiosector Gainclone amp together without a cap between them or after the Gainclone, and have -20mV & -18mV DC at my speaker terminals. Although this is a minus figure, it does not suggest that there is no DC there at all. Correct me if I am wrong, but I would have thought that to have absolutely zero DC in such an event would be most unlikely.
Hi Andrew,
I am kind of aware of the risk I am taking with having no decoupling caps, as I have been warned by both yourself and Salas, but I can't see a reason why I need to incorporate one now everything is working well. I guess a component failure might change things for the worse. Please let me know if I am being really seriously stupid.
I would be more inclined to put a DC protection relay between amp and speakers than start experimenting with decoupling caps. Does that make any sense?
I am kind of aware of the risk I am taking with having no decoupling caps, as I have been warned by both yourself and Salas, but I can't see a reason why I need to incorporate one now everything is working well. I guess a component failure might change things for the worse. Please let me know if I am being really seriously stupid.
I would be more inclined to put a DC protection relay between amp and speakers than start experimenting with decoupling caps. Does that make any sense?
motive
Hi,
to justify his decision to risk his system after adopting a DC coupled amplifier.
And then to convince others that his unwarrantied experiences are worth following.
I know you cannot get zero DC output offset from a DC coupled amplifier over the range of operating conditions that are experienced. Many beginners do not know this. Peter does not tell them this. Even worse, there is no mention in Lohk's comments that component failure can take the output offset to near supply rail voltage and that this is likely to damage a speaker.
I have in other threads said that in my opinion all DC coupled amplifiers need to add some form of protection/correction. I usually state DC servo and DC detect and output isolation as a reasonable way to address the risks of using DC coupled amplifiers.
Hi,
to justify his decision to risk his system after adopting a DC coupled amplifier.
And then to convince others that his unwarrantied experiences are worth following.
I know you cannot get zero DC output offset from a DC coupled amplifier over the range of operating conditions that are experienced. Many beginners do not know this. Peter does not tell them this. Even worse, there is no mention in Lohk's comments that component failure can take the output offset to near supply rail voltage and that this is likely to damage a speaker.
I have in other threads said that in my opinion all DC coupled amplifiers need to add some form of protection/correction. I usually state DC servo and DC detect and output isolation as a reasonable way to address the risks of using DC coupled amplifiers.
protection and/or correction makes a lot of sense. Being aware of these risks and solutions puts you way above Beginners.
Of course you are right that DC offset always exists to some degree and that it can become dangerous to the speakers. I just think calling some guy you have no history with a liar is wrong.
This is definitely a beginners thread and the threat of DC offset is something they should all be made aware of as soon as they start building. I know I slowly fried a speaker and paid dearly for a replacement because I did not know that my DC offset could change over time. Now I check it once a month. Maybe not often enough but its a calculated risk.
Uriah
This is definitely a beginners thread and the threat of DC offset is something they should all be made aware of as soon as they start building. I know I slowly fried a speaker and paid dearly for a replacement because I did not know that my DC offset could change over time. Now I check it once a month. Maybe not often enough but its a calculated risk.
Uriah
Andrew, thanks for that.
In your opinion, what would the net effect on sound quality be with such a DC detect system, with relays to disconnect in the event of contamination?
I am talking of a system as is found quite cheaply on eBay: Speaker protection circuit mirror arranged monobloc ! on eBay (end time 12-Oct-10 17:32:45 BST)
I understand that the effect of the relays is minimal, dependant on their quality, but I am unsure about the signal effect of the detection method.
In your opinion, what would the net effect on sound quality be with such a DC detect system, with relays to disconnect in the event of contamination?
I am talking of a system as is found quite cheaply on eBay: Speaker protection circuit mirror arranged monobloc ! on eBay (end time 12-Oct-10 17:32:45 BST)
I understand that the effect of the relays is minimal, dependant on their quality, but I am unsure about the signal effect of the detection method.
Hi,
a detect circuit simply monitors the output signal from the amplifier and compares the DC level to some pre-determined threshold.
It triggers a relay activation circuit (the output isolation circuit).
The only Sound Quality effect that I am aware of is the addition of the relay into the output lead.
According to some this addition is audible. It does introduce soldered connections and relay contacts and some extra cabling to the speaker output lead.
I would rather have a working speaker, that I can listen to, rather than a damaged speaker that I cannot listen to.
a detect circuit simply monitors the output signal from the amplifier and compares the DC level to some pre-determined threshold.
It triggers a relay activation circuit (the output isolation circuit).
The only Sound Quality effect that I am aware of is the addition of the relay into the output lead.
According to some this addition is audible. It does introduce soldered connections and relay contacts and some extra cabling to the speaker output lead.
I would rather have a working speaker, that I can listen to, rather than a damaged speaker that I cannot listen to.
post1267 gave him opportunity to justify his decisions and explain what's why's and wherefore's.
He chose not to.
I take that to mean he could not justify his statements.
I simply requoted one back to him and made it plain that in my view he is lying. He has still not come back to show why I am wrong.
Well, I have other things to do than to respond to your unnecessary bad comment, Andrew.
You call me a liar, only why you cannot believe that a reduced chipamp without offset is possible? I do not have to prove that to you. Quite a few other persons did already have the same experience - and have posted it.
I should justify my experience? Please come over and measure my amp yourself, if you still have doubts.
I was telling my experience here, my (audisector-based) amp is working beautifully for a year now or two without any problem - and without any offset. With three resistors.
Even more strange is, that another chipamp I have made some years ago, does indeed have some offset - here I need an input cap at least. The offset is not that bad, so I still dod not use a feedback cap even with that amp. It definitely sounds better without.
I have tried the same approach to an LM1876 but this one will only work properly with a feedback cap.
But what are we talking here? The audiosector boards DO NOT provide the option for the feedback cap and they DO NOT have the option for the input cap either. For a good reason: It works perfect and sounds good. End of story.
http://www.audiosector.com/images/lm3875_se_pcb.gif
You call me a liar, only why you cannot believe that a reduced chipamp without offset is possible? I do not have to prove that to you. Quite a few other persons did already have the same experience - and have posted it.
I should justify my experience? Please come over and measure my amp yourself, if you still have doubts.
I was telling my experience here, my (audisector-based) amp is working beautifully for a year now or two without any problem - and without any offset. With three resistors.
Even more strange is, that another chipamp I have made some years ago, does indeed have some offset - here I need an input cap at least. The offset is not that bad, so I still dod not use a feedback cap even with that amp. It definitely sounds better without.
I have tried the same approach to an LM1876 but this one will only work properly with a feedback cap.
But what are we talking here? The audiosector boards DO NOT provide the option for the feedback cap and they DO NOT have the option for the input cap either. For a good reason: It works perfect and sounds good. End of story.
http://www.audiosector.com/images/lm3875_se_pcb.gif
Actually, there is an option for an input cap, as mentioned in post #3: http://www.diyaudio.com/forums/audi...ne-kit-building-instructions.html#post1508778
However, as the title says, it's a kit based on my commercial Gainclone Amp, which was designed for best sound reproduction, even at the cost of some safety options.
The amp has been recognized as Best of 2005 and Favorite Discoveries of 2007 by 6moons:
6moons audio reviews: Srajan's Best of 2005
6moons industryfeatures: 6moons Favorite Discoveries of 2007 - Stephaen Harrell
and also received PFO Writer's Choice Award, 2005
I choose to do it this way, and I will not change anything. As such, I do not recommend it for beginners, if they choose to build it anyway, they have enough info here to be aware of all safety aspects.
However, as the title says, it's a kit based on my commercial Gainclone Amp, which was designed for best sound reproduction, even at the cost of some safety options.
The amp has been recognized as Best of 2005 and Favorite Discoveries of 2007 by 6moons:
6moons audio reviews: Srajan's Best of 2005
6moons industryfeatures: 6moons Favorite Discoveries of 2007 - Stephaen Harrell
and also received PFO Writer's Choice Award, 2005
I choose to do it this way, and I will not change anything. As such, I do not recommend it for beginners, if they choose to build it anyway, they have enough info here to be aware of all safety aspects.
Attachments
Of course I never recommend any speaker protection at the output of the Gainclone. This is against the very idea of simple amp. Besides, majority of other commercial amps don't use such protection.
If you make sure that your source will not produce any excessive DC offsets, you will be fine.
If you make sure that your source will not produce any excessive DC offsets, you will be fine.
I promised AndrewT some measurement data from my audiosector version of the three-resistor-chipamp:
input open (no cable), no speakers: 24mV / 28mV dc offset cold, after some warmup 17mV / 19mV and NO dc hump during switch in
input connected to 1m screened cable plus 10k pot passive on the other end: 9mV / 12mV
with or without speakers, pot closed (shorted) or fully open no difference
with audio signal the offset vanishes to aprrox. 5mV on both channels
(measurements done with multimeter and scope)
I frankly will call that result no offset. There is no audible scratching sound when you plug the speakers etc. There is some very little hum on both channels but this is definitely a construction fault and purely my responsibility.
input open (no cable), no speakers: 24mV / 28mV dc offset cold, after some warmup 17mV / 19mV and NO dc hump during switch in
input connected to 1m screened cable plus 10k pot passive on the other end: 9mV / 12mV
with or without speakers, pot closed (shorted) or fully open no difference
with audio signal the offset vanishes to aprrox. 5mV on both channels
(measurements done with multimeter and scope)
I frankly will call that result no offset. There is no audible scratching sound when you plug the speakers etc. There is some very little hum on both channels but this is definitely a construction fault and purely my responsibility.