My question was about clipping. When an amp clips, the divided-down output falls short of the input and that should trigger the protection. If it doesn't, the protection doesn't do its job.
So you probably would want a delay to make sure that short-time clipping (and EVERY amp clips short-time during music play) doesn't lead to what John calls nuisance tripping.
Are you sure there's no delay of some sort in your protection?
jan
Why would this be the case ? If for example my volume is adjusted such that a 0 dB fs test tone from my CD player is (with some margin) less than the amplifiers input sensitivity, how would it clip ?
My Loudspeakers are 97db/W and my amp 140w. My neighbors don't like too much those peak 120db before it clips!
... And my ears clipping distortion neither.
Please, i would prefer to answer questions about this protection here:
http://www.diyaudio.com/forums/solid-state/221737-ultimate-amp-protection-circuit.html
where i will follow-up your question.
Kill them.
(Funny that was my answer to Jan on the thread)
On my system, clipping occurs at 120db SPL, it kill me before i can kill my neighbors .
Last edited:
So was this one.
gk7 is correct for digital sources (assuming no overshoot from the reconstruction filters), but crazy as it sounds, some people use analog sources where the max levels are rather ill-defined.
That's a philosophy I heartily endorse. When I read about phono preamps with 34dB overload margins being required, I wonder what power amplifier and speakers one deploys to support multi-kilowatt transients!
And with proper design, clippers can be produced that have undetectable distortion outside of onset of clipping.
Brad
World can be crazy. When i was working for TV sound tracks in France, professional digital recorders where set at -18db from full bits for 0db reference level.
In the TV channels, the verification services used peak meters and young guys looking at them all day long to ensure that no soundtrack goes louder than +3db.
How to loose the15 best defined dbs of all the sound channel, including emitters. !!!!
In the TV channels, the verification services used peak meters and young guys looking at them all day long to ensure that no soundtrack goes louder than +3db.
How to loose the15 best defined dbs of all the sound channel, including emitters. !!!!
The point of making a commercial product that high-end would buy (JC's designs) is that YOU, the designer, cannot ASSUME it will not be used where clipping will not happen. And, nusience tripping is not tolerated by consumer for any reason what so ever. The delay in the response time is to be sure it is a serious and sustained over-drive condition and not just playing the music loud on a few occassional peaks.
How this is done in the design varies and the approach discussed here can be adapted to such market demands. For one-off use with knowledgable people to adjust to individual apps is a different story. Thx - RNM
How this is done in the design varies and the approach discussed here can be adapted to such market demands. For one-off use with knowledgable people to adjust to individual apps is a different story. Thx - RNM
And i would not tolerate it neither when i listen music.
...As well as i do not tolerate clipping.
In an other way i remember that, during the 70s, we designed a consumer level amp with a sensible protection and a light changing its color, and, at this time, customers where very attracted by this protection. Like an extra gadget they loved to demonstrate to their visitors, tuning the volume to the max to fire the protection, as well as magazines wrote rave reviews about that..
.
At the end, i believe that customers can love what you design, if there is a good reason for your design and if you argue about. May-be the 70s custommers were more intelligent, to prefer some rare "nuisance tripping" to a burned amp or destroyed loudspeaker ?
Well an indicator showing that clipping occurs is enough and a good educational thing.
I don't see why some would like to spend a fortune to buy a 0.0000000001% hd amp to play it at 20% distortion.
Delay is not a requirement for a protection, only a design flaw. No tripping is the requirement, yes, as well as very fast protection. My protection idea can respond to that with no inconvenience, period.
Last edited:
I am clarifying the situation about what JC's delay refers and why from hard, practical, in the field comments from consumer, dealers (especially) and reviewers.
- not so much commenting on your particular design or its abilities and capabilities.
Comment on clipping lights. From experience, dealers and consumers dont like them. When they play music, that has high crest factor, they feel they must turn the level down until zero clipping, then the average volume is too low. A lot of the time people are listening to their amps clipping. The others cant afford very high power amps which wont clip and speakers that dont compress and can take the power and too few really high sensitivity speakers to choose from. It doesnt solve the problem for consumers and dealers who have to try and sell lower power amps with typical low sens speakers which people will play at high volume. they would just as soon not know their amps are clipping, especially the dealers, rather than listen at faint average levels on low eff spkers.
- not so much commenting on your particular design or its abilities and capabilities.
Comment on clipping lights. From experience, dealers and consumers dont like them. When they play music, that has high crest factor, they feel they must turn the level down until zero clipping, then the average volume is too low. A lot of the time people are listening to their amps clipping. The others cant afford very high power amps which wont clip and speakers that dont compress and can take the power and too few really high sensitivity speakers to choose from. It doesnt solve the problem for consumers and dealers who have to try and sell lower power amps with typical low sens speakers which people will play at high volume. they would just as soon not know their amps are clipping, especially the dealers, rather than listen at faint average levels on low eff spkers.
Last edited:
Since this is about high-end circuits -- JC's -- the best high end approach -- IMO -- is to have amp with enough power that it wont clip at the highest safe listening levels with the speakers of choice in the room size it will be played. And, for sources of a standard output level to never be able to overdrive the power amp into clipping. Then protection is for a truely abnormal condition in which shutdown is warrented. Its a goal. Thx- RNm
Let me respectfully disagree. I give a freedom, a choice to the user, either to ignore the light and listen to compressed on 80W / channel level, or to reduce power below 80W / channel so lights don't flash. Either they choose "Louder", really louder than any other amp that clips on 80W, but much cleaner, or they choose softer, but fair dynamic range reproduction.
Most people would rather have a switch to defeat the lights..... esp annoying to have flashing lights in a darkened room while listening. All these ideas have been seen by dealers and it is usually the dealers who get and give the feedback and that is what I am telling here. Your arguement isnt with me.... its with dealers.
[ You might note the Japanese and Asia brands have dimming on most everything and off for front panel lights.... they have to do this because of dealer/customer feedback.] -RNM
[ You might note the Japanese and Asia brands have dimming on most everything and off for front panel lights.... they have to do this because of dealer/customer feedback.] -RNM
- Status
- Not open for further replies.