clip indicators - how do they work ? |
Post Reply | Page 12> |
Author | |
slaz
Old Croc Joined: 27 November 2009 Location: London E2 Status: Offline Points: 2713 |
Post Options
Thanks(0)
Posted: 06 November 2012 at 10:20am |
Noob question I guess ....
In general - whats happening when the clip light comes on ? Is it just a matter of indicating when input voltage (or somewhere further in to the circuit) is reaching a point where the amp has been measured to start clipping, or is there something more intelligent going on (such as detecting the flattened tops of o/p waveform) ? TIA. |
|
REMEMBER....POLITICIANS AND DIAPERS SHOULD BE CHANGED OFTEN AND FOR THE SAME REASON
|
|
DubSirenSpecialist
Registered User Joined: 21 June 2012 Location: London Status: Offline Points: 301 |
Post Options
Thanks(0)
|
Normally its when the limiters are kicking in above a certain dB threshold on the input gain isn't it? Stops the input waveform from flattening, therefore preventing the amplification of a distorted input signal to a distorted output signal, just the knowledge I have
|
|
The man of many technologies, the technologies of many men.
|
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5482 |
Post Options
Thanks(0)
|
There are loads of different ways to do it, ranging from the very simple and not much use to the much more sophisticated limiting methods of modern amplifiers.
If you really want to see what is happening then there is unfortunately no substitute for an oscilloscope and some dummy loads. |
|
slaz
Old Croc Joined: 27 November 2009 Location: London E2 Status: Offline Points: 2713 |
Post Options
Thanks(0)
|
OK. Was just musing on whether there's any sense in trying to "retrofit" a meaningful gain and/or clip indicator to an amp that doesn't have any ..... in my case car amplifiers for use with battery-powered systems. Car amps usually have screwdriver PCB-mounted pre-set type gain controls on the bottom panel .... I was thinking of mounting a car ampflifier (not one of the things with huge lumps of ally heatsink) in a kind of outer chassis - with sensible connectors for DC in, spkr out (mebbe speakon), but also some proper POT type gain controls, prob a DC voltage indicator, power-on switch etc. Would be useful to have a degree of control over gain e.g. somehow calibrate the pre-set gains so that "front panel" gain is useful - and know (roughly) where you are wrt the amps o/p capability. (Hope that makes sense) |
|
REMEMBER....POLITICIANS AND DIAPERS SHOULD BE CHANGED OFTEN AND FOR THE SAME REASON
|
|
kedwardsleisure
Old Croc Joined: 20 January 2009 Location: Staffordshire Status: Offline Points: 4949 |
Post Options
Thanks(0)
|
Some clip lights are I/O comparators, they indicate when the error signal is above a certain level. Some are output voltage-level dependent and some are driven off the input signal. It is all down to the designer. Some are more meaningful than others.
Edited by kedwardsleisure - 06 November 2012 at 11:50am |
|
Kevin
North Staffordshire |
|
audiomik
Old Croc Joined: 06 April 2010 Location: Bath, UK Status: Offline Points: 2962 |
Post Options
Thanks(0)
|
Slaz
another '3 minute circuit sketch' this time for a simple peak signal indicator: The value of *R1 can be roughly calculated from needing a little over 0.7 Volts peak on the base of the BC182 so as to indicate a peak positive half cycle of signal at the required level Mik |
|
Warning! May contain Nuts
plus springs, washers, screws, etc, etc. |
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5482 |
Post Options
Thanks(0)
|
To expand a little on the above circuit, the NPN first transistor needs approximately 0.7V applied to its base in order to forward bias its base-emitter junction and 'turn on' the current flow from collector to emitter.
So if you had a 1V max P-P signal and you wanted clip indicated at 0.8V you would select R1 such that 0.8x(4700/(R1+4700)) = 0.7 or R1 = approx 700 Ohms (assuming negligible impedance from input capacitor at freq of interest) When the voltage the the base of the first transistor is less than 0.7V, the voltage at its collector is 12V when the capacitor to ground is fully charged. Transistor 2 (PNP) is now off because its base and emitter are both at 12V, we need a difference of 0.7V or more less on the base than emitter to turn it on. When we bias transistor 1 on, then the collector is pulled low to ground via the emitter, the capacitor discharges through transistor 1's junction, and transistor 2's base is lowered to ~6V putting approximately 6V across the LED and resistor as well giving around 13mA to the LED. So you can change the transient response of the LED to detected peaks by altering the value of the capacitor as well. I think that's mostly right! I'm sure someone will correct me if not. I do spend more time thinking about analog electronics than building it. Edited by odc04r - 06 November 2012 at 12:47pm |
|
audiomik
Old Croc Joined: 06 April 2010 Location: Bath, UK Status: Offline Points: 2962 |
Post Options
Thanks(0)
|
Nearly, the BC212 will saturate, Vbe at 0.7V, and the collector will give Vs-Vcesat across the series R and LED, whilst the BC182 is 'on'.
'Slow recovery' is from the time constant provided by the 2 4k7 Resistors and the 2ยต2 Capacitor; so 'fast on' and 'slow off' for the LED. Same circuit with different values to allow for a different supply rail can be used for Amplifier outputs. This Circuit dates back to the late 1960's when it used a low current filament lamp instead of the LED Mik |
|
Warning! May contain Nuts
plus springs, washers, screws, etc, etc. |
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5482 |
Post Options
Thanks(0)
|
Ah yes, I see that now. Fast discharge of the cap through the transistor low resistance junction to pull Q2 hard on and then an RC discharge through the base potential divider to slowly turn the LED off.
I always used to laugh in the first year of electronics when the guy taking lecture would be talking about hard on transistors, probably still would now if I'm honest. |
|
slaz
Old Croc Joined: 27 November 2009 Location: London E2 Status: Offline Points: 2713 |
Post Options
Thanks(0)
|
Aw thanks guys. Really didn't mean anyone to go drawing circuits or writing essays about this.
Dunno about this .... so many variables involved - and no test gear or dummy loads available here (other than an OK analogue test-meter - since my mate's scope died) .... plus I wouldn't dare put continuous sine waves through the amp .... well mebbe at 16R load .... Kind of thinking I may just stick to using a combination of ears and putting back of hand on amp chassis to gauge temperature, as o/p trannies bolt directly to it. Its just the kind of music we play (Jamaican reggae/ska - some from the 60s, some more modern etc. ) is so variable wrt level, media quality, crest factor, recording quality, frequency spectrum etc. .... its quite hard to know "where you are" and set good levels - and even more so when there's someone on the mic having a yack over the tunes ..... but decided (on the mic front) that its better to just train the guy doing the yacking to dip the music level a bit manually when he's about to cut in - or for me to hover over the channel fader when this is going on ...... Just made a folder called "audiomik_sketches" to grab em P.S. guess it would also make sense to assume old DJ mixer with LED meters on o/p (post master fader) is part of the rig (which in fact it isn't always) - and assume that the car graphic EQ after the mixer is unity gain at "full" o/p ..... then take the amp manufacturer's power rating (with current Vibe amp = 80W or so/ channel -> 4R) - work out voltage ( = about 18V) .... set 50Hz sine wave input to mixer to read 0dB on meters, then tweak pre-set gain to get 18V of 50Hz at spkr o/p. 50Hz is all I can measure. Dunno how much lower the voltage would be on load - might be able to try 16R or 24R with speakers in series at this power level. Edited by slaz - 06 November 2012 at 2:19pm |
|
REMEMBER....POLITICIANS AND DIAPERS SHOULD BE CHANGED OFTEN AND FOR THE SAME REASON
|
|
abeltronics
Registered User Joined: 16 July 2009 Location: Norwich Status: Offline Points: 265 |
Post Options
Thanks(0)
|
One way to do it is to compare the output voltage with the amp's supply rail voltage. That's probably easiest from a retrofit perspective. The maximum output voltage of an amp is limited by its supply rails. When the input signal tries to get the amp to go beyond the supply rail, it can't, so it clips, or the output flattens off at near the supply rail voltage. So if you have +/- 80V supply rails and the amp wants to output 90V, it can't just magic the extra 10V out of thin air, so it will flatten off at near 80V. (Yes, there is a voltage drop across the output devices so it'll be nearer 75V).
Comparing the signal to the rail voltage, both positive and negative, is a good approach. When the output signal gets to within 10% of the supply rails, chances are it's clipping or near as damn it. This also takes into account supply rail variations and works well. Here's a good circuit: http://sound.westhost.com/project23.htm
|
|
http://www.abeltronics.co.uk - Electronic Engineers
|
|
Muckerbarnes1
Old Croc Joined: 20 March 2010 Location: Stroud Status: Offline Points: 2654 |
Post Options
Thanks(0)
|
There are still to many variations to take into account for the above to be correct. This is a good general rule only.
The state of amp, PSU, load and o/p devices would have a large variation. Clipping problems would also include the front end gain circuit. A correctly designed and implemented clip circuit can be really simple if the topography and design allows. Mik's circuit for a 'proper' amp with a linear o/p stage and stiff psu is suitable. Edited by Muckerbarnes1 - 09 November 2012 at 1:43am |
|
Billy Dawg.
|
|
Post Reply | Page 12> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |