But they just said it doesnt cause any distortion artifact prior to engagement. Upon engagement, it's distorting the signal, kinda by definition. Would be interesting to do a harmonic spectra for the two cases; "prior to engagement" and "upon engagement" for just the pile of input stage op-amps used to implement, say, at the 0.775V level and 0.875V.
You have to compare leakage at the same reverse voltage and temperature...
https://www.diodes.com/assets/Datasheets/DLLFSD01LPH4.pdf
https://assets.nexperia.com/documents/data-sheet/BAS16_SER.pdf
In fact, leakage current curves are quite similar between the two
https://www.diodes.com/assets/Datasheets/DLLFSD01LPH4.pdf
https://assets.nexperia.com/documents/data-sheet/BAS16_SER.pdf
In fact, leakage current curves are quite similar between the two
Another attempt. An opamp buffer would drive the transistor bases (expressions shown on schematic). When the signal is out of the allowed range, one of the two transistors turns on, becomes a current source, and returns the signal voltage to the clamping limit...
These require a known impedance source, but the amp should have a THAT12xx balanced receiver at the input, so it's okay.
The one on the left could exceed the transistor's Vbeo, so the one on the right adds diodes.
It turns out that the collector of a transistor that is turned off (and bootstrapped) doesn't add much distortion to the signal...
THD in the simulator's noise floor (-220dB) before it begins soft clipping. Ding! We have a winner
These require a known impedance source, but the amp should have a THAT12xx balanced receiver at the input, so it's okay.
The one on the left could exceed the transistor's Vbeo, so the one on the right adds diodes.
It turns out that the collector of a transistor that is turned off (and bootstrapped) doesn't add much distortion to the signal...
THD in the simulator's noise floor (-220dB) before it begins soft clipping. Ding! We have a winner
The way it recovers is different, though. A soft or hard clipper causes gross distortion on the part of the musical waveform that's too loud, a fast-acting limiter only distorts the first part of it and attenuates the rest. If the part that's too loud is too short for the limiter to settle, then it doesn't matter much.
It would be nice if the soft clipper or limiter or whatever could also trigger a monostable multivibrator that drives a clipping LED, so you see that you need to turn down the volume.
Last edited:
Maybe he's just young. When I was in my 20s and 30s I used
to drive the daylights out of my Phase Linear 700. With an amp
that big clipping was insanely loud but still....
G²
I don't need some lights flickering to tell me the music is too loud, my ears already tell me that.
The invention of clipping/limiter indicators is just a marketing adornment.
Besides, as I've mentioned a while back, products are adorned with plenty of superficial lights to make people buy them.
Like mosquitos to a bug lamp.
A 4 db apparent improvement in amplifier power output is nothing to sneeze at, if it's so. If not patented, one would think this would be as ubiquitous as "anti-lock brakes". I wouldnt mind my 20W amp channels having the sonic appearance of, say, 50W by use of a clever circuit - that only engages on those rare occasions when I "turn it up". Once, so far, this year. I'm just old.
Long ago I had a DIY 200W stereo amp but without clip indicator and playing ("soft" clipped guitar) rock music at high level sounded great. But it blew out the tweeters not once but twice. A clip indicator could have saved money and ear wear out as well. So this time I designed a clip indicator, signaling the start of clipping at 1 ppm THD (according to sim). High on the list of projects.I don't need some lights flickering to tell me the music is too loud, my ears already tell me that.
The invention of clipping/limiter indicators is just a marketing adornment.
Besides, as I've mentioned a while back, products are adorned with plenty of superficial lights to make people buy them.
Like mosquitos to a bug lamp.
My amplifier automatically switches off its output relay when the clipping is really bad. The only time that happened (apart from testing it with a dummy load after building it) was when I was playing a DIY nature recording with a lot of wind blowing at the microphone, something like a group of oystercatchers recorded from a 100 metre distance with an ordinary stage microphone. The recording still needed to be high-pass filtered to get rid of the wind sounds, I only heard a lot of loud rumble with very faint bird sounds.
I monitor the error signal of a feedback amplifier to determine if it clips, see pages 140...143 of https://worldradiohistory.com/UK/Wireless-World/90s/Electronics-World-1996-02-S-OCR.pdf Unfortunately that won't work with a soft clipper.
I monitor the error signal of a feedback amplifier to determine if it clips, see pages 140...143 of https://worldradiohistory.com/UK/Wireless-World/90s/Electronics-World-1996-02-S-OCR.pdf Unfortunately that won't work with a soft clipper.
Last edited:
Good idea to switch off output. In the course of time I accumulated a variety of music with a "too wide for comfort" difference in output so either the peak level has to be equalized, or a limiter has to be used followed by lowering volume . With a DSP that would be fairly easy, reading a song in memory, setting the required peak level and playing it. No idea if such an application already exists.
There is a free program called MP3Gain that uses a rough psychoacoustic model to make the apparent loudness of MP3 files roughly equal. It does so by adjusting scaling factors without decoding and re-encoding the audio. Its disadvantage is obviously that it only works for MP3 files.
Apparently there is something called ReplayGain meant for other file types. I don't know anything about it, I only heard the rumour that it exists.
Apparently there is something called ReplayGain meant for other file types. I don't know anything about it, I only heard the rumour that it exists.
Thanks, if such an application exists there might be more so a search could be worthwhile. I've been using wma since it became available and also have much in wav format. The difference can be heard to the extent that I won't use conversion.
It's not difficult to make a very fast compressor-limiter that activates in the first near-clip audio half wave. The basic JFET shunt is typical. A LED driven LDR shunt also has the right attack-release timing and is very simple. Years ago we hung a bridge+LED+LDR+2 resistors on a DJ monitor amp and the effect was barely noticeable. It prevented the DJ at a roller rink from destroying his monitors.
You should be able to clean up the clipping in the power amp with Baker clamps and VAS current limits. You can also add clamping feedback as some have done with old "Tiger" amps.
Bootstraps should be limited with a resistor in series with the bootstrap capacitor or a limited bootstrap voltage by reducing the lower/inner resistor.
You should be able to clean up the clipping in the power amp with Baker clamps and VAS current limits. You can also add clamping feedback as some have done with old "Tiger" amps.
Bootstraps should be limited with a resistor in series with the bootstrap capacitor or a limited bootstrap voltage by reducing the lower/inner resistor.
Last edited:
...on 40mV output voltage. For these circuits the voltage has to be tiny, so SNR suffers...
It's a tag that contains the gain setting the file should be played at, calculated from the average loudness. If you add the tags to your flac library (foobar can do this automatically) then enable ReplayGain in preferences, then it will set the volume accordingly on playback. It works very well. That way, you can make a playlist with well mastered music (high dynamic range) followed by modern compressed to death trash without having to rush to turn down the volume... Very convenient. You can set the gain for the whole album, or track by track.
For example it will turn down LP "Lost on You" by -8dB relative to Dire Straits "Telegraph Road", doesn't make LP sound any better, it was destroyed in mastering which is a shame, but at least it doesn't make ears bleed...
lol, questioning motives?
It's simple: it clips rarely. The reason I want good clipping behavior is that I can then use a lower power supply voltage, which allows to increase the idle bias current for the same idle (wasted) dissipation, and lower distortion at all volumes. Basically clean clipping improves how it sounds at low volume (when it isn't clipping at all) by shifting the forced compromise on bias current versus idle dissipation towards higher bias.
Yeah but I don't want to lower the volume after a peak.
Clamping feedback doesn't work well with multipole compensation.
Your system was obviously mismatched to your speakers.
A common thing.
And your volume levels were unreasonable and a danger to your hearing.
But to profess to me and others that you need some sort of visual alert tells me that your careless useage with your equipment is not the equipment, but your own fault.