Hello everybody,
In the LM1875 datasheet I have see it that the power supply capacitors C6 and C7 have 100uF of capacitance.
But I have see in differents schematics that they often to put higher values like 1000uF, or more.
What is the correct value for this capacitors? I supose that if you put higher value the sound will be better, isn't it?
Thanks in advance to all.
Alcuemar
In the LM1875 datasheet I have see it that the power supply capacitors C6 and C7 have 100uF of capacitance.
But I have see in differents schematics that they often to put higher values like 1000uF, or more.
What is the correct value for this capacitors? I supose that if you put higher value the sound will be better, isn't it?
Thanks in advance to all.
Alcuemar
1000uf is pretty std for clones, the 100uf as per the data sheet is minimal, it will give nice treble and mid range but you will likely find the bass is a bit weak, this is especailly the case with the 1875 chips.
The higher the cap value the stronger the bass, however once you exceed 1000uf you will start to sacrifice the treble, if you want it all then you will need to either build a snubberized power supply as per Carlos FMs, you'll find plenty of info on this forum or use SMPS power, then you can easily use 3200uf to 4700uf and it does sound much better in the lower register like this.
I suggest start with 1000uf std clone then experiment from there as your experience builds, in reality the amp module is the cheapest bit, power supplies etc are the pricey parts.
The higher the cap value the stronger the bass, however once you exceed 1000uf you will start to sacrifice the treble, if you want it all then you will need to either build a snubberized power supply as per Carlos FMs, you'll find plenty of info on this forum or use SMPS power, then you can easily use 3200uf to 4700uf and it does sound much better in the lower register like this.
I suggest start with 1000uf std clone then experiment from there as your experience builds, in reality the amp module is the cheapest bit, power supplies etc are the pricey parts.
As always, your target use has to be considered. Low impedance loads or regular impedance and high volume will put more ripple in the supply line. The caps filter this.
Datasheets often show a very basic configuration to get a chip working, a guaranteed minimum, suggested circuit example. It is not a rule, and may find that use of significantly higher value caps will indeeed help with bass but harm the mid and treble some, but particularly above about 1000uF. Also note that this may only be if one only used the higher value caps, had a one-cap-per-rail simple design instead of the more elaborate decoupling seen in many amps.
CarlosFM has proposed and claimed victory over this situation with his snubberized (everything??), psu and amp additions. Others' opinions seem to support this too.
Often what can be done in general is to put some very small, fast ceramic or film caps decoupling as close as possible to the chip supply pins, then some 100uF (or in this ballpark, some choose a little higher value), and put your largest bulk electrolytics further away, whether that "further" be on the same circuit board or on a separate supply rectification board.
In general I would never think about building an amp with only 100uF caps for the power. Even my headphone amps have a dozen times that.
Datasheets often show a very basic configuration to get a chip working, a guaranteed minimum, suggested circuit example. It is not a rule, and may find that use of significantly higher value caps will indeeed help with bass but harm the mid and treble some, but particularly above about 1000uF. Also note that this may only be if one only used the higher value caps, had a one-cap-per-rail simple design instead of the more elaborate decoupling seen in many amps.
CarlosFM has proposed and claimed victory over this situation with his snubberized (everything??), psu and amp additions. Others' opinions seem to support this too.
Often what can be done in general is to put some very small, fast ceramic or film caps decoupling as close as possible to the chip supply pins, then some 100uF (or in this ballpark, some choose a little higher value), and put your largest bulk electrolytics further away, whether that "further" be on the same circuit board or on a separate supply rectification board.
In general I would never think about building an amp with only 100uF caps for the power. Even my headphone amps have a dozen times that.
Large smotthing affects mid/high???
I have read this many times here.
Knowing the shortcommnigs of "human ears", has anyone actually done any measurements or proper scientific blind A/B comparisons to prove that larger capactiance values do indeed affect the mid/treble range in some way.
Surely, if its audible, its measurable ?
Or indeed made any blind A/B or proper measurements of with/without snubbers ?
And again with Linear/Switched mode power supplies ?
Maybe when I get time, I will.
Any real evidence out there apart from highly subjective opinions ?
Not being negative in any way but science is based on asking the ultimate question: WHY?
Frank
I have read this many times here.
Knowing the shortcommnigs of "human ears", has anyone actually done any measurements or proper scientific blind A/B comparisons to prove that larger capactiance values do indeed affect the mid/treble range in some way.
Surely, if its audible, its measurable ?
Or indeed made any blind A/B or proper measurements of with/without snubbers ?
And again with Linear/Switched mode power supplies ?
Maybe when I get time, I will.
Any real evidence out there apart from highly subjective opinions ?
Not being negative in any way but science is based on asking the ultimate question: WHY?
Frank
Hi Frank
I feel sure that the effect of different caps could be measured, though I have no gear to do this, however I have conducted many A/B test with family and friends, and pretty much every one picked the differences, without being told anything about what was done or to what.
Some things are pretty obvious such as cleaner. less grainy treble, but bass can be harder to nail, for example I felt the bass of the 3886 amps was better than my smps 1875s but after several comparisons I found it was just more pronounced but not as tight, I'm not sure how you'd measure that.
On the whole from my experience and I'm no great authority on this I feel that Gainclones are very power supply dependent, in that changes there seem to cause bigger changes in sound that changes to the actual amp module.
I certainly feel also that the only way to really great sound with a clone is to match it to your speakers and intended usage, It was only when I started playing with high efficency full range drivers that I came to realize just how good these amps can be and more importantly how well they respond to all manner of power supply and layout changes.
Overall I could never have refined my clones to the same sound quality level using any of the other speakers I have as they are jsut not revealing enough and I suspect this could be a problem for many cloners.
I feel sure that the effect of different caps could be measured, though I have no gear to do this, however I have conducted many A/B test with family and friends, and pretty much every one picked the differences, without being told anything about what was done or to what.
Some things are pretty obvious such as cleaner. less grainy treble, but bass can be harder to nail, for example I felt the bass of the 3886 amps was better than my smps 1875s but after several comparisons I found it was just more pronounced but not as tight, I'm not sure how you'd measure that.
On the whole from my experience and I'm no great authority on this I feel that Gainclones are very power supply dependent, in that changes there seem to cause bigger changes in sound that changes to the actual amp module.
I certainly feel also that the only way to really great sound with a clone is to match it to your speakers and intended usage, It was only when I started playing with high efficency full range drivers that I came to realize just how good these amps can be and more importantly how well they respond to all manner of power supply and layout changes.
Overall I could never have refined my clones to the same sound quality level using any of the other speakers I have as they are jsut not revealing enough and I suspect this could be a problem for many cloners.
Bazukaz said:
Correct ! this is the only way that meaningful subjective measurements can be made.
This should be done using a couple of relays which are operated from the listening position.
When I get round to it, I will do this with:
Unreg/Switch mode PSU
3 values of smoothing.
Snubbers
If only work didn't get in the way of this!
I still feel that if these subjective changes are so pronounced, then they should be measurable and quantifiable.
The reason I am so interested in this is I want my next project to perform to its maximum ability. (3 way active speaker system)
Frank
there's not too much subjective about picking the power supply capacitors for a chipamp -- as a circuit draws more current from the power supply the voltage sags -- the repeat charging cycles of the filter capacitor reduce the "potential" of the supply -- less potential (voltage) means less music. (you can demonstrate this for yourself with the freeware from Linear Technology or Texas Instruments) --
my suggestion -- for 20 watts out 10,000uF -- use too small a capacitor and those instantaneous peaks just won't be there.
SMPS and linear supplies -- go back a few years to that halcyon time when Ford made two models of the Mustang GT -- both with the same horsepower -- one was a 5 liter, 8 cylinder with racing headers, the other a turbocharged 2.6 litre 6 -- if you wanted to go fast in a straight line one was always superior to the other. Oh,, the turbo charged unit has been discontinued for about 15 years.
my suggestion -- for 20 watts out 10,000uF -- use too small a capacitor and those instantaneous peaks just won't be there.
SMPS and linear supplies -- go back a few years to that halcyon time when Ford made two models of the Mustang GT -- both with the same horsepower -- one was a 5 liter, 8 cylinder with racing headers, the other a turbocharged 2.6 litre 6 -- if you wanted to go fast in a straight line one was always superior to the other. Oh,, the turbo charged unit has been discontinued for about 15 years.
jackinnj said:
I am questioning the (many) posters here that are saying that larger caps somehow degrade the mid/hf performance somehow, why this might be and if it can be measured.
Some might infer from the above that the gain of a chip-amp depends on supply voltage - It doesn't. Supply voltage sag will only limit the maximum output. If the amp is running at much less than max, supply sag will not matter one iota.
True ?
Frank
1) as far as I can determine, no one has provided empirical evidence to support the assertion that larger caps degrade the sound. it is very easy to demonstrate, either by simulation or on the test bench that an inadequate power supply capacitor will cause high levels of THD%frank1 said:
FWIW, National Semi has said recently that the input coupling cap can have a dramatic impact on sound quality.
2) -- the ultimate gain (of any amplifier) is limited by the supply's ability to maintain its potential -- there are two control loops in a chipamplifier -- that controlled by Rf/Rg and that determined by the impedance of the power supply.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.