Well, those old caps probably have enough ESR to damp the LC resonance formed with the wire inductance and other caps on the amp board... replacing all the caps with ultra-low ESR models could cause problems though.
The purpose of those caps is to supply the amplifier when the rectifier diodes are blocked... it is energy storage, not filtering ; even with some pretty high ESR for a capacitor that size, say 0.1 ohm, well, if a 10 amp current is drawn, the voltage will sag 1 volt, not really a huge problem, ripple is probably much more than that at high power anyway...
Same for inductance... the problem with inductance of these wires is not the added power supply impedance (at audio frequencies) unless the wires are very long, but LC resonances with other caps.
For supply filtering and local decoupling : at HF you want the cap to have the lowest possible impedance, thus the lowest ESR/ESL, which in turn requires low inductance mounting to avoid ringing, hence, planes.
Unfortunately we aren't usually able to fully model the subtle differences, so a lot of intuition and opinion goes into some audio gear. For example signal capacitors that all seem to measure identically, sound drastically different.
Did bypass caps ever do anything? Probably, since electrolytics use to be pretty poor quality.
(For Class-d)
Yes use SMD ceramics as close to load in all circuit (or where voltages makes this practical). I now use 1 uF for all low voltages.
Do you mean to bypass these? ... if not done correct, it will lead to worse performance, so in most cases leave it at only one cap per decoupling point and use same value and type for the whole line
Yes 4 layers woud be better than 2, but I think you can get ok/good performance with just 2 if care is taken
/Baldin
hi , for me the role of a bypass cap is because i have failed somewhere and i'm not happy with the result. so i try small styroflex 10nf , if it 's still bad replace it with 100nf MKP , then 1uf until i find the value that sounds good to me. sometimes it works , sometimes it 's worst.
but ir's an easy and cheap tweak .. so why not playing? it"s jut Diy audio
but ir's an easy and cheap tweak .. so why not playing? it"s jut Diy audio
Bypass capacitors are critical for most designs, but it is critical they are used where they are supposed to be used. For digital circuitry (and class D amplifiers) small value (>100nF) low inductance packaged decoupling capacitors next to power pins are obligatory, for analogue the same goes especially on op-amp circuits.
Bypass caps next to big reservoir caps don't help and as many have said are likely to cause more problems than they fix.
Bypass capacitors supply the instantaneous power when any device is switching, the higher the switching frequency the more critical the bypass requirements.
There is a host of information on the web regarding bypassing (for both analogue and digital) from capacitor manufacturers, chip and active device companies and other interested parties, this is one field (Power Delivery Integrity) where there has been a lot of research over the past few years as circuitry becomes more demanding power wise; for both efficiency and noise reduction, and correct bypassing is one of the critical aspects of this.
Bypass caps next to big reservoir caps don't help and as many have said are likely to cause more problems than they fix.
Bypass capacitors supply the instantaneous power when any device is switching, the higher the switching frequency the more critical the bypass requirements.
There is a host of information on the web regarding bypassing (for both analogue and digital) from capacitor manufacturers, chip and active device companies and other interested parties, this is one field (Power Delivery Integrity) where there has been a lot of research over the past few years as circuitry becomes more demanding power wise; for both efficiency and noise reduction, and correct bypassing is one of the critical aspects of this.
Also they need to study the required layout recommendations for these type of high current switching circuits. Most SMPS and Class D control devices have quite comprehensive layout guides, these should be followed, but if not there are basic techniques that should always be followed to minimise noise coupling to the rest of the system.
Where do people come up with this garbage? The old cap's ESR is most likely incredibly low. When did YOU last actually measure the ESR of any actual amplifier cap? As for your guess of 0.1Ohm- you're not even close- they are 10 to 50 times lower than that. Do you have an ESR meter? I'd say not.
I built a mockup of this guy's circut and scanned it with the network analyzer.
Yeah, I'd need one for quickie measurements and testing caps... But the network analyzer (although a bit more annoying to use) gives much more information.
If you got some old caps with "10 to 50 times lower ESR" (ie, 2-10 mOhm) and which look like karl's caps (ie, PCB mount, not Coke cans with huge screws), good for you ! You can post some pics and measurements and brag about it
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.