Hi All,
I have been building my guitar pedals using a mixture of polyester film caps for NF, N0G ceramics for PF, and Electrolytic for large UF (power filtering). I want to build guitar pedals and preamps for personal use that will last forever, and so am thinking about using MLCC when possible for larger UF values.
I see this MLCC offering from TDK on mouser that has 47UF with 25v rating.
https://www.mouser.com/ProductDetail/TDK/FG26X5R1E476MRT00?qs=vNwBHymccZ/EuAwGgXhUgQ==
I like this because I can buy in bulk without worrying about them expiring and they should basically last forever. I have heard ceramics are bad for audio path due to piezo, but am I correct to assume its okay for the large value power supply filtering with some smaller film caps thrown in as well?
Thank you!
I have been building my guitar pedals using a mixture of polyester film caps for NF, N0G ceramics for PF, and Electrolytic for large UF (power filtering). I want to build guitar pedals and preamps for personal use that will last forever, and so am thinking about using MLCC when possible for larger UF values.
I see this MLCC offering from TDK on mouser that has 47UF with 25v rating.
https://www.mouser.com/ProductDetail/TDK/FG26X5R1E476MRT00?qs=vNwBHymccZ/EuAwGgXhUgQ==
I like this because I can buy in bulk without worrying about them expiring and they should basically last forever. I have heard ceramics are bad for audio path due to piezo, but am I correct to assume its okay for the large value power supply filtering with some smaller film caps thrown in as well?
Thank you!
Just use decent name-brand 105degC-rated electrolytic with a rated life of 3000hrs or more.
Very cheap, widely available; and lifespan estimation closely follows Arrhenius equation: so even at 35degC in the enclosure, vs 105deg C rating: life expectation to ESSR doubling will be about 2^7 x 3000 = 128 x 3000 = 45years of constant use... good enough..?
Very cheap, widely available; and lifespan estimation closely follows Arrhenius equation: so even at 35degC in the enclosure, vs 105deg C rating: life expectation to ESSR doubling will be about 2^7 x 3000 = 128 x 3000 = 45years of constant use... good enough..?
You are correct in thinking that MLCC are in principal OK for bypass, but not good for signal path. NPO/COG can be superb, as some extremely low THD oscillators use them for frequency determination.
But beware--- high K dielectrics can exhibit horrendous drops in capacitance with increasing applied voltage, by a factor of 5 or even almost 10. Some brands/series are better than others. The data is often not easy to discover on the manufacturer's web site--- you often have to search for it, and sometimes it's not even disclosed. The same mechanism is responsible for signal distortion.
I had to characterize MLCC caps for switching power supples and approve only specific mgr's part numbers. I would resort too measuring by hand: put two candidate caps in series to the cap tester and bias the junction of the caps via a high resistance to bench supply. The voltage sensivity was often astonishing!
But beware--- high K dielectrics can exhibit horrendous drops in capacitance with increasing applied voltage, by a factor of 5 or even almost 10. Some brands/series are better than others. The data is often not easy to discover on the manufacturer's web site--- you often have to search for it, and sometimes it's not even disclosed. The same mechanism is responsible for signal distortion.
I had to characterize MLCC caps for switching power supples and approve only specific mgr's part numbers. I would resort too measuring by hand: put two candidate caps in series to the cap tester and bias the junction of the caps via a high resistance to bench supply. The voltage sensivity was often astonishing!
Tantalum capacitors work well in that range.
I know that people don't like tantalums, but the choice is between bad and worse. 😉
Ed
Thank you all. I was worried about the DC bias issue for MLCC... and it seems not always transparent unless testing yourself like BSST said! I will look into Tantalum.
Yes perhaps Tantalum is better for my use case. Basically, I like to collect surplus electronic components and I feel uneasy collecting electrolytic because I don't know how long they have been on the shelf drying out. Tantalum seems to not matter for shelf life, and doesn't suffer from the DC bias degrading capacitance. All of the boards I will be populating have the polarity clearly marked, so that drawback of Tantalum doesn't bother me.
Yes perhaps Tantalum is better for my use case. Basically, I like to collect surplus electronic components and I feel uneasy collecting electrolytic because I don't know how long they have been on the shelf drying out. Tantalum seems to not matter for shelf life, and doesn't suffer from the DC bias degrading capacitance. All of the boards I will be populating have the polarity clearly marked, so that drawback of Tantalum doesn't bother me.
For tantalums, make sure the rated voltage is at least TWICE the voltage they'll operate at. Reference:
https://web.archive.org/web/2015101...Derating Guidelings for Tantalum 2011 (3).pdf
I think I'd stick with electrolytics using Martin's recommendations.
https://web.archive.org/web/2015101...Derating Guidelings for Tantalum 2011 (3).pdf
I think I'd stick with electrolytics using Martin's recommendations.
The more you derate electrolytic capacitors in either or both DC voltage rating and temperature, the longer they will last.
The X5R is a mediocre ceramic dielectric.
The X5R is a mediocre ceramic dielectric.
Just try to buy some 47 uf ceramic caps. Quite available in quantities of 100000. Just wait 26 weeks for the factory to schedule your run.
I found some 10 uf 50 v Aerovox CPO caps in stock at newark 16 years ago, $7 apiece on closeout. Bought 10. Used them to replace some tantalum caps in a Hammond organ that were failing at 2 years of age (by the production dates; some had been replaced in warranty).
Never seen them again in stock, anywhere.
I buy e-caps with service life >3000 hours, from reputable vendors Panasonic, nichicon, rubicon, Vishay, Kemet. Only from authorized distributors. I have not had to replace any of the 800 I've installed since 2008 again. When I was buying TV parts store shelf caps like Sprague atomlytic or beigh CDE, they would wear out in 3-4 years at 2000 hours a year in my ST70. That amp had 4 sets of electrolytic caps.
I found some 10 uf 50 v Aerovox CPO caps in stock at newark 16 years ago, $7 apiece on closeout. Bought 10. Used them to replace some tantalum caps in a Hammond organ that were failing at 2 years of age (by the production dates; some had been replaced in warranty).
Never seen them again in stock, anywhere.
A complete fantasy. True maybe of the aluminum surface. Crap caps have red rubber or worse as a sealant, can leak out the water in weeks or months. I bought a Acme DTV converter from Wal-mart that had wavy lines the first 5 minutes the moment I opened the box. Ie PS caps were expired in the box at the warehouse.
I buy e-caps with service life >3000 hours, from reputable vendors Panasonic, nichicon, rubicon, Vishay, Kemet. Only from authorized distributors. I have not had to replace any of the 800 I've installed since 2008 again. When I was buying TV parts store shelf caps like Sprague atomlytic or beigh CDE, they would wear out in 3-4 years at 2000 hours a year in my ST70. That amp had 4 sets of electrolytic caps.
Last edited:
So we learn nothing is ever lasting. The bigger MLCC sizes can crack mechanically.
This might be relevant if you frequently drop the pedals:
I recently repaired a professional condenser microphone by replacing two multilayer ceramic 1206 SMD capacitors. Someone had dropped the mike and everything survived the drop but one multilayer capacitor. It turned into a kind of short circuit, but one with very poor contact. For reasons of symmetry, I also replaced its colleague on the other side of the balanced output stage of the microphone.
Edit: I see bucks bunny just brought up the same issue.
I recently repaired a professional condenser microphone by replacing two multilayer ceramic 1206 SMD capacitors. Someone had dropped the mike and everything survived the drop but one multilayer capacitor. It turned into a kind of short circuit, but one with very poor contact. For reasons of symmetry, I also replaced its colleague on the other side of the balanced output stage of the microphone.
Edit: I see bucks bunny just brought up the same issue.