X7R MLCC - R U scared of them? Should I be? Should we all?

J

jwin615

Well-known member
What's the consensus of X7R vs C0G in pedals?
I know the audiophile consensus but here's the thing. I have a decent stock of tayda and AliExpress sourced X7Rs because, I have a problem, and also I want to build when I want to/have time to build. I never once thought to myself "I wonder if the electrolyte in MLCCs are different and it matters."
Instead, it was more like "it's a ceramic but less badder for audio."
Then the great and powerful Bean and his wordy build docs reached through the portable document format of the degerator and put the fear of dog in me today. I already didn't like them, to be fair.

Anyway, I do have both an AE and EE background, kinda(tech audio degree and a AAS in EE) so I kinda feel stupid even posting this, but hey, I made an assumption that ceramic sucks, MLCC is like a ceramic Big Mac, better but still sucks. But oh s**t, some Big Macs may be non-linear turd Macs?
Honestly, in fuzzes and such, I don't care. Give me all the non linearities ya got. I'm more worried about the early stages in a 4 stage gain stage, or a digital system or something 'clean'.

So, Fox News me people. Tell me when to worry about the boogie caps coming to take my noisefloor.

Also, I'm not so inclined to keep two types in inventory. I have maybe $20-30 in them. It's worth it to me to donate them to a local youth group and restock if needed to avoid "why is my bonk machine making clown noises when it's hot?" posts. It's a hobby. It's not meant to save money. It's also not meant to deprive sanity either.
 
Last edited:
MLCC are fine for most pedal use, I think.

MLCC's main issues are the variation of capacitance with temperature and with applied DC voltage. The first is not a big factor for things such as pedals that work at room temperature and generate very little heat. The second can be mitigated by selecting caps with a rated voltage much higher than they'll see in the circuit (say, rated at 50V for use in a 9V pedal), and also by selecting the largest package size that will fit (i.e. 1206 SMD will do better than 0805).
 
Thank jtex. One of your posts from earlier this year was one of the few I could find relevant
I did a bit of reading earlier from what seemed reliable sources and also EE stack exchange;) and overspecing voltage seemed to be the way to go. In turn, in the non-sm world, most are already overrated for our needs. I guess the other, further, question is are the taydaexpress ones I have on hand good enough beyond noise boxes?
Also, got the well packed preamp in. Thank you! Hoping to get to installing it and a new bridge humbucker before work travel kicks back up. The whole package in that guitar needs an overhaul though so I may just wait and bite the bullet and get a switch and jack to go with it. It was well abused by it's previous owner. Have all the playability issues sorted now so on to the inards
I don't care for working in(most) cavities for extended time so maybe KO it all at once.
 
Nothing to add on the STEM side, but Joe Gore (a musician with a formidable ear) reported out a test of capacitors several years back, on freestompboxes.org. It was an interesting discussion of interesting data. Here's the link--I think you need to be logged in:

Login

tracing pedals since 2007
www.freestompboxes.org www.freestompboxes.org

My apologies if FSB is still colloquium non grata. :unsure:
 
derevaun said:
Nothing to add on the STEM side, but Joe Gore (a musician with a formidable ear) reported out a test of capacitors several years back, on freestompboxes.org. It was an interesting discussion of interesting data. Here's the link--I think you need to be logged in:

Login

tracing pedals since 2007
www.freestompboxes.org www.freestompboxes.org

My apologies if FSB is still colloquium non grata. :unsure:
Click to expand...
Al Gore, from my home state and creator of the AOLnets told me personally that FSB is cannon.
Thanks.
 
derevaun said:
Nothing to add on the STEM side, but Joe Gore (a musician with a formidable ear) reported out a test of capacitors several years back, on freestompboxes.org. It was an interesting discussion of interesting data. Here's the link--I think you need to be logged in:

Login

tracing pedals since 2007
www.freestompboxes.org www.freestompboxes.org

My apologies if FSB is still colloquium non grata. :unsure:
Click to expand...
Okay, well, that's cool. But it's not very scientific. He did use a preamp box at least...
But that's 15+ yo tech that only averages eq curves. If he'd used SMAART or something more advanced (with white/pink noise at various levels)I might buy it but it's just looking at overall tonality as an average. It looks 'cool' but that's not what that tech was designed to do. You could probably plug 8 different tubescreamer clones in and get them to "null' with that.
Thank you for the contribution to the conversation though and please don't take this as any harshness. I remember using the tonal matching stuff back then. It was just kinda dumb. It tried to match EQ curves between recordings but it didn't take any harmonics or distortion into account. But it was amazing "tech" for about 5 months.

It's kinda like if you fed AI the two datasets, but the only two things in the data sets were singular samples at singular levels. It would output, something. But not necessarily relevant info.
I'd be more interested to see frequency response of the two recorded.andmixed at equal levels with one phase reversed. Then that repeated over multiple spectrums of gain structure. That's when you'd see harmonic difference (non-linear garble). Oh, the. Throw in temperature.

Again, that's for the comment though. And for listening to my rant that probably came off wrong that I'm also too tired to proofread.
But I need to sleep knowing I'm safe in this house with these MLCCs.
 
I take no offense at that! IMHO the interesting part is the discussion that ensued. FWIW Joe also posted a listening test on his site. I trust his ear the way I trust James Hoffmann's palate, and they both tend to argue that some distinctions don't amount to a difference in the outcome.

Back on topic, I share the apprehension toward MLCCs, especially the cheap ones I have a bunch of. I mean, but capacitors are so variable they're practically agricultural products. I don't buy that there's a meaningful difference, but the superstition is kind of fun. I'm not going to take a shit on my friends' insistence that 92 degree water vs 96 degree water affects the flavor of coffee in an Aeropress, so Ima play along with the cap thing too.
(Not implying anyone is taking a shit on anyone's anything here, btw).
 
If you want to do a deep dive, there's an archive of Cyril Bateman's extensive articles in Electronics World on capacitor distortion. He was an EE who worked in capacitor manufacturing if I'm remembering right. In section three we have the test results for ceramic capacitors where he measures 0.00006% THD in his test setup with COG ceramic vs 0.15% with an X7R. Obviously the level of distortion acceptable is circuit and context dependent and Bateman was writing for a slightly different audience than us pedal builders. Douglas Self also discusses Bateman's experiments in his chapters on components and noise sources in Small Signal Audio Design.

My 2¢ on X7R and non COG MLCCs is how I think of most "non-ideal" components, ie, for commissions I usually splurge on components and for my own builds I'm ok using whatever, as long as I know that I am. We're not making high fidelity mixing boards here. I'm also kind of partial to the convenience of X7Rs specifically because they're easy to find in axial packages that I can solder in the same round as resistors and diodes.
 
Last edited:
finebyfine said:
If you want to do a deep dive, there's an archive of Cyril Bateman's extensive articles in Electronics World on capacitor distortion. He was an EE who worked in capacitor manufacturing if I'm remembering right. In section three we have the test results for ceramic capacitors where he measures 0.00006% THD in his test setup with COG ceramic vs 0.15% with an X7R. Obviously the level of distortion acceptable is circuit and context dependent and Bateman was writing for a slightly different audience than us pedal builders. Douglas Self also discusses Bateman's experiments in his chapters on components and noise sources in Small Signal Audio Design.

My 2¢ on X7R and non COG MLCCs is how I think of most "non-ideal" components, ie, for commissions I usually splurge on components and for my own builds I'm ok using whatever, as long as I know that I am. We're not making high fidelity mixing boards here. I'm also kind of partial to the convenience of X7Rs specifically because they're easy to find in axial packages that I can solder in the same round as resistors and diodes.
Click to expand...
This is exactly what I was looking for! Thank you. That response chart in section 2, page 6 is terrifying. That's honestly way worse than I would have imagined. Hope it's okay for me to repost here for the sake of future readers.
Yeah, I can't see myself utilizing these in any "clean" audio path moving forward. In a gnarly fuzz box, maybe just for the added nonlins,.but I'd still question if their additional would benefit the overall tonality. 0.15% THD from a singular passive component is a pretty wild. Even though that measurement is not completely in line with where we typically operate, especially with the higher DC bias, it's still rather indicative of their potential issues.
I'll be reading more into that paper this evening when I have time. Thanks for the link.
Have you read the Self book? Would you recommend it at the $60 price tag, or is there another one you would? I have several, but I circle back to the old Audio Cyclopedia more than most of my others. But it's a little hefty at 1700+ pages and the binding is starting to dry some so I have to baby it a bit. Screenshot_20230802-091415-741.png
 
jwin615 said:
Have you read the Self book? Would you recommend it at the $60 price tag, or is there another one you would?
Click to expand...
I managed to rent some or all of it as an e-book for a week or two, can't remember where though. I thought it was well worth it.

As far as MLCCs, I went through the whole drill when I was working with hand soldering SMD film caps and thought C0G might be a workable alternative in a few filter sections. Sounded fine, but they are still wonky to hand solder as well and I just ported back to through hole film caps, which I mostly spec based on color rather than other forms of mojo.

I do routinely use X7R caps for local bypass at opamp power pins, where I've read that they are the preferred alternative. But I've never bought them in any value but 100nF, so as usual, I know nuthhhink! ;)
 
Last edited:
You must log in or register to reply here.
Back
Top