SMD Film vs Ceramics

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

sonolink

Well-known member
Joined
Feb 15, 2010
Messages
1,279
Location
London-Madrid
I'm building this pedal that is a bit tight and tricky to get into a 1590BB enclosure so I´ve been considering going SMD for this project. At first I thought just the resistors but then I thought why not everything I can?

So my question is: would there be a real noticeable difference in using ceramic SMDs instead of TH Films in the audio path? The pedal is a tube preamp for guitar and bass (based on the Alembic F2B)? Traditionally I've kept myself away from TH ceramics in the audio path cos I was always told they're noisy (and not in a nice way). But I'm also told that that's an old myth.....opinions?

Thanks in advance for your input, suggestions, time and help :)
Cheers
Sono
 
noticeable difference in using ceramic SMDs

There are very large differences in materials which are all lumped generically under the "ceramic" description. The primary description of the dielectric used describes the temperature coefficient, i.e. how much the capacitance changes with temperature, but what matters most for audio performance is how much the capacitance changes with applied voltage (voltage coefficient). The good news is that the dielectric with lowest voltage coefficient is also the dielectric with lowest temperature coefficient. The bad news is that the trade-off for that performance is low volume efficiency, i.e. for a given physical size you cannot get much capacitance.

The dielectric you want to look for is C0G or NP0 (equivalent designations). Note that is a zero, not a letter "O" but you sometimes see them mis-written as COG or NPO. If searching on a distributor site like Digikey or Mouser (err, I see you are on other side, so Farnell?) they probably have it correct, to be sure to type zero not O.

What kind of values do you need? For values near the upper end of what is available it may be less expensive to put multiple lower values in parallel rather than try to get a single device with as much capacitance as you need. Some vendors even offer that as a single part number, multiple capacitors pre-mounted onto small metal leads, but probably less expensive for you to just solder them in parallel yourself.
 
Thanks a lot for your quick reply ccaudle and John Roberts :)

The values I'm using are 120pF up to 2.2uF at 250V, and 4n7 up to 330uF at 25v
I'll check the NP0/C0G caps

Thanks!!
 
You won't be able to get those higher capacities in NP0/C0G dielectric, but then you weren't going to be able to find 330uF in a film cap either.
Wait, I checked and it turns out you actually can get 330uF polyester capacitors. 25mmx45mmx58mm and around US$90. You were going to put that in a guitar pedal?! Or you were just hoping that SMD ceramic caps could replace your aluminum electrolytic caps as well?
 
I'm building this pedal that is a bit tight and tricky to get into a 1590BB enclosure so I´ve been considering going SMD for this project. At first I thought just the resistors but then I thought why not everything I can?
For simple circuits (diy pedals usually are) I would stick with through hole for signals just because it's easier to build and service. Use machined sockets for ICs if it's vaguely possible that they might get zapped. It's a rare thing to build anything and have it come out perfect the first time or not accidentally destroy it with reverse supply. The exception would be non-signal stuff that already necessitates SMD like digital. But that too is mostly optional. What is usually not optional is using SMD for protection circuitry on your ins and outs. Parts like small inductors, protection diodes and ceramics used in RF filtering and protection are frequently only available in SMD.

So my question is: would there be a real noticeable difference in using ceramic SMDs instead of TH Films in the audio path?
In theory, yes. Ceramic capacitors are usually microphonic (vibration will cause a change in voltage that will be imparted into your signal) and vary in capacitance depending on bias voltage. But as others have said, for lower capacitance values, C0G / NP0 ceramics are near perfect capacitors and highly available.

The pedal is a tube preamp for guitar and bass (based on the Alembic F2B)? Traditionally I've kept myself away from TH ceramics in the audio path cos I was always told they're noisy (and not in a nice way). But I'm also told that that's an old myth.....opinions?
If it's all tube, presumably the impedances are high enough that you could use film. For less than 1n, I would use C0G / NP0 ceramic.

However, if you're building a vintage design, I would recommend that you find out everything about the type of parts original units used and try to copy that within reason. If something used film, use film. If something used ceramic, use ceramic and use the same type of ceramic (e.g. not C0G / NP0). If you're doing a "pro-mod" where you take a vintage thing and tweak it with modern parts, that just doesn't make much sense to me. That sort of "X is cool but I'm going to make it better" sort of philosphy is wrong headed.
 
My two cents is that for small value caps, I have not noticed sonic differences, in guitar amp/pedal applications. I have compared ceramics vs mica vs polystyrene, for small values (think treble cap in a tone stack) and the difference was underwhelming. Take that for what it is worth........your ears may be more particular.
 
You won't be able to get those higher capacities in NP0/C0G dielectric, but then you weren't going to be able to find 330uF in a film cap either.
Wait, I checked and it turns out you actually can get 330uF polyester capacitors. 25mmx45mmx58mm and around US$90. You were going to put that in a guitar pedal?! Or you were just hoping that SMD ceramic caps could replace your aluminum electrolytic caps as well?

Oooops you're right, that was an electrolytic....just reading too fast through the BOM hehehe. My top value is actually 1uF/25v!!
 
My top value is actually 1uF/25v!!

You can definitely find that, but it will likely be cheaper to put a few in parallel rather than paying for a 1uF C0G cap. You will just have to check current prices for the various capacities (and compare to film cap prices just in case it turns out better to just use film).
Also, a weird thing about ceramic caps (OK, not so weird, probably applies to everything at this point) is that the price is very dependent on volume, so if you need a 25V rated cap, don't just look at 25V caps, look at everything from 25V up to 100V, you might luck out that there is some high volume use for a higher voltage rated part and it is actually cheaper than a 25V rated part.
 
FWIW back in the bad old days when npo/cog were not so widely available I routinely scaled audio band filters so I could use inexpensive film capacitors (like polyester) that didn't suck.... For DC blocking I used electrolytic caps with the pole frequency set below elephant hearing cut offs.

JR
 
I've been lurking for a while but this thread caught my eye. Gosh I'm a nerd.

I've used SMD film caps for some scientific stuff at work with excellent results. As mentioned above, we try to only use NP0 or c0g ceramic in the signal path. This is for low noise preamp stuff similar to audio.

Component selection is often a design choice in audio land- the lowest leakage, lowest distortion, lowest ESR capacitor isn't necessarily the best sonic choice, especially when subjective opinion and valve overdrive are part of the equation! When I'm building a gain stage for data collection at work, precision and lowest possible distortion are my primary foci, and I tune a circuit with an HP dynamic analyzer and a distortion analyzer. When I'm building a pedal I use my ears! Mic preamps fall somewhere in the middle, for me. Everyone probably has their own opinion and experience.

The alembic circuit is really simple, though. I have a tube pedal build on my bench right now and after experimenting with various configurations I'm using a DC-DC for the higher voltages and running the filaments (and pedal overall) on 12v. Current draw of the filaments with a 12a_7 is nice and low, and the cheap nixie power supply generates a surprisingly quiet 200-300v. The design uses two footswitches with the tube nestled in-between. Everything looks pretty tidy on my mock-up, but it's not nearly finished yet!

My point in all that is that SMD vs through hole is the least of my concerns on my build. The FB-2 is such a simple little preamp; you could probably wire it with a cute little eyelet or turret board and it would look very clean!

Good luck, post pictures!

Jamie
 
Almost two decades ago, I was consulting on a SMD DSP project. I tried to use SMD film caps for the audio path for better linearity. The cap data sheet claimed they would tolerate reflow oven process temperatures, but our contract manufacturer still managed to kill/trash a bunch of caps. Besides value shift some caps were visibly deformed (melted).

The situation for quality SMD caps has improved a bunch since then.

JR
 
Not sure I'm using the quote thing correctly!

My industry is defense related and we've hand soldered a lot of film caps with lead solder- so we've been able to largely avoid the reflow issues! But yeah, solid point, JR.
 
My point in all that is that SMD vs through hole is the least of my concerns on my build. The FB-2 is such a simple little preamp; you could probably wire it with a cute little eyelet or turret board and it would look very clean!
Thanks a lot for your input, Jamie. The thing is that my circuit is just based on the F2B but it has been modified a bit...Here's the final schematic I did. The reason I considered going SMD is to be able to fit everything inside a 1590BB. My first PCB layout had the tube on the outside but after some thought I decided to put it inside, then to be able to use 6NxP tubes too since they're so cheap, and after that I thought how about adding some diode distortion? And then I added a D.I. out,...and a Cab simulator.....so basically it's BASED on an AB763 but it's been modified quite a bit, and especially it has GROWN quite a bit in size :)

I would be very interested in knowing what PSU you are using for your pedal, if you would like to share it :)

Cheers
Sono
 

Attachments

  • F2B proto.pdf
    97.4 KB · Views: 0
Wow, lots of stuff going on there!

I have been using various nixie tube power supplies for tube stuff for years. This is the one I purchased most recently - [SHV5]-5Vin-400Vout(Max) Adjustable HIGH VOLTAGE SUPPLY MODULE-NIXIE TUBE ERA | eBay

Have you breadboarded this circuit? My observation would be that you may find less is more in a design like this. I have also found (with similar designs) that running the jfet section ahead of the tubes is more pleasing. I have also found the jfet stuff to sound more dynamic and "real" running on a higher voltage- maybe 18v? Depends how the stages are biased, and how much interstage attenuation and filtration you're doing.

I have built a few valve overdrives using both 12ax7 and submini tubes. There are some interesting tonal differences and it could be helpful to play with supply voltages and plate/cathode resistors to get different sounds. As an example, a given design with a 6112 rather than a 12ax7, with around 200v supply, sounds crunchier and more aggressive. For the design on question I preferred the submini tube. For low gain/warm stuff the 12ax7 was more my speed.

I would highly recommend breadboarding this stuff before committing to a PCB, but that's just my opinion. Sometimes I'll do a design with turrets or eyelets first, so I can easily rework it.

Jamie
 
Wait a sec- is the jfet part a cabinet emulator? I just took a closer look and a feel silly for not noticing earlier. Disregard my comments about that going first!

Have you ever thought about incorporating an adau1701 for cabinet IR processing? I have one of the cheap $20 boards programmed with a greenback IR and it sounds fantastic. I think I bought the tinysine board and USBi programmer first, then the others to cheaply integrate with effects. And yes, it badly needs the aforementioned high quality SMD caps! The layout isn't terrible, and it has 0.1" headers along the edges so you can stack it with another board.

For several years I have been running a tube preamp into an IR for my "clean" guitar sounds, as a pedal platform. I run all of my stereo effects after the mono IR processing. I could see the adau1701 as a low parts count/small footprint solution for cabinet emulation, and a big sonic upgrade.

The adau1701 is in a lot of commercial products. I'm pretty sure that's the heart of the Suhr PT15IR and the reactive load IR.

Sorry to distract from the main topic, but I thought it relevant.

Jamie
 
I have been using various nixie tube power supplies for tube stuff for years. This is the one I purchased most recently - [SHV5]-5Vin-400Vout(Max) Adjustable HIGH VOLTAGE SUPPLY MODULE-NIXIE TUBE ERA | eBay

Are you sure that PSU would provide enough juice for the whole thing? I'm using a One Spot 9v/1A to feed the whole circuit.

I would highly recommend breadboarding this stuff before committing to a PCB, but that's just my opinion. Sometimes I'll do a design with turrets or eyelets first, so I can easily rework it.

I haven't breadboarded this version of the circuit. I did at first when it was just a tube preamp and built a hand soldered veroboard prototype that sounded great and motivated me to develop it.

Have you ever thought about incorporating an adau1701 for cabinet IR processing? I have one of the cheap $20 boards programmed with a greenback IR and it sounds fantastic. I think I bought the tinysine board and USBi programmer first, then the others to cheaply integrate with effects. And yes, it badly needs the aforementioned high quality SMD caps! The layout isn't terrible, and it has 0.1" headers along the edges so you can stack it with another board.
I could see the adau1701 as a low parts count/small footprint solution for cabinet emulation, and a big sonic upgrade.

That sounds actually very interesting although I'm not sure I would know how to implement it. Can you tell me what board you have? I'd like to take a look and eventually open a new thread about that :)
 
Back
Top