Circuits Designed with Caps In Parallel...Why?

[syscokid]

Something I've been meaning to ask for a long time. What is happening when I see a pair of caps where one is parallel to another cap that is five times smaller in capacitance than the first one?

Carvin Vintage 16 schematic where there are three examples in the preamp circuit, and what's the purpose of the zero ohm resistor in the first example?

Carvin Vintage 16 Schematic JPEG.jpg

[maxkracht]

Usually for better esr, or placing a small decoupling cap closer to what you want to decouple and a large cap where it's more convenient, or because ken fisher did it and someone else wanted to copy. Don't know why in this case, maybe it's a paper town copyright trap sorta thing?

[R.G.]

Something I've been meaning to ask for a long time. What is happening when I see a pair of caps where one is parallel to another cap that is five times smaller in capacitance than the first one?

There are a few reasons that paralleling caps might make sense. Max hit some of them, all right. Splitting for lower parasitics, convenience in local decoupling, even slavish copying or catching slavish copying are all things. There are some variants. Caps of all kinds have not only ESR, but equivalent series inductance (ESL). At some high frequency, the ESL impedance becomes bigger than the decreasing Xc, and above that frequency, you have to use a smaller cap with a smaller ESL to keep things capacitive. This is mostly an issue in RF circuitry, but it has found its way into audio too. Sometimes it's for mechanical layout reasons, or economic: two smaller caps, either side by side or separated, may just be handier to fit. Somtimes it's because the designer needs a value between the stock values. caps don't come in as many different values as resistors do. Sometimes it's for selective population. The boards are laid out with spaces for all the values and only the "right" ones are installed; some models may have different values. Sometimes they're on a service schematic but not on the real amp, so if you build from the schematic, it sounds terrible.

what's the purpose of the zero ohm resistor in the first example?

Maybe as simple as allowing the manufacturer to use a single sided PCB; or to put a resistor in that spot in some models for different voicing.

[LOUDthud]

It's probably a PCB design thing where caps of two different sizes are worked into the layout. This allows the Auto-route and Design Rules Check programs to run without flagging the extra parts as errors. Manufacturing can install whatever caps Purchasing can get at a low price that week. Notice how some resistors are marked as zero Ohms and some caps are marked N/U (not used).

[chaccmgr]

the configuration at V1 also allows to make the left capacitor switchable to alter the frequency response of the first stage. You can populate the resistor to take tha cap out in normal operation and a switch can bridge the resistor to make the capacitor effective. Thus, depending on the capacitor value the switch can act as a boost/bass boost of the first stage.

[Stevem]

Using power supply filters in parallel to reach your needed uf also creates a larger current resuvoir .

[mhuss]

All the above are certainly good reasons, however in this specific case I think the first stage is to control the frequency response via the split cathode bypassing, and the second two were where non-stock values were needed, e.g., .0147uF.

[maxkracht]

Optional bypass switch on v1 makes sense to me, but none of them look like they would make an audible-enough difference to justify the cost. I would be curious to see wether a production amp has all of these caps installed. 1uf vs 1.2 uf and .01 vs .012 are pretty close to the upper end of most cap's tolerance; 1.2uf and .012uf are also standard values. If any of these were made switchable, large value alone then large + small value, I think most people would ask, "what is the point of that switch?".

[LOUDthud]

Cost ? You didn't see the note about how the OT was the one designed for 4 EL84s and you use the 4 Ohm tap to drive an 8 Ohm load.

On the other hand, the amp only has three 12AX7s and a tube driven Reverb.

[maxkracht]

Cost ? You didn't see the note about how the OT was the one designed for 4 EL84s and you use the 4 Ohm tap to drive an 8 Ohm load.

On the other hand, the amp only has three 12AX7s and a tube driven Reverb.

Maybe I’ve been looking at too many newer Fenders… I guess not everyone’s value optimization game was on point in the ‘90s. Still seems odd to add extra caps for what is arguably a subtle change, if tone is the reason for the extra caps.

[syscokid]

Something I've been meaning to ask for a long time. What is happening when I see a pair of caps where one is parallel to another cap that is five times smaller in capacitance than the first one?

There are a few reasons that paralleling caps might make sense. Max hit some of them, all right. Splitting for lower parasitics,

Maybe it's this! Can you explain to me what this means?
.
.

I found a couple old pics of my old Carvin Vintage 16, showing the mainboard and how it confirms the schematic regarding the N/U positions and the parallelled caps.
This pic is before any modding:

IMG_4301.JPG

.

This pic is after some modding was done. C35 originally had a .01 and C8 was a .0022

IMG_4346.JPG

[maxkracht]

R.G. wrote: ↑Wed Dec 11, 2024 8:12 pm
syscokid wrote: ↑Wed Dec 11, 2024 5:19 pm
Something I've been meaning to ask for a long time. What is happening when I see a pair of caps where one is parallel to another cap that is five times smaller in capacitance than the first one?
There are a few reasons that paralleling caps might make sense. Max hit some of them, all right. Splitting for lower parasitics,
Maybe it's this! Can you explain to me what this means?

I'll give this a shot, but i'm no EE so grains of salt. Hopefully RG will pipe in and say how wrong I am. I think splitting for lower parasitics is the example I gave of a large cap in one location and a smaller cap as close as possible to the thing you want to decouple. This way there is a local reservoir that manages transients, and a larger reservoir farther away that manages the slower changes. If you don't have a cap close to what you are decoupling, the resistance and inductance of your pcb trace between the larger cap and the thing you want to decouple can cause a voltage drop. I'm not sure how much this matters in an audio circuit, but maybe it has an audible effect in transient response.

I think in this amp it is notable that the pairs of caps are different types. Film and ceramic or film and tantalum (possible exception of c7, not sure if that is ceramic or film). Maybe they are trying to adjust the characteristics of the larger cap in one way or another, not just adding slightly more capacity.

[TUBEDUDE]

Common in hi fi circles, bypassing a larger electrolytic with a film or foil gives faster transients as electrolytics kinda blow. Their only positive quality is high capacity at voltage in a low cost compact form. Every other quality is inferior to any other type of cap. They should be used as a last resort and replaced whenever possible. Especially in signal paths, like cathode caps.

[maxkracht]

Common in hi fi circles, bypassing a larger electrolytic with a film or foil gives faster transients as electrolytics kinda blow.

Indeed. I was talking location differences in the electrolytic and film or ceramic, not the practice of strapping another cap right across an electrolytic, though I guess the esr/esl of an electrolytic would count as parasitics as well. I am curious if/how much this matters with a low esr electrolytic. I guess strapping a film cap across would still lower esr, but not nearly as much. Does this still improve transient response by adding a smaller cap if the esr is already quite low?

[nuke]

There are many reasons why caps might be placed in parallel.

The simplest is that the values add up to what the designer wanted, say you had a wild hair to have 0.01022uf capacitor for some reason, along with a jar of only green M&M's for the Ozzy show.

Sometimes it is cheaper to parallel a bunch of capacitors instead of buying one very big one, or they just fit better in the box.

The Carvin schematic is a little weird that they drew it with two capacitors, when what they intended was to delete one of the two, depending on which SKU they were building. It is a weird way to do it, but I presume they had their reasons for manufacturing this way. Usually, they're just noted in the BOM as option 1 or option 2, but maybe it was easier for them this way.

Then there's paralleling a large electrolytic with a much smaller capacitor of ceramic or some such, because they expect the presence of high frequency trash that's well above the effective frequency range of the electrolytic. Capacitors aren't perfect devices, some larger capacitors have a lot of inductance and thus, aren't effective at high frequencies. It was kind of all the rage in audiophile stuff for long time. Due to the prevalence of switching power supplies, bulk capacitance and high frequency are more friendly with each other today.