I'm sure this will be a stupid question.

[Rp3703]

Hi,

I’m building a 100 watt 800 clone and after looking at pretty much every version I can find, all of them use the same series/parallel wired dual 50/50 multistage capacitors. Is there not a way to just use one 100/100m cap instead of two 50/50’s?

Thanks,

Richard

[HotBluePlates]

... all of them use the same series/parallel wired dual 50/50 multistage capacitors. Is there not a way to just use one 100/100m cap instead of two 50/50’s?

I don't know offhand if there is such a beast as a dual-100µF cap. But if there is, the - end of both caps are connected, meaning you cannot put them in series; the only available connections are each section separate or both sections in parallel.

So that's why the 2 cap cans. If you could find 2 caps of 100µF each, you could place those in series, but you're still using 2 devices and not 1.

[AZJimC]

Is there not a way to just use one 100/100m cap instead of two 50/50’s?

Yes. The reason for the two caps in series is to utilize lower voltage caps. You can use a single 100mfd as long as you do not exceed the voltage rating on the cap. They use the 350v in series giving a 700v capacity to the pair. 700v caps are expensive and hard to find. If your amp has less than 500v B+ you can use a 100mfd/500v or 525v to replace the two with one.
 That is, if I'm understanding your question correctly.

A 100mfd would be much greater than the two 50's in series however. They make 25mfd, at 700v. For series caps their voltage is additive, but the capacitance is a reciprocal of total capacitance. Thus
1/mfdTotal = 1/50mfd + 1/50mfd, or 25mfd.
 the voltage handling capacity is what doubles when caps are in series.

In other words, if you want the same function, use a single 25mfd that has a voltage rating greater than it would be exposed to in the amp. A 50mfd at 525v would be double the capacitance, and less than 700v voltage handling, but if your amp is running at 450v, then a 500-525v (common values) would be okay.

[jjasilli]

There is another aspect here, though subtle.  Small caps filter better than large caps; and more caps filter better than fewer caps.  This is because caps have inherent physical flaws in their operation; and large caps display more of those flaws.  A series-parallel quad of 50mF caps results in the value 50mF of filtering, at 2X the voltage rating of one 50mF cap.  Not only do you get 2X the voltage rating, but there's also more filtering going on:  The same nominal mF value reduces ripple more, which can have an effect on tone.

[HotBluePlates]

For Marshall, I think the bigger factor was the use of the 50µF+50µF can across their product line.

It would be cheaper to use the common 50-50 part in higher bulk than to use 100µF caps for a potion of their products, than to buy another part in smaller quantity (less bulk-order discount) with differing dimensions (tolling/assembly line changes) and mounting hardware (again, low quantity & higher price).

[jjasilli]

Yes, I agree that component choices for production amps often relate to economic factors.  But multi-cap filtering, or "stacked" caps in the signal path, are options for diy or boutique builds.

[Ken Moon]

In almost all the gut shots using the stack of parallel filter caps with balancing resistors, the builder used very large Mallory or Atom caps, like in this photo:


I like the smaller (.7 x 1.4 in) BC 47uF/450V axial caps, but I'm starting to wonder if anything is wrong with them for this application.

Here is a pic - I put an EH 12AX7 next to it so you can compare the size:



Here is a link to the datasheet:

http://www.vishay.com/docs/28329/041042043ash.pdf

Are these OK? I'm about to use them in a Marshally amp 

[jjasilli]

IMHO they're all OK.  Lot's of very old threads in this Forum.  Some of the new production large-size caps are mostly empty shells, with smaller caps inside!   Photo's have been posted to prove it.  Caps are an analog device - so a larger value cap must be larger in physical size.  However, cap technology has improved so today's caps are physically smaller than they used to be, with higher voltage ratings than in the old days. 

[DummyLoad]

caps are not all created equal. look at the spec sheets. you want devices with longest rated hours at the highest operating temperature with the lowest possible ESR values. caps with decent specs of all three can be very expensive. 


--pete

[jojokeo]

While on the subject of reservoir & filter caps I forget why a smaller non-polar cap is sometimes used in parallel? Is this because of it's ability to filter very high frequencies that the larger ones cannot? And is it parallel to the very last filter cap in the string which feeds the V1 stages? Lastly what is the commonly used value? Thinking it's around 0.1uF or so?

[jjasilli]

Dummyload is right.  But I'm assuming you'll source guitar amp caps from a reputable supplier of guitar amp parts, like Doug, e.g., so this should be OK.


JoJo:  Because large caps-- like any B+ filter cap -- have relatively poor ESR, they can benefit from being bypassed by a small cap.  Personally, I think this improves the tone of an amp, but it's subtle.  See my Vibrochamp & Princeton schematics which are posted:  a small cap bypasses the last filter cap - which feeds the first preamp stage.  The first stage substantially sets the tone for the amp, affecting everything happening downstream.  That's why that stage is the one to bypass as stated.  Hi-fi diy'ers may use such bypass caps for multiple or even all B+ filter stages.

[Rp3703]

Wow, this is the first forum I've posted on and received this many responses. For those who don't know, JJ makes both the 50/50 at 500V and a 100/100 at 500V. I guess it makes sense to use two 50/50's if you're trying to achieve a higher voltage. My reason for asking was mainly just to save space. The 2203 reissues use six 50/50's in them. That's a lot of holes to drill. Thanks again for the responses.

My new problem is trying to justify building an amp from scratch when I just picked up a used B-52 AT100 for $180. I've spent close to $400 on parts to build an 800 clone and I still have to put it together.

[HotBluePlates]

While on the subject of reservoir & filter caps I forget why a smaller non-polar cap is sometimes used in parallel? Is this because of it's ability to filter very high frequencies that the larger ones cannot?

Check out CDE's Aluminum Electrolytic Cap Application Guide. Turn to Page 8; there's a graph of cap impedance vs. frequency.

The cap's impedance (which you'd like to be 0Ω for perfect filtering) drops to a low where it equals the cap's ESR, but rises again above 20kHz due to the cap's series inductance. This is unavoidable and a byproduct of the physical realities of making a cap, though electrolytic caps might be worse in this regard than some caps with a film dielectric.

The impedance rise is beyond the range of hearing and guitar/speaker note reproduction. However, supersonic signals (noise, oscillation) see what amounts to a smaller filter cap (or no filter at all with a high enough frequency). The solution is to bypass the aluminum electrolytic cap with a much smaller-value cap with excellent high-frequency characteristics (and low series-inductance, which is dependent on construction method). Usually, you'll see a polypropylene cap in higher-capacitance values, or maybe polystyrene or Teflon in low-capacitance values.

Lastly what is the commonly used value? Thinking it's around 0.1uF or so?

General rule is no more than 1/100th of the filter cap's value. You might see as-big as a 0.1uF for a 100uF cap, though exuberate audiophiles might have a 1uF bypass cap, with a 2nd bypass of 0.01uF or smaller.

If the filter cap is not positioned right at the circuit sourcing power from it, the bypass is always moved to right at the signal circuit itself.

To date, I haven't tried bypassing a filter cap. That said, my last few amps were built with polypropylene filter caps, and probably don't need a bypass.

[jjasilli]

Well put! Also, re Jojo's question about the last filter stage I should add to my last post along the lines of more caps filter better than fewer caps:  by the time the B+ gets to the last filter stage, it has already been filtered several times.  So the last stage benefits from having the most filtering -- the least ripple-hum reaches the 1st preamp stage.  This is good because that noise would join signal and get re-amplified in later gain stages.  The last filter stage is the most effective one to bypass, having the least work to do because of earlier filtering.  Bypassing prior stages yields diminishing returns.


BTW: though I can hear the difference in an A/B test in the living room, I'm not sure I could tell the difference in a double-blind test, or in a live band situation.

[jojokeo]

Thanks for the info boys! I've had that in the back of my mind for a while now and finally had the oppurtunity to ask. I've often wondered if it could be remotely possible to come through as a slight "high-end hash" of which Tubenit sometimes talks about?

[HotBluePlates]

I've often wondered if it could be remotely possible to come through as a slight "high-end hash" of which Tubenit sometimes talks about?

Maybe, but probably not. I'm not aware of Tubenit using filter cap bypasses, yet he's found other ways to reduce the sound you mention.

"High end hash" could be the result of noise, or it could be oscillation or simply distortion. It might be hard to distinguish what's happening without a scope and careful testing.

[jjasilli]

Dummyload is right.  But I'm assuming you'll source guitar amp caps from a reputable supplier of guitar amp parts, like Doug, e.g., so this should be OK.


JoJo:  Because large caps-- like any B+ filter cap -- have relatively poor ESR, they can benefit from being bypassed by a small cap.  Personally, I think this improves the tone of an amp, but it's subtle.  See my Vibrochamp & Princeton schematics which are posted:  a small cap bypasses the last filter cap - which feeds the first preamp stage.  The first stage substantially sets the tone for the amp, affecting everything happening downstream.  That's why that stage is the one to bypass as stated.  Hi-fi diy'ers may use such bypass caps for multiple or even all B+ filter stages.

Jojo pointed out that my old Vibrochamp & Princeton schematics are gone.  They are attached, and see: http://el34world.com/Forum/index.php?topic=6088.msg55292#msg55292