filter values

[iknowjohnny]

just curious....does anyone pick filters for each stage by swapping out different values till it sounds good to them? And i mean every stage with no regard to uniformity, IE: having the same values everywhere in the pre at least. I see that in most if not all amps the filters seem to be the same from the input tube thru the PI. Well, a month or so ago the last thing i did was to try various values at all 3 pre nodes and i ended with 60UF at v1, 100UF at v2, and 20UF at the PI. I guess this is very unorthodox, but i swear the tone i have now is just as sweet as you will ever hear. I went up and down in value at each node till there was a very apparent happy medium and them moved to the next. granted i could change any of them a bit in either direction w/o messing with the tone, but not much. They're pretty optimal right now as far as tome goes. I have to say doing this has really made a big difference in the tone and feel. I haven't touched the amp since and don't plan to. I have never been 100% happy till now and while it wasn't just this which got it there, it was the final capper. It sounded great before, but that really took it to a level of near perfection.

Has anyone else done it this way or do you all just go with what sounds optimal with the same value at every node? Is there some reaon that what i did would not be considered correct? By the way, regardless of the fact the PI only sports a 20UF node, the tone is no less clean than it was before when i dial it up that way. i would have though a big filter there would sound best for an amp where you want the pre to supply the drive and a clean PI/PA to amplify it only.

[PCollen]

In general, I will use the filter values of the OEM version of the amp I am building. Higher filter values in the pre-amp will generally result in a stiffer-feeling amp, which is why I prefer lower filter values in the pre-amp ( 16-32 uF). I prefer 32-50uF on the power tube supplies (plate and screen) for both diode or tube rectification. I've NEVER considered 100uF in the pre-amp supply rail.

[Cliff Schecht]

The only time you need to be careful about what capacitance you use is the first capacitor after a tube rectifier. The maximum allowable capacitance is usually specified in the datasheet or you can find a recommended value on the internet. After the dropping resistors, you can go as big as you want without any fuss. This isn't an issue with SS rectifiers, unless the inrush current is larger than what the diodes can handle.

I choose my values based off of the amount of filtering in previous stages, the sound I'm going for and usually to hit a certain amount of voltage ripple. I like below 10% (5% or less), but it really depends on what I have on hand and what the stage I'm decoupling is. High gain, low signal level stages can't handle a lot of ripple, either can single-ended output stages. But push-pull output stages tend to cancel common-mode noise so you can sometimes get away with less capacitance on the power tube decoupling. Here is Merlin's article on power supply filtering, some great basic information to get you started. http://www.freewebs.com/valvewizard1/smoothing.html

[iknowjohnny]

yeah, i'm not worried about how much i can use. I have a ss rectifier and the PA has a 100/100uf. But i just found that the amp could be tailored in a way that really made a considerable difference to the good in tone and feel by simply raising and lowering each node to find what sounds best. i was really surprised to find each node required a very different value to sound optimal and that while no one seems to do this it really made a great improvement. Made me wonder why i don't see this more often. I mean, when is the last time you saw an amp with the following filter values from the input to the PI....60uf/100uf/20uf ! Kinda weird eh? I'm absolutely loving it tho. The dynamics in the way it responds to my playing like this is sheer perfection.

[Jana]

One of the variables (and an important one in the eyes of the bean counters) that comes into play with production amps is cost. It is cheaper to buy 50,000 caps of one value than it is to buy smaller quantities of various values. It is also cheaper, and was even more pronounced in the era of "vintage" amps, to use the smallest value possible that they could get away with. Bean counters. Leo was one of them.

[David Root]

Let's be fair, Jana, Leo wasn't a professional bean counter, he was just frugal!

Then CBS came in and they WERE professional bean counters, so then everything went to hell in a handcart. Cheap Italian knockoffs of Cinch tube sockets, Stackpole pots with plastic shafts.......

[iknowjohnny]

Well, true, but that wasn't the point. I was just talking about the fact that i don't see amps with completely different filter values for each node, be it a production amp of firum members. No one seems to do that. Cost isn't the issue in production amps because marshall made a lot of amps that were 50uf across the board. So they could have gone smaller and cheaper. so why didn't they use a 20uf on a certain node because it sounded better? they probably never tried or even thought about it. that was my reason for the thread...does anyone do what i did? I can't be the only one, but it sure worked well for me so it becomes a bit puzzling why no one else. I've questioned things i have come up with in the past only to later realize why. I doubt thats going to happen this time because i've had the chassis in the amp for a long time now and if anything i'm digging it even more now. I should also mention the filter thing really isn't a tonal improvement at all even tho i keep using the word "tone". (out of habit) What it really did was to make the amp super nice feeling and gave it great dynamics and response. It already sounded great b4 that, but when it's this responsive it allows you to take advantage of the tone so much better.

[martin manning]

...Cost isn't the issue in production amps because marshall made a lot of amps that were 50uf across the board. So they could have gone smaller and cheaper. so why didn't they use a 20uf on a certain node because it sounded better?

This illustrates Jana's point. Using the same value across the board means larger quantities of that part number, fewer part numbers to stock, and less time to kit and build the amp. All of these reduce manufacturing cost, and could easily offset the savings from using a smaller cap in one or two places.

[martin manning]

Let's be fair, Jana, Leo wasn't a professional bean counter, he was just frugal!

Then CBS came in and they WERE professional bean counters, so then everything went to hell in a handcart. Cheap Italian knockoffs of Cinch tube sockets, Stackpole pots with plastic shafts.......

Short-term profit motivation with (they thought) negligible impact on the customer's most highly valued characteristics. That works for a while, until the accumulation becomes so great that it becomes obvious to everyone that paradise has been paved over.

[PCollen]

Well, true, but that wasn't the point. I was just talking about the fact that i don't see amps with completely different filter values for each node, (snip).. so why didn't they use a 20uf on a certain node because it sounded better? they probably never tried or even thought about it. that was my reason for the thread...does anyone do what i did? .

You should pose this question to the boutique amp builders: Mike Zaite (Dr. Z), Hogey, Gerhardt,Germino, Rouquefort (sp ?), Howard Dumble (is he still alive ?), and if he WERE still alive Ken Fisher, and others who could make the time and effort to do it if it were worthwile, easily absorbing the cost into their product, but don't/haven't. My guess is that it isn't really worth the effort and the results coan only be appreciated by somone such as yourself who wishes to tailor his/her amp to their specific tone desires. Lastly, you might ask why the amp-mod guru's of this world have not included filter cap mods in their offerings, considering everything else they seem to want to alterer in hopes of making a few bucks.

[iknowjohnny]

Well, true, but that wasn't the point. I was just talking about the fact that i don't see amps with completely different filter values for each node, (snip).. so why didn't they use a 20uf on a certain node because it sounded better? they probably never tried or even thought about it. that was my reason for the thread...does anyone do what i did? .

You should pose this question to the boutique amp builders: Mike Zaite (Dr. Z), Hogey, Gerhardt,Germino, Rouquefort (sp ?), Howard Dumble (is he still alive ?), and if he WERE still alive Ken Fisher, and others who could make the time and effort to do it if it were worthwile, easily absorbing the cost into their product, but don't/haven't. My guess is that it isn't really worth the effort and the results coan only be appreciated by somone such as yourself who wishes to tailor his/her amp to their specific tone desires. Lastly, you might ask why the amp-mod guru's of this world have not included filter cap mods in their offerings, considering everything else they seem to want to alterer in hopes of making a few bucks.

Well, yeah, thats what i was questioning and i WAS posing it to boutique builders, but mostly to those of you on the forum. I never really meant to ask why in relation to production amps at all. But the fact that those to whom a little extra cost makes zero difference like all of you and many boutique builders who are only able to survive if thier amp is above the rest.....they aren't going to worry about a few bucks when they are selling a $3000 amp that no one will buy unless it kills.

But i assure you of this....my amp truly improved in a way i cannot imagine anyone would not like more than how it was before i did that. At least not the typical experienced player to whom an amp's dynamics and responsiveness is very important. It's one of those things like sustain....you could argue that there are people who may want less i suppose, but really, i doubt more than a fraction of a percent if at all. If you did something to improve an amp's sustain, pretty much everyone will like it more like that. I think the same of my filter arrangement. I don't think the improvement is a personal thing thats real subjective. I think most anyone who played it before and after would agree, and if i'm right about that then i stand puzzled as to why only I found this to improve the amp as it did. I discovered it while trying different values. I tried one node and when i hit upon a certain value that improved the amp i then would try that same value on the other nodes, but only to find they were NOT the best value there. thats when it hit me that each node might do well with it's own filter value. And sure enough the difference was not subtle.

Anyways, i gotta tell you this amp is just amazing and i have no desire to touch it again. Every time i play it i marvel at what i know can't be perfection, but one the other hand i can't hear or feel anything in it that isn't ! Sound crazy, but so far about a month into this last tweak i am just floored every time i play it. Even those days where your ears are not right or whatever it is that causes things to sound worse at time, on those day it still sounds and feels great if not quite as good. This is what i was aiming at when i first started building, and till now i feared i'd never reach this point. I'm happy:)

[David Root]

What circuit is your amp? Is it a gain monster? What is your physical playing style, hard & heavy or on the lighter side? Do you have the amp dimed and control it from the guitar?

The filter values on V1 and V2 are certainly unorthodox, while the PI is quite orthodox.

I don't dispute what you are saying, that would be silly, just trying to find out why those values work so well for you.

Another thing that no-one has mentioned yet is the voltage rating of those caps (all of them including the 100+100 in the power amp) compared with the actual operating voltages you run at. I.E. are you running these caps with little or no headroom voltage wise, or only at 60% of rated voltage, for example?

[iknowjohnny]

If i had to describe my style i'd say SRV style dynamics....hard and light depending on the moment and purpose. And i prefer a single channel amp, strat with a HB in the bridge and 2 vintage style singles, I use my guitar p/u selector and volume control to alter gain levels and tone, a clean boost for over the top gain at times, and the amp i would call medium high gain. About as much gain as a peavey classic 30 and maybe 10-15% more than a blues jr with gain knob maxed. And i usually keep the gain at about 2:00. All filter caps are either 450v or 500v.

Like i said, the net result was purely dynamics/feel, not tone.....it already had that. Maybe thats something most people don't strive for when designing. Maybe they strive for tone and if something doesn't sound better they don't bother/don't think about feel. I dunno. All i know is it works for me and i'm pretty sure others would dig it too. I've been playing 40 years and spent 25 of that gigging regularly, so it's not like i'm clueless about tone and what is generally thought of as good.

[AllenBradley]

What circuit is your amp?

[PCollen]

(snip)..
Like i said, the net result was purely dynamics/feel, not tone......

It's well known that lower filtering in the pre-amp/PI will result in a looser feel to the amp, and higher filtering stiffens/tightens things up. I run 16uF on all pre-amp/PI feeds in my JTM50-clone, with 50uF on the power tube plate and screen feeds. I like a 'bouncy' amp. Next time i have it out of the cab, maybe I'll stick another 40uF across the V2 and V3 feed caps to evaluate the results. I doubt very much I'd want to go much higher than 16uF on V1.