40uf vs 50uf filer caps?

[daz]

I know the difference in feel they can make. I have tried different values and noted the difference in how stiff higher values feel. I have a 80uf cap in my preamp's first stages and a 47 at the PI and i tried another 80uf in series with the 1st cap to see how it felt at 40uf. i liked it but felt a bit larger might be perfect. Heres the question...i already have a 40/40/500v JJ cap but because i felt the 1st node could be a little tighter and because i would be losing 7 uf at the PI, i thought maybe i should go for a JJ 50/50 to keep the value about the same at the PI and raise the 1st node a but. Do you think i will feel the difference between a 50/50 and 40/40 or is that just too a difference at both nodes small to feel it? I know this sounds anal as hell but it's not a big deal, i just thought i may as well ask before wasting $.

[g1]

Are you measuring the caps to see what they actually are?
The tolerance on caps is not all that tight. For you to be able to hear a noticeable difference between 40 and 50 may be due to a wider spread in values than they are marked as.

[daz]

Are you measuring the caps to see what they actually are?
The tolerance on caps is not all that tight. For you to be able to hear a noticeable difference between 40 and 50 may be due to a wider spread in values than they are marked as.

Yeah, and they;re very close. The JJ is 39 for both sides. But i have a more pressing Q now. I was going to redo the entire post including the title to ask this but with a reply i felt of course i should leave it. But heres something i'm more concerned with now. I have just 2 preamp nodes, V1 and 2 are on one and PI on the other. But the preamp thru the PI is pretty much the same circuit as my EL34 amp with 3 nodes in the pre just like a JCM. Question is this....will using 3 nodes as in the bigger amp change the sound or feel in the same way as adding bigger UF to the 2 node amp? In other words, will adding a node cause a stiffer feel and if not what will change? I think i tried 2 nodes in the EL34 amp once to see what it was like and it did seem like lowering UF, but it;s been so long it's hard to remember exactly.

[pdf64]

Consider that you may be hearing a difference in the caps' ESR rather than their particular capacitance value.
Also that the values measured using the tiny test signal from a meter with a 9V battery may be different to those exhibited by the cap in a high voltage circuit, especially when things have warmed up and significant ripple current is being drawn.

[daz]

Ok, but that aside, all things being equal, in theory what would the difference be between adding a 3rd node rather than running V1 and 2 off the same node and upping the uf using only one node as far as tone or feel? (feel being the prominent change i'm sure, as tone doesn't really change so much with filters)

[eschertron]

Ok, but that aside, all things being equal, in theory what would the difference be between adding a 3rd node rather than running V1 and 2 off the same node and upping the uf using only one node as far as tone or feel? (feel being the prominent change i'm sure, as tone doesn't really change so much with filters)

As current doesn't generally fluctuate in the preamp nodes, splitting V1 (V1a and V1b) off from additional loads is done primarily for noise rejection. The further the node is removed from any nodes that sag under load, the less they will see the HV change. No. Scratch that. The longer the time constant for the HV to change. I can't say if you'll be able to hear the difference. But in theory, any additional filtering (bigger caps or extra nodes) will make the affected circuit stiffer.

edit: some time spent with PSUDII might show a difference. In theory.

[daz]

I just quickly and sloppily (of course ) tacked it in to see and what i notice is yes, a bit stiffer but not at all in a bad way and very little and by dialing down the highs a bit that goes away. What i really liked and have therefore decided to add it permanently is that it just sounds cleaner. I use distortion and roll the volume down for cleans and now when i do that the cleans sound cleaner and rounder. When i hear about more filtering equals less noise i never really associated it with that kind of noise. More like hum/buzz. The type of noise it seems to eliminate is the nastly bits around the note. Almost a staticy bit of garbage, and thats no longer there.

[J M Fahey]

Different cap values will be felt as stiffer or looser where:
a) current demand varies a lot with music playing, with guitar picking if you wish, so voltage will vary more, and faster, with a smaller cap than with a larger one..
b) said voltage influences gain, such as in a pentode stage, when screen voltage is linked to the power rail.
Both conditions are met at amplfier power tube stage.

None is met at a preamp so 40 uF there will sound exactly the same as 50uF.

Only possible difference will be if the supply is very poorly filtered, has tons of ripple, such ripple gives annoying levels of hum .... but that´s not what we are talking here.

[wizard333]

Ok, but that aside, all things being equal, in theory what would the difference be between adding a 3rd node rather than running V1 and 2 off the same node and upping the uf using only one node as far as tone or feel? (feel being the prominent change i'm sure, as tone doesn't really change so much with filters)

As far as real world application, I've noticed experimenting with power supplies and giving each triode it's own filter, that the response becomes HARD. By that I mean, when you hit a chord, especially anything with low notes in it, it hit's your ears like a hammer. Super, super percussive on the transient, in a really unpleasant way.

The trick is limiting that. If you balance the number of triodes on a node you can give the amp more "punch" without making the transients overly hard, but as noted, the response will be more 'stiff', especially in response to fast transients like tremolo/speed picking. Power chords really hit you in the gut, but playing fast becomes more difficult.

There is always a trade off.

[daz]

As far as real world application, I've noticed experimenting with power supplies and giving each triode it's own filter, that the response becomes HARD. By that I mean, when you hit a chord, especially anything with low notes in it, it hit's your ears like a hammer. Super, super percussive on the transient, in a really unpleasant way.

The trick is limiting that. If you balance the number of triodes on a node you can give the amp more "punch" without making the transients overly hard, but as noted, the response will be more 'stiff', especially in response to fast transients like tremolo/speed picking. Power chords really hit you in the gut, but playing fast becomes more difficult.

There is always a trade off.

Exactly, and thats what i found. At first i liked it but quickly became aware of that hard attack with no give. Same thing you get with higher value caps. I tried a 20uf for the PI and one 20uf cap for V1 and V2 and it got WAY soft and lost some harmonic content too. Yet i am now using one 40uf on V1 and 2 and the total is te same but the hardness has increased to just about the right level. This has me thinking.....you could achieve the degree of harness/softness in the attack that u want 2 ways, with less nodes but higher values, or lower values and more nodes. This begs the experiment it suggests. I should try a separate filter for V1 and V2 but at 1/2 the value of the ONE cap that is now being used for it, a 40uf. So 2 20uf's, one for each tube. Same total amount, but seperated to two nodes with another dropping resistor. Might be interesting to see how the same amount of capacitance but split instead of shared would differ.

[KevinOConnor]

Hi Guys

What you gain by giving each stage its own supply node is better note articulation and lower noise. That is, provided you also fix up the wiring so that each stage sees its own filter first. This tidying up eliminates most intermodulation of the supply and signal, and of signals that shouldn't be mixed.

Where the node voltages are fairly steady in a preamp on a DC basis, there is also a circulating signal current. There are then two currents effectively for each stage. Powering cascaded inverting stages from the same node provides an easy means for oscillation to occur.

The multi-section caps tend to be very low quality, with hgih or unspecified ESR, DF and DA - all deviations from being a perfect cap that all caps share, but electrolytics have more than others, and axials and multi-sections have to a greater extent than radials or snap-mounts. besides, multi-section caps ensure that currents mingle which otherwise should not, making it much harder o have the best performance possible.

I prefer stiff supplies since it is easy to add dynamic effects by adding and R here or there. Most guitar and bass amps are woefully under-filtered. TWs and similar have horrific hum-modulated outputs when they are clipped and this is easily fixed. The Komet is one of the worst amps I've heard. TUT3 shows how to wire an amp properly for best note articulation and lowest noise. It also shows how to soften the sound to taste.

Have fun

[daz]

Interesting, that answers some questions i've had for a long time. Thanks 4 that.

Hi Guys

What you gain by giving each stage its own supply node is better note articulation and lower noise. That is, provided you also fix up the wiring so that each stage sees its own filter first. This tidying up eliminates most intermodulation of the supply and signal, and of signals that shouldn't be mixed.

Where the node voltages are fairly steady in a preamp on a DC basis, there is also a circulating signal current. There are then two currents effectively for each stage. Powering cascaded inverting stages from the same node provides an easy means for oscillation to occur.

The multi-section caps tend to be very low quality, with hgih or unspecified ESR, DF and DA - all deviations from being a perfect cap that all caps share, but electrolytics have more than others, and axials and multi-sections have to a greater extent than radials or snap-mounts. besides, multi-section caps ensure that currents mingle which otherwise should not, making it much harder o have the best performance possible.

I prefer stiff supplies since it is easy to add dynamic effects by adding and R here or there. Most guitar and bass amps are woefully under-filtered. TWs and similar have horrific hum-modulated outputs when they are clipped and this is easily fixed. The Komet is one of the worst amps I've heard. TUT3 shows how to wire an amp properly for best note articulation and lowest noise. It also shows how to soften the sound to taste.

Have fun

[wizard333]

Hi Guys

What you gain by giving each stage its own supply node is better note articulation and lower noise. That is, provided you also fix up the wiring so that each stage sees its own filter first. This tidying up eliminates most intermodulation of the supply and signal, and of signals that shouldn't be mixed.

Where the node voltages are fairly steady in a preamp on a DC basis, there is also a circulating signal current. There are then two currents effectively for each stage. Powering cascaded inverting stages from the same node provides an easy means for oscillation to occur.

While that's true enough in theory, I have hardly ever seen this be a problem in practice.

I prefer stiff supplies since it is easy to add dynamic effects by adding and R here or there. Most guitar and bass amps are woefully under-filtered. TWs and similar have horrific hum-modulated outputs when they are clipped and this is easily fixed. The Komet is one of the worst amps I've heard. TUT3 shows how to wire an amp properly for best note articulation and lowest noise. It also shows how to soften the sound to taste.

Have fun

I have universally found over-filtered amps to play stiff and have a decidedly unpleasant feel. If you know a way around that, can you be more specific on how you could use a lot of filtering and not get that effect? I have yet to ever see that on an actual amp.

[daz]

I have universally found over-filtered amps to play stiff and have a decidedly unpleasant feel. If you know a way around that, can you be more specific on how you could use a lot of filtering and not get that effect? I have yet to ever see that on an actual amp.

I agree. However, i'm not so sure "overfiltered" in the context of how the amp feels applies to more nodes as much as more capacitance. In other words, 2 tubes sharing a 40uf node vs 2 tubes each with thier own 20uf node. I just did that last nite (replaced 1 40uf node for V1&2 with 2 20uf nodes) after reading kevin's post and while it seems to be an improvement in note balance in clarity (albeit small) i didn't notice it being stiffer as i did when i tried a single 80uf node got V1&2. Now maybe i will feel it after a time but so far i can''t tell, tho double the uf overall for V1&2 whether with 2 nodes or one and i can easily feel it. And like you said, i don't like it. It loses the dynamic quality i strive for.

I can't speak for kevin, but if i had to guess on how he would fight a stiffer supply i think there are many things in the preamp at least where certain circuit configurations will add a lot of sag while others will be stiffer. I can't give you specifics as to exactly what will do that because it;s a huge puzzle and one change will affect all other things and often to get there i have to make several changes and cannot say that the one that triggered that looser feel was responsible by itself because if i change any of the other things it then changes the whole feel again. In other words there are a lot of things that together might trigger it so itl;s near impossible to note what exact does it. Also i think larger value dropping resistors will help there.

That said, i'm not sure it;s even possible to counter a stiff supply to the point it feels as loose as it would at 1/2 the filtering (overall value wise that is) as keven suggests. I know approximately 1/1000 as much electronic theory as him on a good day, but i do know what i've experienced and while i agree with him that it can be countered, i don't think it can be brioght back to the same feel as lower filtering exactly. Just to some extent. I may be wrong but like you i have never been able to eliminate it, but i HAVE helped make a stiffer supply more acceptable via other changes, but not that i can say because as i said earlier it was just trial and error and i'm not sure exactly what did the trick. But in the end i have never been able to do it to a point i was happy enough with it to leave it there and always end up reducing the cap value overall.

[Chuck H]

Hey daz... You're a nut and a tweak! I hope you never change, but I also hope you're having some fun with it.