[jameshuls]

Just a general question for all audio components. I figure there are some pretty knowledgeable folks here, particularly in the realm of passive components.

My question: Can significantly increasing power supply capacitance beyond spec have negative sonic consequenes. For example, would replacing the three 10,000 uF filtering caps on the Cornet H+ with 3 x 22,000 uF alter the frequency repsonse? What about the 47uF B+ caps.

It seems to me, in the role of filtering, capacitance can never be too large. It might be an interesting experiment to significantly increase the values on the Cornet. However, I thought if someone had some knowledge on the topic  it would save me wasting time on the matter.

Just some food for thought!

[BobM]

I've found that in a power supply, increasing capacitance by 20-30% is not usually a problem, but by going significantly larger (2x in your case) you have to look at your diodes and probably increase them to something with a larger amperage rating as well to handle the larger caps. Yes, I have blown diodes in the past by doing what you are suggesting.

There could be some sonc benefit (depends on a lot of things) but will probably most noticeably affect the bass impact. You might get some additional sonic benefit by bypassing these caps with something smaller and quicker also.

[jameshuls]

Thanks for the info Bob!

[markC]

If you go too large on the filter cap value you can also stress the transformer. That would give you a real sonic impact, but not one you're looking for.

[hagtech]

Sometimes it can make a difference.  For example, increasing the first 47uF on the CORNET2 / CLARINET / CHIME will cause an issue with the rectifier, reducing it's lifetime.  More often than not, however, it is ok to do as you propose.

jh

[jameshuls]

Thanks Mark & Jim!

[Bernie]

I have increased the capacitance in the heater supply by five times as much on other pre amps. ie (Dynaco,) without any negative effects. It realy smooth the sound out and became quite transsparent.

Bernie.

[amandarae]

Excellent discussion and just the right time for me to ask one of the many things that I do not know the answer.

Pardon me guys, but this is what I do not understand about the Cornet 2.

According to the tube spec sheet of 5Y3 rectifier, if use in a cap input filter, the first cap is 10uF.  In the Cornet, it looks like 47uF.  Will 47uF be putting a lot of stress to the rectifier tube?

If using 5AR4, then 47uF is fine since the spec sheet says for a C input filter, the value is 60uF.

In my limited experienced, I found that increasing the cap value of the last cap on say a CLC filter (the last C is always a part of the signal path) the sound becomes slow (to me like a CJ).  Although I found as many others experienced already that using polyprop or oil cap on this position improves the sound of the amp.   In our case for the Cornet 2, increasing the 47uF cap, per my understanding, will greatly affect the performance of the rectifier tube especially at turn On.  Am I wrong on this?  Am I missing something on the rectifier circuit?

For heaters, I found 10kuF to 20kuF for the CRC filter (on the two preamps I made) is more than enough and I do not hear any improvement for greater C values when use for IDHT tubes at all IME.


best regards

[GRD]

Not much current draw on the 5y3 in the Cornet.  So 47uF is OK.  The specs assume max ratings. 

[tubesforever]

I asked my tube guru friends when I was first looking over the Hagerman as a kit, what do you think of the rectifier circuit.  Can I mount a better tranny or use more capacitance.  These guys know their circuits back and forth.  Every one of them told me the rectifier will work with a maximum of 47uf and not to run it any higher. 

Glad to hear I don't need to experiment with the heater caps. 

The Jim Hagerman recommended RIAA switch for my C2 has me transfixed.  I can literally dial in my C2 to whatever label I have on the table.   Then I reach over to the Piccolo and dial in the loading on the fly.  The sonics are so good, I get goose bumps listening. 

Thanks Jim!

[amandarae]

Not much current draw on the 5y3 in the Cornet.  So 47uF is OK.  The specs assume max ratings.

Uhmmm....okay!    Because the filter is not a Pi filter at all?

Thanks.


Abe

[GRD]

Abe,

Whether it's an CRC or Pi filter (CLC), the first cap is the one that has the most impact on the rectifier.  The bigger it is, the more it wants to draw from the rectifier at each recharge cycle.  These cycles are short and the current demands scale with the size of the cap.  The R or L provides resistance that smooths the recharging demand from the following cap.  Stated another way, the bigger the first cap, the more it looks like a short to the rectifier during the charging cycle.

The Corner draws around 20ma which is pretty low so the cap doesn't need to recharge that much at each cycle.  But I wouldn't go higher (or lower) than the 47uf that Jim specified since he's the expert. 

Grant

[amandarae]

Grant,

I understand.  With two 12AU7 and one 12AX7, the current draw is small.  But what about turn on when the cap is a short?   I am just asking, since it contradicts all I have read from following the spec sheet of a valve(5Y3 in this case).  Yeah, Jim knows more on this for sure.  I just want to learn why.

best regards,

Abe

[GRD]

Abe,

I was wondering the same thing at start-up.  I can only speculate that the resistance of the transformer and tube moderate the initial couple of charging cycles, or perhaps the charging demand is limited during the rectifier warm-up period (as emission builds).  I'll have to defer to other experts.

Grant

[amandarae]

Ok thanks!

regards,

Abe

[hagtech]

tube spec sheet of 5Y3 rectifier, if use in a cap input filter, the first cap is 10uF

Yes, when run at maximum current 47uF would be too much.  But really it is a peak energy rating, as only so much instantaneous power can be safely dissipated by the tube structures.  Such depends on conduction angle and peak current. 

The CORNET2 / CLARINET / CHIME use a lower average dc current on B+.  Plus, you'll notice I add the 220 ohm resistors in series with the plates.  This makes a huge difference by limiting maximum current.  Hence, the data sheet ratings are misleading, as I am not operating under the same conditions.  I cheat.  The greatly reduced peak energies in the CORNET2 / etc. eliminate lifetime and reliability issues .

jh

[amandarae]

Thank you for the explanation Jim.


I posted while back on my interest of using a "two box" solution for my existing Cornet2, but decided strongly against it lately.  It works, sounds great, and quiet as is, why bother.

regards,

Abe

[jameshuls]

I agree Abe. As is, it is both simple and elegant. I think the term would be "refined". Rather than complicate it, my personal goal is just to tweak it to my idea of perfection.