Metal film caps in power section worth the upgrade?

[mloutfie]

Hi All,

 

I've been thinking to upgrade the caps in my phono preamp just as an experiment to see if I can make the sound better. I've asked the designer which capacitor is in the signal path and the one with thick blue lines are the one he said in the signal path but they are already metal film caps so might not need changing. But he also says the 4 electrolytic caps are power supply caps but since it's a single ended design so they are technically in the signal path. Anyone has experience on upgrading capacitor in the power side? and whether it's worthwhile. the caps itself its a 33uF so I can get a decent MKP Polypropylene for around $18 each so not a great cost to try but I want to hear some opinion.

 

Thanks 

 

Edited November 7, 2020 by mloutfie

[Monkeyboi]

A clear photo showing the markings on those electrolytics (to aid in identifying brand and type) would be helpful in providing an informed answer to your question.

 

Cheers,

Alan R.

[muon*]

................

Edited November 8, 2020 by muon*

[rockeater]

9 hours ago, mloutfie said:

but since it's a single ended design so they are technically in the signal path

Technically, there is no such thing as signal path at all.

In each stage signal is sunk to ground and recreated anew on the output of its active element.

This is a wisdom as thought by Dr. Lampizator and he knows a thing or two about it. Particularly the type you like - the valve type

[andyr]

16 hours ago, mloutfie said:

Hi All,

 

I've been thinking to upgrade the caps in my phono preamp just as an experiment to see if I can make the sound better. I've asked the designer which capacitor is in the signal path and the one with thick blue lines are the one he said in the signal path but they are already metal film caps so might not need changing. But he also says the 4 electrolytic caps are power supply caps but since it's a single ended design so they are technically in the signal path. Anyone has experience on upgrading capacitor in the power side? and whether it's worthwhile. the caps itself its a 33uF so I can get a decent MKP Polypropylene for around $18 each so not a great cost to try but I want to hear some opinion.

 

Thanks 

 

 

If you want to keep using the Hagerman case ... I would've thought that using Nichicon ES or KZ 33uF/50v would deliver a good improvement, Mahdie?

 

Andy

 

 

 

[mloutfie]

At the moment the blue caps are Riva phe 426 metal film caps 400v. And the electrolytic is nichicon vz(m) 350v.

 

At the moment a full redone with new case is on the table. Since with the current external power supply and tube combination used proven a bit hard to upgrade from this phono

[mloutfie]

Also here is the psu and preamp schematics

[andyr]

10 minutes ago, mloutfie said:

 

At the moment the blue caps are Riva phe 426 metal film caps 400v. And the electrolytic is nichicon vz(m) 350v.

 

 

 

 

Aah, sorry for my bad advice, Mahdie.  (I hadn't thought about the fact that it has high voltages because of the tubes - so 50v caps are not usable!   )

 

Andy

 

[mloutfie]

@andyr the current candidate is jenzen mkp https://speakerbug.com.au/index.php?route=product/product&product_id=159

Edited November 8, 2020 by mloutfie

[andyr]

45 minutes ago, mloutfie said:

@andyr the current candidate is jenzen mkp https://speakerbug.com.au/index.php?route=product/product&product_id=159

 

Should be great!  

 

But, yes - you will need a bigger case.  

 

(However, that will enable you to implement a 2nd (or even 3rd!) pair of input caps in parallel ... for easy R & C loading.)

 

Andy

 

[muon*]

...................

Edited November 8, 2020 by muon*

[crtexcnndrm99]

If it helps to have used experience, I recently replaced two nichicon PS caps with 3 paralleled silmic II’s each to get the right value. Was in a valve phono as well. 
 

Going to use obbligato PIO caps for the smaller value caps in the PS (need to order another one).

Edited November 8, 2020 by crtexcnndrm99

[mloutfie]

4 hours ago, crtexcnndrm99 said:

If it helps to have used experience, I recently replaced two nichicon PS caps with 3 paralleled silmic II’s each to get the right value. Was in a valve phono as well

Aren't they still electrolytic?

 

Obbligato seems interesting where to get them?

[Monkeyboi]

Considering the power supply is a switch mode operating at a high frequency, C12 and C14 IMHO should be low ESR types instead of the standard Nichicon aluminium electrolytic capacitors used.

 

IMHO I doubt using some exotic capacitors in this filtering role will reap any sonic benefits.  If it was my phono stage I would stick with capacitors that can be accommodated with the existing radial lead spacing to avoid adding any inductance.  Another consideration with film capacitors is to ensure the outer foil is connected to the ground (0 volt rail) to further reduce noise induction.

 

Cheers,

Alan R.

[mloutfie]

9 hours ago, Monkeyboi said:

Considering the power supply is a switch mode operating at a high frequency, C12 and C14 IMHO should be low ESR types instead of the standard Nichicon aluminium electrolytic capacitors used.

 

IMHO I doubt using some exotic capacitors in this filtering role will reap any sonic benefits.  If it was my phono stage I would stick with capacitors that can be accommodated with the existing radial lead spacing to avoid adding any inductance.  Another consideration with film capacitors is to ensure the outer foil is connected to the ground (0 volt rail) to further reduce noise induction.

 

Cheers,

Alan R.

 

No I actually uses teddy pardo LPS for the power supply not the original walwart will that make the Nichihon be ok?

 

One other option is to also upgrade the caps the designer says in the signal path but upgrading Riva metal film to Jenzen metal film will it yield big improvement? I'm new to capacitor specs

[mloutfie]

@muon* were you recommending a different caps before? the comment seems to be edited

[Monkeyboi]

2 hours ago, mloutfie said:

 

No I actually uses teddy pardo LPS for the power supply not the original walwart will that make the Nichihon be ok?

 

One other option is to also upgrade the caps the designer says in the signal path but upgrading Riva metal film to Jenzen metal film will it yield big improvement? I'm new to capacitor specs

 

Well if you look at the circuit diagram you will see the phono stage accepts a 9 volt d.c. supply input via a conventional barrel jack, however to convert the 9 volts to 210 volts to power the H.T. side (plate supply) it goes through a switch mode boost circuit driven by the LM555 timer IC which in turn switches the MOSFET (Q1) on and off at a very high frequency.  This energises  and de-energises the magnetic field of inductor which in layperson's terms creates a high voltage a.c. which is then half-wave rectified to  210 volts d.c. by diode (D1) and crudely filtered by capacitor (C5) before been passed onto the actual amplification circuitry. as the plate H.T. supply.

 

Further H.T. power supply ripple filtering is provided by R25 + C14 + C15  and R24  + C12 + C13 in each channel.  As the frequency of the ripple hasn't changed and is high frequency in nature, low ESR capacitors should preferably be used for C14 and C12 in each channel.  Whilst the existing Nichicon VZ series capacitors are somewhat "doing the job", they are primarily designed for low frequency operation in typical linear mains supplies where the ripple frequencies are 50/60Hz for half wave and 100/120Hz for full wave rectification respectively..

 

As the switch mode convertor in the phono stage is operating at frequencies well above the audio range at approximately 48kHz in this design, low ESR capacitors will provide much superior filtering over conventional aluminium electrolytic types which would have a comparatively higher ESR and therefore less filtering at this higher ripple frequency.

 

Will you hear a difference?  Hard to tell as hearing is so subjective.  Will the H.T. power rail have less ripple?  Yes, and this can be confirmed by measurement using an oscilloscope.  In audio amplifiers of this design, power supply cleanliness (providing clean ripple free d.c. ) will impact on sound quality.  Any ripple appearing on the plate (anode) of V1 will be amplified along with the audio by the next stages of amplification thereby polluting the audio with a supersonic (48kHz) signal.  Now you can't hear 48kHz because it's well above human hearing, but it may "mix" with the audio and potentially produce unwanted byproducts.  Now the RIAA equalisation that takes place between V1 and V2 should suppress most of the unwanted 48kHz ripple present at the anode of V1 but IME and IMHO it's best not to have to rely on this EQ to remedy a poor design.  Just my 2 cents worth.

 

# ESR = Equivalent Series Resistance

*  H.T.  High Tension (a.k.a. high voltage)

 

Just a note on Rifa brand film capacitors.  They developed a bit of a bad reputation in the industry due to poor manufacturing techniques in the past.  Those older Rifa capacitors can be easily identified as they are encapsulated in clear epoxy.  I personally don't use them in any of my designs or for repairs, instead I use Wima brand caps as IMHO they are reliable, good performers and are good value for money.

 

Cheers,

Alan R.  

Edited November 9, 2020 by Monkeyboi
added explanation

[mloutfie]

@Monkeyboi thanks for that lots of things to consider. But switching all to WIMA should be also in my consideration.

 

I might need to read more on the ESR but I think a while back I read on a boost voltage circuit low ESR caps can cause oscillation and make the circuit unstable. But if this is not the case I will put low ESR on the list also

 

[Monkeyboi]

1 hour ago, mloutfie said:

@Monkeyboi thanks for that lots of things to consider. But switching all to WIMA should be also in my consideration.

 

I might need to read more on the ESR but I think a while back I read on a boost voltage circuit low ESR caps can cause oscillation and make the circuit unstable. But if this is not the case I will put low ESR on the list also

 

 

You need to read that Panasonic article in context *.  The meat is in the summary paragraph. -

https://industrial.panasonic.com/ww/ss/technical/lc4#:~:text=If ESR becomes lower%2C the,small%2C it can cause oscillation.

"

Summery

Because switching power supplies and high-speed logic devices can be noise sources, LC filters and RC filters are used as one of the ways for noise management. For both of them, the ESR of the capacitor to be used and filter performance are related, and the use of low ESR capacitors enables the attenuation characteristics of the filter to be closer to ideal. Panasonic conductive polymer electrolytic capacitors, with their low ESR, are effective options.

Meanwhile, for Cout, which is an output smoothing filter for switching power supplies, low ESR is an essential requirement for ripple voltage reduction. Low ESR conductive polymer electrolytic capacitors are an also effective solution to this."

 

Whilst it is possible that low ESR capacitors if improperly implemented in a classic feedback controlled switch mode supply (of which this supply isn't) can potentially cause oscillation it won't happen here because this boost supply in the phono pre-amp isn't a feedback type.  It's more akin to crude inverter converting the 9 volts d.c. input to a 48 kHz square wave which is applied to the gate of the MOSFET which in turn switches the current on and off in inductor L2.  When the MOSFET switches off the magnet field created by the on condition of the MOSFET collapses creating a high reverse voltage.  This high voltage then rectifying by a fast recovery diode (D1) and as I said earlier crudely filtered by a  470nF film capacitor C5.  Unfortunately C5 alone is not enough to significantly reduce the ripple on the H.T.supply, hence the additional filtering and decoupling closer to the actual amplifying stages using C12, C13, C14 and C15 with R24 and R25, 

 

In every switch mode PSU I've ever seen, low ESR  filtering capacitors are the norm because they have a lower power dissipation than conventional aluminium electrolytic capacitors at higher frequencies.  This an electrical fact and not based on audio alchemy.

 

I don't intend to railroad this thread with a tutorial on switching boost converter theory, just tell it how it is.  As I said, the VZ series Nichicon caps are "doing the job" because the current drain in this circuit is very small in comparison to a heavier duty supply typical of a TV or power amplifier.  Most gear is built to a price.  Nichicon low ESR equivalent caps would be more expensive and increase the cost of the BOM, hence if a manufacturer can get away with a lesser component without seriously degrading the performance or safety of of the product they will.

 

Cheers,

Alan R.

Edited November 9, 2020 by Monkeyboi

[muon*]

8 hours ago, mloutfie said:

@muon* were you recommending a different caps before? the comment seems to be edited

I didn't think they were being considered.

 

I recommended Solen PB for the power caps (I use these decoupling in the power feed to the small signal valves in my EL34 amp and used them in the PS of my old modified EAR phono stage when I had that) The guys on the killer dac forum like these in the power supplies in places and that's where I got the idea from.

 

I also suggested as a affordable replacement for the metalised films (blue ones) Jantzen Superior capacitors, more musical and detailed than the Cross Caps. Me I like copper foil caps in these applications but they can cost a bomb and are generally big in size. The Superiors are affordable and a nice cap and smaller for a given value.

[mloutfie]

3 hours ago, Monkeyboi said:

 

You need to read that Panasonic article in context *.  The meat is in the summary paragraph. -

https://industrial.panasonic.com/ww/ss/technical/lc4#:~:text=If ESR becomes lower%2C the,small%2C it can cause oscillation.

"

Summery

Because switching power supplies and high-speed logic devices can be noise sources, LC filters and RC filters are used as one of the ways for noise management. For both of them, the ESR of the capacitor to be used and filter performance are related, and the use of low ESR capacitors enables the attenuation characteristics of the filter to be closer to ideal. Panasonic conductive polymer electrolytic capacitors, with their low ESR, are effective options.

Meanwhile, for Cout, which is an output smoothing filter for switching power supplies, low ESR is an essential requirement for ripple voltage reduction. Low ESR conductive polymer electrolytic capacitors are an also effective solution to this."

 

Whilst it is possible that low ESR capacitors if improperly implemented in a classic feedback controlled switch mode supply (of which this supply isn't) can potentially cause oscillation it won't happen here because this boost supply in the phono pre-amp isn't a feedback type.  It's more akin to crude inverter converting the 9 volts d.c. input to a 48 kHz square wave which is applied to the gate of the MOSFET which in turn switches the current on and off in inductor L2.  When the MOSFET switches off the magnet field created by the on condition of the MOSFET collapses creating a high reverse voltage.  This high voltage then rectifying by a fast recovery diode (D1) and as I said earlier crudely filtered by a  470nF film capacitor C5.  Unfortunately C5 alone is not enough to significantly reduce the ripple on the H.T.supply, hence the additional filtering and decoupling closer to the actual amplifying stages using C12, C13, C14 and C15 with R24 and R25, 

 

In every switch mode PSU I've ever seen, low ESR  filtering capacitors are the norm because they have a lower power dissipation than conventional aluminium electrolytic capacitors at higher frequencies.  This an electrical fact and not based on audio alchemy.

 

I don't intend to railroad this thread with a tutorial on switching boost converter theory, just tell it how it is.  As I said, the VZ series Nichicon caps are "doing the job" because the current drain in this circuit is very small in comparison to a heavier duty supply typical of a TV or power amplifier.  Most gear is built to a price.  Nichicon low ESR equivalent caps would be more expensive and increase the cost of the BOM, hence if a manufacturer can get away with a lesser component without seriously degrading the performance or safety of of the product they will.

 

Cheers,

Alan R.

Thanks for the info. I'll do more reading. But a summary would be you're saying the low ESR will cause less ripple? Will be worth a try. I was just worried about the oscillation problem but your explanation makes it clear thanks. Unfortunately jenzsen caps doesn't have ESR measurements so I might need to take that into consideration against using it for the PSU. so for power the low ESR caps looks more plausible option now

[mloutfie]

15 minutes ago, muon* said:

I didn't think they were being considered.

 

I recommended Solen PB for the power caps (I use these decoupling in the power feed to the small signal valves in my EL34 amp and used them in the PS of my old modified EAR phono stage when I had that) The guys on the killer dac forum like these in the power supplies in places and that's where I got the idea from.

 

I also suggested as a affordable replacement for the metalised films (blue ones) Jantzen Superior capacitors, more musical and detailed than the Cross Caps. Me I like copper foil caps in these applications but they can cost a bomb and are generally big in size. The Superiors are affordable and a nice cap and smaller for a given value.

Where is a local supplier for solen caps? I will give them a look

 

For jentzen standard z-cap yes I agree I use them when possible but having problem finding suitable values for example for 0.47uF I can find cross cap and superion z-cap (red) or even silver z-caps but not blue. the superior red caps are significantly even in low values

[muon*]

I usually order from parts connxion in canada, I like to get the caps i want regardless of where I need to get them from.

I only ever use Jantzen Superior (red ones) if I need to look at budget solution for someone if I'm helping them as they offer the best value without compromising.

 

I needed a good but affordable cap of smaller size for work on this amp and chose the Superiors.

 

 

I can only make suggestion from my experience and views, these suggestions may not always suit someone else.

Edited November 9, 2020 by muon*

[mloutfie]

Thanks all, I think I have enough information to go on and start planning the build now it's a matter of finding a new case, reconsidering the size of replacement capacitor and how many capacitors I'm replacing

[rockeater]

2 hours ago, mloutfie said:

Thanks all, I think I have enough information to go on and start planning the build now it's a matter of finding a new case, reconsidering the size of replacement capacitor and how many capacitors I'm replacing

I myself would advise you against it.

Companies making electronic stuff employ engineers to consider component placement, width of tracks and their alignment and location of different stages within the device. I worked at a place where they specifically did that to increase immunity from external noise (it was not in an audio field where the bar is naturally lower, despite what many might think).

They obviously believe, that this is of more importance than using a specific, higher grade component. Companies like Wadia and Electrocompaniet are making great stuff out of ordinary parts, for the great sound comes from understanding the issues involved and from great engineering, not from better parts. Fitting audiophile parts in the hope of taking the component to another level, is like replacing all the bolts in a V8 Commodore with stainless steel ones with a hope of making a Ferrari out of it. It will never be, for a technical product's greatness starts at the drawing board, not at assembly line.

On more than one occasion people came to my workshop with a DIY project that was soldered properly but hummed, which was for no other reason than lack of understanding of how the different parts (stages) interacted with each other.

So my advice is: if a component does not fit on the board and you are determined to replace it, get a different one that fits there. Otherwise, there is a good chance you will be going backwards.