In talking with a friend, he suggests using low DCR chokes for the powers supply, on the order of < 10 ohms per choke.
I am currently using 10H/225mA/51 ohm chokes. In your experience, will changing chokes to a 5 ohm DCR result in better transients and dynamics compared to the 51 ohm chokes? I would use the same brand/specs, Monolith Magnetics but just lower the DCR to 5 ohms. I am trying to get a bit more snap / liveliness to the sound.
Thoughts?
I am currently using 10H/225mA/51 ohm chokes. In your experience, will changing chokes to a 5 ohm DCR result in better transients and dynamics compared to the 51 ohm chokes? I would use the same brand/specs, Monolith Magnetics but just lower the DCR to 5 ohms. I am trying to get a bit more snap / liveliness to the sound.
Thoughts?
Attachments
The C74/10H220 choke has the largest core that Monolith uses in "standard" choke series.
If you see the picture
http://www.monolithmagnetics.com/sites/default/files/ps chokes datasheet.pdf
the bobbin is full, so on this core unable to winding with thicker wire (aka lower DCR).
Lower DCR and same inductance (and same saturation current) requires designing new choke with another core (another saturation density).
If it's would be possible easily, Monolith would have already done... but it's not easy.
If you see the picture
http://www.monolithmagnetics.com/sites/default/files/ps chokes datasheet.pdf
the bobbin is full, so on this core unable to winding with thicker wire (aka lower DCR).
Lower DCR and same inductance (and same saturation current) requires designing new choke with another core (another saturation density).
If it's would be possible easily, Monolith would have already done... but it's not easy.
To fill a same bobbin window with thicker wire, the choke has less inductance/Idc/Vrms capability. Mathematically this is linear and it's up to the constructor how he will distribute these parameters.
Most of my clients report positive results from lesser Rdc chokes, although I find it has to do more with increased core and copper mass. Heavier inductors tends to shift tonal balance towards lower frequencies, giving a tonality of a fuller, bigger and authoritative sound.
Small chokes can be tricked to sound somewhat big using specific impregnants. Each impregnant has different tonalities, so it's a long research.
Most of my clients report positive results from lesser Rdc chokes, although I find it has to do more with increased core and copper mass. Heavier inductors tends to shift tonal balance towards lower frequencies, giving a tonality of a fuller, bigger and authoritative sound.
Small chokes can be tricked to sound somewhat big using specific impregnants. Each impregnant has different tonalities, so it's a long research.
Last edited:
I have reached out to Monolith for their thoughts on a custom choke, so my expectation is that this can be achieved. The question is, what, if any, sonic impact should be expected?
50AE - thanks for your thouhts on tonal impact. Any impact on "liveliness" / transient "snap"?
50AE - thanks for your thouhts on tonal impact. Any impact on "liveliness" / transient "snap"?
Sorry if all y'all are going to choke on this, but except for the E=IR voltage drop across a power supply choke, the dynamics limitations of the power supply is 99% due to the final (last stage) reservoir capacitor size. Quite simply, without the 'output reservoir' capacitors, the chokes in the power supply would ridiculously squash amplification itself. Because a choke's dynamic impedance is a linear reationship:
Which of course (sigh, again…) brings up the surprisingly enormous advantages of having active regulation instead of chains of chokes and resistors, both to suppress rectification noise, and to endow the power supply with line (wall) voltage variation independence. Cheap, simple, reliable, much smaller footprint.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Z = 2πFL
. Higher the frequency, the higher the impedance (L×F). Contrarily, for capacitors, Z = 1/(2πFC)
, which follows the inverse of an inductor. The higher the frequency, the less a capacitor imposed impedance onto the downwind devices demanding power. Sure, the series resistance of a choke also imparts its own power-supply droop on big current transients. With the judicious choice of a big fat can on the output, most all of that is mitigated. Which of course (sigh, again…) brings up the surprisingly enormous advantages of having active regulation instead of chains of chokes and resistors, both to suppress rectification noise, and to endow the power supply with line (wall) voltage variation independence. Cheap, simple, reliable, much smaller footprint.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Diyaudio member "Blitz" has a lot experience with different chokes and can give you, if he wishes, a thorough accumulation of subjective experiences with different values of Rdc chokes on his system.
As for my years of experience with cored inductors and transformers, "liveliness" and "snap" are a result of everything - core material, gap material, wire copper grade, insulation grade, interwinding insulation type, eventual impregnant, mounting elements, etc.
Of course, I need to understand your definition of liveliness and snap. Adding mylar will bring your medium to high frequency brightness together with a specific sense of speed and rock n roll. Overdoing with mylar can result in harsh timbres, especially when mixing with HiB and grade 2 enamelled copper wire.
Amorphous on the other end can sound "rolled off" and "dark". Mixing it with Nomex 410 can bring additional darkness. But some clients love it, especially when it counter balances a harsh sounding system.
So, things are getting complex. My favorite choke is the nanocrystalline + soft copper wire + impregnated paper. Although I haven't played yet with nickel cored chokes.
Last edited:
GoatGuy - I have a 50uF Russian PIO as my last cap in the power supply.
50AE - For reference, I listen 99% of the time to chamber music. Solo violin, viola, cello. String quartets. Trios and duos. My system has a very evenly balanced tonality, top to bottom. Actually, I find the current tonal balance to be the best I have heard in any system in 30+ years of hi-fi. Strings have great harmonics and tonality. So, what is missing? The aggressive snap of live, acoustic instruments. Take for example a violin. In person, a violin can be very aggressive (yet sweet sounding at the same time) and keep you on your seat. A pizzicato string pluck should be alarming. fffforte should be in your face, yet smooth or raw depending upon the intention of the musician. A live snare drum has a "snap" to its sound. Non-compressed. My amps are very good, but not great at the non-compressed sound. I am seeking that last bit of dynamic snap to sound lifelike.
As a frame of reference, my OPT is a Monolith Magnetics SX-11 non-crystalline. Fantastic. I use Dave Slagle's great sounding nickel interstage transformers. Love the IT tonality. Nickel is great.
Keep the input coming folks.
50AE - For reference, I listen 99% of the time to chamber music. Solo violin, viola, cello. String quartets. Trios and duos. My system has a very evenly balanced tonality, top to bottom. Actually, I find the current tonal balance to be the best I have heard in any system in 30+ years of hi-fi. Strings have great harmonics and tonality. So, what is missing? The aggressive snap of live, acoustic instruments. Take for example a violin. In person, a violin can be very aggressive (yet sweet sounding at the same time) and keep you on your seat. A pizzicato string pluck should be alarming. fffforte should be in your face, yet smooth or raw depending upon the intention of the musician. A live snare drum has a "snap" to its sound. Non-compressed. My amps are very good, but not great at the non-compressed sound. I am seeking that last bit of dynamic snap to sound lifelike.
As a frame of reference, my OPT is a Monolith Magnetics SX-11 non-crystalline. Fantastic. I use Dave Slagle's great sounding nickel interstage transformers. Love the IT tonality. Nickel is great.
Keep the input coming folks.
I am with Alex: size matters. Low henry/low dcr/large amorhous cores work for me. Russian pios do not.
Member
Joined 2009
Paid Member
And what’s the DCR of the OPT primary, is it similarly low?
And the first choke is in series with the rectifier, how is the dynamic resistance compared with the choke DCR?
And the first choke is in series with the rectifier, how is the dynamic resistance compared with the choke DCR?
Member
Joined 2009
Paid Member
The OPT reflected 8R load is 5k, but that’s not the DCR.
My thinking is, why kill yourself for even lower DCR of the choke (I personally think your existing choke is fine) if the output tube is drawing current through an OPT with a DCR of say 80R?
You may want to explore SS rectifiers, very low dynamic impedance but B+ will be higher and that makes comparisons difficult.
My thinking is, why kill yourself for even lower DCR of the choke (I personally think your existing choke is fine) if the output tube is drawing current through an OPT with a DCR of say 80R?
You may want to explore SS rectifiers, very low dynamic impedance but B+ will be higher and that makes comparisons difficult.
Usually nanocrystalline delivers the goods, snap + transparency and clarity. HiB also sounds snappy, but a bit smeary in the midrange microdetails.
I start to think your situation is a bit specific and you don't have to necessarily play with the Rdc of chokes. You might try adding some bright sounding alloys to your system, such as aluminum alloys or brass in the role as feet, cable lifters or pieces on top. Try playing with single stranded hard copper cables, such as UP-OCC Neotech. Such stuff brings "in your face". If you have one of these bamboo cutting boards in your kitchen, try placing one beneath your source/preamp.
Last edited:
Capacitors, REALLY big ones.
Thank you, Esteemed sir.
I know I'm well on the way to being declared Anathema, but I feel on this subject similarly to how I feel on hand-braided silver-over-single-crystal Litz wire speaker cables.
It is so much better to forego the absurd expenses, and just throw 'more wire' at things.
Same goes for the output reservoir.
The author concedes a 50 µF 'final'.
Myself, I would have hunted down a nice big 1,300 µF 500 V KEMET (ALS70A132KE500) part, $25 or so apiece, QTY 1. Could put the PIO in parallel with it. Thing tho' is, with a substantial capacitor (holding about 100 joules), just about any train of acoustically well-recorded transients will come out crystal clear and unadulterated by IMD or most other parasitic power-supply related shortcomings.
Capacitance rules, folks.
Inductance's sole role is to be a much less 'lossy' and 'frequency squashing' to hum, buzz and noise. Much less lossy as compared to a resistor, and unlike a resistor, also self-scaling with frequency.
Thank you, Esteemed sir.
I know I'm well on the way to being declared Anathema, but I feel on this subject similarly to how I feel on hand-braided silver-over-single-crystal Litz wire speaker cables.
It is so much better to forego the absurd expenses, and just throw 'more wire' at things.
Same goes for the output reservoir.
The author concedes a 50 µF 'final'.
Myself, I would have hunted down a nice big 1,300 µF 500 V KEMET (ALS70A132KE500) part, $25 or so apiece, QTY 1. Could put the PIO in parallel with it. Thing tho' is, with a substantial capacitor (holding about 100 joules), just about any train of acoustically well-recorded transients will come out crystal clear and unadulterated by IMD or most other parasitic power-supply related shortcomings.
Capacitance rules, folks.
Inductance's sole role is to be a much less 'lossy' and 'frequency squashing' to hum, buzz and noise. Much less lossy as compared to a resistor, and unlike a resistor, also self-scaling with frequency.
Z = 2πFL
Again, thank you, RayMa
Might as well also say that in a tube amp, a large power supply inductor has always degraded the sound
in my experience, and by a lot. Silicon diodes and large filter capacitors (~1000uF) sound much better.
in my experience, and by a lot. Silicon diodes and large filter capacitors (~1000uF) sound much better.
Last edited:
I was just thinking about this, the music will find most of its return path through that last capacitor, it isn't going the go crawling up to the choke right? Too much impedance when it can just zip through the last can.
Maybe this is an argument for using the relatively new DC link capacitors for the final cap. They have amazing specs and wont "get stale" in 10 years like electros or have the heat issues.
https://www.mouser.com/ProductDetail/KEMET/C44UOGT7150M31K?qs=Jslch3jnSjm5mIVLBQWAdw==
https://www.mouser.com/datasheet/2/212/1/KEM_F3119_C44U_M-1830208.pdf
There are other nearby values in stock, or lower values in parallel for even lower ESR and higher ripple current. But they are items for that "money is no object" project.
