Building an Unregulated Linear Power Supply

[Gavin1977]

Parallel Hammond chokes… anyone tried? 🤪

 

I’m guessing it would allow more current headroom?

[MarcelNL]

I'm using Lundahl 2773 chokes (5A version) in parallel mode now so capable of 10A. Problem with paralleling the Hammonds is that the core is not really big enough to prevent saturating at 5A IMO (I'm not sure they can carry 5A continuous). Keep in mind the inductance halves when paralelling two identical chokes (well you'd need 4 Hammonds).

 

A comparison between the Lundahls with 16mH @10A versus paralleled Hammonds could be informative...intuitively I'd expect the Lundahls to outperform the Hammond chokes.

 

 

[Gavin1977]

Just now, MarcelNL said:

I'm using Lundahl 2773 chokes (5A version) in parallel mode now so capable of 10A. Problem with paralleling the Hammonds is that the core is not really big enough to prevent saturating at 5A IMO (I'm not sure they can carry 5A continuous). Keep in mind the inductance halves when paralelling two identical chokes (well you'd need 4 Hammonds).

 

A comparison between the Lundahls with 16mH @10A versus paralleled Hammonds could be informative...intuitively I'd expect the Lundahls to outperform the Hammond chokes.

 

 

Good stuff!  You planning on trying this comparison?

[MarcelNL]

Just now, Gavin1977 said:

Good stuff!  You planning on trying this comparison?

No I'm not, simply because I think the Hammond will underperform....I;d be willing to do the experiment but for the pain of connecting up double 14 AWG wire to the Lundahls again...using some jumper wires and provisional connections it could be done on the bench and if there would be gatherings like ETF geared to computer audio I'd be game to do this as a listen and learn session.

[Gavin1977]

No worries - I’ll check out the Lundahl  2773

[MarcelNL]

3 minutes ago, Gavin1977 said:

No worries - I’ll check out the Lundahl  2773

https://www.lundahltransformers.com/wp-content/uploads/datasheets/2773.pdf

[Gavin1977]

I can’t see where you’ve written up any impressions of single vs parallel Lundahls… what subjective difference did you hear?

[MarcelNL]

6 minutes ago, Gavin1977 said:

I can’t see where you’ve written up any impressions of single vs parallel Lundahls… what subjective difference did you hear?

correct, as I changed the circuit THINKING my Taiko ATX module was toast I changed the chokes from CMR to parallel, started using the Taiko active rectifier and added the Taiko custom 220.00uF cap...so any subjective observation is pointless with  this many changes.

 

I am currently still waiting for the new LPSU to break in,, now some 550 hours into it and the sound is still changing for the better by the day..I suspect it's the 220.000uF monster cap breaking in....

Once that is broken in I plan to try one Taiko ATX for the ATX supply of the MB and the other ATX module on the CPU supply, initially both fed from the same 36V rail but if that proves to be an improvement I'll likely build another LPSU.

[MarcelNL]

on that same note; it sounds GREAT, I've never been closer to the live experience with regards to natural sound and leading edge trailing edge and dynamics. Currently there is nothing not to like, to the point where sleep deprivation is a reality.

[bit01]

On 10/12/2022 at 3:52 PM, Gavin1977 said:

Parallel Hammond chokes… anyone tried? 🤪

 

I’m guessing it would allow more current headroom?

I have the Hammond 159 ZJ (10mH/5A/.16Ohm rating) in parallel with their 159 ZG (15mH/4A/.25Ohm rating). I figure this gives the equivalent of 6mH/9A/.10Ohm. The i9-10900K/Audiolinux LXQt  based server would not operate otherwise (unless I tweaked to limit power setting perhaps, which I did not fancy). SQ is great, using J River MC29 with no DSP- no issues that I can report.

The configuration is in the form of the 'Nenon V3' with slight differences, feeding the Taiko DC DC ATX. The server currently has 16GB G.SKILL F4-3200C16 RIPJAWS V (a fav of mine) and a JCAT USB XE powered separately.

 

ATB

b.

[drjimwillie]

This is a go figure and keep the faith note.

Then, a question.

 

Back when the Taiko rectifier and big 220,000 uF became available I ordered it and it shipped 8/10. 
 

I do not like important packages left on my front porch so I always arrange to pick them up at the local distribution hub in New Jersey. I called FedEx and made my request and I could see on the website that they had changed the destination to my local distribution hub. however, after a couple of weeks I could see that my package had not made it out of Holland. When I contacted Taiko they said FedEx told them that I had requested my package be held in Holland.😳

I called Fed ex every couple of days and they told me that they had just scanned my package and they were bumping up my case and they would try to get it moving. Every time I called them it was a different story sometimes positive, sometimes doom and gloom. I was told more than once to put in a report that my package was lost and ask the supplier to reimburse me. 
I kept the faith and finally, two months later, my package arrived.

has this happened to anyone else?

 

The question I have is the big cap does not have the screws for the terminals that I need. The Taiko website says it is a M6 Bolt.  I am assuming I need a bolt that is the same material as the terminal, otherwise I will have dissimilar metals, which will cause electrolytic action and corrosion. Although, I did see someone use what looks like a brass screw.

 

What should I use and where can I get it in the US, please?

 

Thank you,

Will

[MarcelNL]

FedEx sucks, just as much as any other courier, and all of them will tell you anything if they messed up....

 

 

I am using brass but have red copper bolts on order, galvanic corrosion is an issue since there is no water to act as conduit.

https://en.wikipedia.org/wiki/Galvanic_corrosion

 

Aliexpress for red copper, brass should be easier.

 

[MarcelNL]

BTW, it;s ridiculous how this circuit is breaking in, I have never experienced breakin in this strong, day to day the sound is improving. NICE

 

[drjimwillie]

Hi Marce 

Thank you for the information.

 

I’m sorry, I am a little bit confused. 
 did you mean to say:

21 hours ago, MarcelNL said:

 

I am using brass but have red copper bolts on order, galvanic corrosion is an issue since there is no water to act as conduit.

https://en.wikipedia.org/wiki/Galvanic_corrosion

 

 

Isn’t there always water in humidity?

 

why wouldn’t you use the same metal that The terminal is made out of?

can we get them from Mundorf? I wonder why they did not come with the cap?

 

Thank you 

 

it is encouraging to hear how wonderful these parts sound!

These are the last parts I needed for the V3.

you can understand how I felt when the FedEx people were telling me that they did not know where the package was. They told me to have a another one sent. I made sure they understood that there were no more of this product and they needed to find it, ha ha.

i’m going to build the V3 in the Taiko chassis.😁 I have everything I need for the DIY Sage; except the CPUs. I have to save my pennies for a little while or sell some thing. it is going to be a little while before I get to it, because I am building TL-2 speakers. 
 

it’s all exciting. The wonderful thing is, I love how my system sounds now! I am listening to Joni Mitchell, Miles of Isles. I feel like I can touch the edges of her voice.

 

Please enjoy your day.

[Superdad]

7 hours ago, drjimwillie said:

I am listening to Joni Mitchell, Miles of Isles. I feel like I can touch the edges of her voice.

Nice!! 

[MarcelNL]

I have not ever seen galvanic corrosion on Elco screws, and I have used quite a few screw connector elco's for years in a Le Monstre amp.

In this build I've lapped the nuts and bolts and washers to ensure max surface availability and cleaned with IPA afterwards, when i get the red copper I will do the same and will report on any hint of corrosion.

Little moisture means little potential for galvanic corrosion, as long as you do not put the LPSU out in 90 or so % of humidity it'll all be OK I reckon, the alternative is using aluminium bolts and washers AND wiring(shudder). 

[Nenon]

I would not obsess too much about the bolts. 

Something like this should work just fine - https://www.mcmaster.com/92095A239/

Stainless steel is good for fighting corrosion and is also almost non-magnetic.

[Rovo]

As far as I can judged, the screws for the Mundorf MLytic HC are galvanized steel (a good quality, but nothing special nevertheless)

[Exocer]

Hi All,

So I remember a while ago I said I wanted to compare the Saligny power rectifier and Taiko rectifiers. In my own build, I got a bit lazy and jumped straight to the Taiko rectifier once I got my hands on it instead of trying the Saligny power first. The installation method of the Saligny power is why I chose not to try it first, since I would need to waste two high quality ring terminals to conduct the test.

 

Well, I am happy to report I have now performed the test, but in addition I got to compare V2 and V3 within minutes of each other in my system. I built the following V2 ULPS for a forum member who will step forward if they feel the need to:

 

This was before I shrank the shrink wraps, but let me tell you, this thing exceeded my expectations. Angelique wire used throughout (just like my personal build).

 

To summarize what i'm hearing:

 

• I personally do NOT consider the Saligny power rectifier to be a significant downgrade from the Taiko rectifier after hearing both back to back. I suggest anyone who missed out on the Taiko rectifier not to lose sleep over it.
• V2 is almost as dynamic as V3 (in my system).
• Yes, there is more clarity with V2. I would consider V2 to be more "direct" sounding.
• V3 sounds bigger and has more authority down low. The soundstage of V3 is never ending, but the soundstage of V2 is still sizeable and large. V3 has a bit more of a "grand" vibe to its presentation.
• I went with GX-16 4 pin output on this V2 build and I am hard pressed to hear any negative influences from this connector. My own build goes direct to Taiko DC-ATX without a connector and I while we all know the GX-16 4 pin isn't an "imperfect" connector I would use it again in these types of builds since I now truly consider its impact to SQ to be minimal or none at all in my system. I can send a direct link to the connector I used if anyone is curious.
• Both LPS's sound GREAT.

 

Would I be happy with V2 if I hadn't heard V3? absolutely.

 

Hopefully those on the fence find this useful. If you need to save some space, go with a V2 build. I used this chassis for the recent V2 build: "Slim Line 03/230 10mm BLACK front panel - 3mm aluminium covers"

 

Cheers all,

-Rob

[MarcelNL]

50 minutes ago, Exocer said:

Hi All,

So I remember a while ago I said I wanted to compare the Saligny power rectifier and Taiko rectifiers. In my own build, I got a bit lazy and jumped straight to the Taiko rectifier once I got my hands on it instead of trying the Saligny power first. The installation method of the Saligny power is why I chose not to try it first, since I would need to waste two high quality ring terminals to conduct the test.

 

Well, I am happy to report I have now performed the test, but in addition I got to compare V2 and V3 within minutes of each other in my system. I built the following V2 ULPS for a forum member who will step forward if they feel the need to:

 

This was before I shrank the shrink wraps, but let me tell you, this thing exceeded my expectations. Angelique wire used throughout (just like my personal build).

 

To summarize what i'm hearing:

 

• I personally do NOT consider the Saligny power rectifier to be a significant downgrade from the Taiko rectifier after hearing both back to back. I suggest anyone who missed out on the Taiko rectifier not to lose sleep over it.
• V2 is almost as dynamic as V3 (in my system).
• Yes, there is more clarity with V2. I would consider V2 to be more "direct" sounding.
• V3 sounds bigger and has more authority down low. The soundstage of V3 is never ending, but the soundstage of V2 is still sizeable and large. V3 has a bit more of a "grand" vibe to its presentation.
• I went with GX-16 4 pin output on this V2 build and I am hard pressed to hear any negative influences from this connector. My own build goes direct to Taiko DC-ATX without a connector and I while we all know the GX-16 4 pin isn't an "imperfect" connector I would use it again in these types of builds since I now truly consider its impact to SQ to be minimal or none at all in my system. I can send a direct link to the connector I used if anyone is curious.
• Both LPS's sound GREAT.

 

Would I be happy with V2 if I hadn't heard V3? absolutely.

 

Hopefully those on the fence find this useful. If you need to save some space, go with a V2 build. I used this chassis for the recent V2 build: "Slim Line 03/230 10mm BLACK front panel - 3mm aluminium covers"

 

Cheers all,

-Rob

 

Just a question out of curiosity, how many hours break in on each? (My experience is that that V3 needs an incredible amount of time and swings wildly in the first 500 or so hours of breaking in, more than I have ever experienced breaking in)

[Exocer]

1 minute ago, MarcelNL said:

 

Just a question out of curiosity, how many hours break in on each? (My experience is that that V3 needs an incredible amount of time and swings wildly in the first 500 or so hours of breaking in, more than I have ever experienced breaking in)

Good question. Thousands of hours on V3 since it is my daily driver. V2 has about 25 or so hours on it and already sounds amazing.

[Exocer]

Hi All,

Another user reached out after my previous post to request a V3 build so this gave me an opportunity to listen to 2 relatively similar V3 builds in my system within moments of each other. We have all learned that the 220,000uF cap takes an eternity to break in and I tried my best not to factor that into my impressions. This is a very educational test for those who are curious about whether the Taiko rectifier is substantially better than the Saligny power rectifier, all else equal.

 

This build and my personal build use Angelique wire, the same lugs, same Odams, chokes, transformer etc.

 

There was one difference though. This build uses the Aviation GX-16 4 pin plug and my personal build does not use a chassis connector. It is wired directly to the Taiko DC-ATX. Given the nature of these builds, it was very simple to add the chassis connector to my build as well to make the test more fair.

 

Here are my takeaways:

 

1. The Aviation plug receives a lot of flak here, but it did not detract much if anything from the sound quality. Perhaps my system is not transparent enough to recognize the distinction between going direct (better than any connector) and the aviation connector. For those using these connectors, try not to let the nervosa get to you. It is a solid sounding/functioning connector that will do the job.

 

2. The Taiko and Saligny power definitely sound "different", all else equal. Again, the Saligny power is more than up to the task.

 

Without getting too wordy: I don't know if I would be able to distinguish the two of these ULPS's if this were a blind test and both 220,000uf caps were broken in.

 

Since the test, I had added a Duelund .47uf bypass cap on my personal build, at the Mundorf AG cap. I like this A LOT more than the .10uF ODAM in that position. I am no sure if it due to larger size of the Duelund or its own sonic characteristics.

 

Happy holidays all!

-Rob

 

Here is that build: 

 

 

[Tubeman66]

16 minutes ago, Exocer said:

Hi All,

Another user reached out after my previous post to request a V3 build so this gave me an opportunity to listen to 2 relatively similar V3 builds in my system within moments of each other. We have all learned that the 220,000uF cap takes an eternity to break in and I tried my best not to factor that into my impressions. This is a very educational test for those who are curious about whether the Taiko rectifier is substantially better than the Saligny power rectifier, all else equal.

 

This build and my personal build use Angelique wire, the same lugs, same Odams, chokes, transformer etc.

 

There was one difference though. This build uses the Aviation GX-16 4 pin plug and my personal build does not use a chassis connector. It is wired directly to the Taiko DC-ATX. Given the nature of these builds, it was very simple to add the chassis connector to my build as well to make the test more fair.

 

Here are my takeaways:

 

1. The Aviation plug receives a lot of flak here, but it did not detract much if anything from the sound quality. Perhaps my system is not transparent enough to recognize the distinction between going direct (better than any connector) and the aviation connector. For those using these connectors, try not to let the nervosa get to you. It is a solid sounding/functioning connector that will do the job.

 

2. The Taiko and Saligny power definitely sound "different", all else equal. Again, the Saligny power is more than up to the task.

 

Without getting too wordy: I don't know if I would be able to distinguish the two of these ULPS's if this were a blind test and both 220,000uf caps were broken in.

 

Since the test, I had added a Duelund .47uf bypass cap on my personal build, at the Mundorf AG cap. I like this A LOT more than the .10uF ODAM in that position. I am no sure if it due to larger size of the Duelund or its own sonic characteristics.

 

Happy holidays all!

-Rob

 

Here is that build: 

 

 

Which model of Duelund did you choose?

[Exocer]

8 minutes ago, Tubeman66 said:

Which model of Duelund did you choose?

The one that was on sale at PartsConnexxion. Duelund Tinned Copper.

[MarcelNL]

ever try a Jupiter copper beeswax? IMO that is one of the nicest regularly available caps.

I'm now using a vintage Sprague Vit Q high voltage PIO in that position and I like that a lot (they are pretty rare these days)