Taiko Audio SGM Extreme : the Crème de la Crème

[romaz]

Just to share my experience with the Tranqulity Bases:



I own several and so I've tried them under various amplifiers including a Soulution 711 (which has a SMPS) and they made no worthwhile difference. Under various preamps (both tube and transistor), I heard small differences. They seem to make the biggest difference under digital gear like DACs and servers. Not unexpectedly, they make a difference under networking equipment such as an EtherREGEN and my cable modem but the difference isn't huge.

 

[Steve Williams]

You might want to consider putting the footers between the Extreme and the platform. you might even want to listen with just the footers alone

 

[romaz]

@romaz and @Ultrafast69, could you share your experience in terms of using the Extreme with the Stromtank (please specifiy model) vs directly? Thanks

As for the quality of power delivery, this is where I find it most effective to spend time and resources with the Extreme. I have found the Extreme to respond more favorably to my Sound Application TT-7 power line conditioner (PLC) than any other component I own. After that, it's my DAC then my preamp and finally my monoblocks.

Back in 2016, I heard a demo of the Stromtank S5000 HP at RMAF in the D'Agostino room and I was impressed by how the noise floor dropped, dynamics improved, and soundstage became more expansive. I brought in an S2500 convinced it would improve my system in the same way.



I brought in the Stromtank S2500 before I owned the Extreme but I tried it with various pieces of equipment including my DIY server at the time, an Aurender, an Ayon Spheris III tube preamp, D'Agostino HD preamp, Soulution 711, and D'Agostino M400 monoblocks. As the S2500 can output about 550 watts continuously, at low volumes, this small device handled any of these pieces just fine but compared to connecting this same equipment into my TT-7 PLC, I heard no improvement. Zero. It appeared to me and to a group of audiophiles I had over that this unit performed no better than my TT-7. If anything, with respect to transient response, the transients through the Stromtank sounded a bit soft to me and so overall, it was a step backward and this was the case whether I had everything plugged in including my amps and preamp or if I had just the server or DAC or server + DAC plugged in.

The problem with batteries is that they discharge current through a chemical reaction but that chemical reaction takes time and so battery-based power supplies traditionally are hampered by high output impedance resulting in a soft sound and so you potentially trade low noise for soft transients which in my case is unacceptable.

The solution to counter this is to combine multiple battery cells in parallel and 2 batteries in parallel (2p) will have half the output impedance as a single cell and so you can go to 4p, 8p, etc. to drop your output impedance further. Having now built a few of my own DIY Stromtanks, what's interesting is that it isn't my current giant Block Audio SE Mono Blocks that respond most favorably to low output impedance power, it was my DAC followed by the Extreme and then my preamp and this may be because my monoblocks have a 500,000uF bank of energy storage capacitors to aid them. Regardless, it turns out my digital gear has more fluctuant power demands than my analog gear and benefit more from low impedance power.

I use the storage capacity of a battery to give me a relative idea of the output impedance of a battery array and so the S5000 HP that I heard and liked back at RMAF has a storage capacity of 5000Wh. The S2500 I brought into my home which I didn't find useful has a storage capacity of 2500Wh and so this unit likely has 2x the output impedance of its bigger brother. The S1000 is even worse (5x higher). Even if you were going to plug in only a low powered DAC and the Extreme into a Stromtank, I would personally bypass the lower models and go for the S5000 HP.

At the present time, despite my DIY efforts at building my own battery-powered inverter / pure sine wave generator, my best sound comes from my TT-7 PLC plugged directly into mains power although I have fairly big 6awg in my walls and my panel outputs a rock steady 123V. Despite using an array of batteries with an output impedance as low as 10 milliohms, I am still getting a touch better dynamics with a TT-7 plugged straight into the wall and so nothing is guaranteed with a Stromtank (be sure you have the option to return it). If you are plagued by unstable or low voltages or wildly varying sound quality at various times of the day or night, the upsides of a Stromtank may be worth any downsides.

 

[romaz]

You might want to consider putting the footers between the Extreme and the platform. you might even want to listen with just the footers alone

Yes, I have tried the CS2 footers between Extreme and the SR Tranquility Base, a maple plinth, an HRS base, and a Daiza platform and they all sound different with the CS2 + Daiza sounding the most natural in my system. However, the EMI impact of the Tranquility Base is great enough to trump what the Daiza offers. When I find the time, I will order another Daiza and stack it on top of the Tranquility Base but right now, these are low priority issues for me.

 

[romaz]

BTW I should add the whole BPS idea actually came to fruition after discussions with a few WBF members we should credit, @romaz , @austinpop , @nenon

Having commissioned Sean Jacobs to build me a "one off" battery-powered DC4 supply for my Chord DAVE DAC in July of 2020, I was smitten by the novelty that I could now take my DAVE to remote areas that had no power. Using a Li ion battery pack, compared to the DC3 mains-powered PSU I was previously using for my DAVE, I was impressed by the lower noise floor but dismayed by the softer transients. It's the reason that a battery-powered laptop functioning as a music server does not sound good to me. I'll take mains noise over soft transients any day.

Collaborating with @nenon, @ray-dude, and @austinpop, we began playing with various lithium-based battery arrays and realized that as we paralleled more and more battery cells, the output impedance of the battery array dropped and dynamics improved and the soft transients diminished and to our surprise, it was DAVE's 5V digital rail that benefited more than its pair of 15V analog rails. With the aid of a bank of Mundorf capacitors, we were able to get output impedance close to 1 milliohm for DAVE's 5V rail and we were thrilled with the results. It was this effort that formed the foundation for what is now the ARC6 DC4.

This lead me to wonder what a battery-powered Extreme would sound like especially when mated with Emile's Ga-N (gallium nitride) DC-ATX converter that he had developed for the DIY market and so back on March 27, 2021, I privately asked Emile if the Extreme could be powered by a giant 24V battery and his response was "Maybe! That would be really giant..."

On April 19, 2021, I told Emile I was entertaining the idea of ordering an Extreme without a power supply with the intention of experimenting with batteries and to my relief, he expressed interest in taking this on and he took it on in a way that only he can, with insane passion.

Emile experimented with different battery chemistries and put together massive arrays:



He found batteries have their own noise spectra and what was the most damaging was the BMS (battery management systems) implemented by most battery manufacturers. BMS is responsible for protecting the battery cells against over and under voltage situations, overcharging, and everything else that could result in instability that could lead to thermal runaway. Having experimented with raw battery cells without a BMS, I can attest that these batteries can catch fire or at the very least "puff up" and start to smoke which can be quite scary. The problem with traditional BMS is they generate noise in the millivolt (mV) level and so they mask the low noise benefits (at the nanovolt level) of lithium batteries. This may be one of several reasons why I found a Stromtank to not sound "blacker" than my TT-7 connected to mains and so Emile devised his own BMS to his own standards and so his upcoming BPS for the Extreme is supposed to have noise in the nanovolt range. To give you perspective, 1 millivolt of noise is 1,000,000x more noise than 1 nanovolt of noise. Further to that and just as important to me, the BPS for the Extreme is supposed to have an output impedance of about 1 milliohm including the DC leads which is just phenomenal.

As for ease of use and monitoring, I think this graphic on a smart phone pretty much says it all:

 

[DW101]

...my mind has gone there too. May I ask what did you use for the cages? Copper mesh looked to have potential to me, providing shielding and air flow. For mission-critical applications, they use gasket, grommets, etc. to completely block radiation, but that never looks reasonable to do.

Does anyone know: if there are gaps in a Faraday shield, does it concentrate energy? Must a cage be 100% "sealed" or can we take what we can reasonably get without extraordinary methods.

I am no expert, I just used cheapo 'rodent proofing' wire mesh, I am sure your copper would be good.

IMO I wouldn't worry too much about perfection to start with - I advocate just giving it a go and see if it helps. At 5GHz the wavelength is about 6cm / a couple of inches (299,792,458 m/s divided by 5,000,000,000 Hz) so poking a cable though a small slit is OK.

As per @oldmustang I found rapid prototyping with aluminium foil to be a good idea too, to identify targets before figuring out something more domestically and psychologically acceptable, whilst quickly being able to rid yourself of the evidence if you start to feel crazy!

I use Audiowise products (OPTO DX, DC BLOCK, and SRC DX - I'm pleased with all of them), but after trying a quick 'aluminium foil wrap' type experiment on my DAC I didn't hear anything to warrant exploring their wrap or boxes at this time in my set up.

For disclosure, my Extreme hasn't arrived yet (soon?) so please take everything I say in that context.

 

[SwissTom]

The problem with traditional BMS is they generate noise in the millivolt (mV) level and so they mask the low noise benefits (at the nanovolt level) of lithium batteries. This may be one of several reasons why I found a Stromtank to not sound "blacker" than my TT-7 connected to mains and so Emile devised his own BMS to his own standards and so his upcoming BPS for the Extreme is supposed to have noise in the nanovolt range.

Very interesting post. Thank you.

This suggests the BPS Extreme will outperform the Stromtank in terms of noise level, if not in raw power delivery/peaks.

I'm already preparing its spot on the rack

 

[matthias]

This suggests the BPS Extreme will outperform the Stromtank in terms of noise level, if not in raw power delivery/peaks.

Further you get rid of two additional stages:
DC to AC in the Stromtank and AC to DC in the Extreme
The upcoming battery powered Extreme is very different to a combo of Stromtank and existing Extreme, it is much more than a "better" Stromtank.

Matt

 

[MarcelNL]

nice experimenting! DO read up on the saga of Jean Hiraga's 'Le Monstre' class A amplifier (my first DIY audio project way back)

The Class-A Amplifier Site - Hiraga 'The Monster'

sound-au.com

power is king in audio, it's no different in digital

 

[Kris]

Hi @Kris. I recall you have advocated NOT using wifi for a while now. I looked at the adapter/converter pieces to run the TAS "control" tablet wired instead of using wifi, but decided against it.

As I recall, you don't stream, which makes eliminating wifi a bit more practical. Do I recall that correctly?

I'm not opposed to the bother, but I can see so many other wifi signals on that tablet, that I thought it might not be worth the trouble, as the wifi soup is everywhere. I felt I would have to put the front end in a Faraday cage. My wife is a scientist, but I'm not sure that would fly (although I have actually looked into sourcing copper mesh to build a cage, in the interest of science, of course).

So boiled down: Is the wifi I use for the control tablet audio system worse than the general wifi beaming around?

I stream a lot .
simply everything is connected with ethernet cables.
also to the modem and router. Router is wired only.
I can connect access points to this network , but prefer not to.
I used wired ipad and wired phone as well as an remote unit,

 

[Kris]

Hi @Kris. I recall you have advocated NOT using wifi for a while now. I looked at the adapter/converter pieces to run the TAS "control" tablet wired instead of using wifi, but decided against it.

As I recall, you don't stream, which makes eliminating wifi a bit more practical. Do I recall that correctly?

I'm not opposed to the bother, but I can see so many other wifi signals on that tablet, that I thought it might not be worth the trouble, as the wifi soup is everywhere. I felt I would have to put the front end in a Faraday cage. My wife is a scientist, but I'm not sure that would fly (although I have actually looked into sourcing copper mesh to build a cage, in the interest of science, of course).

So boiled down: Is the wifi I use for the control tablet audio system worse than the general wifi beaming around?

i created my listening room as a Faraday cage .
no wifi and no phone network inside that room.
also it has exclusive ground that I made in my backyard,
every component has its own balanced trafo to isolate it
do I miss anything .?

 

[VPN]

I used stromtank S5000 for 6 months.
Extreme was actually the only grear that benefited from powering from it.
Amps DAC power supplys to switches router modem WORSENED with S5000 .

So my prediction was that SQ of the system benefited because of isolating the extreme from other very sensitive devices.
powering extreme from different socket / exclusive line to extreme also improves SQ in a simmilar way.

But I was checking also at the same time BALANCED trafos for each of the devices and this is the best solution .
every gear is isolated form the others having its own 2.5 kva trafo
So I sold Stromtank and keep balanced trafos. Its much Better solution and so much cheaper !

Thank you! Which model/brand of balanced trafos did you purchase?

 

[Kris]

It is interesting, I have zero doubts Kris is having these results. Worthy of further investigation. Here I have no significant influence of 5G Wi-Fi, there is a minor difference between running the remote wired versus wireless, aswell as a minor difference between Android and Apple tablets. But completely disabling wireless is inaudible here. Since we added more access points to have a strong signal everywhere in the building (no retransmits), 5G outperforms 2.4G which did better before but apparently due to its longer range causing less retransmits with a worse access point coverage.

I wish I could have that .
unfotunate there must be something here that adds a lot of bad noise when wifi is in use,
I am willing to test everything.

one thing that came to my mind is the grounding that is gigantic structure in my backyard
and I remember you mentioned it might pick everything as an anthenna.
other think is the power that is very poor in my area with everyday turn off for a second or two,
Stromtank was not inmproving the SQ , but balanced trafos do improve SQ.
I know you have excellent commercial power lines . I dont.
I will list potential issues and test them ,
thats the only way I believe.

 

[Kris]

Thank you! Which model/brand of balanced trafos did you purchase?

I purchesde Airlink trafos from UK.
each is 2,5 kva.
big and heavy.
added DC traps as they tend to make a noise from time to time.

 

[cat6man]

I purchesde Airlink trafos from UK.
each is 2,5 kva.
big and heavy.
added DC traps as they tend to make a noise from time to time.

Do you have a diagram of how you use the transformers in your system?

 

[Kris]

This is very easy.
It is listed on the sticker of the transformer whan it arrives.
So no way to connect it wrong.

Phase incoming is in my case 240v.
Primary are :
There are 3cables to connect . Brown (phase) Blue (neutral) , yellow ( ground). As for any device.

The secondary are
Red 120v + in phase
Black 120v - in phase
Yellow is Zero ( can be connected to ground)

So you have 3 cables on the primary side and 3 cable on the secondary.

If you have differet voltage they can make any trafo you want.
I ordered mine with M4 core and EC + EMI screens+ varnish impregnation that is very good in audio applications.

 

[Taiko Audio]

This is very easy.
It is listed on the sticker of the transformer whan it arrives.
So no way to connect it wrong.

Phase incoming is in my case 240v.
Primary are :
There are 3cables to connect . Brown (phase) Blue (neutral) , yellow ( ground). As for any device.

The secondary are
Red 120v + in phase
Black 120v - in phase
Yellow is Zero ( can be connected to ground)

So you have 3 cables on the primary side and 3 cable on the secondary.

If you have differet voltage they can make any trafo you want.
I ordered mine with M4 core and EC + EMI screens+ varnish impregnation that is very good in audio applications.

Please be careful with that, especially with how to configure the secondary side. The equitech website has a good easy to read piece on the balanced power topic here:
https://equitech.com/history/

 

[Onepoint5]

This is very easy.
It is listed on the sticker of the transformer whan it arrives.
So no way to connect it wrong.

Phase incoming is in my case 240v.
Primary are :
There are 3cables to connect . Brown (phase) Blue (neutral) , yellow ( ground). As for any device.

The secondary are
Red 120v + in phase
Black 120v - in phase
Yellow is Zero ( can be connected to ground)

So you have 3 cables on the primary side and 3 cable on the secondary.

If you have differet voltage they can make any trafo you want.
I ordered mine with M4 core and EC + EMI screens+ varnish impregnation that is very good in audio applications.

The yellow from the secondary, must be connected to the incoming earth wire. In addition, all the shields at the transformer to the same earth point as well. Can't emphasise enough for the balancing to work and for safety requirements, the centre point on the secondary needs to be at earth potential. Connect the equipment earthing points to the same earth terminal.

If you use two transformers to split the load, perhaps to sources and a different transformer to power amps, connect the secondary earth point to each others, keep this wire short or the same length to avoid voltages differences between the two connections. Star earthing always works, in this case, works well.

When finished on the install, measure the red to earth and black to earth, should be very close within 1-2 V, the lower the better for noise cancellation.

For protection, a GFCI/ELCB/RCD 2 pole device needs to be wired from the secondary to the loads, since the primary ground fault interrupter from the house can't see the secondary.

 

[Markus87]

Thanks a lot @romaz for giving us more insights on the battery investigations, very interesting read!!
Emile, that reminds me of the (product) roadmap you have been writing here around Munich HighEnd. Is there already a first (draft) version you could share with us? I am really curious and excited about the upcoming releases!

 

[Taiko Audio]

Thanks a lot @romaz for giving us more insights on the battery investigations, very interesting read!!
Emile, that reminds me of the (product) roadmap you have been writing here around Munich HighEnd. Is there already a first (draft) version you could share with us? I am really curious and excited about the upcoming releases!

It's quite elaborate to catch in an easy to read overview due to the sheer number of projects, I estimate it to be ready in about 2 weeks.