Gallium

[Ralf11]

I've posted a few times about the advantages of Ga- based semiconductors, but it looks like we are getting close to something worthwhile for those of us who do not launch packages on satellites ...

 

There's this:

http://epc-co.com/epc

 

and... 

https://rbtereport.com/solid-state-power-amplifiers-market-2018-2023-analysed-by-business-growth-development-factors-applications-and-future-prospects/

 

not to mention:

https://www.theverge.com/circuitbreaker/2018/10/25/18022106/anker-powerport-atom-pd-1-27w-usb-c-brick-gallium-nitride

 

yes, it is only $30.

 

Besides greater efficiency for the Class D aficionados, Ga-x could ultimately lead to a nice amplifier based on a Nelson Pass design but capable of driving real speakers...

 

I know @jabbr and some others have First Watt or Nelson Pass amps...

 

maybe now, the lowly neonatal care docs can get something too...

[jabbr]

1 hour ago, Ralf11 said:

 

Besides greater efficiency for the Class D aficionados, Ga-x could ultimately lead to a nice amplifier based on a Nelson Pass design but capable of driving real speakers...

 

I know @jabbr and some others have First Watt or Nelson Pass amps...

 

You could try building a variation of a “Zen” amp using s GaN FET ... or see if anyone on the Pass subforum at DIYAudio has looked at that —- building amps with an unusual FET is right up their alley.

[sandyk]

1 hour ago, Shadders said:

Yes, and you have never been the same since...... 

 

Perhaps. It was quite a revelation at the time to see an Esaki diode in a very simple setup be tuned into an Analogue VHF TV station and demodulate decent quality FM audio..

[Shadders]

1 hour ago, sandyk said:

 

Perhaps. It was quite a revelation at the time to see an Esaki diode in a very simple setup be tuned into an Analogue VHF TV station and demodulate decent quality FM audio..

Hi.

Gallium was always high frequency implementations - i wonder if they produce transistors applicable for audio. Is there a reduced base width modulation at audio frequencies ?

Regards,

Shadders.

[jabbr]

1 hour ago, Shadders said:

Hi.

Gallium was always high frequency implementations - i wonder if they produce transistors applicable for audio. Is there a reduced base width modulation at audio frequencies ?

Regards,

Shadders.

 

The FirstWatt series of amps, most of which have published schematics, as well as build threads over at DIYAudio contain many of Nelson Pass experiments using Power Switching transistors for audio. I have SemiSouth SiC transistors in my J2 and SIT-1 and the DEFSIT/SIT-3 uses unique combo.

 

So GaN probably could be made to work but I haven’t studied the data sheet enough, I suspect someone will try it if for no other reason than bragging rights...

[Ralf11]

8 hours ago, Shadders said:

Hi.

Gallium was always high frequency implementations - i wonder if they produce transistors applicable for audio. Is there a reduced base width modulation at audio frequencies ?

Regards,

Shadders.

 

high freq. like... Class D??

 

Ga-N & Ga-arsenide are both expanding rapidly into the ... ah... silicon space, so we will see where it goes, and how quickly

[Panelhead]

  I have seen solid state single ended amps using “doorknob” antique gallium transistors. Not everyone likes doped silicon transistors.

 

[marce]

14 minutes ago, Ralf11 said:

 

high freq. like... Class D??

 

Ga-N & Ga-arsenide are both expanding rapidly into the ... ah... silicon space, so we will see where it goes, and how quickly

Texas are pushing SMPS devices with gallium, may also be class D devices I'd have to read back through the endless emails they send out.They allow higher switching frequencies so noise etc is in higher frequency ranges. Layout is critical, very high Di/Dt, very short, wide copper pours in the switching loops (of course not to large otherwise you can capacitivly couple the noise to nearby circuitry) to minimise inductance and get the minimal hassle from skin effect, current density will be greatest at the bottom of the copper shape nearest the ground plane (next layer down, though switching currents will have their own ground on the top layer often a group of vias underneath the device linking this to the main ground plane.)

[Shadders]

1 hour ago, Ralf11 said:

 

high freq. like... Class D??

 

Ga-N & Ga-arsenide are both expanding rapidly into the ... ah... silicon space, so we will see where it goes, and how quickly

Hi,

I should have added in RF area - predominant for the last 25 years. When i was at Uni in the early 90's, 10GHz clock speeds were implemented using silicon on sapphire or silicon on diamond substrates.

Regards,

Shadders.

[phosphorein]

3 hours ago, Shadders said:

Hi,

I should have added in RF area - predominant for the last 25 years. When i was at Uni in the early 90's, 10GHz clock speeds were implemented using silicon on sapphire or silicon on diamond substrates.

Regards,

Shadders.

We used GaAs fet amplifiers in high speed microwave pulse circuits, but were tricky in minimizing noise at high bandwidths.

[avta]

I believe Merrill's new Element Amp uses Gallium.

[Ralf11]

Looks like  it: Using a zero feedback, open loop proprietary design, code named ZXOL from Merrill Audio, the Power amplifiers get the most details available and the fastest response. The open loop design takes advantage of the very latest exotic transistor technology using Gallium Nitride. Extracting and keep this accurate speed and huge amount of detail in the musical reproduction required a circuit board that has almost all its parasitic inductance and capacitance reduced to near zero. Using expertise developed over a decade the extra special Circuit boards were designed and a unique pure copper onboard heatsink for very fast thermal transfer and stability. The pure copper heatsink the power supply and amplifier weighs a hefty 16 pounds.

 

$36,000

[mansr]

1 hour ago, Ralf11 said:

extra special Circuit boards

Ooh, I'm impressed.

[Ralf11]

well... marce noted some design restrictions, and who would want to divulge trade secrets?  OTOH, it could be BS (unlike all other audiophool products)

[mansr]

4 minutes ago, Ralf11 said:

well... marce noted some design restrictions, and who would want to divulge trade secrets?  OTOH, it could be BS (unlike all other audiophool products)

It's just an audio amp. Nothing that special. In fact, I'd be surprised if this thing measured as well as a good $2000 amp. Intentionally added distortion is the only way to differentiate these days.

 

Also, why did they cheap out and use copper for the heatsink rather than silver? Or diamond.

[Ralf11]

I like heat pipes...

[marce]

15 hours ago, Ralf11 said:

well... marce noted some design restrictions, and who would want to divulge trade secrets?  OTOH, it could be BS (unlike all other audiophool products)

Gonna have a look, but firstly you cannot reduce the parasitic's to near zero, there will always be some inductance and capacitance. Copper heatsink could be bottom mounted surface mount devices or a copper slug in the board (the latter used in RF quite a bit, so nothing new there.

[mansr]

9 minutes ago, marce said:

Gonna have a look, but firstly you cannot reduce the parasitic's to near zero, there will always be some inductance and capacitance.

Depends on your definition of near, I suppose.

[marce]

Would like to see a pic of the PCB, havn't found one... Are they using Gal transistors or just Gal Fet drivers in the switching stage. Its a class D design in a nice case, of course being class D it will suck...

[marce]

1 minute ago, mansr said:

Depends on your definition of near, I suppose.

LOL

The trouble is with PCBs you lower one and increase the other, though in a class D or any design especially PSU's the parasitic's can be used to improve the design...

 

[Ralf11]

ELEMENT Audio has announced a new Ga based amp at a 'moderate' price - only $15k, which is half the price of their top of the line Ga amp.  The 114 is a stereo amp with 200 W/channel.

They are claiming zero distortion, and have an interesting power supply design.

 

No info yet on MQA...

[lucretius]

On 10/26/2018 at 6:40 PM, Ralf11 said:

I've posted a few times about the advantages of Ga- based semiconductors, but it looks like we are getting close to something worthwhile for those of us who do not launch packages on satellites ...

 

https://www.technics.com/us/products/r1/se-r1.html

[DuckToller]

I have seen Java has eGan-Fet in their JAVA integrated amplifier (from US$7995)
http://www.javahifi.com
Best, Tom

[marce]

1 hour ago, Sam Lord said:

 

GaN power transistors.  They didn't show the PCB of a new design and first implementation of GaN power transistors in consumer audio?  Shocker.

 

It will suck?  How about you listen and tell us how much? 

 

Well thank you for your reply, why would I bother listening to a class D amp, class D sucks... And thank you for re-opening a year old thread just to have a misguided dig!

Around the time of this post, we were discussing the fact that class D sucks on numerous other threads, so you do not understand the sardonic note of my reply...

https://audiophilestyle.com/forums/topic/54656-powered-speakers-suck-because-of-class-d/?tab=comments#comment-891970

https://audiophilestyle.com/forums/topic/31671-class-d-turns-out-it-does-suck/?tab=comments#comment-657693

 

So the comment was a continuation of and a reference to a wider conversation... So maybe you should look at my comment in context of the wider conversation.

Oh not only is it class D, but it is also in a price bracket where only mediocre reproduction is possible...

[firedog]

One of the main criticisms of Class D is that it must suck b/c the amps switching frequency is too low to produce a true audiophile sound. These amps work at a much higher frequency and so supposedly solve that problem.