Impressions: SMSL PA200 GAN FET Class-D Power Amp

[Theboar]

If you do a google search you can find an interview wherein Bruno and Lars discuss GaN fet's and why they haven't felt the need to use them. There are some serious technical reasons why their advantages over silicon fets aren't easily exploited in audio amplifiers.

Thanks, I will do the search and listen to their arguments.

There's another issue I try to figure out for myself as well, I have 4 rooms all with different equipment build in the last 20-30 years. All different amps and speakers and frequency responses and Sinad and so on.. all different colors, solid state, 300b vs el tube amps, to NOS and OS r2r dacs to the newest AKM chips, different cables, net filters , grounding, chrysek vs accusilicon and so on.

At one day in one of the setups I changed an >300 euro expensive hdmi cable from a wellknown brand for a specific so called 'tailormade' cheap chinese 99.99 silver I2s cable from Ali. I know this sounds like psychiatry but the music started to become very interesting in that I wanted to hear all my favourite tracks again and again. Since this eye /ear opening moment I am starting to become a little frustrated with expensive audiophile recommendations and marketing of expensive stuff.

this said, If you ask me to choose between timing , frequency or resolution, at this moment in my journey I would say it's 'timing' worth doing the further research.

and this braught me in the hands of the marketeers of the SMSL PA200 with their 'short dead time nanoseconds' but I now get the feeling I might be misled, or just don't understand how it works from an engineer's perspective.

Does this dead time shortening has any correlation with something our brains pick up?

 

[goryu]

Dead time is related to the switching frequency of the amp. To take advantage of GaN fet's advantage here, they need to be operated at very high frequency, which apparently is not trivial. Most of the GaN fet amps on the market do not run at frequencies material higher than the silicon fet based class d amps. Thus, this advantage is not exploited. The PA200 is said in the marketing literature to run at speeds "up to 1MHz!" yet they don't mention the actual frequency. Yes, it can run at up to 1MHz, the question is, what is it actually running at?

 

[Roland68]

Dead time is related to the switching frequency of the amp. To take advantage of GaN fet's advantage here, they need to be operated at very high frequency, which apparently is not trivial. Most of the GaN fet amps on the market do not run at frequencies material higher than the silicon fet based class d amps. Thus, this advantage is not exploited. The PA200 is said in the marketing literature to run at speeds "up to 1MHz!" yet they don't mention the actual frequency. Yes, it can run at up to 1MHz, the question is, what is it actually running at?

It should run at 500 kHz, as that's the default.
Since there are two FETs per channel, marketing probably made it 500 kHz + 500 kHz = 1 MHz.

 

[goryu]

That's the problem- they need to run at much higher frequencies as I understand it if the dead time advantages are to be exploited. 500Hz is actually slower than what Hypex and Purifi run at...the Purifi 9040 for example idles at 600kHz...

 

[Julf]

At one day in one of the setups I changed an >300 euro expensive hdmi cable from a wellknown brand for a specific so called 'tailormade' cheap chinese 99.99 silver I2s cable from Ali. I know this sounds like psychiatry but the music started to become very interesting in that I wanted to hear all my favourite tracks again and again.

Not psychiatry, just psychological. It is how the human brain works.

this said, If you ask me to choose between timing , frequency or resolution, at this moment in my journey I would say it's 'timing' worth doing the further research.

and this braught me in the hands of the marketeers of the SMSL PA200 with their 'short dead time nanoseconds' but I now get the feeling I might be misled, or just don't understand how it works from an engineer's perspective.

Does this dead time shortening has any correlation with something our brains pick up?

No. Don't be mislead by the word "timing". The nanosecond dead time of the FETs does not produce any sort of audible delay. It pretty much only affects the efficiency of the switching circuit if you want to use really high switching frequencies.

Any time anyone uses terms like "timing" and "resolution" when talking about audio, they should define what specific timing or resolution they talk about, and what they mean by it.

 

[Loba]

As a follow up, I left the amps on overnight with nothing playing. This morning I checked with my hand and the amp casing is indeed warm to the touch (not crazy hot, but warm).

-Ed

Hi Ed, if you left the amps on and then you play something, does it start playing or do you have to turn it on again from "standby" mode if any?

 

[Deleted member 72219]

Hi Ed, if you left the amps on and then you play something, does it start playing or do you have to turn it on again from "standby" mode if any?

There’s no automatic standby. They are always on if you don’t put them into standby. Now, if you leave them plugged in and then put them into standby either with the remote or by pressing in and holding the knob until the red LED is lit, the amp still remains curiously warm (though not as warm as when left fully on, even with nothing playing). This is why I decided to get the Panamax trigger-activated power strip.

-Ed

 

[Loba]

There’s no automatic standby. They are always on if you don’t put them into standby. Now, if you leave them plugged in and then put them into standby either with the remote or by pressing in and holding the knob until the red LED is lit, the amp still remains curiously warm (though not as warm as when left fully on, even with nothing playing). This is why I decided to get the Panamax trigger-activated power strip.

-Ed

Ah, I think I will keep them on then. I did that for my previous smsl ao200 and ao300 which I sold off for v3 monos.

The warmness is probably from the internal psu continuously drawing power....hopefully not much compared to my v3 monos.

I decided to get pa200 for cheap from china, about US$275 each brand new.

 

[Deleted member 72219]

Ah, I think I will keep them on then. I did that for my previous smsl ao200 and ao300 which I sold off for v3 monos.

The warmness is probably from the internal psu continuously drawing power....hopefully not much compared to my v3 monos.

I decided to get pa200 for cheap from china, about US$275 each brand new.

That is a shockingly good price! Congratulations!

-Ed

 

[KL....]

If you do a google search you can find an interview wherein Bruno and Lars discuss GaN fet's and why they haven't felt the need to use them. There are some serious technical reasons why their advantages over silicon fets aren't easily exploited in audio amplifiers.

Thank you, is the Lars, Lars Clausen of the LC Audio (pre 2000 days, edit: before Hypex and I also remember their creations/implementations)? If yes, ah, I remember his DIY creations/implementations (and used them), quite well, how about that. I also remember that he had decided to give more attention/interest to wave/wind tech (edit: perhaps, also and because at that time, Hypex DIY classD implementations were improving and beginning to surpass the LC Audio classD implementations (well done to Hypex, especially because they have continued to improve) :=)

If you still know/remember the link, of that Bruno and Lars discussion, do you mind posting the link in this thread, thank you....

 

[Roland68]

Ah, I think I will keep them on then. I did that for my previous smsl ao200 and ao300 which I sold off for v3 monos.

The warmness is probably from the internal psu continuously drawing power....hopefully not much compared to my v3 monos.

I decided to get pa200 for cheap from china, about US$275 each brand new.

I hope you checked the shop.
The price is actually too good, especially for such a new device.
Currently, Aliexpress seems to be having a problem with newer shops offering goods at too low a price. They provide a false tracking number. The shipment is then supposedly sent to another state, which is unfortunately a common scam.

 

[Loba]

I hope you checked the shop.
The price is actually too good, especially for such a new device.
Currently, Aliexpress seems to be having a problem with newer shops offering goods at too low a price. They provide a false tracking number. The shipment is then supposedly sent to another state, which is unfortunately a common scam.

I got 1 of them from a personal seller, unopened amp as a part of a promo bundle direct from smsl, 1 of only 3 guys. He kept the speakers but sells the amp.

the other is a shop selling about 10% off current promo price from SMSL, these are online middleman, they sell direct from factory with no storefront overheads, being around for 3 years with good ratings.

Prices in china are about 20~30% below overseas retail prices. Of course in this way i bypass the distributors at the risk of warranty claims that I have to bear the costs of shipping to and fro.

 

[KL....]

Ah, I think I will keep them on then. I did that for my previous smsl ao200 and ao300 which I sold off for v3 monos.

The warmness is probably from the internal psu continuously drawing power....hopefully not much compared to my v3 monos.

I decided to get pa200 for cheap from china, about US$275 each brand new.

Thank you, until the ASR measurements, what are your (objective/context) impressions of the PA200 (if you are ok with doing so)?

 

[Theboar]

Thank you, until the ASR measurements, what are your (objective/context) impressions of the PA200 (if you are ok with doing so)?

I left the PA200 'on' for a couple days now and the unit is warm. Not as warm as my Dartzeel d5pro clone but for some reason it's indeed warm. To be honest if I have to choose listening music for fun I more and more go back to the NAD c298 again now the 'new ness' of the PA200 hast lost it's glare a little bit. Though for the money I am still very impressed by it. But that's all subjective psychology...

The good thing for me is that is it completely silent when no music is playing and when you switch on your power supply the unit goes on and off as well so you don't have to push for another button.

12v trigger would be nice but for me it also works fine with: https://produkte.kopp.eu/de/produkt...ngsfilter-14m-zuleitung-farbe-silber-schwarz/

 

[Loba]

Thank you, until the ASR measurements, what are your (objective/context) impressions of the PA200 (if you are ok with doing so)?

I will let you know when it arrives though, objective/context impressions will be against V3 monos driven by topping pre90 from e70v from cambridge audio cxn v2

 

[KL....]

the 'new ness' of the PA200 hast lost it's glare a little bit

Thank you, I have the Sabaj A30a and your experience (and @Melba59, which implys 3 (relevant) observations) has also been my experience/observation with the A30a. To be precise, from a little shouty/irritating/beamy (bearable) to musical/enjoyable/real (much closer), over about 200hrs/elapsed of playtime/unit powered on time (not contiguous but with short Off/Sleeping periods. Further and since 200hrs, I have noticed that the A30a is continueing to improve (although it still seems to have its (occasional) less/a little less musical/enjoyable/real moment/s but is continueing to improve. Overall I am enjoying the A30a very much, especially that it is a Digital/Analogue Integrated Amplifyer (DACless but could be called a Digital/Analogue Integrated Power DAC, couldn't it? and a even a Devialet style of implementation, couldn't it?). Ah, we could (objectively) ask ourself, why so?, in this case and as best we can with available information/data/context, couldn't we?

Ah, no need to discuss further, other than for consideration, and the PA200 is a nice example of material tech, via GaN (which allows/produces certain Electrical propertys), isn't it? A proposed consideration/suggestion is that is there a type of Conditioning occuring. To clarify, there are types of Conditioning (such as Annealing/Cryo/even Electrical tech), isn't there? The application of Annealing/Cryo is to change a materials property(s)/condition(s) to a desired Condition where one is a Heat/cool/etc and the other is a Cold/warm/etc, isn't it? A question/suggestion is, can Electrical Conditioning occur when applyed to materials (material tech).... a reasonable question/suggestion, isn't it? I am not in a position to do this, at this time, but to be objective, before and after measurements would need to done, wouldn't they? If done and the measurements showed differences then this would need to be considered as part of Setup, prior to measurement, wouldn't it? If no, (applying Sherlock Holmes Principle) then it must be something else and such aspects/considerations are Anticipation Bias (and I have witnessed/seen this many times and we cannot exclude ourself, can we?), Placebo Effect, but other/ALL Aspects* need to be considered/pondered, don't they? For consideration, there is a whole lot of Maths/Modeling considerations/components occuring inside each Device, isn't there?

edit: other/ALL Aspects*.... suggestion/consideration, all devices/units/components/speakers/connections need to not just talk to each other but with a, 'really looking forward to communicateing with you (Compatability/Feed/Feeding), thank you for feeding me and I will transparently pass it on (transparency), all the way to the listener (presentation/impression)'. This is your mission/task to make it your objective and make sure it happens, isn't it? do you accept this mission? Also refer to these threads....

• Does Op-amp Rolling Work?
• Big test of DIP dual op-amps for audio: LM4562, NE5532, OPA2134, TL072 and LM1458

Overall, it is a Musical experience (this is the desire/Intent), isn't it?.... enjoy :=)

 

[Theboar]

Thank you, I have the Sabaj A30a and your experience (and @Melba59, which implys 3 (relevant) observations) has also been my experience/observation with the A30a. To be precise, from a little shouty/irritating/beamy (bearable) to musical/enjoyable/real (much closer), over about 200hrs/elapsed of playtime/unit powered on time (not contiguous but with short Off/Sleeping periods. Further and since 200hrs, I have noticed that the A30a is continueing to improve (although it still seems to have its (occasional) less/a little less musical/enjoyable/real moment/s but is continueing to improve. Overall I am enjoying the A30a very much, especially that it is a Digital/Analogue Integrated Amplifyer (DACless but could be called a Digital/Analogue Integrated Power DAC, couldn't it? and a even a Devialet style of implementation, couldn't it?). Ah, we could (objectively) ask ourself, why so?, in this case and as best we can with available information/data/context, couldn't we?

Ah, no need to discuss further, other than for consideration, and the PA200 is a nice example of material tech, via GaN, isn't it? A proposed consideration/suggestion is that is there a type of Conditioning occuring. To clarify, there are types of Conditioning (such as Annealing/Cryo/even Electrical tech), isn't there? The application of Annealing/Cryo is to change a materials property(s)/condition(s) to a desired Condition where one is a Heat/cool and the other is a Cold/warm, isn't it? The question/suggestion is, can Electrical Conditioning occur when applyed to materials (material tech).... a reasonable question/suggestion, isn't it? I am not in a position to do this, at this time, but to be objective, before and after measurements would need to done, wouldn't they? If done and the measurements showed differences then this would need to be considered as part of Setup, prior to measurement, wouldn't it? If no, (applying Sherlock Holmes Principle) then it must be something else and such aspects/considerations are Anticipation Bias (and I have witnessed/seen this many times and we cannot exclude ourself, can we?), Placebo Effect, but other/ALL Aspects need to be considered/pondered, don't they? For consideration, there is a whole lot of Maths/Modeling considerations/components occuring inside each Device, isn't there?

Overall, it is a Musical experience (this is the desire/Intent), isn't it?.... enjoy :=)

I think you are right , we cannot excluse ourselves from things as anticipation bias, nocebo an placebo effects... And when `i'm tired after work my limbic system receives input different compared to a morning coffee moment.

That's why it's good there are engineers on this platform active who help us to at least understand the marketing pittfalls.

As a general rule I think when you enjoy your system don't spend to much money on trying to get it better. To be happy with a system is priceless. If a system gives 'stress' (after weeks of listening in different moods) instead of the 'healing effect' music can have the first and second advice is to look inside first (psyche) before buying new stuff. The third advice for me personally (after my subjective experience with the I2s cable) was to start listening to these 'time domain smearing' stories some people are telling online. But that's just a personal journey. I think there are a lot of ways to get to this point were you will enjoy music again.

 

[Loba]

I think you are right , we cannot excluse ourselves from things as anticipation bias, nocebo an placebo effects... And when `i'm tired after work my limbic system receives input different compared to a morning coffee moment.

That's why it's good there are engineers on this platform active who help us to at least understand the marketing pittfalls.

As a general rule I think when you enjoy your system don't spend to much money on trying to get it better. To be happy with a system is priceless. If a system gives 'stress' (after weeks of listening in different moods) instead of the 'healing effect' music can have the first and second advice is to look inside first (psyche) before buying new stuff. The third advice for me personally (after my subjective experience with the I2s cable) was to start listening to these 'time domain smearing' stories some people are telling online. But that's just a personal journey. I think there are a lot of ways to get to this point were you will enjoy music again.

I am never happy with my system, it's either too hot (300B and fosi v3 monos!), or not warm enough in mid range (v3 monos), muddy bass (300Bs) and both sucks power.

hence an upgrade to try to maintain similar performance while achieving less heat, using less power. The last closest one I had was SMSL AO300, before that AO200(not enough power to kick bass), NAD D3020 v2 (this was sweet sounding but hot!) , yamaha wxc-50 (this is dry), older Arcam (too polite, no bass) and Cambridge audio (hot and sounds crass) class AB mid range early 2000s amps. which are objectively lower performance in measurements and audibly.

the last frontier to cross was my wife who is the daughter of an audioman in the media industry who had done recording in the music industry, built his own speakers (15" drivers!) and does manual crossover of his speakers. She has trained golden ears who will tell me immediately if the amp is rubbish or too harsh. Sometimes I will just swap in new amps without telling her and ask how's the music hence how a Class A jeff rowland amp got retired to my office due to the unbearable brightness to her. I had secretly turned them on a few times but her mood will immediately turn for the worse asking me to turn off the music citing the music being extremely irritating.

 

[Julf]

hence an upgrade to try to maintain similar performance while achieving less heat, using less power.

My suggestion is to go for efficient class D amps and a good DSP unit so that you can dial in whatever "sound" you want.

 

[Julf]

As a general rule I think when you enjoy your system don't spend to much money on trying to get it better.

Indeed. The point is to listen to and enjoy the music, not the system. Many people enjoy the music even on a cheap transistor radio or ghetto blaster.