Yep, part of the reason why I use it. Mogami makes good wire I am not an interconnect swapping kind of guy, more get something good and run it. Will be experimenting with Mogami speaker cable shortly. Sorry for small high jack
OK. Apparently I misunderstood- I thought you had an amp with a spdif input; hence all my questions. BTW I've heard a lot of good stuff about your amps. For that to happen GaNFETs would have to be superseded. Hard to know if that will happen- MOSFETs are nearly as fast these days so that advantage went away. But GaNFETs are a lot easier to drive, having about 1/10th the gate capacitance of a MOSFET. IMO their weak spots might be their reverse conduction characteristic, which often means the output filter choke plays a role in how they switch and the other issue might be the thermal impedance of the heatsinks that are possible with them. GaNFETs are thinner than a credit card and often smaller than a mini secure digital card. They can heat up very quickly on that account- faster than a heatsink might be able to more the heat away. But because of their low on resistance they don't dissipate a lot of energy and so often don't need a heatsink at all. But if you're going for a higher power design a heatsink will be important. There are other newer switching devices out there. But they have to gain wide acceptance so there's more than just one manufacturer in order for a new kind of device to get acceptance. Right now GaNFETs are made by several vendors so its 'safe' to include them in a new design. I don't think they are a flash in the pan.
I was referring to any alleged performance advantage with regard to sound quality, like whether any of these GaN amps sound better than the Purifi modules, for example.
I was a moderator at AudioCircle when the NCore 400 amps were hitting the market. According to the early adopters, that amp was the absolute end of class A, A/B, tube, or any other amp. The fad died down after about six months. I've owned several class D amps and they always impress the heck out of me upon initial listen, a month later, the wrinkles start to reveal themselves. So yeah, let's wait and see.
I understand what you are saying, I have not lost interest in mine, however I do not believe there is an end to any of the designs. A good design is just that. Each one has its benefits and sound. Some people tire of one particular sound and then go for another until they tire of that and continue on. I am not saying that is you but we all know people who buy and sell all the time and are always with some new component. This is why it is nice to have all these choices. For me tube, A A/B, A and D all have the ability to please me depending on what I am after and I am glad they are all there to choose from. Currently and this is over a year I am still happy with my configuration and hope to be for the foreseeable future. I have learned a long time ago that nothing lasts forever, so one day I might want an incredibly efficient horn speaker and a SET amp, who knows.
GaN HEMTs - they are HEMT structures and called FETS to sound similar to MOSFETs, are actually harder to drive, not easier and they 100% need heatsinking. Being extremely fast devices, they can generate unwanted noise, so fast is not always good. Their main benefit is power efficiency improvements.
I notice that GaN Systems doesn't list any reliability information on their website or in the GS61008P datasheet. In fact, for hi-rel applications, they recommend contacting their "valued partner", Teledyne, who probably burn these in for hi-rel customers. If I were building one of these amps, I'd use the re-badged Teledyne TDG100E90BEP over the GaN Systems GS61008P, even though it'd likely be 3-5X the cost.
If you are stating that GaNFETs are actually something else (GaN HEMTs) and that this was done for marketing reasons, then your statement is entirely incorrect. For example, having a much lower gate capacitance translates to less current required to switch the gate. In most books that would be described as 'easier to drive'. And they really don't need heatsinking unless you are pushing them hard and I know that from listening to a GaNFET amp for the last year that does not have heatsinks. Finally, they are no more noisy than MOSFETs and might be a bit less, owing entirely to less stray inductances compared to the MOSFETs used in the Hypex and Purifi modules, which have leads, while GaNFETs do not.
Considered your thoughts, thus made contact with Teledyne! Mont Taylor, VP Business Development, responded with these interesting comments: "Hi Ted, we are a partner of GaN Systems and offer versions of their devices to the aerospace and defense market. These parts go through additional qualification and screening for defense and space applications. If you have a commercial application, then GaN system parts would work better for you. If you need parts for a ruggedized or high altitude system then Teledyne devices might offer what you need, but are more expensive than the equivalent GaN System device...Regards, Mont" Believe Orchard Audio's design implementation with the GaNFET GS61008P transistors will be an end-game audio power amp component for me, as I have no reason to go with it into space or on the battle field!! Ted
I work with Teledyne a lot for my normal job with CACI (CACI Photonic Solutions and Laser Communications ). The up-screened parts cost orders of magnitude more than the un-screened counterparts it is not practical to use them for commercial applications. One transistor will cost you almost as much as an amp.
Here is a TI example of a regular op-amp and the same exact part after it has been packaged and screened for space application.
Lets see if my math is correct?!? For the price of "1" Space Ready Op Amp ($8.32 usd), you can have over "1500" Regular Op Amps (.00549 cent)...almost Unbelievable!!! Ted
Well, not quite. The space ready price is $831.60 each in quantity of 100, while the commercial price is $5.26 each in quantity of 1000 (or $7.89 each in quantity of 100), so space ready costs a little over 100x as much in quantities of 100.
Dave, You’re correct, but with a different pricing volume selection of the regular units! I figured "Price Per Unit" for the Given chart examples...($.00526=Regular Commercial Unit VS $8.31=Space Ready Unit)...that's 1580 to One! Ted
But these transistors, absent any quality screens, has a higher potential for failure in the field, which is not good for future business. Could you at least get a qual report from GaN Systems to see what their expected life is? So Teledyne is charging $1,000 per amp for an upscreened GS61008P transistor? Really? So it would be $4,000 just for the 4 Hi-Rel GaN transistors in your amp? Wow. That seems quite excessive. Makes me wonder even more about the quality of those unscreened parts.
Actually, there are 8 of them per amp, each module is a full bridge so each module has 4 of these fets. The Stereo amp has 2 modules.
More precisely, they are AlGaN/GaN HEMTs - GaN bulk, then layers of AlGaN then GaN. Look up the HEMT structure on Google. GaN devices don't have an oxide layer, therefore are not like MOSFETs, but they do use the field effect. GaN Systems use the HEMT name now for their normally off devices. They offer two package types, bottom cooled or top cooled. Bottom cooled you design the heat spreading into the PCB, top cooled you have to attach something. Call it what you will but it's heatsinking. Unless you've actually designed a drive circuit you haven't a clue mate.
Third parties have been performing that service for various transistors since the 1970s, perhaps even earlier.
Yeah, did exactly that; we've been making a class D of our own design for over a year now and I agree with your last statement above! You are correct about top or bottom cooled, however if you have a top cooled device and no heatsink is present, as long as you don't push the amp too hard the devices will be fine. This is as I stated earlier when you challenged me on this topic. The amps I'm playing in my system have been there for about a year. We do make heatsinks for our amps of course; they have to be carefully machined in order to work right, but since my speakers at home are 16 Ohms and 98dB 1 Watt/1 meter, my amps don't have heatsinks in them (they were one of the first pair from our first production prototype run). If you want to include the foils of the boards as heatsinks I understand, but FWIW they don't have a lot of area as there was no intention of the board being the heatsink. Very different from transistor requirements in a AB design!
Starkrimson Stereo Ultra Amp four GaNFETs "underside heat sink"...after all-day listening sessions, case barely gets up to 85deg! Ted
