No lateral SiC FETs exist are available though, no?
https://ieeexplore.ieee.org/document/9440439 Interesting article about some samples though.
https://ieeexplore.ieee.org/document/9440439 Interesting article about some samples though.
Not aware of any lateral SiC Fets available on the market at this point. I can say that SiC Mosfets have been and are being used in linear audio applications.
Although the intention was for them to be used in high power switching applications, they can handle 30 to 50 or so Watts of continuous dissipation without problems. Even the smaller ones have a max rating of 700 Volts. How do they sound ? well again it depends on the rest of the circuit interaction.
Designing a really awesome sounding MOSFET power amp for guitar is beyond my capabilities rn haha
oh wait i severely miscalculated currency conversion lmao. Could you replace Q1 and Q2 of the higher wattage KMG designs with a different pair of transistors? Seems like they're only there for buffering.
I haven`t done a output transformer coupled SiC, but involved in a couple of other designs..can`t talk much about it as it was a consultative role.
Here`s a ballpark approach, rough estimates B+ at 375 V , P-P transformer with about 6K primary. SiC gates DC biased for around 10 W at idle (keeping in mind the max DC current of the transformer). At max input signal maybe you`d want about 40W to keep the SiC far enough from it`s maximum dissipation rating.
Here`s a ballpark approach, rough estimates B+ at 375 V , P-P transformer with about 6K primary. SiC gates DC biased for around 10 W at idle (keeping in mind the max DC current of the transformer). At max input signal maybe you`d want about 40W to keep the SiC far enough from it`s maximum dissipation rating.
It will work, but put out way too little power for such a large $$$ , weight and effort invested.
100V supply means 90V peak swing .
200 mA peak current means 180 mA peak swing , in both cases discounting 10V for saturation and losses, and 20mA idle current.
So we have 90V * 0.18= 16.2W peak so meager 8.1 W RMS
You will need an 1800 ohm plate to plate OT, such as:
https://www.amplifiedparts.com/products/transformer-marshall-replacement-output-100-w
1700 ohm P/P , 8 lb weight, U$144 + shipping.
Not sure about others but I'd rather have a custom 8W transformer wound (what KMG did) instead of repurposing a standard Tube OT.
To avoid visual clutter, modern Forum software generally does not allow quoting immediately above post, since it's visible anyway.
You can work your way around it if strictly needed, but it takes manual copypasting, typing [ quote ] commands, etc..
Yes, if you use a Tube OT at typical or nearby Tube voltages, it will work fine, provide tons of output, etc.
Not sure about SiC MosFets biasing and curves, never checked them, fear they are similar to switching types, horribly nonlinear at low currents if "used as tubes" in a common source architecture with OT on drains.
To get any transformer color you need to use it in a Tube Amp architecture:
Phase splitter > power devices > OT loading plates-drains-collectors > OT > speakers.
Preferably little or no NFB.
If you just add it as an autotransformer between a "big OpAmp" such as most SS amps and speakers, the amp low impedance "any current you need" characteristic will swamp most of OT's quirks.
They work pretty nicely with direct speaker coupling, but that's running higher currents. There is a lot of gate capacitance to deal with (similar to power mosfets). About NFB, some local source degeneration is helpful and I think adding some "global" NFB from the OT secondary back to the phase inverter, similar to some tube amps would be possible, but not necessary. Just to be on the safe side I would include it.
DC Biasing with some temperature compensation is fine.. like adding series diode (s) and linking it to the heatsink.
Not saying that these devices are ideally suited to driving an OT, but they can withstand much higher operating voltages.
I wanted to get back to looking at the ACA Mini, and originally the amp design was made available with a single PCB: 2 channel stereo version.
But very recently Nelson Brock on the DIYA forum did a layout for separated (mono block) PCBs. They look good, are not too expensive and then you could build up the power amp for a (mono) guitar amp..and maybe keep the extra board as a spare. Check out the ACA MinMax. I would probably consider incorporating a thermal switch onto the heatsink just as an additional safety.. as you know live guitar signal isn't quite the same as played back recorded material. Maybe lateral Mosfets could be used on this? Maybe some limiting diodes as well.. it would depend on the overloading character of the power amp.
IMHO, 10 watts is a real sweet spot for a nice portable combo amp. I am thinking Princeton style preamp based around Jfets. Simple, somewhat well behaved, not an over the top distortion machine, but something you get nice progressive distortion/breakup from both the preamp and the power amp sections.
But very recently Nelson Brock on the DIYA forum did a layout for separated (mono block) PCBs. They look good, are not too expensive and then you could build up the power amp for a (mono) guitar amp..and maybe keep the extra board as a spare. Check out the ACA MinMax. I would probably consider incorporating a thermal switch onto the heatsink just as an additional safety.. as you know live guitar signal isn't quite the same as played back recorded material. Maybe lateral Mosfets could be used on this? Maybe some limiting diodes as well.. it would depend on the overloading character of the power amp.
IMHO, 10 watts is a real sweet spot for a nice portable combo amp. I am thinking Princeton style preamp based around Jfets. Simple, somewhat well behaved, not an over the top distortion machine, but something you get nice progressive distortion/breakup from both the preamp and the power amp sections.
I'm lost on using a class A hifi amp for guitar. I see amp design as analogous to cooking; multiple carefully selected ingredients that taken together yield an out of this world flavor (sound). A class A amp built to not distort at all, ever, would be like using distilled water as a cooking ingredient; as transparent (odorless, tasteless, colorless) as possible.
Depends on how a class A amp is built, with little or no global feedback it will have low order relatively progressive harmonics going into higher power, primarily 2nd and 3rd for a push pull like the ACA mini. You could bias it lower into Class AB operation I think..if that is preferred. ''Distilled water'' would be to apply a lot of global feedback which straightens out the transfer curve, but then it clips more suddenly when at the supply limit. I think the goal of the ACA series was to design simple amplifiers, that sound good/perform well and not necessarily adhere to a rule that any distortion is undesirable for ''hi-fi''.
Last edited:
Many simple class A amplifiers, like the jlh amp, the amps by Jean hiraga, or the Nelson Pass amps, have several of the attributes that you want in a guitar amp: moderate to low global negative feedback, gentle clipping behaviour, highish output impedance.
I would say it is the conventional class AB hifi amplifier topology that does not work very well as a guitar amp. They just sound horrible when overdriven. It seems a bit silly to start with such an amp, and then try to fix things with diode clippers and so on.
It makes more sense to me to follow the classic tube guitar amp recipe with a phase splitter and a push pull transformer coupled output stage.
Me too. Crazy Joe - me again - would even use a transformer phase splitter, like those old Japanese radios of the 60s. Why? Because it's been done, its well understood and you can - as a hobbiest. Plus people like the sound of audio transformers for some reason, you can find evidence of that "here and there".
I was going to say try telling that to Mr Fahey, but he's already liked your post.
The Univox U65RN is an early example of a signal transfo inverter/splitter driving a pair of PNP power transistors. I think back in the day, PNP had higher current capability and hfe than their equivalent NPN. So it probably was more effective to run a "balanced" push pull PNP pair.. it's a small interstage transfo which as you mentioned was pretty common in many discrete audio circuits.
But nobody used such a thing driving output tubes (maybe in the 1920s...battery radios) - besides the occasional SET HiFi amplifier design. Why would they, when you can just use a tube - so much easier, cheaper? It'd be my guess that such a double-transformer topology would have an identifiable sound; perhaps implemented out of necessity in the 60's, but no longer practiced outside of niche designs. Why, you can go to the vendor page right here and find "transformer input amplifier" (I think...) - why would someone do that and think whatever that does is good enough to sell, unless...it actually does something audible and for some listeners, it is good enough to sell.
It's so much cheaper and easier to build amps without signal transformers or output transformers, or tubes for that matter. These designs developed over many years depending what parts were commonly available on the market. Even now we're discussing discrete linear solid state, which of itself became a very niche market, regardless if it`s listeners of hi-fi or users of musical instrument amps. So on a DIY hobby level, we don't necessarily have to budget a design for economy (but we mostly do!). So building something yourself that is enjoyable to listen to may overtake the ''need'' for it to be commercially viable.
Having said that, there are plenty of examples of new products copying the NEVE 1073 studio preamp..a discrete design with 2 signal transformers in the chain.
Having said that, there are plenty of examples of new products copying the NEVE 1073 studio preamp..a discrete design with 2 signal transformers in the chain.
Was that design not with two 3055's, in antiphase driving a symmetrical transformer?
And from there, the next progress was an OTL.
And now we're stuck with this DIYaudio.com website.
I'm guessing there's a market for that particular "ingredient", as supposedly the Sound City console was full of them and carried the original signals that comprise many famous recordings. And I'm sure like black, white, red, Chinese "pepper" it has to be the right transformer, versus just a transformer in an amplifier design to give some captivating sound.
I mentioned the 1920s radios as I used to buy them at flea markets and antique stores when I was much younger, maybe still riding a bicycle. Almost all had audio interstage and output transformers, again I assume because they had to, using tubes that had internal structures where you could see why they called it a "plate", because that's literally what it was. None of the 360 deg wrap around elements then...
Been there done that. Was going to put together a JLH but Chinese knockoffs were so cheap I bought a ready built board. Slap a heatsink on and a laptop supply, used a distortion pedal as the preamp. Sounded like a guitar amp and I did not find any drive level objectionable. One day I will get back to it. I figure the power wasted in a 5W SS Class A amp will not be any worse than a 5W tube amp.