Toshiba MOSFETs with a twist

[nige2000]

No, this gain stage was designed to be just that - a stage for providing gain in front of a unity gain buffer (either for high impedance headphones like DT880s or for speakers). It could be up-biassed though to drive 32ohms - it would need the output CCS winding up to be higher than the maximum output current demanded.

is that biased by lowering the r value on the k339 source?
i lowered to 10 r but after 30 secs it gets uncomfortably warm
lt spice suggests a r of under 3 ohms and 133 mwatts for a 30 ohm load
that has to be miles too hot

Changing to a balanced supply would negate the constant current draw design philosophy. It will still work but I reckon dynamics will suffer.

because the output return is between the ccs?

[abraxalito]

is that biased by lowering the r value on the k339 source?
i lowered to 10 r but after 30 secs it gets uncomfortably warm

R86 (62ohm) on the version I have. If you run enough bias to drive 32ohm to the rail (24V), it'll overheat for sure. But 32ohm cans won't need more than 1VRMS so the trick would be to lower the power supply (no more than 6V), then you can run 40mA bias without overheating. 40mA would be 15ohm for R86.

because the output return is between the ccs?

Sorry, cannot follow your question. The return current sometimes has to go to the +ve and sometimes to the -ve rail when you run with balanced supplies.

[nige2000]

is that biased by lowering the r value on the k339 source?
i lowered to 10 r but after 30 secs it gets uncomfortably warm

R86 (62ohm) on the version I have. If you run enough bias to drive 32ohm to the rail (24V), it'll overheat for sure. But 32ohm cans won't need more than 1VRMS so the trick would be to lower the power supply (no more than 6V), then you can run 40mA bias without overheating. 40mA would be 15ohm for R86.

because the output return is between the ccs?

Sorry, cannot follow your question. The return current sometimes has to go to the +ve and sometimes to the -ve rail when you run with balanced supplies.

oh i thought if the return went to 0v gnd it would ruin the pssr......

just trying to learn about mosfets thanks for the info
ill go ahead and build the tk13 buffer and then hopefully a v2 compact gainstage version

have you come across any mosfets in larger packages suitable for loudspeaker that can take a heavy bias?

[abraxalito]

There is this one, not sure if I've mentioned before - http://toshiba.semicon-storage.com/us/p ... 0J30D.html.

It has a 410W dissipation spec but realistically I can't see any heatsinking method beyond running it under a cold tap being able to get that.

[nige2000]

There is this one, not sure if I've mentioned before - http://toshiba.semicon-storage.com/us/p ... 0J30D.html.

It has a 410W dissipation spec but realistically I can't see any heatsinking method beyond running it under a cold tap being able to get that.

if it dont sound good it could make a good electric kettle :)
i dont see any availability at the usual suspects have u found a source?

[abraxalito]

I have 8 in a bag on my desk - they're from Taobao. Haven't examined them electrically yet to see if they're genuine ones, but they look fine physically. I plan to try them out in the same case as the TO220 TKs, the amp using these will only need 2 per channel.

[nige2000]

how far did u push the bias of the to220 tk

[abraxalito]

At present its running a total of 2.1A for the two channels with 24V supply. Its slightly under-biassed but I rarely hear anything like distortion due to current limiting. The case can take more current I think, its not too hot to touch and this is summer where indoors its around 32oC.

[abraxalito]

v3.0a.jpg (95.26 KiB) Viewed 610 times

Fixed up the FR bumpiness by increasing the impedance of the filter ( = higher valued Ls). Now the design works with all stock inductors. An interesting change is that now the whole board is less microphonic - tapping it barely registers anything in the headphones. Could be because the caps themselves are less microphonic at lower values - I'm using mostly 100nFs now, before it was made from 1uFs. From preliminary listening the good news is it hasn't lost its addictive quality, perhaps a smidgeon less ultimate bass slam in comparison with the lower impedance filter. There will probably be some upside connected to the reduced microphony - like improved attack.

[nige2000]

Got the hotbuff working
Well one channel
Sounds good so far on headphone