Toshiba MOSFETs with a twist

[abraxalito]

I shall be putting the DAC schematic up on Hackaday - here's the link to the project : https://hackaday.io/project/27001-audio ... t-no-money

[nige2000]

is this dac going to be something we (i) will be able to build?
is there not hand made inductors in it?

heres the speaker amp in building (well....more thrown together than built, made with hot snot and some stuff i had lying around)
just to make it a little transportable :)


sounds pretty good so far, very energetic
similar to the hot buff
dual mono too

these heatsinks are just a little warmer than id like at 40v so ill likely drop another five v
set the gain stage at 25v

[abraxalito]

is this dac going to be something we (i) will be able to build?
is there not hand made inductors in it?

The publicized version I'm calling 'lingDAC' and that's using TDK SLF7045T inductors so yeah you can build it so long as you don't mind paying inflated (Mouser) prices for those coils. This is a NOS DAC with the addition of passive filtering prior to the (discrete) I/V stage.

The 'by invitation only' version is more complex - its 2X oversampling and can use handwound coils (for guilding the lily) or next-best thing, Coilcraft MSS1260s. I'm even working on a cut-down version of this one's filter using the same range of TDK inductors. So its possible to mix'n'match a bit.

On the amp, yes 'energetic' is a good way of putting it, the thing has drive. But the dynamics with my two-in-parallel version (TO220 FETs) don't match up to what I'm getting with headphones.

[jrling]

is this dac going to be something we (i) will be able to build?
is there not hand made inductors in it?

The publicized version I'm calling 'lingDAC' and that's using TDK SLF7045T inductors so yeah you can build it so long as you don't mind paying inflated (Mouser) prices for those coils. This is a NOS DAC with the addition of passive filtering prior to the (discrete) I/V stage.

The 'by invitation only' version is more complex - its 2X oversampling and can use handwound coils (for guilding the lily) or next-best thing, Coilcraft MSS1260s. I'm even working on a cut-down version of this one's filter using the same range of TDK inductors. So its possible to mix'n'match a bit.

On the amp, yes 'energetic' is a good way of putting it, the thing has drive. But the dynamics with my two-in-parallel version (TO220 FETs) don't match up to what I'm getting with headphones.

I had much the same question about building the DAC and the speaker amp, but I am assuming technical R&D is still ongoing and it's great to observe that happening in real time on Tir (and Richard's website). Thanks for the clarification above. It sounds like it they may be projects for the more simple-minded lower skilled operators like me.

Jonathan Ling
P.S. Love the new name for your DAC! Just in case you're wondering, I'm not Chinese and certainly do not speak Mandarin. Wish I did though.

[nige2000]

On the amp, yes 'energetic' is a good way of putting it, the thing has drive. But the dynamics with my two-in-parallel version (TO220 FETs) don't match up to what I'm getting with headphones.

dont think any of my dacs excel at dynamics enough to challenge it enough for a test, especially the soekris dac
in saying that bass can be taut and fast so id be hopeful that the extra muscle will help the dynamics

would the parallelled tk13's not be running out of steam at 8 ohm at speaker voltage levels?

[randytsuch]

Thanks for the link Richard

I'm now reading your fiddle thread nige. I skipped it before because I knew it meant more work to do lol.

Trying to work out a mouser order now, need to buy some mosfets and transistors, probably a few other things too. Luckily I have some heatsinks in my junk drawer, I think I'll have something suitable.

I'm sure there will be another Mouser order for the DAC, this will be mainly to build the HP amp, but hopefully the active stuff for the DAC too.

I guess I have to buy the 1387 chip from china, I avoid buying IC's from china if possible because I worry about counterfeits, but I'll cross my fingers and hope for the best.

Anyone have any advice for the 1000uf output cap?


Randy

[nige2000]

theres 3.5ish v between the tk30j25d gate and v- what sort of cap should i put across it?

tk30j25d are not available in usual reliable spots either if your looking for those, i got some from utsource that seem perfectly fine,
but i also got 2sk170s there for 10c each too last yr that looked suspect, so..... at your own risk

dunno about output 1000uf cap,
i used some nichicon bi-polars i got last yr i haven't experimented

[randytsuch]

theres 3.5ish v between the tk30j25d gate and v- what sort of cap should i put across it?

tk30j25d are not available in usual reliable spots either if your looking for those, i got some from utsource that seem perfectly fine,
but i also got 2sk170s there for 10c each too last yr that looked suspect, so..... at your own risk

dunno about output 1000uf cap,
i used some nichicon bi-polars i got last yr i haven't experimented

The tk30j is not on my shopping list, I didn't see it in the HP amp or the DAC circuit
Did I miss it?
I was going to pick up some nichicon bi polars and some oscon types for the 1000uf cap and see what I like.

[abraxalito]

I guess I have to buy the 1387 chip from china, I avoid buying IC's from china if possible because I worry about counterfeits, but I'll cross my fingers and hope for the best.

Having bought all my TDA1387s from China (which themselves are recycled) I've yet to encounter one that doesn't work according to the DS. I can't see a way they could be faked, in any case they're so cheap (on Taobao the lowest price I've found is $0.03) that there could barely be any money in a faking business, even if it were possible. OTOH I've encountered a few fake MOSFETs when ordering Toshibas. The first bore different markings, the second batch had what seemed to be the correct codes on them but performed slightly differently.

[abraxalito]

theres 3.5ish v between the tk30j25d gate and v- what sort of cap should i put across it?

Are you asking about improving the PSRR? In which case a cap between gate and v- isn't what I have been suggesting. It would probably improve PSRR but at the expense of worse output impedance.