Page 17 of 23

Re: Toshiba MOSFETs with a twist

Posted: Fri Sep 08, 2017 10:56 pm
by nige2000
how much current should i push through the output stage ?
i started at 1.3 amps moved to 1.9 amps with much improvement sonically
it will take a bit more its about 40 degrees c atm

Re: Toshiba MOSFETs with a twist

Posted: Sat Sep 09, 2017 12:19 am
by abraxalito
It depends on the supply voltage and load impedance. For 24V supply and 8ohm load its about 1.5A per channel, 3A totally. If you're only hitting 40oC you can raise the volts and the current.

Re: Toshiba MOSFETs with a twist

Posted: Sat Sep 09, 2017 8:17 am
by nige2000
I had it at just under 30v and 1.9amps
What's the upper limit for voltage ?
35v?

Re: Toshiba MOSFETs with a twist

Posted: Sat Sep 09, 2017 11:42 am
by abraxalito
30V and 1.9A sounds just about perfect for 8ohms. There's plenty more voltage you can apply beyond 35V to the poweramp - the big FET's rated at 300V. But the gain stage is using K339s isn't it? So that'll be limited to 40V. If you increase the volts to the gainstage might be better to raise the CCS biassing resistors to save the smaller transistors getting too hot.

Re: Toshiba MOSFETs with a twist

Posted: Sun Sep 10, 2017 10:06 pm
by nige2000
due to the constant current nature of the output stage i assumed that does not need much in the way of capacitance or regulation ?
i cobbled together from some stuff i had to hand a 35v traffo with a bridge rectifier and a 1000uf cap to see what would happen
and it seemed perfectly fine as good if not better than my bench supply which usually does a good job
is there any point going to greater lengths with this output stage supply ?

i have 10x5v secondary traffos spare here i could try probably could get 60v dc out of them

Re: Toshiba MOSFETs with a twist

Posted: Sun Sep 10, 2017 11:27 pm
by nige2000
Is it possible to run the output stage from rectified mains power?

Re: Toshiba MOSFETs with a twist

Posted: Mon Sep 11, 2017 3:40 am
by abraxalito
Definitely possible to run from ripply mains but I'd add an RC to feed the CCS bias as the PSRR will be limited by the bias feed.

Re: Toshiba MOSFETs with a twist

Posted: Mon Sep 11, 2017 9:22 am
by nige2000
Is that a cap between the tk's gate and gnd or across the power resistor

on ltspice if i try to lower the power resistor value below the 1.9 amps and raise voltage to 250v dc it begins to clip on the negative wave
suggesting that amperage needs to be circa 2 amps into a 8 ohm load regardless of voltage?
and voltage is only required to provide enough swing for output?

if thats the case realistically over 40 v is going to be hard to dissipate the heat

Re: Toshiba MOSFETs with a twist

Posted: Mon Sep 11, 2017 1:30 pm
by abraxalito
nige2000 wrote: Mon Sep 11, 2017 9:22 am Is that a cap between the tk's gate and gnd or across the power resistor
No its a cap to -ve supply (I think, need to refresh myself of the schematic) to filter AC from the bias feed to the CCS. The CSS has a high-ish value resistor between GND and the TK's gate, this is the sensitive path.
suggesting that amperage needs to be circa 2 amps into a 8 ohm load regardless of voltage?
and voltage is only required to provide enough swing for output?
That's it yep. 2A into 8ohm comes from 16V swing, so no voltage above 32 is going to be useful unless you also increase the bias.
if thats the case realistically over 40 v is going to be hard to dissipate the heat
Yes, that's where paralleling comes in with multiple enclosures...

Re: Toshiba MOSFETs with a twist

Posted: Tue Sep 12, 2017 6:28 am
by randytsuch
I'd like to build this dac and hp amp.

Is the schematic for the hp amp in this thread the one I should use?
Also worried that without any voltage gain, that is will drive my hd-700's that are 150ohms.
But it looks pretty easy and inexpensive to try it.

How do I get the latest DAC schematic?

Thanks
Randy