Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Millett "Starving Student" hybrid amp
New Posts  All Forums:Forum Nav:

Millett "Starving Student" hybrid amp - Page 471

post #7051 of 7056

As long as your source can drive a 5k impedance (worst case input impedance scenario with a 20k pot), the concern with using a different pot value is the input capacitance of the 19J6 and whether it will create high frequency roll-off.

 

Let's do the math!

 

Cmiller = Cg-k + Cg-p * (stage gain + 1)

 

From the 19J6 datasheet, Cg-k is 2.6pf and Cg-p is 1.5pf. Let's call stage gain Mu (though it's much lower in reality).  So Cmiller is about 60pf.  But we have two 19J6s in parallel for each channel so double this to 120pf.

 

f = 1 / (2 * pi * Cmiller * Z)

 

We'll use the full 20k impedance from the pot as Z and remember to express Cmiller in farads, not micromicrofarads (uuf or pf). 

 

1 / (2 * pi * 20,000 * 120 x 10^-12) = a little over 60khz

 

The 50k pot PM spec'd gives a roll off at about 26khz using the same Cmiller as above. You'll actually have a little more high frequency bandwidth. I don't think we have to worry about oscillation here (60khz is still pretty low in the grand scheme of things), but you could add a 300-1k grid stopper if it worries you.

post #7052 of 7056
Quote:
Originally Posted by Sodacose View Post
 

As long as your source can drive a 5k impedance (worst case input impedance scenario with a 20k pot), the concern with using a different pot value is the input capacitance of the 19J6 and whether it will create high frequency roll-off.

 

Let's do the math!

 

Cmiller = Cg-k + Cg-p * (stage gain + 1)

 

From the 19J6 datasheet, Cg-k is 2.6pf and Cg-p is 1.5pf. Let's call stage gain Mu (though it's much lower in reality).  So Cmiller is about 60pf.  But we have two 19J6s in parallel for each channel so double this to 120pf.

 

f = 1 / (2 * pi * Cmiller * Z)

 

We'll use the full 20k impedance from the pot as Z and remember to express Cmiller in farads, not micromicrofarads (uuf or pf). 

 

1 / (2 * pi * 20,000 * 120 x 10^-12) = a little over 60khz

 

The 50k pot PM spec'd gives a roll off at about 26khz using the same Cmiller as above. You'll actually have a little more high frequency bandwidth. I don't think we have to worry about oscillation here (60khz is still pretty low in the grand scheme of things), but you could add a 300-1k grid stopper if it worries you.

holy crap, thanks for the awesome reply and explanation. :))

post #7053 of 7056
You are welcome!
post #7054 of 7056

Here's a Starving Student variation that I finished recently:

 

 

It's a different tube, but same basic overall topology (tube voltage gain feeding MOSFET source follower) and same power supply. The tube is resistor loaded single-ended and it's direct coupled to the MOSFET, which is CCS loaded. All Class A, of course. The 7370 tubes are pretty cheap on the 'bay. The rest of the parts are very inexpensive and easy to source. If you are having trouble finding 19J6s or otherwise want to try something other than 12AU7, I highly recommend this amp.

 

Note both the MOSFET and LM317 should be heatsinked. I used one heatsink at about 3" x 5" x 1" and bolted all the chips to it with insulating mounting kits (it only gets a little warm). The amp dissipates about 5W per channel in the TO-220 devices (48V @ 100mA per channel) so it doesn't need a ton of heatsink, but you're better off with too much than not enough.

post #7055 of 7056

I am ordering parts to do a Starving Student build using Fred_fred PCB. I want to use a Hammond 1455 enclosure. How tall does the enclosure need to be fit this setup?

post #7056 of 7056

I really like the 1455nhd.  I"ve not seen board specs yet, but I hope one fits in that enclosure

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Millett "Starving Student" hybrid amp