Millett "Starving Student" hybrid amp - Page 415

sounds good! if i ever get around to it ill be sure to let you know how it sounds

i am a tube newbie, want to build this as a first tube project. whats a good source for these tubes. which tubes are good, I see russian, chinese etc on ebay. any recommendations? dont want to spend too much though. any particular brand i should be looking at.

 

I'll asume you're talking about 12AU7 tubes. An ultra quick Google search yielded this. It also gave links to www.tubeworld.com, www.thetubestore.com and www.tubedepot.com.  Remember, Google is your friend.

 

As to brands, pretty much any 12A_7  you get from those places will do. My own 12AU7 build was done with a couple of JJ tubes from tubedepot.com. I also tried a couple of new 'Mullard' (Russian made) and Sovtek 12AX7 tubes.

 

Your local musical instrument shop might also carry the Sovtkek 12A_7 tubes as they're popular for guitar amplifiers.

 

Good luck with your build.

 

cheers!

Actually I am not sure which tube i should go for, as long as its low voltage. I will go with whatever recommendations I get here as I am a total tube newbie. Yes, google is a friend as long as you know what you are looking for.

Edited by doors666 - 10/3/12 at 7:39pm

 

The amp design is low voltage, and there are several tube options to implement it:

 

• 19J6 - (if you can find any) this is the tube used in the original design by Mr Millett
• 12SR7 - details posted by user Logistic in this same thread
• 12A_7 (usually 12AU7) - details start with this post.
• 17EW8 - same pinout as the 12AU7 build, only a couple resistor values change.

 

Any of them will do nicely, I recommend you pick the one that is the easiest/cheapest for you to get.

 

cheers!

Thank you, that surely helps.

19j6 tubes are so hard to find these days. i was hoping to get some different models to compare sound signatures on my ssmh, but the ones i found are pricey. probably just better to go with a 12AU7 build

Just starting my build now, wont be able to finish it for a couple weeks though, as a couples of caps the mosfets are coming from overseas.

@the_equalizer: quick question. im putting together my bom for a ccs for my 19j6 ssmh. would a value of 1.2Kohm for RA8 and a value of 12Kohm for RA9 be alright? that should give me a current of about .542mA through the 19j6. also, is it ok to use the same transistor you selected? or will any pnp bjt transistor with the right tolerances be ok? 

 

also, what value trimmer would be good for adjusting the bias of the tubes? 

 

i am also concerned that this will constrict the use of my volume pot. with a stock build and my hifiman he-4's, i already have limited use of the volume pot. should i put some additional resistance between the pot and the input for each channel? if so, how would i go about calculating these values?

Edited by kchapdaily - 10/6/12 at 3:24pm

 

Those resistor values seem quite alright, yes. If you want to be absolutely sure and set the CCS to the idle current in your 19J6 build, simply use your meter to measure the voltage across the cathode bias resistors R5 or R11 (no music playing, volume pot fully down) and use Ohm's law to calculate the current passing through the resistor; that's the current passing through the tube. Oh, while you're at it, get some sockets for your CCS resistors, you'll surely want to experiment with other current values and desoldering gets old quickly :)

 

Your question about the transistors got me thinking. I'd suggest you use BC557 or BC560 at the very least; 2N5087 (same as used in the MiniMAX) should be better and BC556 would be best, I think. The reason is that the BC327 has an emitter to collector voltage (Vceo) of 30V which is low compared to the supply voltage of 48V. I think, the transistor should be OK given the voltages it most likely sees but it's better to play it safe. I'll try measuring the voltage across the BC327 at amp startup if I have some time later today and report back.

 

According to the datasheets I checked, the BC557 and BC560 have a Vceo of 45V, the 2N5087 has a Vceo of 50V (just adequate) and the value for BC556 is 65V (nice safety margin). All of these transistors (including the BC327) have an Hfe (current gain) greater than 100, which according to my limited understanding is a requirement for constant current sources; with the 2N5087 having a value of 250.

 

Whichever you choose, be sure to check the datasheets to get the correct pinout for your part.

 

About the trimmer, I think twice the value of the bias resistors is OK, so go with 4K. Those will still give you fine control if you decide to try higher currents (1 or 2 mA maybe).

 

Finally, I can't help much with the padding resistors as I have never used them myself. Still, remember they operate as a voltage divider so Ohm's law is again your friend, maybe 100K resistors will do. However, I suggest you try the amplifier without them first just to see how it handles; then add them later if you feel you can't tame the little amp. :)

 

Good luck with your mod.

 

cheers!

thanks for the detailed response! i have an exam tomorrow, so i wont be able to open up my amp and start measuring values until tuesday. value of 4kohm sounds good for the bias pot, and i will be sure to check out the BC556. my microelectronics professor was actually just talking about using 2 transistors to create a constant current source last week, so maybe ill stop by his office hours and see what he thinks. ill be sure to post my results! and thanks again equalizer

hi there,

nice to see theres still something going on in this thread!!

 

since my first try on the ssmh 12au7 has been hiding in a box for way over a year or even longer, last week i decided to redo it with some perfboard.

Mostly because my old build (not working) was just a mess of wires.

I ordered all parts (around 8 Euro's) and tried again, re-using the already wired pot/sockets/power switch and the rca's.

 

I fired it up at the weekend and luckily : no sparks and stuff, tubes light up fine, mosfets get warm,

so i decided to hook up a cheap mp3 player and some old sony headphones.

And then : NOTHING ... not even a hint of sound, noise or whatever, just pure silence :)

 

Btw, I wired my pot like beftus did in post #5326 (should be the same one) and the switched neutrik jack just to try if this might be the fault, on the other side, too. nothing.....

 

So i measured :

48V ---r13--- 44,7V

Zero V at both inputs and zero V at both outputs at r6 & r12

48,1 V at both Mosfet Pin 1 after r3 + r9

1,4 V at both sides of the tubes pin 3 + 8

21,6 V at the tubes pin 1 + 6 both sides , and 18,7 V at mosfet pin 3 both sides

 

Whats wrong ?

Any help is much appreciated :)

so i opened up my amp today to measure the cathode bias resistor current, resistors 5 and 11. one channel reads about 2mA, the other reads 1.21mA. now i am a bit worried that the values are not closer to the .5mA. are there any other values i can measure to check if it is functioning correctly? could this be because i used .22uF caps at C2 and C4 and 220uF caps at C1 C3 and C5?

 

How did you measure the current across the cathode resistor? 

I must ask because the "obvious" way of setting your meter to current and plugging away is totally wrong. 

 

The correct way is to measure the DCR of the cathode resistor with the amp off. This is a bad place to say "well I know its a 6Kohm resistor, plug that into the calculator" MEASURE! Measuring verifies what you already know - which is a failsafe if your trying to figure stuff out. 

 

Then fire up the beast, and measure DCV across the resistor. 

 

If you just want to verify that the 2 channels work the same just look at the voltages. 

If you are looking for a specific current do the math for: I=V/R. Respect units. 2Kohms=2000ohms NOT 2ohms...

 

If you did it right, I got nothing. 

Edited by nikongod - 10/8/12 at 2:23pm

I measured the current with the amp on, volume pot all the way down, no music and no load. Am I right in assuming that the bias resistor cannot have current without the power on? Also where should I measure voltage to see if each channel performs similarly?