• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

12AU7 Hybrid Headphone Amp

thefragger

thefragger

Certified Crazy.
So I've been looking to build this headphone amplifier for some time now.

Looking at the schematic closer recently, as I've started to gather parts, I've noticed that there's only one channel shown. Tom Bavis pointed out that I have to duplicate the parts for the other channel and that the single 12AU7 tube handles both channels, but when pins on the tube do I use for each channel? Which pins are shared and which are inclusive to the R and L channels?

Additionally Tom made the suggestion that the heater coil is 20ohm and could be used to replace the 20ohm 5W resistor in the schematic.

I'm stoked-- this'll be my first tube amp build ever-- it's a modest start, I think, and being party SS I'm a bit more comfortable with it. I've had a couple stutter-starts that stalled on me, but I've got a good feeling about this one ;)


Thanks guys,
Philip.


P.S. I picked up a nice piece of Red Oak in the 75% bin at Rona today that I plan on making the case out of.
 
Register to hide this ad
Philip, pins 4 and 5 are your heaters, 9 is the center tap for the heater. Other than that, each triode is completely seperated. One half uses pins 1, 2, and 3 for the plate, grid and cathode respectively. The other have has it's own distinct plate, grid and cathode at pins 6, 7 and 8.
 
dnewma04 said:
Philip, pins 4 and 5 are your heaters, 9 is the center tap for the heater. Other than that, each triode is completely seperated. One half uses pins 1, 2, and 3 for the plate, grid and cathode respectively. The other have has it's own distinct plate, grid and cathode at pins 6, 7 and 8.
Click to expand...

Newman, thanks for that info!

What does the center tap for the heater do?
 
The center tap lets you connect the two sections of the heater in parallel for 6.3V instead of 12.6V. Connect the heater to pins 4,5 for one terminal, 9 for the other.
 
The heater coil in series is 12.6 volts @ 0.15 amps, which is 84 ohms. Each side of the heater coil is therefore 42 ohms. In parallel, 6.3 volts @ 0.30 amps, it WOULD be 21 ohms. HOWEVER, this is only 1.89 watts. Using it in a place where a 20 ohm resistor would be dissipating 5 watts will blow up the tube! Also since according to the schematic you should use a non-inductive (not wirewound) resistor there, anything with the word "coil" in it should raise a red flag. Coils are inductive, and although in reality the amount of inductance presented by a resistor isn't going to be enough to matter at anything but radio frequencies, it isn't actually the exactly correct part, regardless of how much power it would need to dissipate.
In addition to this, the amount of current a filament will pass depends on how hot it is, and it will stabilize at some point but will be very unpredictable otherwise due to the fact that varying current wouldn't necessarily fully heat it very quickly. When you consider that 5 WATTS of varying current would simply blow it out, and that a tube heater will want the correct constant voltage to work properly, it's not such a good idea to use a 12AU7 filament in the place of a 20 ohm 5 watt resistor, especially if you need the tube to be heated consistently.
 
Inductance of the heater is negligible at audio frequencies. The circuit calls for a 20 Ohm 5W resistor, which will be dissipating about 2W (6.3V squared / 20 Ohms). Since the tube can be heated by 6.3V AC, which is changing from zero to 8.9V 60 times a second, it will work fine on 6.3V DC with some audio frequency variation from that value. Average voltage will remain 6.3V regardless of audio.

A 20 Ohm load is needed for each channel. If the center tap is grounded, and each half is paralleled with a 39 Ohm, 2W resistor, it will provide the two 20 Ohm loads, and reduce the total current drain of the circuit by 150 mA since the heater is doing double duty.

I can't claim credit for this idea - Pete Millett and others have used this in headphone amps.
 
Oh, never mind then, it will work. What I hadn't realized, as I didn't really look too hard at the actual circuit, was that it was only dissipating 2 watts. What I saw was that in the notes it called for "5 watts minimum" and I assumed this must mean the resistor was dissipating something in the area of that much power.
 
Alright, last night I stayed up into the wee hours of the morning. I wired up the circuit exactly as shown in the schematics and I'm getting some 60Hz hum in one channel (sounded the same as the lighting) and nothing out the other channel.

Going to take a closer look at it, but I think one MOSFET is dead, as it didn't heat up as well as the working channel did.

It sure as hell looks pretty cool, though!


Philip.


Photos;


​
Green is ground, red is +12V.

​
Black wires go to the MOSFETs and the red / black / yellow go to the plate / grid / cathode and the green wires are for the heaters. The wires are long for testing.
 
Nice looking soldering there!

The amp is set up so that you can use the same power supply for the heater and the plate. There is virtually no noise rejection, but if you can use a switch mode power supply i/e laptop or printer power supply around 12v 2A you will have virtually noise free operation. If you have access to a 12v 7 or 9 AH battery the amp will be dead silent.

These are very common in the thrifts and you can spot them easily as they are usually flat.

Ground pin 4 and apply 12v to pin5.(or opposite doesn't matter) Leave pin 9 disconnected. You can do what Tom suggested, this is used in Pete Milllet's starving student amp. I suggest starting simple.


Mosfets are very static sensitive and will die easily, and remember the tab is at the power supply voltage.

Here is an update to this amp with a smaller part count, but must use a battery PS. http://diyaudioprojects.com/Forum/viewtopic.php?f=10&t=126

I put this amp together for new builders (like myself), this amp sounds decent and costs less than a standard hammond transformer. True sound comes from amps with much higher plate voltages, but as you see little mistakes wont bring the fire department :).
 
Roggom!!

Thanks for the compliment-- I've been into electronics for about a decade now, but doing more things related to robots and no so much audio. Never could get a CMOY up and running :(

Quick question; that 100k pot has both one end and the wiper connected?

I've been using a Canon wall wart and still have some wicked hum. No audio signal is passed. The 5W resister (R5) starts to smell like it'll burst into flames after 5 minutes with power applied.

I might have to be more careful with the MOSFET (good thing I got 20!)-- I substituted the IRF612 with the IRF610 because I couldn't find the '12s anywhere. Should they be mounted up on the same heatsink to keep their temps matched?

I could run the heater on a battery and the rest with the wall wart to see if that kills the hum?



So maybe that was a couple questions... didn't realize you were a member here! Would love to pick your brain. I'm trying to get into tubes but feel more comfortable with SS so I'm working on getting this guy up and running to experiment with.


Philip.

P.S. how would I impliment noise rejection? Throw a couple chokes on the power supply lines?
 
The pot will have pin one to the power 12v and pin 2 and 3 connected together then to the 200 ohm gate resistor. It is a good idea to write the pinout of the mosfet on paper while you are soldering, because we solder upside down so connections can get crossed.
Looking at the front of the IRF610 the pins are gate--drain--source. You have to use mica insulators or equiv to ensure the tab of the mosfet does not ground out.

As far as the resistor, it will heat up. It is running class A and is on constantly, I use a small cpu fan on top of my amp to keep it cool.

Rogers
 
Alright, I did some troubleshooting today and found that the hum is introduced at the plate of the tube, in front of the 1uF film cap.

Tried all the tubes I've got with no changes.


:(

Even with the tube out of the circuit the hum remains...

With the tube hooked in but without the heaters, the hum is fractionally quieter; with the heater hooked up the hum sounds 'fuller.'


Hmm, now I'm starting to think that maybe it's the socket...
 
Dont lose hope, try another power supply you can go up to 13v if you have to anything more and you will have to regulate the filament power supply. If you built just the second half of the circuit without the tube there is potential for hum because the mosfet stage is always on, it will amplify anything. Also if the amp is too close to the wallwart you can get some interference.


If you notice most hybrid tube amps have opamps in them that have differential amps built in. This would take the 60hz hum invert it and cancel it out. Also make sure that the ground is lifted from your wallwart. Most switching power supplies have a power cord to the adapter then another cord coming out for DC.
42W_Series_Fixed_AC_Cord_Switching_Power_Supply.jpg


If there is a ground pin on the pwer supply try a cheater plug for testing
 
I'll try running the signal straight to the MOSFET.

My wallwart came from a Canon colour printer and has no ground plug. I'm running it through an isolation transformer for good measure just to be sure that I don't screw my scope up.

That supply puts out 13V but states 12V on the body.

I need to get a load so I can test the noise coming from the supply...

I've tried a 12V 1A supply and a 4.5 800mA supply and all three so far don't change the charectoristics of the hum.


Philip.
 
Then you have a ground loop somewhere, carefully inspect all of your ground solder joints, the input connector ground the volume pot, the headphone jack etc. and try a portable mp3 or cd player as a source. Be careful where you inject signals the 12v can fry a sound card if you are using your laptop as a source. when testing I use an old cd player and a 1uF axial cap (non polar). Other devices close to the amp can bleed over like cel phones and a variac.

I sent you a pm
 
Alright. Dug out an old HP power supply and the hum is GONE, but for some reason I can hear the music on my brothers computer from two stories up on this amp during quiet parts of the music.
 
Well great that you got it working!

They say tubes are micro phonic, but... Wow. Are you playing through your laptop? are you both hard wired to a hub? I have had issues with getting my cordless phone too close and catching the stray RF.

One more thing before you fully case it up, if you add a small double throw switch , a lm317 (heatsinked) and a 4 ohm (2w) resistor you can switch the filament supply to run on 6.3 volts for use of 6dj8 and 6cg7. The 6dj8 has a little more punch, and the 6cg7 has a warmer sound for vocals. I drew it with a double pole but you should get the idea.
By all means enjoy the amp for a while first.



Credit to Greg Szekers and fa-schmidt.
 

Attachments

  • dual filaments.PNG
    dual filaments.PNG
    15.1 KB · Views: 80
roggom said:
Well great that you got it working!

They say tubes are micro phonic, but... Wow. Are you playing through your laptop? are you both hard wired to a hub? I have had issues with getting my cordless phone too close and catching the stray RF.
Click to expand...

My test source is my trust ol' Sony MP3 player (the headphone jack doubles as a line out source, switchable in software!), so, that's weird.

I'm thinking that it's the wiring on this side of the house. I picked up the speakers at a Future Shop for a song because they were missing the correct power supply, so I picked on up for $5 at a surplus shop gave that setup to my brother to use. Could it be possible that the music is getting into the my amp by way of the power supply?

Anyways, when those speakers are off the music sounds brilliant... on the channel that's working. I think I've isolated the problem to a defunct pot on the other channel.


roggom said:
One more thing before you fully case it up, if you add a small double throw switch , a lm317 (heatsinked) and a 4 ohm (2w) resistor you can switch the filament supply to run on 6.3 volts for use of 6dj8 and 6cg7. The 6dj8 has a little more punch, and the 6cg7 has a warmer sound for vocals. I drew it with a double pole but you should get the idea.
By all means enjoy the amp for a while first.

Credit to Greg Szekers and fa-schmidt.
Click to expand...

Good idea-- future proof it in case I want to roll a few tubes around later on! I'll probably throw a recessed switch underneath and label it accordingly. Also, before I close it in, which would generate lass heat-- the 20 ohm resistors or the earlier post about using the tube filaments?


I couldn't use this as a pre, could I, or would I just use the first part without the MOSFET follower for the tubey sound? The pre on my main amp has no more left channel and I'm looking for a quick fix because I have no time to tear apart my rig and diagnose the problem. I think I'm going to try giving it a TT as the input and see how the gain is-- I know full well that I don't have an RIAA EQ circuit in there, but I just want to see... :)


Lastly, I absolutely love this design, now that I understand it a bit more. Thanks so much for keeping up with me on this build!


Love this place!! :banana:
Philip.
 
You can use this as a preamp, but it will not sound correct without an RIAA stage, but for cd and mp3 it will work fine. I harvested a TT preamp board from one of my Marantz part donor units and cased it up in another similar Lansing case. It runs on 24volts and feed that into my hybrid.

Post some pics once you have it all cased up.
 
As it turns out, if you build a mirrored circuit, that doesn't mean that the transistor will work backwards.

I'm not feeling too bright right now, amp sounds amazing, though!

Any idea where I could get a passive RIAA EQ to add-on to this li'l guy?


Philip.


Post script; when the volume pot is in the 'middle' I get some wicked hum coming through. At the extremes it's loud music or silence.

Post post script; it turns out the volume pot hum was caused by my touching the bare metal potentiometer. Realizing this, I went on to touch different wires in the test setup and found that I could inject humming at different stages in the amplifier. At this point I'm going to treat humming as a non-problem until I finalize the circuit and cut wire lengths down as much as possible.
 
You must log in or register to reply here.
Back
Top Bottom