Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › Woo3 Modified
New Posts  All Forums:Forum Nav:

Woo3 Modified - Page 16

post #226 of 240
2/18/13 at 7:36am
Appears so in the driver section. The output tube circuit looks consistent across the evolution. Which leads me back to the CCS.
I mentioned this in the 6AS7 tube rollers thread as well. In my WA3, the PS sits around 190V depending on the tube. What I'm noticing is substantially different operating points in my amp when various tube types bias up in the output circuit...which I found interesting based on comments people make on how the different tubes "sound". The 6080 and 6N13S bias up around 38mA with 57V on the cathode resistor. 5998A and 7236 settle around 21mA with 32V on the cathode resistor. The 5998 is at the other extreme with only 18.6mA current and 28V on the cathode resistor. Some of these currents are pretty low and infringing on the nonlinear portions of their respective graphs..... Im thinking I'd like double the current in the 5998 to get it up higher on the curve. I think the CCS may be able to do this without pushing the 6080 into the danger zone. Now I've got to decide on an operating current to set the design at. smily_headphones1.gif
Quote:
Originally Posted by Capt. Z View Post

If I understand you correct the Woo 3 has gone through some major changes, not just some tweaks. Is that correct?  blink.gif

 
Reply

Head-Fi's Best Sellers

post #227 of 240
2/23/13 at 1:23pm
So, I've been messing about with various CCS circuits, trying to squeeze the size down so I can get it to fit where I want and use the amplifier plate/internal chassis as a heat sink. Using a piece of proto board, it will fit on a 1x1 square. I've wedged the incompete prototype in my WA3 for fit check and marking for the mounting screw. I'm still waiting on my low current LEDs to come in...so I can't do any testing yet. I plan to do one side 1st, then mirror the circuit to fit on the other side for the other channel.

Edited by GrindingThud - 2/23/13 at 1:25pm
Reply
post #228 of 240
2/23/13 at 1:24pm
Well, the parts came in and the prototype is now functioning in my WA3. I'll tidy up the wiring once I'm done messing about with it. Good news, it works. Better news, it may have made an improvement to dynamic 'slam' and possibly soundstage. Details are all very clear. I'll be listening to this for quite a while before I make any additional changes. I think I understand what the Crack/Speedball crowd raves about now (thank you Doc B.). smily_headphones1.gif
]

On the parts front, the LEDs are DIALIGHT 551-1107. They are low current LEDs that drop about 1.6 volts across them at a few mA current. TIP50 is bulk Motorola from eBay. The 2N2222s are some old sleazy stuff laying around in my junk box (think 1988 Radio Shack bulk pack). Resistors all nice quality from eBay. The values in the schematic set the current close to 40mA, which is what the stock design cathode resistor sets up for a 6080 and more than double what the 5998 settles in at.

Update: The WA3 definitely warms up now.....I'm going to need to watch case temperature for a while.

I'd like to credit 'pinkmouse' over at diyAudio and their ancient posts on the subject of CCS board layouts/design as well as John Camille who I believe designed the original basic circuit. http://www.diyaudio.com/forums/tubes-valves/72740-ccs-tubes-valves-pcbs.html
Their document describing the operation of the circuit can be found here: http://bildarkiv.hififorum.nu/Ryssen/Pic/Rör/diyAudio-CCS-beta2.pdf
Edited by GrindingThud - 3/1/13 at 5:13pm
Reply
post #229 of 240
3/1/13 at 5:12pm
Update 2: Temperature does not seem to be a problem, amp can run for hours and just get mildly warm to touch, so I decided to clean up the wiring and keep it. Also, I left out details on the install of the CCS. The CCS is added to each cathode in place of the 1.5K cathode bias resistors. To make the mod as simple as possible (and easily reversible) I just clipped the grounds off the cathode resistors and left them hanging in the circuit. The CCS board then has three easy connections to the amp: B+ (right off the tube socket), ground from a nearby existing post, and a wire to the circuit side of the abandoned cathode resistor. I used the ground post that the output cap bleeder resistors are attached to. Note that these posts were just grounded to the chassis with a screw....I added a separate ground wire back to circuit ground for each one to ensure a clean current path.
Reply
post #230 of 240
3/16/13 at 4:50pm
Tried rolling 6N6P into the input stage as an experiment...hummed terribly. Thought these would have worked. frown.gif
Back to the 6N1P-EV.
Reply
post #231 of 240
3/16/13 at 6:13pm

While that tube can be used as a driver, it is frequently used as a power tube.  The Valhalla uses it as the driver (and the the 6N1P as the driver).  

 

Quote:
Originally Posted by GrindingThud View Post

Tried rolling 6N6P into the input stage as an experiment...hummed terribly. Thought these would have worked. frown.gif
Back to the 6N1P-EV.
Reply
post #232 of 240
3/16/13 at 6:38pm
Yea, I was hoping it would bias up properly in the WA3 in the driver section. I've read people sing high praises of the tube in the driver role. I took some measurements and it appears to be drawing too much current and sagging the driver power supply below operable limits. I'm not willing to change the bias point of the driver section to accommodate given the large variety of tubes that do work very well.
On a good note, the 7802 is a great tube in the amp and adds a startling increase to dynamic range....I think mostly due to the lowered output impedance. smily_headphones1.gif
Quote:
Originally Posted by hodgjy View Post

While that tube can be used as a driver, it is frequently used as a power tube.  The Valhalla uses it as the driver (and the the 6N1P as the driver).  
Reply
post #233 of 240
3/16/13 at 7:22pm

Oops, meant to say "power" instead of "driver" in my last post, but you understood.  I am enjoying reading your WA3 journey.  I have a WA3 but it's in stock form.

Reply
post #234 of 240
3/17/13 at 2:06pm
That being said, when I was in there investigating the 6N6P flop, I noticed that the heater circuit is single ended. Two separate windings (one for driver tube, one for power tube), each having one side grounded. I know many (including me) have hum issues with some output tubes...ok, maybe more than some tubes. When we find a good one, its awesome. I decided to change up the filament heater ground scheme and built a virtual center tap out of a pair of 330R resistors for each winding. The 660R pair is placed right on the tube socket, across the filament leads and the center of the pair grounded. Too soon to say if this has helped any of my problem child tubes, but it certainly has not hurt anything. smily_headphones1.gif
Quote:
Originally Posted by GrindingThud View Post

.....I'm not willing to change the bias point of the driver section to accommodate given the large variety of tubes that do work very.

Update: Yes, the heater mod indeed makes a significant hum reduction. Not sure why Woo did not do it this way to start. It's 4 cheesy resistors. Highly recommended for anyone with soldering skills that is dissatisfied with hum performance. Just lift the ground wire and stub onto a resistor pair.

Yes.....eventually those Dayton caps may go....but wow, choices choices. rolleyes.gif
Edited by GrindingThud - 3/18/13 at 5:00pm
Reply
post #235 of 240
4/1/13 at 6:19pm
Caught eye of an interesting mod concept here: http://www.head-fi.org/t/652270/output-coupling-cap-and-transformer-mod
Idea is to impedance match the output using the Hammond 119DA like the Mapletree MAD ear does: http://www.veiset.net/mad/manual_EP.pdf
This would be a parafeed cathode follower configuration.
Reviews of the EAR+ seem to be very good and their measured frequency response quite flat. I'm encouraged that this mod will make an already awseome OTL amplifier an equally good low impedance unit. Difficulty is that there is no room in the WA3 to fit a pair of these inside. An auxiliary enclosure that sits under the unit may make an interesting pairing. Parts are inbound and I'll post the outcome as it comes together.

I've also ordered larger power supply capacitors. Cornell Dubilier makes a 820uf low ESR cap that looks like it may fit the WA3 PCB. Goal is to lower ESR and drive the PS noise down some more. The original caps are 330uf.
Reply
post #236 of 240
4/1/13 at 6:37pm

Your willingness to mod your WA3, and your ability to do so, are impressive.

Reply
post #237 of 240
4/1/13 at 6:50pm
@hodgjy, thanks! My wife thinks it's some kind of affliction. smily_headphones1.gif
Quote:
Originally Posted by hodgjy View Post

Your willingness to mod your WA3, and your ability to do so, are impressive.
Reply
post #238 of 240
4/1/13 at 7:13pm

The WAF (wife acceptance factor) is big part of this hobby.

 

Quote:
Originally Posted by GrindingThud View Post

@hodgjy, thanks! My wife thinks it's some kind of affliction. smily_headphones1.gif
Reply
post #239 of 240
4/4/13 at 4:18pm
The Cornell Dubilier 381LQ821M200J032 are in and they just fit. Note, the WA3 power supply board is double sided and you will need patience and a solder sucker to release the old caps. Noise floor seems much lower listening with earbuds (no noise with HD558). Transformer experiment ongoing here: http://www.head-fi.org/t/652270/output-coupling-cap-and-transformer-mod/30#post_9326874
I prefer the sound of OTL, but the transformer sounds exceptional.

.....forgot to mention I bought a quad of stealth diodes to replace the bridge. Plan was to put them on a daughter board and mount to the PS board, but the new caps were rather largish and ate my planned space. Diodes are on hold until I can figure out a clean install.
Edited by GrindingThud - 4/5/13 at 7:07pm
Reply
post #240 of 240
4/4/13 at 5:13pm

As always, impressive work!

Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Headphone Amps (full-size)

Head-Fi's Best Sellers

Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › Woo3 Modified