[PinkFloyd]

Quote:

Originally Posted by 1UP Thanks, Mike, yes that explanation really helps. Begs the following questions, though: is it possible to set a gain of 5 or 5.x? What clock-position is your volume knob when listening to rock music at a comfortably enjoyable volume level? I find I need to get to between 12 and 1 o'clock with my DT990s.

The DT-990 is a pretty hard phone to drive from all accounts so you may want to do two things:

1: Up the quiscient current by 20% this will give the amp a bit more grunt. To do this remove the four 10 ohm resistors (R7 & R8) and replace them with four 8 ohm resistors. If you do this ensure you fit heatsinks onto TR1 and TR2 as they will give off more heat which needs dissipated... having said that I've got my amp running with 8 ohm resistors in R7 & R8 and the transistors (TR1 & TR2) don't get too hot without sinks but to be on the safe side fit clip on heatsinks to TR1 & TR2 if they are not already fitted.

2: Up the gain to six. This will ensure that your comfortable listening level is dropped down from 12 o'clock to around 10 o'clock on the volume control. To adjust the gain to 6 remove the 3.3K resistors (R3) from both channels.. fit 4.7K resistors in their place and that will give you a gain of around six.

You can experiment with the gain and there are no rules written in stone, adjust it until you find a volume control setting your ears are comfortable with.

Gain is calculated thus:

G=(1+R3/R2)

In the stock guise:

R2 = 1K
R3 = 3.3K

So: 1 + 3.3 divided by 1 = 4.3 gain

Here's another example:

R2= 1K
R3 = 5K

So: 1 + 5 divided by 1 = 6 gain (1 plus 5K divided by 1K = 6)

Always calculate using 1 + the value of R3 and divide it by the value of R2 that will give you the gain.

Hope this helps.

All the best

Mike.

EDIT: My comfortable listening level with HD-600's is 9 o'clock postion if I went up to 12 o'clock it would be a lot louder.

 

[individual6891]

Does anyone have any pics of the plug-top power supply's specs? Or of the supply itself? How's the power supply with regulation? LM317 regulated outputs?

 

[PinkFloyd]

Quote:

Originally Posted by individual6891 Does anyone have any pics of the plug-top power supply's specs? Or of the supply itself? How's the power supply with regulation? LM317 regulated outputs?

I believe David supplies the FRIWO switched mode power supply with the kit amp... it can be used in any country and is supplied as a "get you going" PSU if he were to provide an exotic PSU with the kit the price of the kit would surely double. I've never heard the WNA with the FRIWO but according to David it is surprisingly good!

Details of the FRIWO here

The starter PSU I use is the regulated Stontronics PSU but it's really only suitable for UK users as it comes with 3 pin connection as standard... details of the Stontronic PSU here

 

[PinkFloyd]

Quote:

Originally Posted by null Oh no! Pinkie, I'm so tempted now to order a kit! Should I!? Should I!? (EDIT) Is it hard to complete Pinkie?

Hi Null,

Question is do you need an amp, are you up to building a kit amp? If you answered yes to both then you should!!

 

[PinkFloyd]

It's just come to my attention that there is a major upgrade to this superb amp taking place and I will announce it as soon as I can in a new "WNA PCB1 PCB2" thread shortly as soon as I have evaluated the latest revisions ........... This is looking VERY promising indeed.


Mike.

 

[damitamit]

 

[PinkFloyd]

I hope Dr. White doesn't mind me leaking this:

"There was never going to be a good time to do this but there are now two new versions of the WNA headphone amp

The first version ( type 1 pcb ) is similar to the previous one except the pcb is a little larger; there are now two pairs of output devices per channel to deliver more grunt into low impedance headphones; the rail splitter is preceded by a capacitance multiplier to make performance less dependent on the quality of the external power supply; the output caps now mount on the pcb; and each channel is now completely independent ( i.e. the amp is dual mono ) so that you can have one power supply per channel.

The latest version of the cascode power supply accomodates this because it consists of two independent regulated outputs ( which you connect in series if you need a split rail power supply ).

The price for the kit is now £160 and the assembled and tested version is £210. The case is the same as previously but the pcb mounting holes will be in different positions.

The second version ( type 2 pcb ) is as for the first (pcb1) except that the pcb is larger again because the volume control and headphone jack now mount on the pcb - which can still be shoehorned into the standard case.

I'm thinking of offering this pcb in a kit with a high quality case. I have a prototype built into a very substantial heatsink case but nothing is cast in stone at present. If I went ahead with this prices are likely to be £200 for the kit and £270 for the assembled and tested version.


I realise this is tough on people who have already bought headphone amps, particularly those who have purchased recently. So to lessen the pain I will offer existing owners type 1 pcbs for £15; or a type 1 pcb plus the extra parts necessary to implement the new design for £25. This assumes that you desolder and reuse all of the components on your old pcb.

I can also upgrade assembled and tested units to the new standard but this isn't going to be as cost effective because of the labour involved. If anyone is interested in this option they should contact me directly.


You can post all of the above if you want.

Regards, David"


Looking GOOD more details to come as soon as I get them

 

[Alick]

When??? I feel a new build coming on ...

 

[aeriyn]

In reading some of tangent's papers on opamps, I read a bit on the LM6171. Do you feel that the LM6171 presents enough of a sound quality increase as opposed to something like the AD8610, which by virtue of being an FET input opamp is easier to work with?

I gathered from tangent's writing that the LM6171 wasn't in the same quality class as the AD8610 and OPA637, etc.

And this is totally unscientific and subjective, but what kind of sound does the LM6171 have? Is it mellow or lively?

What brand of film caps does the WNA use as its output caps?

 

[1UP]

Sorry, I'm not clear what the benefits of the second PCB option are - is it cosmetic, solely down to a new nicer enclosure? Or are their sonic benefits, too?

 

[Alick]

Off the top of my head (I've seen neither of the new boards); as the volume pot and headphone jack mount on the board, there are a few advantages:

1. Easier assembly (no need for headphone and volume control hook up wire and no danger of incorrect connections to pot or socket).
2. Signal routing intergrity is fixed and does not depend on cable quality/length/route.
3. Fewer solder joints in signal path. Hook up wire needs a solder joint at each end. Eliminating the wire eliminates one solder joint per connection, 9 in total if my arithmetic is correct.
4. Allows for more front panel options; you don't have to use panel mounting nuts as the PCB solder joints provide alternative fixing points.

Downsides include:
1. May restrict choice of pot and socket to those which physically fit the PCB location (although you can still use flying leads but would then be better with PCB 1)
2. If you drill your own front panel, you'll have to be much more accurate with the pot and socket hole locations with respect to each other and the PCB mounting holes on the base.
3. More difficult to upgrade to a stepped attenuator (though not impossible).

Are the input sockets and DC power socket PCB mounted too?

 

[Syzygies]

Quote:

Originally Posted by PinkFloyd and each channel is now completely independent ( i.e. the amp is dual mono ) so that you can have one power supply per channel.

This question has come up several times recently elsewhere, but I haven't seen it answered. Apparently, some dual mono designs don't mind being used with shared ground headphones, and some do.

My read of the above is that WNA isn't forcing everyone to rewire all their phones, they're ok with tying the grounds at output.

Can anyone explain the salient differences in amp design that allow some dual mono amps to tolerate shared output grounds, with others explicitly warning against this? What is the WNA choice that allows this?

In shared "ground channel" amps like the Pimeta and the PPA, the ground channel doesn't use a feedback loop that looks anything like the signal channels. It looks to me like the signal channel feedback loops would fight each other, sounding ok but wasting current, if one plugged a mono jack into the output, tying together the two signal outputs. In a true balanced dual mono amp, wouldn't the same thing happen with shared ground headphones?

Maybe what some people mean by "dual mono" is separate ground channels, each like the single Pimeta and PPA ground channel, allowing the grounds to be tied on output. And "balanced dual mono" means identical feedback loops on both signal and ground, for a design that's grumpy when the grounds are tied on output?

Am I missing something here? To recap, I've heard both answers to "Can you tie grounds on a dual mono amp?", I believe both answers apply to their respective amps, and I don't understand why.

 

[Alick]

My reading of Mike's post makes me tend to believe that the signal grounds will be tied together at the headphone socket. I'm sure Mike or David will explain shortly.

 

[PinkFloyd]

Quote:

Originally Posted by Syzygies This question has come up several times recently elsewhere, but I haven't seen it answered. Apparently, some dual mono designs don't mind being used with shared ground headphones, and some do. My read of the above is that WNA isn't forcing everyone to rewire all their phones, they're ok with tying the grounds at output. Can anyone explain the salient differences in amp design that allow some dual mono amps to tolerate shared output grounds, with others explicitly warning against this? What is the WNA choice that allows this? In shared "ground channel" amps like the Pimeta and the PPA, the ground channel doesn't use a feedback loop that looks anything like the signal channels. It looks to me like the signal channel feedback loops would fight each other, sounding ok but wasting current, if one plugged a mono jack into the output, tying together the two signal outputs. In a true balanced dual mono amp, wouldn't the same thing happen with shared ground headphones? Maybe what some people mean by "dual mono" is separate ground channels, each like the single Pimeta and PPA ground channel, allowing the grounds to be tied on output. And "balanced dual mono" means identical feedback loops on both signal and ground, for a design that's grumpy when the grounds are tied on output? Am I missing something here? To recap, I've heard both answers to "Can you tie grounds on a dual mono amp?", I believe both answers apply to their respective amps, and I don't understand why.

Good question.

I'm not in a position to answer it as I haven't got a clue what the new boards entail but I posed your question to Dr. White and here is the reply I received:

"If you've tied the earths ( grounds ) together at the output or used a jack socket with the mounting bush in contact with the jack sleeve, AND the phono input sockets are bolted directly to a metal case then you've made a tight earth loop which will cause problems with hum and noise under some circumstances. I have Never had any problems with the WNA amps ( which all have a shared output earth ) although as far as I can see most people use insulated phonos and insulated headphone jacks and the case is earthed at one point only. The new amp is no different from the old ( which was dual mono except for a shared power connection ) in this respect. So you just use your headphones in the same way as you did with the older version. Dual mono to me means that each amplifier channel is electrically and physically separate with no shared components.

The Pimeta and PPA use a shared active virtual earth ( driven by the third channel ) isolated from the real earth which solve this problem comprehensively. With careful pcb layout and sensible wiring up the active virtual earth shouldn't be necessary but it does provide an insurance policy against unintended sloppiness. Whether there are other advantages to an active virtual earth or whether it benefits the sound I couldn't comment on at this point.


There are three ways of driving headphones. (1) Two amplifier channels with a shared real earth return (2) Three amplifier channels one of which serves as a shared active virtual earth and (3) Four amplifier channels providing a differential output with no output earth. Methods (1) and (2) don't require you to modify your headphones, but method (3) usually involves cutting off the headphone jack and replacing it with a 4 pin connector of some kind ( xlrs are popular ). If your headphone lead is 3 rather than 4 wire ( the majority ) then you need to make up a new headphone lead. With Senn HD600s or 650s for example you just need to cut off the jack and replace it with a 4 pin xlr. I made up a new lead for my Senn HD600s so that I could replace the existing wire with decent silver plated, ptfe insulated wires which are shielded with a silver plated copper braid. Balanced headphone amps completely isolate each channel ( no shared earths either virtual or real ) and sound better to most people who've tried it. But its quite an expensive route to go down."


I trust Dr. White's reply helps you

All the best.

Mike.

 

[Syzygies]

Thanks!

What I read into his answer is that he isn't using balanced mono or ground channels in his design. He sets up two independent circuits starting with separate power sections, but he allows you to tie the grounds together when you plug in your earphones.

This of course forces an interaction between the power supplies, they are no longer truly independent, but my "feedback loop" concerns are irrelevant, the feedback loops are separated from ground by the impedance of the headphones.

Unless someone else wants to weigh in, I guess I have to stare at some circuits to figure out my question completely.