Single ended class-A headphone amp using two transistors: T2

Status
Not open for further replies.
People will hear differences all day long even where there are none, or misattribute an existing difference to a different cause. Our ears are not exactly precision measurement devices by a long shot. Anyone can waste their time however they want, but maybe they don't actually want to, you know?

Swapping out coupling caps usually takes a few minutes and is not unlikely to involve bumping the volume pot. If you know anything at all about listening tests, this pretty much ruins any chance of valid results with the kind of minute differences to be expected here.

If you want to make sure that the effect is real, here's what to do:
Make an amp that includes an extra 3-position (2p3t) selector switch, 3 sets of sockets for coupling caps, and have someone put in 3 sets of coupling caps of different values randomly, with said person noting down what's what but not showing it to you. (Shielded wiring of non-excessive length is advised.)
Now see whether you can identify which is which by flipping through the different switch positions while listening. Note down your results. Have second person shuffle around the caps and note down their actual positions, and try again yourself.
Repeat maybe a dozen times. Finally, compare your notes with the actual positions and see how much correlation there is, if any.
For easier evaluation, run it as an ABX test: Put known values in position A and B, and second person puts a second set of either A or B values in position X and again notes down what it is. You note down what you think it is after listening and flipping through switch positions at your leisure. Second person swaps out the caps in position X (or doesn't), and you try again. Rinse and repeat a dozen times or so. Afterwards compare notes and calculate your hit rate. Pure chance = 50%. If you get 80-90%, yup, that's an audible difference. 56%... ehhhh, probably not - give it a few more tries perhaps.

All of this is at least 10-year-old hat - Foobar2000 has had an ABX comparator plugin since at least 2008. (You are officially allowed to read this post in a bored and monotonous voice.) So you could even record your amp output to the PC, sync up the A and B versions, and not require a second person, if at the expense of not being able to choose playback material on the fly (maybe record a whole playlist).
 
Member
Joined 2011
Paid Member
Ouch! I hope that gets corrected quickly.

Sometimes the US Postal Service package tracking tells me that a parcel was delivered to my house on Tuesday, however it actually shows up on Wednesday. The general consensus among our neighbors is: they didn't finish their delivery route on Tuesday BUT wanted to avoid punishment for slow work, so they marked everything delivered, in the computer, even if it wasn't. Then they'll catch up on Wednesday and the supervisor need never know.
 
Grrr just checked my Mouser order and its showing as having been delivered yesterday. However it wasn't delivered to me!

I've had this happen with both USPS and FEDEX deliveries from the US to me in the UK. If it is from USPS then it reports as having been delivered a day or two before a postman brings you a card stating that the item is held at the local Postal Depot awaiting the payment of Import Duties. FEDEX just say they have delivered when they have not and sometimes report the address as having been wrong and decide to return it to sender rather than actually deliver it. If yours is from Fedex I suggest you call their UK customer services to investigate and call you back with details.

John
 
Alls good thanks the package ended up with a neighbour who dropped it off this evening.

Thanks for the heads up on that USPS Fedex funny enough I have built maybe a dozen bottlehead kits which ship from the USA via USPS there is always a customs and handling fee to pay. The strange coincidence (or not I think) is the postman always leaves the card on a Friday saying the item can be collected from the depot on Monday but if I pay a extra fee they can deliver it Saturday. Happens each and every time!
 
I was checking the fixing size for the standoffs in the bom in preperation to drill the holes required for mount the pcb and noticed the 3mm fixing hardware for mounting the pcb using the standoffs in the BOM looks to be missing.

Its not a problem here as I have a selection of mounting hardware to hand but might be handy to add the 3mm machine screws and a few washers to fit the standoffs required for mounting the pcb in the Galaxy case..
 
Member
Joined 2011
Paid Member
JamieMcC: you're right. The M3 hardware is not mentioned in the BOM. Neither is the Volume knob for that matter.

Esteemed moderator 6L6 is also right: I've been editing the "Build Instructions" .pdf file, and the newest Release Candidate does include a section that discusses the necessary M3 nuts and bolts and washers. Mouser carries some but not all of them, and Mouser's prices for these little bits are very expensive, compared to most other sellers. So I recommend you buy them either from your local hardware store, or get one of those 500 piece assortment boxes of M3 nuts and bolts from Amazon. They're not expensive, and if you build another piece of DIY audio gear you'll probably use the assortment on it too. And then the one after that; DIY can become addictive.

To build the headphone amp, you'll need these little hardware bits:

  • 6 pieces metric M3 hex nuts
  • 6 pieces M3 star washers
  • 4 pieces M3 flat washers (1/8" works too)
  • 4 pieces pan-head Phillips M3 bolts, length 20 or 25 or 30 mm
  • 2 pieces pan-head Phillips M3 bolts, length 10mm (MOSFET to heatsink)
  • 8 pieces Cap Socket (hex head) M4 bolts, length = 10mm see footnote
  • 1 piece volume control knob to fit ALPS RK27 pot shaft. eBay has hundreds to choose from

FOOTNOTE
The cap socket M4 bolts are only needed if you decide to use the PCB material, pre-drilled front and rear panels. If you decide to drill the normal front and rear panels that are supplied with the 1U Galaxy chassis kit, you can use the flat head, countersink-fitted M4 bolts that come with the chassis kit.
 
Enjoyed stuffing my pcb I hadnt noticed until actually starting but the mouser BOM includes a reference for each component postion R1 C2 etc etc on each component pack.

This was a very helpfull saved a heap of time compared to keep refering to the scematic very much apreciated thanks

I'm thinking about wire for hooking up I have two options at hand.

1/ Use some of the 28AWG as twisted pairs left over from stripping a 1 meter length of cat 6 cable for the ferrite

or

2/ Strip down a short length of some Mogami miniature star quad which is 4 x 26AWG I have a bit left from another project.

Any preference?
 
People will hear differences all day long even where there are none, or misattribute an existing difference to a different cause. Our ears are not exactly precision measurement devices by a long shot.

Agreed. It's also worth noting that the output of the various synaptic circuits in the ear are quite far removed from the auditory stimulus that enters the ear. The neuronal output is not coherent in time relative to the auditory stimulus. In the end, our brain does most of the work piecing everything together. That stereo image that you hear is an auditory illusion. :)
The upshot here is that since the brain is a major part of our auditory experience, we can train our hearing. Harman has a program that you can go through: Harman How to Listen: Welcome to How to Listen!

Anyone can waste their time however they want, but maybe they don't actually want to, you know?

Yeah... I'm a bit more of a build-it-and-be-done kinda guy myself.

Swapping out coupling caps usually takes a few minutes and is not unlikely to involve bumping the volume pot.

True, but your auditory memory is only a few seconds long, so by the time you have swapped the cap, the memory of the previous signal is gone.

If you want to make sure that the effect is real, here's what to do:
Make an amp that includes an extra 3-position (2p3t) selector switch, 3 sets of sockets for coupling caps, and have someone put in 3 sets of coupling caps of different values randomly, with said person noting down what's what but not showing it to you.

I dunno man... That sounds an awful lot like SCIENCE! Can't have that... :) :devilr:

Afterwards compare notes and calculate your hit rate. Pure chance = 50%. If you get 80-90%, yup, that's an audible difference. 56%... ehhhh, probably not - give it a few more tries perhaps.

Actually, if you end up in the latter situation, you conclude that there is no effect or difference. It's bad science to modify your experiment just because your data didn't support your hypothesis.

Tom
 
I dunno man... That sounds an awful lot like SCIENCE! Can't have that... :) :devilr:
And really, where'd that be getting us? Like, on the moon or something.
Actually, if you end up in the latter situation, you conclude that there is no effect or difference. It's bad science to modify your experiment just because your data didn't support your hypothesis.
Bad wording always keeps coming back around and biting me in the rear end for some reason. What I meant was: If you have something like 10 tries and still can't see a clear trend, obtain more data to work with, and run the analysis again with previous and new data. When the calculations are done as quickly as in this example, I don't see anything wrong with that. (Maybe the calculations should be done by the second person and the results not revealed until enough data is collected. Even so, would someone really be listening differently if they knew they had a 56% hit rate?) The only thing decidedly wrong you could do is tossing the whole shebang in a corner if the result doesn't match your expectations. Politicians have a tendency of doing that at times, usually not to the best of results.
 
HI, Jamie....

I read all that as well....my experience has been a very good one with the T2 here regardless of measurments. I do have some turn/off thump but being an experienced user I always make sure my stuff is turned on and off in a logical manner. That said I have had headphones plugged into into it and never had an issue damaging things.

In my 2 ch stereo system I have a Van Alstine amp and a seperate pre-amp and the assoicated downstream sources. I always make sure volume controls are down and turn on the pre-amp first, then the amp and on shut down turn off the amp then the preamp.

Lots of my other stuff has muting and turn on delay relays etc...I guess one could be added to the circuit rather easily if desired.

Alex
 

6L6

Moderator
Joined 2010
Paid Member
Measurements are not a good indication of your perception of sound.

Measurements are a fantastic tool, and absolutely, unquestionably have a place in audio design, that’s not at question.

There are plenty of things in audio that measure bad and sound great, just as there are plenty that sound great and measure poorly. I can’t explain it, and people have been grappling with the same question for many decades.

All I know is that in my opinion the T2 sounds much, much better than it measures. I’ll be damned if I know why.
 
Measurements are not a good indication of your perception of sound.

There are plenty of things in audio that measure bad and sound great, just as there are plenty that sound great and measure poorly. I can’t explain it, and people have been grappling with the same question for many decades.

Science disagrees with you on that. There's overwhelming evidence that listeners (trained, untrained, professionals, audiophiles) prefer equipment that measures well in blind tests. See Olive & Toole's work for examples.

Tom
 
Status
Not open for further replies.