Post pics of your builds....
Apr 30, 2010 at 10:05 PM Post #6,856 of 9,811
Random blurry cell phone shots of the mess that is my senior design project (work in progress).







Behold the mysterious black box...

 
Apr 30, 2010 at 11:29 PM Post #6,857 of 9,811
I have had so bad difficulties making good cases for my amplifiers. I don't have measuring tools for doing holes in exactly right places, I have only 135W multi-tool for grinding and drilling. I estimate by eye and then drill through. That either works or not.

I managed to stuff a good CMOY into very small case that was for op amp samples from Intersil.

With AD8620 and OPA1642 and Intersil HA-5002 buffer on the ground sound is very good. Nice texture and strong bass. OPA1642 is excellent on ground channel but not so good on left and right, there is now good synergy between op amps.

I first tried Sijosae's Class AB core there, but I couldn't get it to work right and got mad and changed it to my special SMD breadboard Cmoy. Feedback loop is only couple millimeters long (one resistor between -I and Output)
This was pain in the ass to get it working right. So small space means that something seems to go wrong as I crammed the stuff into the case, maybe some wire got loose and so on. It took two days!! Now I am listening to it and offset is high still, 10 mV on both channels. Runs on one 9V external battery. I don't yet know if that will work in a shaky bus trips. Any tips appreciated!

10y4cc4.jpg


i1h3k0.jpg
 
May 1, 2010 at 8:01 PM Post #6,859 of 9,811
Here's some Stax jacks I just made. I used 3/4" Teflon round stock. I drilled the holes on a drill press using a template I made.

They are meant to be mounted on a PCB, which is then mounted to the front of the amp chassis.

staxjack004.jpg


I'm pretty happy with the way they turned out. They accept my Stax headphones very well.

staxjack005.jpg
 
May 1, 2010 at 8:23 PM Post #6,860 of 9,811
Quote:

Originally Posted by Spacehead /img/forum/go_quote.gif
10y4cc4.jpg


i1h3k0.jpg



two more pictures of it:
qyd7k3.jpg

Please give me some comments out of this layout. It is for high speed op amps.

2hmq1xg.jpg

I feel always that I need to be sorry for showing this mess. Forgive me.

The sound improved a lot now that I added 47 ohm output resistor and closed the feedback loop after it! Now output impedance is 0.01 ohm with OPA2613.

The improvements have got the DC offset down to 0.1 mV. Input impedances are now matched to 12K.
With OPA2613 and Intersil HA-5002 on the ground current output is very high and it can drive any headphones. Designed for 32 ohm though. Sound is tight and strong. Need to find best sounding op amp to left and right...
Reliability is a question still.
Should I attach the wires somehow? I have never made a portable amplifier.
 
May 1, 2010 at 9:50 PM Post #6,861 of 9,811
Quote:

Originally Posted by Spacehead /img/forum/go_quote.gif
Please give me some comments out of this layout. It is for high speed op amps.

I feel always that I need to be sorry for showing this mess. Forgive me.



Admittedly it is a bit of a mess, yes I know it's a bit harsh and yes you are using smt, but it could be a lot cleaner. Having everything crammed into that box like that is going to make it harder to fix/change any problems/bugs that might arise.

Wouldn't it be better to draw up a schematic for that amp and figure out a small form factor layout, rather than adding bits here and there? Normally I'd prototype a circuit on a breadboard first until I was happy with the circuit, then draw up how I'd connect it on a proto board then finalise it.

Even if you had the whole board laid out properly but still undecided on certain components, you could make them swappable, but normally that would be done in the prototyping stage.

Just my two pennies worth, feel free to argue anything with me.
smily_headphones1.gif
 
May 1, 2010 at 10:17 PM Post #6,862 of 9,811
Quote:

Originally Posted by Pageygeeza /img/forum/go_quote.gif
Wouldn't it be better to draw up a schematic for that amp and figure out a small form factor layout, rather than adding bits here and there? Normally I'd prototype a circuit on a breadboard first until I was happy with the circuit, then draw up how I'd connect it on a proto board then finalise it.

Even if you had the whole board laid out properly but still undecided on certain components, you could make them swappable, but normally that would be done in the prototyping stage.

Just my two pennies worth, feel free to argue anything with me.
smily_headphones1.gif



I have still yet started using schematic programs. Simple op amp circuit doesn't need one, I have it on my mind.

That is my first implementation of cmoy with only SMD resistors. It is pretty cool that I got it working. I now have EL1516 on there, that is 350 Mhz op amp and it isn't running hot and there is only about 20 mV offset. So I guess it isn't oscillating.

I have thought about doing it into one single piece of breadboard. But it is working fine, so why would I bother? (I will some time, I know). I have been changing the resistor values and added some decoupling capacitors. I am in a point where I don't need too much modifying.

What I am missing is the ground plane! I have read some Analog's and Linear's documents and as I don't have ground plane, performance suffers a lot. But is that audible, I don't know?

Now I hope that the layout is small enough so that I don't hear cell phone network changing beeps when I am traveling with a bus. Feedback loop is about 1/32 inch long. Is that small enough?

I am enjoying music through this amplifier now. It sounds pretty hifi actually. Quite close to PPA v2 performance with proper high current op amp. Very silent too with battery.
 
May 1, 2010 at 11:39 PM Post #6,864 of 9,811
The Only reason I mentioned drawing up the schematic for it is so you could optimize the layout. I did it by drawing the circuit on some squared paper and positioning all the components to make sure channels didn't cross, keeping it tidy and generally keeping a small profile. A lot of those wires inside that box aren't necessary and will only pick up RF interference easier, which leads onto the noises you get from phones through amplifiers. I've reboxed my CmoyBB into a plastic project box and that is a very optimized circuit and yet that still picks up phone signals. As for the ground plane, you won't be able to implement it into that circuit properly. Though I would love to see the schematics for it so I could see how it could be better arranged, I have a few prototyping ideas that might be beneficial to your project.
 
May 2, 2010 at 12:50 AM Post #6,865 of 9,811
Quote:

Originally Posted by Pageygeeza /img/forum/go_quote.gif
Though I would love to see the schematics for it so I could see how it could be better arranged, I have a few prototyping ideas that might be beneficial to your project.


I am interested

Left and right:

nfgpox.jpg



ground channel:

35k02mp.jpg
 
May 2, 2010 at 11:32 AM Post #6,866 of 9,811
Need a bit more info on that setup Spacehead. I need to know your virtual ground setup too. I also need to know how you are attaching the inputs and outputs to the channels.

Just one thing to clarify, I'm a relative noob too, so I might need correcting.

Is the feedback loop on the L+R channels really necessary? The way I'm reading it is that you're feeding the signal from the end of the amp into the headphones but returning it back to the input stage of the amp? If that is correct then the signal returns from each channel via the ground not the channel it's feeding.
 
May 2, 2010 at 11:47 AM Post #6,867 of 9,811
Quote:

Originally Posted by Pageygeeza /img/forum/go_quote.gif
Need a bit more info on that setup Spacehead. I need to know your virtual ground setup too. I also need to know how you are attaching the inputs and outputs to the channels.

Just one thing to clarify, I'm a relative noob too, so I might need correcting.

Is the feedback loop on the L+R channels really necessary? The way I'm reading it is that you're feeding the signal from the end of the amp into the headphones but returning it back to the input stage of the amp? If that is correct then the signal returns from each channel via the ground not the channel it's feeding.



that ground channel op amp gets virtual ground signal from TLE2426 through R1

input jack left and right is connected to C1 and continues from there on. signal ground is driven by ground channel op amp before buffer

output jack is connected to left and right channel op amp after the 47 resistor

output jacks ground is connected to ground channel buffers output

feedback loop sets the gain
it is necessary that R3 is connected to ground, it is reference point
op amp sees ground 15K away and output 47K away
it tries to catch up these differences and amplifies the signal so that these points are reached. but it can only amplify to its output, so R3 stay referenced and R4 gets the output signal back to -Input so that op amp knows it is doing the right thing
In simple words, it tries to make both inputs the same voltage
 
May 2, 2010 at 10:47 PM Post #6,868 of 9,811
Ok, think I've got my head around it now. I'll need to know your power supply/virtual ground supply setup so I can put it onto the same board. Or I might be able to design it as a plug-in module.

Just one other thing, though this might be up for debate..... Which way round does the ground channel go? Going by current flow the input would be from the headphones, so it'll going the opposite direction to the channel amps themselves. Is this correct?
 
May 2, 2010 at 11:49 PM Post #6,869 of 9,811
Quote:

Originally Posted by Kerry /img/forum/go_quote.gif
Here's some Stax jacks I just made. I used 3/4" Teflon round stock. I drilled the holes on a drill press using a template I made.

They are meant to be mounted on a PCB, which is then mounted to the front of the amp chassis.

I'm pretty happy with the way they turned out. They accept my Stax headphones very well.



Where did you source the pins?

I make the backer pieces (just made 2 yesterday), where I use some 5/8" rosewood (or whatever scrap I have in 7/8").

The process I use is:
1. Start with 1.125" or 1.25" square material
2. Draw a cross from the corners to get a center point
3. Use a circle template (or compass) to mark the screw holes at 1.125"
4. Use a 7/8" Forstner bit to drill to depth
5. Drill the balance with a 3/4" Forstner (5/8" is marginally ok, but rubs the pins).
6. Drill the screwholes.

Should take no more than about 15 minutes once you have the sized stock. This is exactly why I save every scrap of wood from all my projects. Delrin or something more exotic might be better, but wood is expedient.

The attached are walnut 7/8" stock, but I like the rosewood better.
 
May 2, 2010 at 11:50 PM Post #6,870 of 9,811
Quote:

Originally Posted by Kerry /img/forum/go_quote.gif
Here's some Stax jacks I just made.


So that's where all those connex pins that you bought went
wink.gif


Great work~ They look fantastic!
 

Users who are viewing this thread

Back
Top