| lgreen |
My 1st real DIY project, well my first project of any kind in 15 years. Its bold, but not as bold as my BOSOZ, in progress.
This is a differential input (buffered) inverting GC with a ultra low noise power supply for the INA134 used as a buffer.
 |
|
|
| lgreen |
Its an old Kyocera CD 301 cd player, got 2 (one of which was working) on ebay for $30.
faceplate was redone and interior was gutted.
 |
|
|
| lgreen |
Here is the interior:
Top board in the middle is the power supply and buffer.
Bottom board on the right is for controlling the voltages across every LED. Believe it or not the blue ones are set below spec and still light up.
Amp module is around the heat sinks. Two LM3886TF chips.
PCB, I don't need no stinking PCB for my module!
More details on my web site here:
http://home.san.rr.com/john1in2/DIY/GC_1/
 |
|
|
| lgreen |
Close up of LM3886 Module
 |
|
|
| roibm |
I can't see any picture attached...
Am I the only one? :confused: |
|
|
| lgreen |
I used parts (black gate caps and holco/riken resistors) from Peter Daniel's inverting configuration posted here a long time ago, back when inverting was the rage. Now noninverting is the rage, but looks like buffering is acceptable too. Whew.
Kimber speaker wire and pure silver wire used in the signal path. Power supply caps are Panasonic FC.
You can see from the inside pic (above) that I added 4 filter capacitors to the diode bridge, the DC just had too much ripple using the 1,000 uF caps per rail (4 rails). Had to do it, now bass is much better and HF is clear and articulate.
All diodes are high speed soft recovery, MUR1520 in the main supply (8 heatsinked unnecessarily) and MUR820 in the low noise supply.
These sinks in operation get only slightly warm. If cranking max output into a 5 ohm resistor they get much to hot to touch. If cranking max output into speakers, they get very hot, but you can still touch them. So they seem adaquate, but I am going to drill some vents into the top cover to let them cool via convection easier (holes are already in the bottom of the case for this purpose).
 |
|
|
| lgreen |
Here is the back panel.
One switch selects RCA or XLR inputs, it grounds the positive terminal of the differential input should RCA be selected. That way the balanced input passes through no switches.
The other switch turns off the Aux (low noise) power supply so that external batteries (another project in the works) can be used. You see the miniplug under the speaker binding posts for this input.
 |
|
|
| lgreen |
Here is bottom of the low noise power supply board. It uses Linear Technologies LT1964 and LT1962 surface mount ultra low noise voltage regulators mounted on bottom of board. You see that they are small, but they seem to be working well. I get some hash on the highest setting on my scope, backgrounds are silent, things are good!
 |
|
|
| lgreen |
Here is a close up of the differential receivers on the low noise PS board. This is to be used with my BOSOZ so I wanted differential inputs.
2 more LEDs indicate power is applied.
 |
|
|
| lgreen |
Another view of the low noise board.
 |
|
|
| digi01 |
very good!
have you make power supply pcb boards use paint paper and FeCl3 ?? |
|
|
| lgreen |
This board sets the LED voltages, each LED has its own trimpot, and I used diodes and base resistors for saftey. Even if you trim to 0 ohms you are still ok, and even if you reverse polarity, you won't blow the LEDs.
This board is fitted for the long cermet trimpots as well as the short stubbier type, just happens that I used all the long ones here, but for BOSOZ, I need two of these with different types of trimpots.
One reason I started this DIY stuff is that I wanted to make my own PCBs, and here in this project we have two. Very interesting even though they can be easily hardwired. I did it for the learning experience. I used Cadsoft Eagle PCB layout, and emailed my designs to www.custompcb.com.
 |
|
|
| lgreen |
The front panel was designed by me and made by www.frontpanelexpress.com. I used their free software and just emailed my design, they made it of 3mm thick anodized aluminum and it cost around $30. I have a much more complicated one for the BOSOZ which was around $60.
The lettering is etched and then filled with ink by front panel express (I have black and blue ink fill on this faceplate). Black is shown below.
 |
|
|
| S.C |
| It looks really pretty, and how dose ur not average soccer mommy GC scream like?:D |
|
|
| lgreen |
I have had this up and running for about 2 weeks, and used it without the differential buffer for another 2 weeks. The amp has very tight bass, and smooth and articulated highs. It does not sound weak at all. And yes, its stable into my Snell E's with 25 feet of speaker wire, as well as into my other speakers and into 5 ohm 100 W resistors. There is no hum at the tweeters at normal levels.
I very much enjoy the sound. Even though I have a ARC tube amp (very old), a BAT VK-200 and a Pass Aleph 3, I get a certain sense of DIY enjoyment out of this, and don't feel like I have sacrificed very much at all. I'm using a huge 500 VA transformer and the amp sounds effortless at all reasonable volume levels. It good to know the design, nothing in there to roll off the high frequencies, which I like, and it has really good extension on the top end.
As I said, this is literally my 1st real DIY project, so please give me all your comments, I really would like to hear from you here or by email (check out my web page). I am an EE by training but went in another direction when I graduated from school 15 years ago, and I didn't know a "via" from a "header" when I started this. Learned about both during the project, bought a soldering iron and a scope. I did this for fun, not to save money, this was not inexpensive.
I hope to have measurements up and posted in 2 or 3 months. I have to work on this stuff on the sly if you know what I mean. This took about 10 months of a little here a little there (in parallel with my BOSOZ, 4 months away from being finished), and the wife says it looks like its from the '70s! Hey, the case is from the '80s get it right! |
|
|
| Nuuk |
Nice work lgreen! ;)
I was always lead to believe that LED's were noisy items best kept away from hi-fi, even to the point of removing them from the mains distribution strips.
Now, if anybody can say if they are noisy, it should be you as I have never seen so many LED's in one project!
I don't think that I will ask you to disconnect them all though to see if it makes any difference - we may never hear from you again. ;) |
|
|
| roibm |
ah, finally I see them...
considering the number and the brightness of the LEDs you won't need that much energy for your bulbs, so I guess it's ok...
not to speak about the signal path which is not really optimized, but nice front tho(except the LEDs thicket)
BTW, what kind of wire did you use? |
|
|
| lgreen |
Too many LEDs? I'm sorry for finishing this project first, my BOSOZ is going to have at least twice as many LEDs (however they will be able to be turned off via front panel switches...switches with LEDs on them of course). I would hate to give you guys the wrong impression that the 8 used on the front panel here are adaquate. My apologies!
I can actually disable these LEDs quite easily, I just unplug either of the 2 receptacles from the 9 pin headers on the LED board, takes about 1 second and it can easily be plugged back in. Doing this, it does not appear that the LEDs actually add noise. So don't ask me to remove the LEDs, I like them! Ok maybe I should have used the smaller size ....hey, this was my 1st project, still learning.
Moreover, if you look at the LED board, each LED is after a diode, which prevents electrical noise created in the LED from going backwards to the power supply that it is powered from. This is good for isolation, but if you are thinking that LEDs create noise that is radiated into the air, well nothing I can do about that except locate them remotely from the circuit, and most are very distant.
To respond to the question about the type of wire, signal wire is DH labs 54927 or 63005 silver cable; speaker cable (the blue cable) is Kimber 51981 4TC (from www.partsconnexion.com).
Here is my 1st go at the BOSOZ panel with openings for illuminated switches, LEDs . This is designed to match, assuming that one is able to see both stacked on each other through the blinding blue and yellowish light. Hmmm.... I wonder if I should start looking for welding masks to protect visitors? |
|
|
| tiroth |
| quote: | Originally posted by Nuuk
I was always lead to believe that LED's were noisy items best kept away from hi-fi, even to the point of removing them from the mains distribution strips. |
That is complete rubbish. LEDs are often used as voltage reference in hifi regulated supplies and CCS. While not as silent as some references, they are obviously very quiet. Besides, INA13X and LM3886 have tremendous PSRR. |
|
|
| Nuuk |
| quote: | | That is complete rubbish. LEDs are often used as voltage reference in hifi regulated supplies and CCS. While not as silent as some references, they are obviously very quiet. |
Yes, I have to admit to having some discrete regulator circuits here that use LED's and they do work well.
This looks like one of those rumours that has been around so long, people like me believe it. At least, my take on it was to leave out LED's if I could as they certainly wouldn't improve the sound, and if there was some negative affect, it was one less thing to worry about.
I did use a couple of blue LED's to light up the tubes in my VBIGC and that also sounds just fine. ;) |
|
|
| lgreen |
| Nuuk, and anyone else who has GC galleries, if you think this is worthy, feel free to post my pics in your gallery (full version and more pics are on my temporary web site linked in this thread and "www" below). Just send me an email so I know. My pics take time to load due to slow web server, just hit "refresh" to get them all. |
|
|
| Nuuk |
| quote: | | Nuuk, and anyone else who has GC galleries, if you think this is worthy, feel free to post my pics in your gallery |
Certainly worthy! I'll pop it in when I get time. ;) |
|
|
| soundNERD |
nice work, actually, excellent.
i have never seen such an amazing looking first project!! i think you might be the next Peter Daniel.
keep up the good work!
i like the LEDs. i dont get all this noise stuff, its not like you'd notice it with a 3886 or most speakers.
-mike |
|
|
| AJ Bertelson |
| are those the heatsinks that were posted for sale on electric goldmine about a month ago? |
|
|
| Nuuk |
| quote: | | I note that they do ring somewhat when dinged with your finger. So dampening is somthing you might want to conisder. I've tried felt pads underneath, but interwoven rubber bands seem to work the best. |
Full story
here. :smash: |
|
|
| cpemma |
| quote: | Originally posted by Nuuk
I was always lead to believe that LED's were noisy items best kept away from hi-fi, even to the point of removing them from the mains distribution strips. |
One possible source for the bad LED publicity...
I've a NAD CD-player and one thing I remember from reviews around the time is that NAD said they'd used an LCD for the info panel rather than the usual LED panel to reduce noise from the rapid multiplex switching of segments carrying several mA each.
However, an indicator led on steady current is a totally different kettle of fish. ;) |
|
|
| arniel |
I seem to recall reading that NEON indicators generate noise on the supply lines...
Arnie |
|
|
| Doovieman |
Very nice work. I like the look. I was wondering if the inside of the amp gets hot. I didn't see any slits in the top of the case for the warm air to escape. Are they not needed or did you come up with a cleverl solution to let the warm air dissipate? Thanks,
Doovieman |
|
|
| lgreen |
| quote: | | I was wondering if the inside of the amp gets hot. I didn't see any slits in the top of the case for the warm air to escape. |
OK, the heatsinks don't get very hot unless you really work the amp, but I do so I was concerned about this. I drilled some ventilation holes in the cover.
I don't know if I said this before, but I have a very very minimal set of tools.
So I created a hole pattern on the computer that would overlap the heat sinks, printed it out taped it to the cover and drilled away. Here are the results. You see that some of the powder coating on the chassis cover kind of stuck and didn't drill away. Anyway, the ventilation seems ok and I can make the holes bigger if necessary. Also, if you look at the web pages you will see that I planned for this in that one of the first things I did was to drill vent holes in the bottom of the chassis to allow for convection through the top holes when I got around to drilling those.
This ventilation is pretty good, but not great. When playing this very loudly, the sinks get hot enough to burn your hand, and the vents don't really evacuate air that quickly. I'm going to do a kind of thermal study to see what gives, but it certainly works just fine and I've had no problems with extended high volume runs. |
|
|
|
