[f] AMB on Facebook

My γ3 build, and adventures in DIY panelwork (long)

gamma3, gamma24 (plug-in for the γ3)

My γ3 build, and adventures in DIY panelwork (long)

Postby Rabbit » February 10th, 2017, 9:28 pm

This is a recap - I have been working on my γ3 since May of 2016 or so. The photos below, and others, can be found here https://flic.kr/s/aHskDbfQwu

Ordering parts
I decided to go with the AMB reference build because I was overwhelmed with the amount of options. I had not fully wrapped my head around the expanse of this project when I decided to take on the project. As a result, I had to place a few different orders, including at least 3 at AMB's shop, as the build proceeded. This is, by far, the most complicated DIY project I've undertaken.

Many, many bags
I took my time to label all of the parts. I used a gallon ziploc for each board. I found this to work very well. Labeling the parts took a long time, but it was definitely worth it.

Populating the boards
The σ11/σ22 builds went smoothly. I did populate an extra resistor on the σ11s. I discovered it when my voltage was not correct. Had I paid more attention to the instructions, this would not have happened. I tend to get very excited and go nuts ordering everything on the BoM, and populating everything.

ImageIMG_9638.jpg by PScal, on Flickr

The γ3 and all its child boards took a bit longer to build. This was not my first encounter with surface mount soldering - I have built a γ2 and also enjoy making DIY LED flashlight drivers which use surface mount components. I recently bought a cheap hot air station, which earned its keep when I positioned not one, but both of the DAC chips crooked. I should have waited until I was fresh to work on the difficult chips, but it worked out in the end. Somehow, I nailed the SRC chip on the first try. I used solder wick and liquid flux to achieve my best results.

ImageIMG_9682.jpg by PScal, on Flickr

ImageIMG_9706.jpg by PScal, on Flickr

ImageIMG_9681.jpg by PScal, on Flickr

At this point, given the massive amount of time and money invested in the build, I took a break to gather my courage. This is, by far, the most expensive and time consuming DIY project I have taken on, and I was very nervous. I spent a lot of time staring at AMB's reference photos. It was at this point in the build when I began to understand more about the γ3 and its various components. In retrospect, I should have overcome my excitement and done this research before I started ordering parts.

I wired up the power supply, drilled the enclosures, and mounted the boards. Many of my holes were off, but the tiny round file helped me out. The tiny round file is on the top of my list of most useful tools for this project (more on that later.) I built the umbilical power cord.

ImageIMG_9660.jpg by PScal, on Flickr

The initial check was next. I discovered an error in my power umbilical. I did not mirror the cable, so the entire thing was backwards. I am SO glad I went through the entire initial check procedure because this would have been disastrous had I not found the problem. I gathered my courage once again and prepared to flip the power switch.

With a tiny pop and the slight smell of magic smoke, all of my worst fears came true when I flipped the power switch for the first time. The minutes after it happened were filled with denial. Once I got over the denial, I began troubleshooting. I knew that the large scale of this project would most likely present complexities when it came to making it operational. That is part of DIY (for me, at least.) I quickly located an upside-down chip in its socket in the LCDuino, where the smell had originated (Amateur hour!!!). It was the port expander - I forget the part ID. $1 part. I went ahead and checked all of the voltages while it was powered on and everything looked good.

I ordered 2, and also another copy of the motor pot controller chip, just in case the upside down chip took anything else down with it. When I installed the new chip, the γ3 powered on. Sweet.

My main source is USB. USB not detected. A small amount of research led me to re-flow the 2 crystals under the ζ1. Once I did that, the ζ1 fired right up and was recognized by USB. I hooked up the RCA cables to my computer speakers, crossed my fingers, and clicked play. Music. Sweet, sweet music. The sense of accomplishment and pride I felt was great. My wife, working at her sewing desk, thought something was wrong because I was so aghast at that moment. Success!!

The next (and ongoing) phase of this project is panel work. I decided to DIY the panels. After having trouble reaching par-metal, I ordered 4 pieces of 1/4" aluminum (for the fronts) and 4 pieces of 1/8" aluminum (for the backs) from online metals dot com. 2 for the γ3, 1 for my M cubed (so they match), and 1 extra. I plan to paint these and possibly use my wife's vinyl cutter to make a stencil for the labels. At this time I am not sure exactly how they will be finished.

ImageIMG_3718.jpg by PScal, on Flickr

The parts arrived and looked OK. They were scratched, and had been cut by a brake, so there are crease marks and shearing on the cutoff sides. I cleaned up the sides with 50 grit sand paper on a flat surface. This kept me busy for several nights.

I used a par-metal front panel from my M cubed as a template for the corner holes. that went pretty well. I sanded the front of one of the 1/4" panels to give it a brushed look for the time being, and screwed it onto the front of the PSU, totally blank. I may drill a hole for a power LED, but it's pretty obvious when it's powered on because of the LCD display on the γ3. Panel 1, stage 1, done.

The first cuts I made were for the rear panel of the PSU. I absolutely butchered the square hole for the IEC inlet. Luckily, the lip covers the hack job. I did not bother with the D cutout on the umbilical connector - drilled it round. Panel 2, stage 1, done.

The next panel to tackle is the front panel of γ3. I bought a big pack of 20 tpi blades for the scroll saw I inherited but had never used. I took it very slow and stayed inside the lines. The cut turned out OK after I cleaned it up with the file. While drilling one of the LCD mount holes, I felt the bit hit the other side of the panel. CRAP. I immediately stopped. Sure enough, there's a tiny hole on the front side. I sanded the front with 50 grit and it smoothed out the hole to be barely visible. I think it will be unrecognizable after it's pained. A couple round holes and a countersink later, and panel 3, stage 1, is done. It did not come out perfect, but I did not expect it to. I am satisfied with how it turned out.

(Sorry for the crappy photo. The volume knob is temporary.)
ImageIMG_20170210_104351059 by PScal, on Flickr

For the rear panel of γ3, I decided to follow AMB's reference design. This turned out to be a mistake. Because I was using 1/8" material and not milling into the panel from behind, there were several points of contact between the panel mount jacks and the γ3 circuit board. I had several hours in that panel, so it was a painful lesson. I'm glad I got spare panels. At this point, if I fail at another attempt, I will just leave my M cubed as it is for the time being and use the panels previously intended for it as spares.

That's where I am right now with my γ3 build. If anyone out there is thinking about building a γ3 and has read this, I hope my experiences aid in your build. I will post an update after my next attempt at the rear panel. Until then, I will be enjoying the fantastic γ3 DAC. Many thanks to AMB for the excellent design, documentation, and support he has provided for this high end DAC he has created.

3 people like this post.
Rabbit
 
Posts: 58
Likes: 28 posts
Liked in: 22 posts
Joined: April 7th, 2010, 6:50 am
Location: Little Rock, AR
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby amb » February 10th, 2017, 11:08 pm

Very nice! You are doing a great job. γ3 is ambitious, complex and challenging for me to design. As a project to build, it also has those characteristics. Compared to some other people, your experience is actually relatively smooth. A few errors are understandable in a project like this, but after all is sorted, you will be very proud of what you accomplished!
User avatar
amb
Site Admin
 
Posts: 10380
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby poseidonsvoice » February 11th, 2017, 8:23 am

Doesn't the ζ1 come fully assembled/tested? Which crystals needed resoldering?

Best,
Anand.
User avatar
poseidonsvoice
 
Posts: 109
Likes: 68 posts
Liked in: 9 posts
Joined: April 21st, 2012, 11:09 pm
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby Rabbit » February 11th, 2017, 8:31 am

X1 and X2 on the γ3 backplane were the culprits. The ζ1 was indeed fully assembled and functional.
Rabbit
 
Posts: 58
Likes: 28 posts
Liked in: 22 posts
Joined: April 7th, 2010, 6:50 am
Location: Little Rock, AR
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby Rabbit » February 14th, 2017, 7:30 am

I haven't had time to have another go at the rear panel, but in the meantime I've been living on the edge and enjoying my γ3 with the connectors dangling. I've also been reading up about XLR connectors - they are new to me. I have a question about the shield, pin 1:

I currently have the 3 wire harnesses on my γ3 connected like: Pin 1 to IG, Pin 2 to output +, Pin 3 to output -. On my downstream amplifier, the instructions say Pin 1 should be connected to the chassis, which is also connected to the safety ground. Will I be creating a ground loop with this wiring scheme? My intuition tells me I should attach Pin 1 to the chassis of γ3, and leave IG disconnected on the 3 pin harness.
Rabbit
 
Posts: 58
Likes: 28 posts
Liked in: 22 posts
Joined: April 7th, 2010, 6:50 am
Location: Little Rock, AR
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby MisterX » February 14th, 2017, 9:16 am

Your intuition is correct.

http://pin1problem.com/
1352.5 and counting.
User avatar
MisterX
 
Posts: 1143
Likes: 52 posts
Liked in: 36 posts
Joined: March 31st, 2010, 2:00 am
Location: Kalamazoo, MI
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby amb » February 14th, 2017, 3:17 pm

The γ3 has galvanically isolated sections, so you won't run into ground loop issues by wiring the analog output XLR connectors as instructed on the website (pin 1 to IG). If you do wire pin 1 of the analog outputs to the chassis, you actually defeat the isolation, and create a ground loop scenario!
User avatar
amb
Site Admin
 
Posts: 10380
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby Rabbit » February 14th, 2017, 5:44 pm

Now I am confused! MisterX's answer corroborates with my intuition but AMB's seems to not. For the record, I am only talking about attaching the XLR shields to the chassis and not to IG; RCA (-) will remain connected to IG and the RCA jacks will be isolated from the chassis.

Here is my justification - please correct anywhere that I am mistaken. Again, I am new to XLR and balanced, shielded cables, so please pardon my lack of knowledge.

It seems to me that it would be undesirable to attach the isolated ground to the XLR shield, which connects to the chassis of my amplifier, and because my amp is grounded, the safety ground. Since IG is isolated from the chassis, it seems preferable to connect the chassis to the XLR shield, keeping IG isolated and only utilized for signal purposes and not for shielding the interconnect.

The worst case would be to connect IG to the XLR shield, and the XLR shield to the chassis. That's definitely asking for trouble.

It could also be that I'm overthinking this. Please let me know if I am! :ugeek: :? :idea:

Please bear with me as I wrap my head around all this. Thank you for the information and support so far.
Rabbit
 
Posts: 58
Likes: 28 posts
Liked in: 22 posts
Joined: April 7th, 2010, 6:50 am
Location: Little Rock, AR
Country: United States (us)

My γ3 build, and adventures in DIY panelwork (long)

Postby amb » February 15th, 2017, 12:14 am

Ideally, and in simple circumstances, your intuition would be correct. I.e., pin 1 should be connected to the chassis only.

But the reality is that many (most?) balanced devices have their pin 1 connected not only to chassis, but also to their signal ground too. This includes just about every balanced device I've come across, from the Benchmark DAC1 to the Behringer DCX2496. This has to be taken into consideration.

Also , many balanced amps are not fully differential, but are implemented with two ground-referenced amplifiers per channel to handle the hot and cold signals.
User avatar
amb
Site Admin
 
Posts: 10380
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: My γ3 build, and adventures in DIY panelwork (long)

Postby Rabbit » February 15th, 2017, 7:11 am

Thanks for the information, AMB.
Rabbit
 
Posts: 58
Likes: 28 posts
Liked in: 22 posts
Joined: April 7th, 2010, 6:50 am
Location: Little Rock, AR
Country: United States (us)

Next

Return to γ3 high resolution DAC and γ24 DAC core module

Who is online

Users browsing this forum: No registered users and 0 guests

Quick Links

AMB audio DIY resources
AMB audio shop

cron