2/29/2012

The building of a Beta22 part 1

Somewhere in October I decided to build a 2-channel Beta22. It has been almost five months since then, and a lot of money. Currently I have spent about  €540 ($725) and about 50-60 hours on this project.
The best part is: I'm not even done yet.

In this blog post I will describe what I have done up until now, including the painfully long hours of troubleshooting.



The whole project started with ordering the parts, which unfortunately took quite long. I ordered parts from Glass Jar Audio and from THL Audio. Jeff Rossel from Glass Jar Audio was usually quite quick to respond, but it still took about three weeks to get the ordering done because I was very much unsure about how I would do the grounding of all the jacks in the beginning. With Ti Kan's help I eventually sorted it out, and I was ready to order.
Correspondence with THL Audio on the other hand went far from smooth. This is the company where I ordered my chassis and knobs from. They are based in Taiwan, which is unfortunately quite noticeable in their terrible English. They usually took three days to a week to respond to my emails, and as a result it took quite some time before I finally got the order up and running.
There was also a final third company from which I ordered parts, this was an eBay seller named Parts Pipe. I needed a total of four DPDT switches (two poles, two states). As expected from those Hong Kong based companies everything went without a hassle, and to my surprise they were actually the first parts to arrive.

When finally all the parts were in my possession, I was ready to start populating the boards.
My workbench prepared for the hard work that was about to come

All the resistors populated on the first board

Small capacitors and resistors

Large capacitors and trimpots
Unpopulated second channel

It takes a lot of time to solder all the parts, so eventually I decided to just  almost all of them in one go like this. 

All resistors and transistors populated on the second board.  In this channel I actually messed up the orientation of the JFET's, which set the project back quite a bit of time.

Both boards almost fully populated. In both cases some resistors are still missing because Jeff from Glass Jar Audio made a mistake in the order and gave me some wrong value resistors. Luckily he corrected his mistake by sending the right resistors very quickly without extra charge.
Also visible in this picture is the fact that I accidentally put the trimpots (the blue parts in the middle with the screw) the wrong way around. It was incredibly difficult to remove them, and I can call myself lucky that I didn't destroy them in the process. 

Both channels fully populated, except for the six missing resistors.The boards are actually quite heavy at this point.
After this I went straight on with the Sigma22 (the power supply)

About half-way populated Sigma22. A lot less (hard) work than the Beta22.

The fully populated Sigma22 board. 

Starting on the chassis. I measured and marked all the points which I needed to drill. In the middle of the crosses I drilled  small pilot holes, which were later enlarged to the correct size with a larger drill.

All the holes except for the one for AC have been drilled on the backpanel. Note that I made a small mistake on the position of the top right hole in the middle. 

Since I didn't have access to any reliable method to drill rectangular holes, I decided to make little holes in the corners and saw myself a rectangle using a hacksaw. I then filed of the remaining bits until I was left with a rectangle exactly the right size. 

The front panel ready to be drilled. 

All the holes in the front panel drilled. I didn't have a drill large enough for the XLR hole, so I had the manually enlarge it using a file with rather terrible results. I am now of the opinion that it's not humanly possible to file perfects circles. I eventually filled up the hole with putty, which will hopefully be invisible after painting.

Populated front panel. Only the potentiometer is still missing, as Jeff forgot to include it in my initial order. He sent it to me along with the missing resistors with no extra charge.

The back panel. Two RCA jacks are still missing since Jeff ran out of them. I still have not received them, and I believe I will have to order them from eBay.

The three boards sitting inside of the enclosure waiting for the missing resistors and potentiometer.

The initial try of the wiring. very messy, and ultimately did not work either. In this picture the two windings of the transformer are of inverted phase, and the grounding is messed up for all sorts of reasons. I never imagined it to be this difficult to do properly, but it is.
When I started it up in this configuration the fuse immediately blew. Oops. After I ordered new fuses and corrected my mistake in the transformer wingdings the amplifier still didn't work. The voltage across a certain resistor on the left board was way to high, indicating that JFET's were blown because I got the orientation wrong. After checking how I had them on the right board, I accidentally destroyed one of the JFET's by brute force. Since these JFET's are matched to a very high degree, I needed to replace all the 8 JFET's, even though only 5 were damaged. Ouch!

Removing the JFET's wasn't an easy job either. And even after I had removed them I still needed to put the new one, which meant I had to remove pretty much all of the solder from the hole. Since the only solder sucker I have is very weak, I solved the problem by driving a small screwdriver through the hole while heating it, and then removing the screwdriver. 

This is a picture showing one of the many examples in which I messed up the wiring. It took me an endless amount of time to get the wiring to working order. I thought I was following the schematics perfectly, but in the end I wasn't. I wish to thank Ti Kan, the designer of this amp, for helping me through this painful state in which the amplifier 'almost worked' every single time I fired it up.
One of the problems in the wiring is that the input RCA jack need to be grounded to the potentiometer, and completely isolated form the chassis. For this I also had to drill bigger holes for the jacks in the chassis, such that there would be a layer of air between the chassis ground and the body of the jacks.
Another problem is that the signal ground is way too long, and connected at the same point as the ground of the AC receptacle. I fixed this by grounding the ground loop breaker much closer to the power supply.

The current status of the project. The amplifier is in full working order, but produces an incredible amount of noise. This is most likely caused by the proximity of the power transformer and the mains lines. I will try to fix this by putting these two things in a separate enclosure.




Geen opmerkingen:

Een reactie posten