Closing Out January

This week wasn’t quite as challenging as last week. There were some successes and some frustrations.

Finishing the Dynakit

I got the Dynakit sorted out. It turned out to be an embarrassingly simple solution: one little loose solder joint.

That’s much better - 30 watts per channel, per the specification.

The CM Labs Preamp

I mapped out the connections and voltages, and got the boards out. Replacing the capacitors and touching up the solder did the trick.

The cabinet needed a little sprucing up with organic wax furniture polish.

The Dean Costello

After going in circles for much, much too long, I found the problem: a flaky resistor. I replaced it.

There’s the new resistor, 2.7k, 1/2 watt.

There’s the Costello on the bottom, the Dynakit in the middle, and the CM Labs on top. The Dynakit has a nice new (used) cover, which is good for safety. I was surprised to find that Dynakit still has this cover available new.

The Black Lion B173 Preamp

This preamp has been here for… well, a while. I found the transistors I had ordered when I was looking for something else, and I decided it was time to just try it and see. The issue is that the preamp has some crackle, and the transistors are the likely culprit. The company was unresponsive, so I was on my own. There is no schematic available - it’s based on a Neve 1073, but the original schematic was not very illuminating. Unfortunately the transistors are in pairs in tiny surface mount devices with a three letter identifier that doesn’t directly point to an obviously appropriate component - the closest was an NPN/PNP pair, so I ordered some.

I extracted the Black Lion from the 40’ shipping container which is my overflow storage and set out confirm that the new transistors lined up with the old transistors, because orientation and connections matter with transistors. These initial tests showed that the little 6-lead device was made up of two NPN transistors, and I rolled my eyes a bit at myself for not having ascertained this earlier, because putting a PNP in place on an NPN would not have worked.

I wanted to confirm that my initial findings were correct, so I floated one of the tiny 6-lead devices off the board. I then realized that the leads were much too tiny for connecting my transistor analyzer. I soldered the little device to an adapter board so I could take the readings. (I’m avoiding calling it an IC, because it’s not an integrated circuit, it’s just two tiny transistors combined into one little package)

The transistor analyzer (connected up as shown in the photo) confirmed that the device is, in fact, a pair of NPN transistors. I ordered NPN/NPN transistors with (hopefully) the right pinouts.

The Godwin

What I need is a power supply that provides +21V, +15V, -5V, and -15V. Dedicated three-terminal voltage regulators are available for the last three. The +21V is the problematic supply, because there’s no regulator dedicated to that voltage. There are variable regulators that need a few more components around them to dial in the voltage. I drew out the circuit and tested it on the breadboard.

The +15V supply on the existing board is working properly, as is the rectifier and filtering. I removed all the other parts and installed the variable regulator, the -5V regulator, and the -15V regulator. Next I connected it to the new 36V center-tapped transformer and powered the whole thing up.

The +21V supply doesn’t need to supply much current. This reading from the programmable load shows that the voltage stays stable at almost +21V while it’s providing 50mA. This is a good result. The +15V and -15V and -5V supplies are also stable.
The next step will be to install the transformer and the upgraded power supply in the Godwin and continue with the troubleshooting. I would have done this, but I was out of space on the bench.

The Fisher

The Fisher belonging to my trainee remains recalcitrant. We replaced some of the electrolytic caps, which didn’t help, but had to be done anyway. We traced the signal through it, and adjusted the DC offset of the output. We ended up focusing on the FET (Field Effect Transistor) pair at the beginning of the signal chain. The part with the back-to-back transistors is no longer available, so the plan is to replace it with a pair of similar FETs and bond them together so that they maintain temperature. I’m hoping this does it, because this passed from Repair Project to Electronics Lesson In a Box some time ago.

The Pioneer SA-1050

I had all but given up on this, but I woke up early one morning and had an inspiration, which resulted in my finding a shorted capacitor, caused in turn by a short in the speaker selection switch. I replaced the cap and cleaned the switch.

I also replaced those two crispy resistors that feed the headphone jack.

It now has a satisfying 103W/channel output. Whew! The AP analyzer makes this test so easy, I am really trying to remember what it was like before I had it.

The Collaro

I finally got the Collaro record changer in a cradle for testing. The motor works, the switch works, but pretty much everything else is frozen.

Both of these wheels are supposed to be able to contact the spindle, and the spindle is supposed to move up and down. The wheel on the left is the idler which turns the platter, the one on the right drops the record and moves the tone arm. The wheels are vintage rubber and have lost all their flexibility and grip. It’s all fixable, but it’s a big job.

Thanks For Reading

There were some other projects this week also, so there may be a supplement. Whether there is or not, thank you very much for reading. If I said I would get your thing on the bench this week, please accept my apologies. I’m pedaling this unicycle as fast as I can.

As always, stay safe, stay warm, and tell the people you love that you love them.

Also as always, this content is 100% Alden and 0% HAL9000 or any of its friends.

Bright blessings,

Alden