Author Archives ⇒ daryl

TPS-4000 Replacement Parts

· daryl

The new parts arrived to repair the Topward TPS-4000 power supply I had diagnosed in a previous post.

The Motorola MJ2955 was replaced with a MOSPEC branded part as originals are in short supply and popular with audio amplifier enthusiasts.

The 2SA1015GR was replaced with a BC557C as I have many of these.

Being the inquisitive type, I decided to cut open the old power transistor to see what was inside.

Using a digital microscope I managed to capture significant damage on the silicon.

Completed power supply.

Topward TPS-4000 Power Supply Repair

· daryl

The Topward TPS-4000 is a linear bench power supply with two independent adjustable outputs and a third fixed 5V output. It is this 5V output that has failed.

Original schematic (incorrect)

The original schematic seems to be missing the pass transistor and has the other transistors connected in an odd manner which suggested something might not be right.

Reverse engineered schematic (correct)

The transformer/bridge rectifier/filter capacitors have omitted for clarity.

LTspice Simulation

Q2, BD912, would have originally been a 2SA699 transistor.

Failed components – Motorola MJ2955 pass transistor and 2SA1015 sense transistor.

The MJ2955 has a short circuit between the collector & emitter and the 2SA1015 has some odd junction behavior when tested under forward & reverse bias.

I’m waiting for a replacement MJ2955 and the 2SA1015 will be replaced with whatever jellybean transistor I have at hand.

New components and completion

· daryl

The following pictures show the new components and the completed board.

Capacitors

The above image shows the new and old Rifa capacitor. You can see the crazing and discolouring on the left hand side.

TRIAC

There was some difficulty in finding an alternative TRIAC. From the part numbers on the package, TAG D2 450 800 , I took a guess on the typical specifications as I could not find a vintage TAG Semiconductors parts catalogue. I used a ST BTA16-800CW with a blocking voltage of 800V and an RMS current of 16A.

Finished PCB

The motor is now working well unloaded and is surprisingly loud at max speed (28,000 rpm).

Watch this space for a reverse engineered circuit diagram.

PCB Cleaning

· daryl

The following components were removed after inspection and shall be replaced:

2x 22pF Class Y
3x 0.1uF Class X
2x Resistors
2x 220uF Electrolytics
1x TRIAC – This was not faulty but the legs snapped on removal of the clamping bolt.
1x Inductor – Just for inspection.

Bare PCB

The failed resistor had become quite hot and burnt the top of the PCB. I cleaned this with a fiber glass pencil and applied a couple of coats of clear nail varnish afterwards.

Back of PCB

The back of the PCB shows some signs of the burning but it is still intact.

Damaged Capacitor

Above is the damaged capacitor next to it’s new replacement.

Strip Down

· daryl

It soon became obvious once the lid was off where the magic smoke had come from and where the problem may lie.

Motor controller

The controller is full of the dreaded ‘Rifa’ capacitors, these are paper capacitors with resin dipped casing that breaks down with heat and age allowing moisture in. Once this happens a bang soon follows.

You can see the 0.1uF capacitor in the centre of the image has split open horizontally and in the background a resistor has also burnt.

Brake upgrade

· daryl

The original cantilevers would have been sufficient if they were not broken. The plastic collars hold adjusters which soon crack when exposed to the elements.

Cantilever Brakes

I looked for a modern alternative but after learning about brake pull ratios I decided the new levers would work better with Shimano ‘V-brakes’.

V-Brakes

These were fitted without issue and provide significantly better brake feel.

Brake Pivot
Fitted V-Brake