Dunlap-Clarke Dreadnought 1000
This thing is a monster!
Front view of the amp. This thing wieghs around 80 lbs, and is called the 1000 because it will dump 1000 watts into 2 ohm loads. What was wrong? It blew fuses at power-up. The customer stated that a repair shop had looked at it and given up. He was, however, diligent and had accumulated a lot of information from online sources. What we did not have was a complete schematic.
The output circuit is fully complementary symmetry, and is most like a servo amplifier. The input goes to an op-amp and the feedback to that op-amp comes from the speaker output. The simplified schematic I had showed that much. The speakers are DC coupled to the output stage through fuses. Those fuses were not blown.
Of course, with blown fuses the power supply was checked first. The customer stated that the filter capacitors (huge blue cans) had leaked in the past and been replaced. He also mentioned that the output devices had been replaced. The power supply was disconnected from the amplifier circuit and brought up on a variac. No problems. So, the driver boards were removed, and each bank of output devices was checked for resistance to the positive speaker terminal. If output devices fail short, this test will show that short. In fact, the left side of the amp had shorts and an odd mix of output devices. Upon removal, some originals present were failed short. Some incorrect replacement devices were failed open. And they were mixed, which is bad practice.
According to our information, the original output devices (2N5631/2N6031) were selected by Dunlap Clarke for this amplifier based on a forward breakdown voltage of 180V, from devices specified at 160V. This is a vital clue, because those incorrect replacements were 150V devices.
At this point the repair moved into the realm of restoration. How to get the best results, in the least time for the least cost? Knowing a little about why prior replacements had failed, all new output devices were selected. ON Semiconductors’ MJ21193 and MJ21194. These are 250V, 250W, 16A devices. This is far beyond the originals and 10 of each came in just under $100. New mica insulators, replacements for 8 electrolytic capacitors and 4 39K 1W resistors on the driver boards, and a bag of 8A fuses were also ordered.
On each driver board were the 4 electrolytic capacitors, which looked OK but were old, and 2 burned 39K resistors. These were replaced as routine. Then the output devices were replaced. This is where I found that only the right side output devices had been successfully replaced in the past with some Motorola devices. All were replaced with the newly selected devices. Each new device was installed with a new mica insulator and fresh silicone grease. In fact, 16 devices are output and 4 devices of the same type are used as drivers, which are mounted separately directly to the chassis under the heat sinks.
The above, a view from one side, shows everything in place.
The result? I only have an 8 ohm test load, so I tested each channel with that. Specification is 250 Watts into 8 ohms, and I measured approximately that on each channel. Nailed it! So I have to ask: How can an established professional repair shop fail at this? They’re the pros, right? This challenges that assumption!