You are reading a single comment by @Jonny69 and its replies.
Click here to read the full conversation.
London Fixed Gear and Single-Speed is a community of predominantly fixed gear and single-speed cyclists in and around London, UK.
This site is supported almost exclusively by donations. Please consider donating a small amount regularly.
@Jonny69
The power supply on a class A amp is a funny thing. It runs flat-out all the time, so it’s not subjected to sudden current demands like on a class B or later amp. That means it doesn’t suffer from the associated voltage fluctuations that happen when the current demands vary, so the output and amplification don’t vary either. What that basically means, if I’ve understood it correctly, is it’s quite suited to a regulated voltage. The original paper suggested this as a design:
According to the notes, that should have made a circuit with variable voltage. I couldn’t get it to work even when loaded, and swapped the position of VR1 and the 2.7k resistor to make a voltage divider:
That worked just fine. I used a 3904 NPN and a 2N3035 in place of the MJ480, but the gain/outputs should be the same.
However, any ripple is replicated through the circuit, just at a higher current, which sort of defeats the object. I want the amp to maintain vintage components but I’m not all that bothered by the power supply. I just want it to hold its voltage steady at the correct voltage. I’d bought five MJ481s so I had the fifth one for the power supply, but actually a later dedicated voltage regulator would probably be better. In steps the LM350:
The advantage is that the LM350 makes a reference voltage internally and bases its output on that rather than as a percentage of the input voltage, so it doesn’t matter what goes in, the output voltage is always constant.
The basic circuit worked fine and I wanted to know what sort of heat it would kick out under load. So I repurposed my home made battery discharger that I’d cobbled together to test the hybrid battery in my Honda Insight because that had a 50W 15R resistor sat on a piece of aluminium extrusion.
Short answer was it kicked out a lot of heat and it’s going to need a lot of heatsink to keep it cool. No worries, I have heatsinks.
The final part of the puzzle was the PNP input transistor. I said before I wanted to use old components but I couldn’t find a 3906 PNP transistor in an old metal package, only the more modern plastic ones. I narrowed it down to a 2907 which looked close enough, but I wasn’t 100% confident. I spotted that the modern JLH circuit diagrams actually have a modern 2907 specified, so that gave me a bit more confidence that I was interpreting the data sheets correctly.