Rambly and slightly botched LED driver repair

I got a couple of salvaged substation lights and they both had the classic strobing issue. I thought it would be interesting to see if we can fix them, and I'll probably install one of them in my garage workshop, since they kinda suit that application. It's also worth mentioning that these lights are old enough that they aren't grilling the LEDs. They have a very large heat spreading area inside. My initial thoughts were that it was probably just old dry capacitors, but changing them didn't resolve the issue. It turned out to be a classic culprit - the bootstrap diode that powers the primary side circuitry. They had used a GS1G diode, and replacing it with a fast recovery UF4003 diode solved the issue. I only changed the diode in the other strobing power supply and it fixed that one too. It makes me wonder if the bootstrap capacitor used to power the primary side circuitry is maybe too low a value, and possibly doesn't allow the secondary side capacitors to charge fully before running out of juice and rebooting the supply. If these give any more issues I may up its value to 33uF, which will give a slightly longer initial start time, but allow the secondary side to stabilise. The crustiness of the solder was annoying. In hindsight I should have paused and bumped the soldering iron temperature up a bit higher. Here's how a typical bootstrap circuit works in a power supply. The drive circuitry needs power to operate, and is effectively powered by its own winding on the transformer, which also allows the circuitry to monitor what's happening on the isolated side. At power up, a string of resistors trickles a very small current into the power supply's own little capacitor. When it reaches a high enough voltage the power supply starts up, and if everything is working the control circuit's capacitor will be kept topped up by the feedback winding and a diode. If the secondary is short circuited there will not be current coupled back to the control circuit's capacitor and it will discharge to the point the circuitry locks off and the starting process starts again. That can manifest as a seemingly dead power supply making a quiet clicking noise each time it tries to start. If the secondary is open circuit the feedback winding has a much higher voltage as the secondary side smoothing capacitors charge to their open circuit voltage. That is usually sensed by a two resistor divider that signals to the chip that there is an issue and it causes the chip to shut down and restart. That manifests as defective and flickery LED lights that pulse and dim. Other things that can cause continuous rebooting are failed capacitors that don't absorb the charge due to their higher resistance, and let the voltage float too high. That manifests as lights strobing at full brightness. In this case the fault was with the feedback circuitry itself. The diode tested OK with a simple diode test, but may not have been reacting fast enough to charge the control circuitry's capacitor. Generally speaking it's faster to just buy a new power supply, but in some instances it makes sense to repair existing units. Especially if they are better quality or hard to find units. If you enjoy my videos, you can help support the channel with a dollar or two for coffee, cookies and gadgets, and keep it independent from the quirks of the YouTube algorithm by supporting me on Patreon. This also lets me link to my content on other platforms if YouTube uses the ban-hammer (again!) I release content for critique and feedback on Patreon as soon as it has been made.   / bigclive   Alternatively, for a single contribution you can use PayPal:- https://www.paypal.com/paypalme/bigclive #ElectronicsCreators