10-04-2012, 04:25 PM
Looks good - nice and simple 
Does that 470 ohm resistor provide enough current? Worst-case, a 2n3055 might have an HFE of only 20, implying that 50mA of base current would be needed at 1 amp. The 470 ohm resistor might be running at around 14mA, depending on the unregulated input. Something like 120 or 150 ohms might be better, but of course would warm up a bit. Although you're probably not suffering any problems with the circuit you've built, it's probably worth mentioning this as a potential issue for anyone that builds their own version in the future...
The 2n2219 is one of my favourite TO5 transistors. No idea how widely available they are these days, but it's probably worth saying (again, for the benefit of potential future constructors) that almost any NPN transistor will work there. I bet the humble BC108 would work well.
In a similar vein (for other constructors), it's probably worth pointing out that the diodes used in the regulator and crowbar circuit aren't critical, and any silicon diode will probably work. I like the 1n4148 as they are smaller and easier to handle - the 1n540x range are rather "chunky" and their leads fit tagstrip but not 0.1" Veroboard...
Should anyone want to improve the ripple performance of this then the 470 ohm resistor can be split in half, with a capacitor decoupling the mid-point. Even better; add another transistor to make the 2n3055 a Darlington pair, then larger values of resistance can be used - probably 4K7 or more - which will reduce the size of capacitor necessary. Also, a capacitor between the base of the 2n2219 and the output rail will reduce the ripple by increasing the AC feedback.
I've attached the schematic of the Hacker VP408 external power supply - you'll note a lot of commonality
Again, good work
Mark
Does that 470 ohm resistor provide enough current? Worst-case, a 2n3055 might have an HFE of only 20, implying that 50mA of base current would be needed at 1 amp. The 470 ohm resistor might be running at around 14mA, depending on the unregulated input. Something like 120 or 150 ohms might be better, but of course would warm up a bit. Although you're probably not suffering any problems with the circuit you've built, it's probably worth mentioning this as a potential issue for anyone that builds their own version in the future...
The 2n2219 is one of my favourite TO5 transistors. No idea how widely available they are these days, but it's probably worth saying (again, for the benefit of potential future constructors) that almost any NPN transistor will work there. I bet the humble BC108 would work well.
In a similar vein (for other constructors), it's probably worth pointing out that the diodes used in the regulator and crowbar circuit aren't critical, and any silicon diode will probably work. I like the 1n4148 as they are smaller and easier to handle - the 1n540x range are rather "chunky" and their leads fit tagstrip but not 0.1" Veroboard...
Should anyone want to improve the ripple performance of this then the 470 ohm resistor can be split in half, with a capacitor decoupling the mid-point. Even better; add another transistor to make the 2n3055 a Darlington pair, then larger values of resistance can be used - probably 4K7 or more - which will reduce the size of capacitor necessary. Also, a capacitor between the base of the 2n2219 and the output rail will reduce the ripple by increasing the AC feedback.
I've attached the schematic of the Hacker VP408 external power supply - you'll note a lot of commonality
Again, good work
Mark
