I have been thinking for a long time (too long) what to do about the 24V (µPC7824H) regulator in my B-2102 which simply gets too hot in my opinion. The heatsink is too hot to touch after about 15 minutes, and the radiated heat threatens to 'cook' the Left Channel driver board. I checked the input voltage was within spec limits and it was, I also checked the current through the regulator - well within limits.
I changed the regulator for a new one, same problem. A way to reduce the heat is to reduce the input voltage, but this could cause regulation issues, and there is very limited space to mount a larger heatsink. I thought about mounting the regulator remotely from the PSU board on a bigger heatsink, and even thought of putting it on the left channel main heatsink. Both of these ideas have drawbacks too great for further consideration.
I had a careful look at the currently available PCB mount heatsinks looking for a bigger (taller) replacement - tough job. Eventually I found one which is 25% bigger and is also black anodised, the original is plain aluminium. I also bought another one (bigger, which only just fits), in case that wasn't enough. So I was thinking, "what if neither of these results in an improvement"? So plan "B" would be the inclusion of a rectifier diode in series with the feed to the regulator to cut the input voltage by 0.7V - this was about as much I felt I should reduce it by.
I am pleased to report that the combination of the new heatsink and the series rectifier diode have provided a most satisfactory solution to the heat problem. The heatsink is warm, but never too hot to touch, and the 24V regulation is fine. (input voltage is now 28.7V, regulated output voltage 23.7V).
LED's for the power meter backlight?
In parallel with this I was trying to implement the replacement of the 6 incandescent power meter display backlight bulbs with LED's - but after trying 3 or 4 different ones, and never getting an even spread of light, I bought some replacement standard bulbs and I am happy with the end result.
So, I can now get on with other stuff, knowing the B-2102 won't cook itself.
I changed the regulator for a new one, same problem. A way to reduce the heat is to reduce the input voltage, but this could cause regulation issues, and there is very limited space to mount a larger heatsink. I thought about mounting the regulator remotely from the PSU board on a bigger heatsink, and even thought of putting it on the left channel main heatsink. Both of these ideas have drawbacks too great for further consideration.
I had a careful look at the currently available PCB mount heatsinks looking for a bigger (taller) replacement - tough job. Eventually I found one which is 25% bigger and is also black anodised, the original is plain aluminium. I also bought another one (bigger, which only just fits), in case that wasn't enough. So I was thinking, "what if neither of these results in an improvement"? So plan "B" would be the inclusion of a rectifier diode in series with the feed to the regulator to cut the input voltage by 0.7V - this was about as much I felt I should reduce it by.
I am pleased to report that the combination of the new heatsink and the series rectifier diode have provided a most satisfactory solution to the heat problem. The heatsink is warm, but never too hot to touch, and the 24V regulation is fine. (input voltage is now 28.7V, regulated output voltage 23.7V).
LED's for the power meter backlight?
In parallel with this I was trying to implement the replacement of the 6 incandescent power meter display backlight bulbs with LED's - but after trying 3 or 4 different ones, and never getting an even spread of light, I bought some replacement standard bulbs and I am happy with the end result.
So, I can now get on with other stuff, knowing the B-2102 won't cook itself.
Last edited:
