Zener tolerances will do that, and dynamic impedance is not zero. You just have to trim these things, you get what you get and prepare to adjust. 13 diodes is a lot, I aim for about 5. I get far better result dropping the current (and dissipation) down and using a pass transistor. Support circuitry required is minimal.
I've used some like this before:
https://www.mouser.com/ProductDetail/Aavid/501100B00000G?qs=YTENqGpblNY5ZvYHsvH8xg==
Just put some epoxy on each diode and span several diodes as far as the heatsink length will span.
https://www.mouser.com/ProductDetail/Aavid/501100B00000G?qs=YTENqGpblNY5ZvYHsvH8xg==
Just put some epoxy on each diode and span several diodes as far as the heatsink length will span.
I'd suggest you solder each zener part 'directly' to the copper sections, so that the zener junction heat transfers directly to the copper 'heatsink', rather than having to go via a long zener wire lead. Then stand the board vertical, so that air flow over the copper sections is like a thermal chimney. Then solder fins to each copper section so that the heat more easily convects into the air. You can also use a small fan to further assist.
There are many diy ways to improve thermal transfer away from a parts junction. The datasheet often provides some indication of the thermal resistance presented by different mounting and cooling setups.
There are many diy ways to improve thermal transfer away from a parts junction. The datasheet often provides some indication of the thermal resistance presented by different mounting and cooling setups.
Hey. So You also sugest an effective heat transfer between plastic case and a heatsink is possible please ? I don't have quick access to the heatsink nerdorama suggested above but I can align the diodes between two copper stripes and make firm contact with screws, so the diodes will be "sandwiched" between copper stripes with ,eventually some thermal paste in between. If the heat transfer from plastic case is effective it worth...Carefully mounting will be necessary when heatsink is shared as now any faulty contact between first and last diode will carry full 400V potential !!!
Last edited:
No, my suggestion was to solder each zener such that it has minimum lead length between the package and where it starts to be soldered to each copper pad. The zener datasheet may show a table of thermal resistance to ambient for certain lead lengths. The shorter the lead length, the lower the thermal resistance, and the lower the junction temperature for a given copper pad temperature. The copper pad then becomes the heatsink, and so soldering copper flaps to each copper pad improves the thermal transfer to air from the copper pad.
I agree that adding a copper cover to the package of each zener is prone to bad accidents, and imho is not a good thermal path due to the plastic package of the zener.
I agree that adding a copper cover to the package of each zener is prone to bad accidents, and imho is not a good thermal path due to the plastic package of the zener.
Thanks. Yes I know what You mean and I was thinking the same but wasn't sure how to 'balance' this problem: The Zener leads have own capacity to dissipate from its leads surface and its own leads thermal conductivity. On the other side, between zener leads and copper pads there is the solder conection which act with its higher thermal resistance (God knows how well it do). This was the reason I preferred to dissipate as much from length of Zener leads. Basic my dilemma was : better dissipation vs better higher thermal conductivity.
I will try with shorter zener leads but using a screw bolted contact between Zener and copper leads later.
Now, the ideea to dissipate from plastic case is not completely worthless as I also believed (apologies to nerdorama).I don't know how much efficient is ( don't have a temperature probe) but for sure is effective- it dissipate a lot of heat and from voltage measurement it shows a clearly difference between with and without case heatsinking. I.ll give a shot late please. Thanks.
I will try with shorter zener leads but using a screw bolted contact between Zener and copper leads later.
Now, the ideea to dissipate from plastic case is not completely worthless as I also believed (apologies to nerdorama).I don't know how much efficient is ( don't have a temperature probe) but for sure is effective- it dissipate a lot of heat and from voltage measurement it shows a clearly difference between with and without case heatsinking. I.ll give a shot late please. Thanks.
Last edited:
A 25W transistor heat sink won’t be any smaller, but it will be more convenient to work with. If you use a fullpack encapsulated TO-220 and heat shrink on the leads you wont have to worry about high voltages and user’s fingers. Can’t mount that zener monstrosity anywhere it might come in contact with people or pets (Which means INSIDE the enclosure, where air circulation is limited).
You will also only be pulling 65 mA peak, not 65 mA all the time.
You will also only be pulling 65 mA peak, not 65 mA all the time.
