| Electronics > Beginners |
| Questions regarding diodes in parallel |
| (1/3) > >> |
| GadgetBoy:
Ok, so I'm working on a Cockroft-Walton ioniser, and I know there some problems with efficiency on NA voltage. So I've been thinking of ways to build a more efficient ioniser, and besides adding more stages (I have a 15 stage one currently, and I'm thinking of adding 5-10 more), I've been thinking about the effects of low-esr caps and changing diode types. So, I was thinking about schottky diodes, and their fast response and low forward voltage, but they're generally rated at no more than 40v. If I put multiple schottky diodes in parallel, will their voltage rating add (i.e. put three 40v schottky diodes in parallel for 120v rating)? |
| Zero999:
No, putting diodes in parallel doesn't cause their voltage ratings to add, but remain the same, since the voltage across parallel components is always equal: Kirchoff's voltage law. Connecting diodes in series increases their voltage rating, but it does the same for the voltage drop, so doesn't gain anything efficiency wise. For a relatively high voltage Schottky diode, try the RB168L150. http://www.mouser.com/ds/2/348/rb168l150-774152.pdf There's also the issue of voltage and current sharing, with the above proposals, because the components aren't perfectly identical. To boost efficiency, cut down on the number of stages and use a transformer. |
| TimFox:
When connecting diodes in series, the problem is with the reverse voltage, since the leakage currents are not identical and the voltage will not divide equally across the diodes. Adding high-value shunt resistors across each diode is the usual solution. When connecting diodes in parallel, the problem is with the forward current, since the forward voltages are not identical and the current does not split equally between the diodes. Adding low-value resistors in series with each diode is the usual solution. Both of these solutions are usually a bad idea, compared with buying a single diode with proper ratings. |
| Zero999:
--- Quote from: TimFox on June 07, 2018, 11:37:09 am ---When connecting diodes in series, the problem is with the reverse voltage, since the leakage currents are not identical and the voltage will not divide equally across the diodes. Adding high-value shunt resistors across each diode is the usual solution. When connecting diodes in parallel, the problem is with the forward current, since the forward voltages are not identical and the current does not split equally between the diodes. Adding low-value resistors in series with each diode is the usual solution. Both of these solutions are usually a bad idea, compared with buying a single diode with proper ratings. --- End quote --- Yes that's true. Another solution is a two diode package, which won't have the same issues with voltage and current sharing, since the two diodes will be on the same piece of silicon and therefore well matched. |
| David Hess:
--- Quote from: TimFox on June 07, 2018, 11:37:09 am ---When connecting diodes in series, the problem is with the reverse voltage, since the leakage currents are not identical and the voltage will not divide equally across the diodes. Adding high-value shunt resistors across each diode is the usual solution. --- End quote --- Some diodes are avalanche rated so this is not a problem for them and they may be used in series without equalization resistors. Usually they are constructed to prevent surface breakdown and to have a uniform junction to prevent partial breakdown which would produce hot spots. High voltage diode stacks take advantage of this but there is no reason separate avalanche rated diodes cannot be used the same way. --- Quote ---When connecting diodes in parallel, the problem is with the forward current, since the forward voltages are not identical and the current does not split equally between the diodes. Adding low-value resistors in series with each diode is the usual solution. --- End quote --- The usual rules about thermal runaway apply so if the forward voltages are matched well enough and the thermal resistance between the diodes is low enough, no series resistors are required. Sometimes I see switching power supply designs which take advantage of this by using both diodes in a dual diode package in parallel. |
| Navigation |
| Message Index |
| Next page |