It seems that I am the only one here that thinks there is no need for voltage dividing resistors. Yes, one diode will bear more voltage than the other, if the reverse capacitor or the reverse leakage current is not the same. But who cares? when one diode bear near-break-down voltage, the other one will automatically take the rest voltage, provent the first from actually breaking down. External capacitors may be useful to take away spikes, but external resistors have no use.
Vax,
Would you want the diode to breakdown? In this case, its all acedemic 'cause he has suitable diodes. In a real case, the issue becomes finding a resistor that can take the voltage... again, stacking becomes the norm.
Mrshow,
Its called "The Rectifier Handbook" (I think) Motorola... 1990? I have it here somewhere.
lndm,
1200 Volts with 2000 PRRV diode is fine... relax. MOV on the pri is fine, but remember... they are consumable... they wear out and go poof when they do. So don't size them too close to the line voltage, or they will wear out early.
Would you want the diode to breakdown? In this case, its all acedemic 'cause he has suitable diodes. In a real case, the issue becomes finding a resistor that can take the voltage... again, stacking becomes the norm.
Mrshow,
Its called "The Rectifier Handbook" (I think) Motorola... 1990? I have it here somewhere.
lndm,
1200 Volts with 2000 PRRV diode is fine... relax. MOV on the pri is fine, but remember... they are consumable... they wear out and go poof when they do. So don't size them too close to the line voltage, or they will wear out early.
Look here at the top of the second column.
Just a little bit about reverse voltage sharing and snubbers.
http://www.ixys.com/t03232ka.pdf#search='diodes%20in%20series'
Just a little bit about reverse voltage sharing and snubbers.
http://www.ixys.com/t03232ka.pdf#search='diodes%20in%20series'
Poobah,
But these diode will not "break down". This term sometimes fools people. A simple question here for you guys who are still suspecious, with some make_up parameters,
Q. Two diodes in series, A and B. Diode A has reverse breakdown voltage of 1001V, leakage 10uA. Diode B has reverse breakdown voltage of 1002V, leakage 20uA. Now apply 2000V reverse voltage. What is the total reverse leakage current? What is the voltage on A and B, respectively? What is the power on A and B, respectively? Will A or B, or both, break down?
Now my stand is more clear.
But these diode will not "break down". This term sometimes fools people. A simple question here for you guys who are still suspecious, with some make_up parameters,
Q. Two diodes in series, A and B. Diode A has reverse breakdown voltage of 1001V, leakage 10uA. Diode B has reverse breakdown voltage of 1002V, leakage 20uA. Now apply 2000V reverse voltage. What is the total reverse leakage current? What is the voltage on A and B, respectively? What is the power on A and B, respectively? Will A or B, or both, break down?
Now my stand is more clear.
,,,I am getting it. If 1 diode reverse breakdowns first, it will zener at low current pulling up the 2nd diode up. Thanks for Poobah regarding the IXYS paper. I see where resistors are used for static voltage dividing, and caps for dynamic voltage dividing. It doesn't seem like the "snubbing" capacitances have to be large either. And cap tolerances are certainly close enough that each cap/diode will recieve a similar voltage drop.
I guess it won't break either, I don't know. Stray capacitances, short duration transients, derating with repeated PIV, and stuff like that sometimes worry me. I just hate when HV stuff blows up. I repaired an AR D-150. It had arcced through one of it's filters. It sounded like Frankensteins lab. I put my left hand behind my back and the yanked the power cord with my right. Sorry for the tangent, but I respect HV and try do make it break as little as possible. I think if I was required to stack diodes and hand the PCB space, I'd design in a spot for a cap and a resistor. ....just in case
I guess it won't break either, I don't know. Stray capacitances, short duration transients, derating with repeated PIV, and stuff like that sometimes worry me. I just hate when HV stuff blows up. I repaired an AR D-150. It had arcced through one of it's filters. It sounded like Frankensteins lab. I put my left hand behind my back and the yanked the power cord with my right. Sorry for the tangent, but I respect HV and try do make it break as little as possible. I think if I was required to stack diodes and hand the PCB space, I'd design in a spot for a cap and a resistor. ....just in case
Vax,
R, C, and RC snubbers are common practice when stacking diodes to buy PRRV. I didn't invent the practice (although I would have), I merely attempted to explain for those that were curious.
Diodes are often stacked within single discrete packages for the same reason. In these cases, many device parameters are matched to insure voltage sharing. Some of these are: reverse recovery charge, reverse capacitance and leakage.
But... I can't get an "A" on your quiz.
First, we don't what the forward current was prior to reverse bias.
Second, we don't know the dv/dt of the reverse voltage... you will recall I mentioned that would affect which whether the leakage or the capacitance mechanism would domnate the voltage sharing issue.
Third, we don't know the individual reverse capacitance of the diodes.
Fourth, we don't the individual reverse recovery charge of the diodes. This may or may not apply as we don't know the value of the previous forward current.
New quiz:
1) Why will the diodes breakdown or avalanche?
2) Where will the avalanche current come from?
3) Why must avalanche be avoided?
4) And laterally why are "bleeder" resistors placed across large 'lytics stacked in series in linear supplies?
R, C, and RC snubbers are common practice when stacking diodes to buy PRRV. I didn't invent the practice (although I would have), I merely attempted to explain for those that were curious.
Diodes are often stacked within single discrete packages for the same reason. In these cases, many device parameters are matched to insure voltage sharing. Some of these are: reverse recovery charge, reverse capacitance and leakage.
But... I can't get an "A" on your quiz.
First, we don't what the forward current was prior to reverse bias.
Second, we don't know the dv/dt of the reverse voltage... you will recall I mentioned that would affect which whether the leakage or the capacitance mechanism would domnate the voltage sharing issue.
Third, we don't know the individual reverse capacitance of the diodes.
Fourth, we don't the individual reverse recovery charge of the diodes. This may or may not apply as we don't know the value of the previous forward current.
New quiz:
1) Why will the diodes breakdown or avalanche?
2) Where will the avalanche current come from?
3) Why must avalanche be avoided?
4) And laterally why are "bleeder" resistors placed across large 'lytics stacked in series in linear supplies?
Poobah said:
:New quiz:
:1) Why will the diodes breakdown or avalanche?
Because reverse voltage is too high.
:2) Where will the avalanche current come from?
Supplied by the source who supply the reverse voltage.
:3) Why must avalanche be avoided?
There is no "must". We use avalanche in voltage regulators. Avalanche will not damage the diode if you control the reverse current.
:4) And laterally why are "bleeder" resistors placed across large 'lytics stacked in series in linear supplies?
This is because when one capacitor breaks down, all voltage will be placed on the other. This is different from the case of two diodes.
:New quiz:
:1) Why will the diodes breakdown or avalanche?
Because reverse voltage is too high.
:2) Where will the avalanche current come from?
Supplied by the source who supply the reverse voltage.
:3) Why must avalanche be avoided?
There is no "must". We use avalanche in voltage regulators. Avalanche will not damage the diode if you control the reverse current.
:4) And laterally why are "bleeder" resistors placed across large 'lytics stacked in series in linear supplies?
This is because when one capacitor breaks down, all voltage will be placed on the other. This is different from the case of two diodes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.