Peak repetitive forward current??

[Nixie]

Looking at http://www.fcisemi-mag.com/products/...ltage/hv05.pdf
I'm using the HV05-15 in a capacitor-input filtered linear supply, and I want to know if 600 mA DC is possible from the supply using these rectifiers in a bridge of four. Their rating is 550 mA average, which is fine, but due to the capacitor-input filter, the peak current is a few amps. The datasheet doesn't have any repetitive forward current rating, only a graph of non-repetitive which goes up to a hundred cycles. What about the hundreds of million cycles in the lifetime of an average amplifier? I can't see that I can extrapolate that from the graph.

[peranders]

600 mA out with 550 mA diodes not what I would choose. 1 A or more would be my choice.

[Nixie]

That's per diode. I'm using four in a bridge, and the 600 mA is the total draw from the power supply. Clearly, the average current rating is not exceeded, and that was not my question. Also, I don't have any 7+ kV diodes that will do 1 A on hand.

[AndrewT]

Hi,
your bridge will be using two diodes in series.
The series pair then hand over to the other pair, getting a short rest on each alternate half waveform.

Peak forward and repetitive is down to try it and see, but at high kV the answers could be spectacular.

Try the manufacturer.

[peranders]

Originally posted by Nixie
That's per diode. I'm using four in a bridge, and the 600 mA is the total draw from the power supply. Clearly, the average current rating is not exceeded, and that was not my question. Also, I don't have any 7+ kV diodes that will do 1 A on hand.

Still, if you have 600 mA DC out the diodes (each one) should be more than 600 mA, preferable 1 A or more.

[Nordic]

Sorry a related question, how do I know what a diode's forward voltage drop is, I'm looking for a replacemnet to OA91 which has a drop of about 200mV..

[AndrewT]

Hi,
having a look at the datasheet, I am inclined to guess that temperature sets the limit for these diodes.
The Vf @ 8V when passing 550mA gives a dissipation of 4.4W.
18C/W gives a junction temp of 114degC when ambient is 35degC.
The temp derating starts at 60degC and this equates to 140DegC junction temp (just above max storage temp).

If any of that conjecture is close to truth then how about measuring operating temperature of the diodes? Safely?

If they are hot then you are using a high average dissipation.

The repetitive forward graph has gone asymtotic @ 5.5A by the time it reaches 100cycles.
Again guessing, if you can guarantee les than 50% of this as a peak repetitive value, then you might be safe in the long term.

Note, that capacitor input has to be derated to 80%.

[jneutron]

Originally posted by Nixie
Looking at http://www.fcisemi-mag.com/products/...ltage/hv05.pdf
I'm using the HV05-15 in a capacitor-input filtered linear supply, and I want to know if 600 mA DC is possible from the supply using these rectifiers in a bridge of four. Their rating is 550 mA average, which is fine, but due to the capacitor-input filter, the peak current is a few amps. The datasheet doesn't have any repetitive forward current rating, only a graph of non-repetitive which goes up to a hundred cycles. What about the hundreds of million cycles in the lifetime of an average amplifier? I can't see that I can extrapolate that from the graph.

Be very careful extrapolating from manufacturer supplied graphs. Sometimes they are rather creative, many times they are just hand drawn to look good.

The rep surge graph asymptote being one.

Also, the numerical data has 200 amp surge, the graph has 20 amps at one pulse.

Then the data has 100 ma rep surge, but on a 550 ma device??

Either ask the manu, or just try it.

I worry about the inrush. Put a resistor in series to try to lower the peak to less than say, 5 amps. Can't tell ya what R to use, though.

If you use those, use big guage wire to the device to help draw heat out, they rely on the axial conduction path for the majority of the heat transfer.

Good luck

John

[richie00boy]

Bad design IMO. Due to power factor, I always work on the AC current being 1.62 times the DC, so I would be using diodes rated for 972mA at least. Even ignoring power factor, using 550mA diodes for 600mA load is just not a good idea.