Schottky rectifier

[wdariusw]

Here i have few questions (about diodes) . I need to rectify small AC voltage ~800mV . And here is question - what diode bridge i should use. I know that schottkey here are best, but... Making simulations with TINA-TI software. I have bought few this diodes http://ezphysics.nchu.edu.tw/prophys/ael/File/Datasheet/hsms2822..pdf . And here is another that I simulated http://www.diodes.com/datasheets/ds23001.pdf .
Can anyone tell my, why the results are like this :



From datasheets i see that forward voltage
HSMS2822 min 0,34V max 0.5V
1N5817 min 0,45mV - max ....

Why after rectification 1N5817 gives better results? Thanks !

 

[KrisBlueNZ]

Are you sure your simulations are exactly the same? Have a good look at the V_OUT curve for the second simulation, the one using the HSMS282.



The little increments I've marked with arrows, and the overall gradual curve of the green line, show that the output capacitor is not fully charged until after many cycles of the input waveform. Compare that to the 1N5819 graph where the output reaches nearly its full voltage on the first half-cycle.

This indicates some kind of resistance in the circuit. It's possible that the HSMS2822 just has a high incremental resistance, i.e. a very gradual slope to its V_F vs. I_F graph, but check that there are no other sources of resistance in your simulation circuit. You could post that circuit too.

So compare the V_F vs. I_F graphs in the data sheets of the two diodes. Also, perhaps run a simulation on each diode type to graph the V_F vs. I_F characteristic of each model, to make sure the models are accurate.

 

[wdariusw]

Yes, i'm sure that the simulations are the same, same circuit copy-paste, only diodes are changed. Simply bridge rectifier with 100uF capacitor.. From datasheets graphs as i see 1N5817 forward voltage is little bit lower than HSMS2822. But only few mV.

 

[KrisBlueNZ]

It might be worth checking the V-I graphs of the simulation models then.

 

[hevans1944]

It would be most helpful if you would post the simulation parameters and the model used for each of these diodes. The static resistance of the HSMS-2822 is a relatively HUGE 6Ω and the typical dynamic resistance is an even larger 12Ω. This may help explain the simulated charging behavior when charging a 100 μF capacitor in your model.

IMHO, the late, great, analog integrated-circuit designer Robert Pease was more than justified in his lack of faith in SPICE models. Since you have the components in hand, wire them up and measure what you get with a real oscilloscope.

 

[KrisBlueNZ]

The static resistance of the HSMS-2822 is a relatively HUGE 6Ω and the typical dynamic resistance is an even larger 12Ω.

Hop, could you explain that terminology please? I think I understand the dynamic resistance (aka AC resistance, incremental resistance, slope resistance etc) and how it can be calculated at various points on the curve, but how can you define a meaningful "static" resistance for a diode?

 

[hevans1944]

The "static" resistance, or equivalent series resistance Rs, comes from the Agilent datasheet, linked by the OP, for their SPICE model. The model has Rs in series with Rd which is in parallel with the forward-biased junction capacitance. Rs in their SPICE model is specified to be 6 Ω. More information on SPICE modeling of these SMD packages at frequencies up to 6 GHz can be found on the attached application note. Scroll down near the bottom of the OP's link to see the SPICE model and the parameters used in the model.

Perhaps I was wrong in assuming that Agilent's use of Rs meant "static" instead of "series" but in any case it does not appear to depend on frequency. If I were to try to define a "static" resistance for a forward biased diode (of any kind) it would have to be defined as the limit of the ratio of forward voltage divided by forward current as the forward voltage approached zero. Pretty much the same as the definition for dynamic resistance, since this will be the slope of the V vs I curve at zero.

I can't imagine why anyone would deliberately package a diode die to create a 6 Ω fixed series resistance, but what the heck do I know about SMD die packaging? I apologize for the confusion in terminology.

 

[KrisBlueNZ]

I guess I should have looked at the data sheets at the start! The HSMS2822 is not a "normal" Schottky diode and is not intended to be used as a rectifier. The first clue is the description in the heading of the data sheet: "Surface Mount RF Schottky Barrier Diodes" (the clue is the "RF" in the title).

Then there's the comment that "Typical applications of these Schottky diodes are mixing, detecting, switching, sampling, clamping, and wave shaping." Rectification is strangely absent from that list! This is further confirmed by the I_F specification from the table, which gives a limit of 1A for "Forward Current (1 µs Pulse)". The equivalent specification for the 1N5819, I_FSM, is "Non-repetitive Peak Surge Current, 8.3 ms Single Half-Sine Wave" and it's 25A!

Also, the HSMS2822 is specified with a junction capacitance of less than 1 pF, whereas the 1N5819 has a typical junction capacitance of about 100 pF at -4V. And the HSMS2822 has a maximum rated reverse voltage of only 15V - not appropriate for rectifying a voltage derived from the mains, which will have spikes and surges on it that could easily exceed 15V. Even using a 1N5819 in that application is potentially risky and it would be wise to put a varistor or other clamp across the transformer secondary and capacitors across the diodes.

Finally, here are the V_F vs. I_F graphs for the two diodes with some important information highlighted in red!

 

[wdariusw]

My generator only have AC smaller than 1V, so 15V reverse voltage is not a bad thing. Thanks for the information, KrisBlueNZ ! HSMS2822 do not fit for rectification. Found this one (in my country 0.32 € each) http://www.nxp.com/documents/data_sheet/PMEG1020EA.pdf . I think with this i will get best results

 

[KrisBlueNZ]

http://www.nxp.com/documents/data_sheet/PMEG1020EA.pdf

Yes, very nice! I'll hang onto that data sheet. In fact I have an application in mind for it already. It's a bit of a shame that it's only rated for 10V reverse voltage, but you can't have everything you want.

 

[hevans1944]

The main (and maybe only) advantage of Schottky diodes, AFAIK, is their really fast recovery when reverse biased. Since there is no PN junction, per se, there are no charges (minority carriers in a PN junction) to clear out when transitioning from forward conduction to reverse blocking, leaving only junction capacitance to slow things down.

You need some very fast switching waveforms to appreciate the benefit, but.this is why Schottky diodes are often specified for magnetically-stored-energy converters like buck-boost circuits, or RF mixers and detectors, as mentioned above by @KrisBlueNZ., You pays your money and you makes your choices.

There are alternatives, such as synchronous rectification involving MOSFET switches, but at the expense of increased circuit complexity. If a Schottky diode will do the job, they are no longer prohibitively expensive, thanks to modern semiconductor processing technology.

There is an advantage to the relatively low forward-bias voltage of Schottky diodes in power-switching and rectification applications, if you can tolerate the pitiful reverse breakdown voltage.

 

[wdariusw]

Very bad thing here... For now it doesn't matter, but for the future.. The device should be as small as possible, and here DIP component like 100uF or more capacitors are very big... :/

 

[hevans1944]

Very bad thing here... For now it doesn't matter, but for the future.. The device should be as small as possible, and here DIP component like 100uF or more capacitors are very big... :/

I am pretty sure a low-voltage 100 μF SMD capacitor is available somewhere. Try this link for starters. For the Schottky diode, try a 1N5819HW. See attached data sheet

 

[(*steve*)]

Silicon Carbide Schottky diodes are very interesting (having high voltage ratings, low leakage, etc), but are not the best for the use discussed here. If you're interested in amazing devices, they're worth a look.

However for this use, I would look at the 19Q015. The forward voltage is *very* low. It does have significant leakage and this may be an issue.

 

[hevans1944]

Silicon Carbide Schottky diodes are very interesting (having high voltage ratings, low leakage, etc), but are not the best for the use discussed here. ...

Steve, I didn't see anything on the datasheet I uploaded that indicated the 1N5819HW is a silicon carbide Schottky diode. Are you sure that it is?

The maximum operating temperature of 125 C is way below the typical maximum temperature for silicon carbide devices, and the plastic (rather than ceramic) package is also an indication that the device is not suitable for high temperatures. OTOH, it has a high reverse voltage rating (40 V), a fairly low reverse leakage (75 μA maximum) at 6 V and room temperature (25 C), and a fairly low forward voltage of 0.45 V maximum at 1 A forward current, which is the maximum rated average current.

Why is the 1N5819HW not suitable for the OP's application of full-wave rectification of ~800 mV into a 100 μF filter capacitor? What am I missing here?

 

[KrisBlueNZ]

I don't think Steve meant to imply that the 1N5819HW is a silicon carbide diode. He was just mentioning silicon carbide diodes as another topic.

 

[hevans1944]

I don't think Steve meant to imply that the 1N5819HW is a silicon carbide diode. He was just mentioning silicon carbide diodes as another topic.

Oops. I sure hope the OP comes back and tells us what they are trying to accomplish. Apologies to Steve.

 

[(*steve*)]

Yeah, silicon carbide diodes are just cool devices. I looked up their specs just in case any have a low Vf, but they don't. :-(

The diode I referred to has a Vf of between 0.1V and 0.2V at 1A. It may be worth a look at. At moderate currents like this (and at low voltages) it may be a better option. At low currents it is not as the reverse leakage is 1mA at 25C, and 100mA(!!!!) at 125C.

 

[hevans1944]

Well, we still haven't explored synchronous rectification using MOSFET switches. I know such circuit exist, but I have no knowledge of the details. I suspect their main use might be for RF mixers and detectors rather than power rectification. Maybe someone will jump in here to enlighten us!

 

[KrisBlueNZ]

AFAIK synchronous rectification with MOSFETs is used in switching supplies to improve efficiency by avoiding the forward voltage drop of the catch diode or rectifier diode. I guess the phrase might be used in other contexts as well.