Author Topic: STmicroelectronics ultra-low Vf Schottky  (Read 3247 times)

0 Members and 1 Guest are viewing this topic.

Offline MustardManTopic starter

  • Contributor
  • Posts: 44
  • Country: au
STmicroelectronics ultra-low Vf Schottky
« on: August 31, 2020, 06:14:46 am »
I am very intetrested in a part that STmicroelectronics manufactured, and I don't know if they still do, or if there is a replacement.

The part is a ulta-low Vf schottky diode, with a quoted (datasheet) Vf of 120mV at 10A. I have not come across a diode with anywhere near this Vf (at any reasonable current, 1A being reasonable).

Most suppliers list this as an obsolete part, but Digikey lists it as still active (zero stock) with a minimum of 3000. I want less than a tenth of that number! And even if I order, they may come back and say no-dice.

It would appear that a TI part (the SM74611) has a lower Vf (claimed 26mV at 8A), BUT this is not a real diode, it is a so-called "smart" diode. Digikey has that part if you wish to look up how it works. Needless to say that although it is rather unique, it does not have the characteristics of a genuine diode.

Trying to contact ST has been a nightmare, and I still haven't managed to, to ask if there is an equivilant.

Is anyone aware of a replacement, or of another ultra-low Vf schottky?

Cheers, MM
 

Offline exmadscientist

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
  • Technically A Professional
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #1 on: August 31, 2020, 07:01:28 am »
Am I to assume you're talking about the SPV1001N30, or one of its close relatives? (You didn't give a part number.)

Because those aren't diodes either, they're MOSFETs with built-in driving circuitry, like the TI part.

In general the physics governing diode barriers does not permit the existence of such a thing, at least not as a true diode. For a Schottky barrier, the lower the forward voltage, the more the barrier leaks... until the barrier reaches zero volts, at which point the thing becomes perfectly leaky, because you've got a wire. The takeaway here is that forward voltage is fundamentally linked to diode characteristics. Advances in diode construction are about better approaching ideal performance, and figuring out how to trade off things we don't care about against things we do care about.
 
The following users thanked this post: thm_w

Offline jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #2 on: August 31, 2020, 07:43:45 am »
I used the SM74611 to get myself out of a jam a while ago. It starts off with about 0.6V drop across MOSFET body diode, charges up an internal capacitor and then turns on the MOSFET as an ideal diode below 100mV. The capacitor runs down and the cycle repeats.

Average power loss is good but the fluctuations in forward voltage could be an issue for some applications.

I was curious, so I added a Schottky in parallel. The reduced voltage drop stopped the internal circuit from starting up, so it actually made the losses worse :D

If you need better behaviour than a standard Schottky, you can drop an ideal diode controller and MOSFET.
 
The following users thanked this post: thm_w, BrianHG

Offline jbay

  • Newbie
  • Posts: 7
  • Country: ca
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #3 on: September 01, 2020, 01:54:25 am »
Ideal diodes are usually good for slow or DC applications or perhaps for 60 Hz rectification, but generally are too slow for application in a lot of faster applications where you might want to use them, like for reducing losses in a DC-DC converter. So I can see why the OP would prefer a real diode.

Good luck!
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #4 on: September 01, 2020, 02:11:41 am »
It would appear that a TI part (the SM74611) has a lower Vf (claimed 26mV at 8A), BUT this is not a real diode, it is a so-called "smart" diode. Digikey has that part if you wish to look up how it works. Needless to say that although it is rather unique, it does not have the characteristics of a genuine diode.

That TI part is designed exclusively for solar cells.  When switched on, it actually needs to open that switch for a small amount of time to charge an internal capacitor to drive it's internal mosfet's gate.  It is designed to protect solar cells in the shade & pipe through the most power with the least loss.

Yes, that so called diode literally opens for around 3.5us 32 thousand times a second.
And you cannot cap smooth that thing out, it needs that open time's voltage drop to build a negative charge to drive it's mosfet gate.
 

Offline MustardManTopic starter

  • Contributor
  • Posts: 44
  • Country: au
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #5 on: September 01, 2020, 06:02:38 am »
The part number of the device with the claimed Vf of 140mV is SPV1512. From looking at many schottky datasheets over the last couple of weeks, it would appear that for nearly all of them the reverse voltage tolerance is closely related to the forward voltage drop. For the SPV1512, Vr is 12 volts. Not the lowest I've seen, but certainly getting there!

There is no mention in the SPV1512 datasheet that this is a switched FET device, but after your mention of the SPV1001 (I could only find the obsolete SPV1002), I went back and re-read the SPV1512 datasheet, and the line "... is the significant evolution with respect to the traditional standard schottky diode" (my highlighting). And its' use is intended for solar panel/cell bypass.

So, it would appear that the SPV1512 is no longer being produced because it (1) is not what it claims to be - ie: it is NOT a diode, and (2) there are now other devices that do what it does with significantly better performance (like the SM74611).

Looks like 220mV (eg: LSM115) is what I am going to have to live with. Thanks for the replies.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #6 on: September 01, 2020, 07:18:24 am »
https://www.mouser.com/datasheet/2/389/spv1512-974148.pdf

They certainly don't advertise it as such; it is a "switch" of some sort.

The Vf apparently doesn't vary much with If; I wonder if rather than switching on and off, it's more of a shunt regulator?  (I mean it's still going to be switching somehow internally, but it's going to be a high-ratio charge pump, not up and taking a big gulp of air.)  That is, they've made a control circuit that's able to operate down to 100mV, but no lower, and so it's able to hold it just at 100mV, almost independent of current (and higher at high current, just because of Rds(on)).

Another clue is the low reverse leakage.  A low-barrier schottky would have many mA of leakage at high temperature, and sloping up aggressively towards breakdown.

Still another is the complete and utter omission of AC parameters -- switching speed, forward or reverse recovery, charge, capacitance?

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9340
  • Country: fi
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #7 on: September 01, 2020, 02:07:55 pm »
Why would you think it is a schottky diode when the title reads "bypass switch", and the words "schottky" and "diode" exist only once in the datasheet, in a sentence where they explicitly compare it to a schottky diode?

It looks like you can't get below about 200mV if you need a diode, with good switching characteristics. In any case, be very careful with the reverse leakage specs with low-Vf schottkys. Note it's both temperature and Vr dependent, and they often give curves only for "typical unit". The risk of thermal runaway is real. Sometimes, a higher Vf (say 400mV) device dissipates less and performs better than, say, 250mV unit, if you run the numbers; depending on your application of course. This is especially the case with switchers where the diode produces heat during on-time, then (sometimes unexpectedly) produces some more heat during off-time as well due to leakage, heating up more than expected, which in turn increases the leakage even further, until destruction.
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #8 on: September 01, 2020, 05:31:40 pm »
STmicroelectronics also created 'field effect rectifiers'.  Some have a very low voltage drop.


See here:
https://www.st.com/en/diodes-and-rectifiers/field-effect-rectifiers.html

Voltage drop is around the same of 0.24v at 1amp, but, these are far superior when going above 1 amp compared to your LSM115.  Like 0.3v @5amps or 0.35v at 10amps VS 0.5v @ 5amp for the LSM115, not to mention the LSM115 cant sustain 5amp while the field effect rectifiers go way into the 20 amp or more range depending on the part #.  (I was comparing to the FERD40U45C)

The LSM115 wins at very low currents below 1 amp.
« Last Edit: September 01, 2020, 06:22:27 pm by BrianHG »
 

Offline MustardManTopic starter

  • Contributor
  • Posts: 44
  • Country: au
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #9 on: September 01, 2020, 11:09:25 pm »
I came across those FERD (Field Effect Rectifier Diodes) only a few days ago, and they look very interesting. I did a search on them last night hoping to find out how they were made and got hold of the patent application. Unfortunately it was a really poor B/W scan of what was obviously a colour original, and lots of parts of the diagrams (and graphs) did not turn out - in some cases at all. But the text was informative enough - invented by ST, so they are putting a bit of thought behind low Vf diodes.

I've now got a couple to test, and I will be comparing them to the LSM115. My application is low voltage (max Vr is ~3 volts) and at a moderate If of 10mA to 100 mA. The goal of my circuit is low voltage full wave low frequency power rectification and the reverse leakage (~1mA) will be significantly outweighed by the forward current of the complementary diode. Losses of 5% give-or-take are not a big issue in my circumstance.

Cheers, MM.
 

Offline MustardManTopic starter

  • Contributor
  • Posts: 44
  • Country: au
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #10 on: September 02, 2020, 01:52:00 am »
In case anyone is interested I did some rough-and-ready tests. My instrument was not calibrated so the absolute value of the readings could be off, however, they were all performed within the hour so the relative figures are good. The test was only performed on one device so I would expect some variance.

I don't know why I got a bump on all the reverse readings, but it is consistent amonst all of them (although the reverse characteristics are important, I am more interested in the forward performance).

As expected the reverse leakage of the LSM115 is bad, but not as bad as I had expected and with a forward performance only marginally different from my test FERD... I have two contenders...

« Last Edit: September 02, 2020, 02:07:52 am by MustardMan »
 
The following users thanked this post: thm_w, KE5FX

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #11 on: September 02, 2020, 02:13:29 am »
If you are going for full wave bridge rectification, have you considered the LT4320-1 Ideal Diode Bridge Controller?

I mean, if you really need that 0v drop, though, it needs at least 9v to work.

There are also smart Smart Diode Rectifier Controller which emulate the solar cell diodes with the mosfet of your choice.  Use one of those with a cheap SOT23 N-Channel 5amp mosfet with those ridiculous small RDS-On figures and only a 2.5v VGs on and you will have a small voltage drop.

« Last Edit: September 02, 2020, 02:19:33 am by BrianHG »
 

Offline MustardManTopic starter

  • Contributor
  • Posts: 44
  • Country: au
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #12 on: September 02, 2020, 05:07:30 am »
I am working with a really small voltage in this project: +0.6 to -0.6 swing. With the LSM115 (at approx 50mA) I am loosing ~175mV off each peak, giving me a rectified voltage in the order of 850mV.

I will be using a TPS61220 (or similar, 0.8 v input) to step it up to either 3.3 or 5 volts. A couple of manufacturers do very-low voltage switchers, some claiming they work down to 0.65 volts, but deeper reading of the datasheets indicates that once started that is what they will run down to. Usually the start voltage is a bit higher.

I have been giving a lot of thought to syncronised rectification with a FET, but unless I get some voltage to work with first, I can't drive the gate - and once I get enough voltage (eg: the 0.8 volts the TPS61220 needs) then I don't need syncronised rectification. My input will either be zero, or +-0.6

I saw the syncronised FET bridge in my travels, and it is a device I will certainly keep in mind. But it won't work for this particular project. Thanks for the suggestion though!

Cheers, MM.
« Last Edit: September 02, 2020, 05:11:49 am by MustardMan »
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #13 on: September 02, 2020, 05:34:44 am »
If you are not extracting power, you do realize that with an op-amp, you can get a 0v ideal switch with an opamp and normal diode.

Also, if the signal is always oscillating, you and use 1 diode and caps to multiply that voltage with the loss of a single diode drop instead of 2 in full bridge rectification applications, however, you double the current across the diode.

Also, for low power higher frequency apps, a small transformer can multiply your source voltage 10 fold.
 

Offline MustardManTopic starter

  • Contributor
  • Posts: 44
  • Country: au
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #14 on: September 02, 2020, 06:14:05 am »
I am using the device to suck power, and the power it vampires is the only power it will have to operate. Once bootstrapped I could use something better (eg: syncronised FET), but once I have power available, then everything is hunky-dory anyway.

I had looked at a half-wave multiplier with more than one stage to boost the voltage to a decent level, but with a half wave doubler (or tripler, or n-er), the ripple becomes very large as soon as I start to vampire power out of the thing - hence my (so far) deciding a full-wave is better.

My input will be close to a square wave. I had not thought about a transformer to boost the voltage, although at 50Hz (my operating frequency) I could run into size problems.
 

Online BravoV

  • Super Contributor
  • ***
  • Posts: 7549
  • Country: 00
  • +++ ATH1
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #15 on: September 02, 2020, 06:51:28 am »
 
The following users thanked this post: BrianHG, MustardMan, Vovk_Z

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: STmicroelectronics ultra-low Vf Schottky
« Reply #16 on: September 02, 2020, 09:57:35 am »
If you have an auxiliary supply with enough voltage to run useful functions, you can build a charge pump (or there's probably an energy harvesting chip that starts/runs lower than building blocks) to keep itself running, and also operate some switching circuitry.  Downside: such a scheme usually needs extra voltage to start up.  Which if it can't be made to self-start, it could be as simple as adding a coin cell, or maybe a supercap (make sure the circuitry shuts down with low leakage, when the main source is unavailable!).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf