Parallel Diodes Voltage Drop / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Parallel Diodes Voltage Drop

previous 1 2 3 next

Move Thread

LAN_403

cduma

Fri Jul 29 2011, 12:08PM

Registered Member #1822 Joined: Fri Nov 21 2008, 08:04PM
Location:
Posts: 300

Darn, you assume wrong. It is a SMPS...

Dr. Slack

Fri Jul 29 2011, 01:12PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

If you use silicon diodes, you will always get a threshhold of 0.7v min, plus some resistive drop. Putting a large number in parallel will get you closer to 0.7v. Use an inherently high current diode is much the same as using several smaller ones in parallel.

If you use schotky diodes, all the above applies, except use 0.4v instead of 0.7v.

If you use a controlled FET, then your drop is the product of the Rds(on) and the current flowing, there's no threshhold. The disadvantage is that you have to control the FET actively. Some people use a diode and FET in parallel (or a FET with an intrinisic body diode). The diode carries the current before the FET has turned on, the FET controller senses the forward diode drop, usually against a forward several 10s of mV threshold, and turns the FET on. This non-zero threshold is necessary as you have to ensure that the FET turns off if the diode is needed to block, so it's a safety margin to ensure that reverse voltage is detected properly.

You can do much the same with a relay instead of a FET, but it won't turn off as fast when it's needed to reverse block.

Can you hack the SMPS to give a little more voltage, or remote sense at the other end of the diode, so that you can just use a simple series diode?

cduma

Fri Jul 29 2011, 01:27PM

Registered Member #1822 Joined: Fri Nov 21 2008, 08:04PM
Location:
Posts: 300

I was thinking of using relay. I dont think the slow speed of a relay would be an issue. Can the average SMP handle 200mS of back feed?

I havent checked but, I may be able to increase the voltage. How would this effect battery charging? Wouldnt it eventually match the actual voltage the SMP is outputting regardless of the drop on the diode?

Neet Studio

Fri Jul 29 2011, 03:59PM

Registered Member #4037 Joined: Fri Jul 29 2011, 03:13PM
Location:
Posts: 86

At 50A, then you are likely looking at something like 0.7V diode drop for schottky.
A N-MOSFET e.g. IRL3716 is around 0.004 ohms which would be around 0.2V when fully enhanced.

You are better off using "Active Or'ing Diodes" - basically MOSFET with an IC that has the biasing and control circuits.

For example: TI's N+1 and ORing Power Rail Controller: TPS2412/TPS2413

Similarly, Linear Tech, IRF and possibly has their own Or'ing diode controllers.

Note:
TI chip regulates the drop at 10mV, so unfortunately that is too low a drop to do load sharing among multiple controllers.

Dr. Slack

Fri Jul 29 2011, 06:08PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Before knocking yourself out to solve a problem, does the problem really exist? What happens if you "back feed" a SMPS? Does it burst into flames, drain the battery, or neither?

Arcstarter

Fri Jul 29 2011, 08:07PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

As was previously said, a Schottky rectifier is a good choice for this. It is not hard to come by one that can handle 50 or more amps, and some have pretty impressive Vf characteristics.

Most ATX PSUs have schottky power rectifiers for increased efficiency, and the speed is also a plus. This 250w PSU board has 2 of these 45v 2x30 amp to-247 schottky rectifiers. The 2 diodes in the packages can be paralleled, and they should share current very well. That gives you 60 amps, and then you could parallel another one if you needed to. Even if they dont share current perfectly, it would probably still help somewhat. Here is the datasheet:

If you don't have a PSU to take apart i will happily ship these 2 to you.

Neet Studio

Fri Jul 29 2011, 09:59PM

Registered Member #4037 Joined: Fri Jul 29 2011, 03:13PM
Location:
Posts: 86

The first question is : What do you know about the power supply?

Normally the correct way is to consult the datasheet or the manufacturer to see if the power supply can be back driven.

Second question: Is there a over-voltage crowbar to short the outputs? Would it be triggered at the maximum battery voltage?

Third question: What type of power supply it is?

If it is from surplus or unknown source, it is a matter of opening it up and look up part numbers of semiconductor parts mounted on the low voltage side of the heatsink. Those are usually large MTO-3P parts for rectifiers. There is a clear separation on the PCB due to isolation requirement.

If it is a SMPS with regular schottky diodes, then you can get away with back driving it.
If it is a power supply with synchronous rectifier output and if it was designed to be back driven, then you are fine.

If it is a power supply with synchronous rectifier output and not designed to be back driven... !!!

What usually happens is that the synchronous rectifier circuit becomes a power converter and feed power back to the primary side (without feedback!). I have seen this happened a couple of times.

I have even heard of a bus rail power supply said that they would purposely slam the output rail to discharge output on shutdown. I would not spec such a power supply for a bus rail.

Dr. Dark Current

Fri Jul 29 2011, 10:32PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

The problem with back driving eg. a flyback type SMPS might be that the clamp voltage is transformed on the primary side, and the switch sees higher peak voltage, unless the controller completely kills the PWM pulses. However, I don't think this would usually be a serious problem.

Pinky's Brain

Sat Jul 30 2011, 10:23AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Marko

Sun Jul 31 2011, 06:59PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Dr. Slack wrote ...

Before knocking yourself out to solve a problem, does the problem really exist? What happens if you "back feed" a SMPS? Does it burst into flames, drain the battery, or neither?

Yep, as said, there's just no problem, I just can't see from glance what kind of "backfeeding" could occur. SMPS's can charge lead acid batteries just fine, and if it's a high power supply there should be a small trimmer resistor somewhere with which you can adjust the voltage (and with it the rate of charge) of the battery.

Marko

PS. on a second thought, there may actually be a small problem with synchronous rectifiers after all. After one thinks carefully about it, active rectifiers are often made with mosfets driven from the extra windings on the power transformer. If it is fed in reverse it could turn into a blocking-style royer oscillator and cause the mentioned backfeeding into the input. This will charge the input cap but it may not be a problem at all.

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob