Changing the Output Voltage of a Power Supply with a Diode?

The main problem with paralleling diodes is the forward voltage drop is quite different between devices unless purchased as a matched set.

This means they will not share the current equally between them and you will risk destroying them as one heats up more than the other which reduces its forward drop even more and eventually this thermal runaway condition causes a catastrophic failure.

You could put a small ohm value, high wattage, series resistor inline with each diode help balance the load sharing and thermal runaway, but this creates even more losses.

The diode solution basically just wastes energy and produces heat, which as Salvo has stated, is not a very practical solution.

Yes, that's right. It didn't even occur to me consider the different temps.

Thanks for all your help Salvo. I am going to consider all you've shown me.

In the meantime, I am going to mess around with my little gophert 1620, and see how fast it raises the voltage to about 13.8 volts, at various soc's. It can be easily hooked up to the bank via a pair of clamps onto a pair of studs inside my inverter cabinet. Not ideal, but it would only take another couple minutes of my time to hook it up and then pack it back away.

I'm also going to see if I can find a reasonably priced electronics guy to look at both the Turnigy and my Peak 40 amper. I'd like to modify the voltage output of both, in a manner like smk suggested. And Landyacht did some similar mods on his Mean Well.

I think 4 diodes will work even when battery is at 12.2V. Surface charge should quickly bring up the voltage. When the diodes are cold (at turn-on) they will drop 1.8V, not 1.3V (when hot).

In addition (if it doesn't work) you could get a spool of 12 awg house wiring and get the extra voltage drop. 12 awg has a fusing current of over 200A, but it will get hot.

As I said, it's not going to be elegant nor pretty.

jrnymn7 wrote:
Thanks Salvo,

So, if I understand this correctly, by paralleling 4 diodes, I can distribute the current amongst them, thus divvying up the heat dissipation, but will only end up with a total of 1.3 volt drop, the same as with an individual diode at the same temp.

This would only get the Turnigy down to 12.32 volts; still well above the 12.2 volts of a 50% dod bank.

But what if I was to use forced air cooling, along with a heat sink? (keeping in mind, the ambient air could be more like 35C).

No, smk, I have not. And I had no plans to do so. I have neither the tools nor experience to get into modifying the internal components. If it comes down to that, I'd rather sell it to a hobbiest, or to an rv'er who needs a decent high amp converter to supply power to their rig's 12v system.

Personally, I believe the 12v distribution system and the battery charging system should be independent of one another, if for no other reason than not throwing 14++ volts at the various 12v fixtures during charging. And in the case where the batts are a long way from the 12v panel, it makes more sense to add a charger to the mix, than upgrading all the wiring in between.

But yes, for someone with the know how and equipment, it could very well be an easy fix. And I've already ordered a pm3-75, which I plan to modify with a simple turn of a screw, so this thing has become more of an educational process than an absolute necessity. But if I do figure out an easy fix, I will keep the unit as a back up, and apply what I have learned to getting other out-of-control chargers back in line.

jrnymn7 wrote:
The Turnigy 1080w outputs 13.62 - 18+ volts, and I would like to decrease its overall output range to anything less than 12 volts at the lower end, but no less than 14.8 volts at its upper limit

I would be more inclined to update the existing control circuit such as add some resistance to a variable potentiometer.

Have you opened the unit to review the control circuit?

Thanks Salvo,

So, if I understand this correctly, by paralleling 4 diodes, I can distribute the current amongst them, thus divvying up the heat dissipation, but will only end up with a total of 1.3 volt drop, the same as with an individual diode at the same temp.

This would only get the Turnigy down to 12.32 volts; still well above the 12.2 volts of a 50% dod bank.

But what if I was to use forced air cooling, along with a heat sink? (keeping in mind, the ambient air could be more like 35C).

Yes, it's possible to lower the voltage by adding a diode. The PS current will decrease and you won't get to the 60A foldback.

The fix is not pretty. The objective is to keep the junction temperature under 150C.

data sheet

Looking at the 100 C performance curve, we get a 2.1V diode drop at 60A. The max power the diode dissipates is 2.1V * 60A = 126W.

The thermal resistance of the diode is 0.8C/W. We need to calculate the heatsink thermal resistance so that the junction temp is at 150C when dissipating 126W.

The diode package as a thermal resistance of 0.8C/W. The temperature rise within the package is 0.8C/W * 126W = 100C.

If ambient temperature is 25C, that leaves just 25C for the heatsink. The thermal resistance of the heatsink is:

Theta_HS = 25C/126W = 0.2 C/W

That is extremely small. This won't work, the diode will burn up. You'll have to parallel multiple diodes together.

Let's take 4 diodes in parallel. Each diode conducts 60A/4 = 15A. The diode drop is 1.3V at 15A, at 100C. Power is 1.3V * 15A = 20W.

Temp rise in diode package package is; 0.8C/W * 20W = 16C

Theta_HS = (150C - 25C - 16C) / 20W = 5.5C/W

That's doable.

Google 5 C/W heatsink

You need to wire up 4 of these in parallel. You'll get about 1.3V drop.

"It sounds like your power supply may have foldback current limiting versus straight line current limiting, based on your description of the issue. If this is true diodes won't help. "

Diodes will not serve to replace constant current limiting, but they may provide a way to do a soft start, and thus eliminate the need for ccl.

"Best to use a power supply without foldback limiting."

Yes, in fact it's absolutely essential when using the PS with such a large load, as with charging a 12v battery bank.

smk,

The issue is not with the voltage, it's with the type of protection. In many cases, such as with powermax switching power supplies, the unit has constant current limiting. But it turns out the Turnigy does not. I could not find sufficient literature on the unit, and therefore should have made further inquiries into the matter before buying it. My bad.

Apparently the same issue exists with some of the Mean Well models, so one must be careful which one they buy.

Now I'm trying to figure out a simple fix. And if I do, that fix may prove quite useful for other applications as well.

I explain the situation, in detail, in my reply to Salvo, three posts above yours.