Do I need an additional diode in this use case

[littlebill]

I am using 2 of the attach designs in parallel. I am going to use droop resistors, but was also going to add another set of diodes for reverse protection, but I am correct in assuming the diagram indicates there is already a diode here to protect against that and another one is not needed?

[littlebill]

after additional studying it does appear I need them since PIN5 feedback is not protected I also assume voltage current here can make things go haywire. it appears the D1 is used to offload the current of the internal Diode in the xl6009 module. ( was getting conflicting information)

would love some confirmation

[Konkedout]

First of all, you really ought to post a datasheet link.  Is this it?

https://www.haoyuelectronics.com/Attachment/XL6009/XL6009-DC-DC-Converter-Datasheet.pdf

Secondly, I cannot think of any good way to "connect two boost converters (or two boost converter ICs) in parallel" unless they have specific features to allow that.  Malfunctions are almost definitely going to result.

[tooki]

after additional studying it does appear I need them since PIN5 feedback is not protected. it appears the D1 is used to offload the current of the internal Diode in the xl6009 module. ( was getting conflicting information)

would love some confirmation

Ummm, assuming that by “module”, you mean the little boost converter boards commonly found on AliExpress and co, then no, D1 is the diode in the module. The module contains the circuit shown. The XL6009 IC doesn’t contain a diode.

I highly suggest reading up on how boost converters work, so that you find out what each component does.



Why do you want to parallel them anyway? If you need more current, just get a boost converter that can supply what you need. If you want two for redundancy (i.e. each converter alone is capable of fully supplying your load), then you could use diodes to do it. There are special load switch chips that do it better though.

I wouldn’t just parallel the outputs through resistors, you don’t know how the converters will react. As Konkedout said, converters that are safe to parallel typically have special features to synchronize the chips. But usually you just use a beefier converter (or one that uses an external MOSFET so that you can select it to whatever current you need).

[littlebill]

I have done more reading than you can imagine. I am not strong here. I did not expect to go down this path.

I have a size and efficiency limitation at this point, another module would almost make me start over. I am not doing , it can burn to the ground at this point.

https://soldered.com/productdata/2016/12/Soldered_XL6009_datasheet.pdf

based on this. there is an internal diode, unless I am not understanding the function block on page 4. unless those symbols don't mean diodes?


in regards to the resistors, i have read the theory on it, im using .1ohm. I don't really care if its not perfectly balanced. I just need a bit more capacity to get the thermals down to a normal level.

[Vovk_Z]

Secondly, I cannot think of any good way to "connect two boost converters (or two boost converter ICs) in parallel" unless they have specific features to allow that.  Malfunctions are almost definitely going to result.

+1 to this. I am almost sure they want work togther in parallel, or with a battery at it's output.

[littlebill]

well I guess were going to find out. with diodes on the output . I am not entirely sure why it won't, the load just takes from the highest voltage. This is no different then in a rv where there are multiple chargers in line sometimes up to 4 at once, and I have never seen the magic smoke.

[tooki]

I have a size and efficiency limitation at this point, another module would almost make me start over. I am not doing , it can burn to the ground at this point.

https://soldered.com/productdata/2016/12/Soldered_XL6009_datasheet.pdf

based on this. there is an internal diode, unless I am not understanding the function block on page 4. unless those symbols don't mean diodes?

The function block (figure 3) is on page 3, and it shows no diodes at all.
The typical application schematic (figure 4) is also on page 3, and it shows one diode. It is NOT inside the chip.

Again, when you say “module”, do you mean the little boards? Or are you using the word “module” to mean the XL6009 chip itself?

I don't really care if its not perfectly balanced. I just need a bit more capacity to get the thermals down to a normal level.

So your problem is that you have one and it’s getting too hot?

well I guess were going to find out. with diodes on the output . I am not entirely sure why it won't, the load just takes from the highest voltage.

Right. And since all the current to the load is flowing from the supply with the higher voltage, then the supply with the lower voltage won’t see any load at all, and the higher-voltage supply will continue to get hot.

This is why I said diodes might work for redundancy (= power continues even if one converter fails), but not for balancing.

This is no different then in a rv where there are multiple chargers in line sometimes up to 4 at once, and I have never seen the magic smoke.

Without knowing how those chargers are designed (and what they are being used for, and how) it’s impossible to draw any conclusions from that.

[littlebill]

I am using a module that is identical to  the typical application circuit design on page 3. The function block confused me a bit, I am not sure what are the discrete components inside the IC, again didn't expect to have to get to this point

yes its too hot, these are listed at 4amps, obviously with forced cooling after some testing. I can't do active cooling.

The resistors are suppose to keep the balance in check

Those chargers are generic battery chargers nothing special, and is done everyday by millions of people. Think solar, generator, converter, and alternator. etc..

[tooki]

I am using a module that is identical to  the typical application circuit design on page 3. The function block confused me a bit, I am not sure what are the discrete components inside the IC, again didn't expect to have to get to this point

There are no discrete components inside an IC. “IC” means “integrated circuit”, precisely because it is not composed of discrete components. But in terms of functional blocks, you see a bunch of opamps/comparators and a MOSFET forming the core of it.

yes its too hot, these are listed at 4amps, obviously with forced cooling after some testing. I can't do active cooling.

The chip is rated for a 4A switching current. First off, the switching current ≠ the maximum output current. The maximum output current is also limited by the choice of external components. Those little modules are often… let’s just say “optimistically marketed” and aren’t capable of their rated specs.

The typical application schematic says right on it: output 18.5V/2.5A. Unfortunately the datasheet provides no engineering guidance whatsoever on how to design a whole circuit, so we can’t really calculate what the true output power of your module might be. (It might be possible to use the datasheet for the original part from TI that the XL6009 is a copy of. That happens to be the LM2577. The XL6009 is basically an improved version of the LM2577 with a 4A MOSFET instead of a 3A BJT in the original. Here’s the LM2577 datasheet, which is 4 times as long because it actually has some basic design guidance: https://www.ti.com/lit/ds/symlink/lm2577.pdf )

I wouldn’t even bother trying to parallel them. They aren’t designed for that and the behavior might be unpredictable, and almost certainly won’t accomplish what you want.

Either buy a converter with the output power you need (plus headroom!), or if your load can be divided, split it up and give part to each module. (E.g. if your load is 100 lamps, divide them into two groups of 50, and put each group onto its own module.)

The resistors are suppose to keep the balance in check

You might get lucky and it works. Or you may get unlucky and blow up everything. Hard to predict since they’re not designed for it.

Those chargers are generic battery chargers nothing special, and is done everyday by millions of people. Think solar, generator, converter, and alternator. etc..

That doesn’t tell us any of the information we’d need to know. There’s no such thing, from an engineering standpoint, as a “generic” “nothing special” charger.

[littlebill]

So to close this out.Tested it out, .56A output on module1 .95A output on module2. 23% different. My calcs were slightly off, prolly could have got it tighter. but the modules are now just warm vs burning. Glad I finished this out. I have several other modules on order, with different chips. but this was an odd ball design which put me in a bad place. the xl6009 are still smaller modules then everything else I looked at, and there was a size limitation that these barely fit, so outside of a TPS61088 which may be a better fit for future stuff. also ordered a LTC3780 sync rectifier board which should also run cooler.

I would not have needed droop resistors if i bought boards with Constant Current control, just the diodes, these could have been dialed in much better, but eh we got there.

Hope other readers find something useful from this.

below thermals were taken at 25 minutes in. they had stabilized.