DC-DC boost converter circuit

[timelessbeing]

I bought two of these ubiquitous XL6009 DC-DC converters from the internet.

Specimen A.
Advertised voltage output: 5-35V output
Of course when it came, it didn't match the photos in the description; A bunch of component were missing. Presumably, this was done to cut manufacturing cost.
My tested output voltage: 5 - 50V

Specimen B.
Advertised output 5-50V, 2.5A, 15W
This one came fully populated.
Tested output: 5 - 48V

Question 1
The missing components in spec.A were a diode near the input (reverse polarity input protection I assume), a few SMD ceramic caps on the input and output(smoothing I guess), and an inductor in series with the output right after the freewheeling diode.  I've never seen an inductor on the output of DC-DC schematics like that before. What is it's purpose?

Question 2
The boost output of the 35V rated module with missing components was actually higher than the fully populated module rated for 50V. Why the difference in advertised and measured outputs? The feedback resistor divider is identical on both.

Question 3
The application circuits in the XL6009 datasheet use feedback resistors greater than 1k, but these boards have resistors less than 1k. What's up with that?

[NiHaoMike]

The second one has noise filtering while the first is basically a bare bones design that just manages to work.

[ocset]

wHEN YOU GOT MORE OUTPUT VOLTS THAN YOU EXPECTED, MAYBE YOU WERE ON NO LOAD THEN? (sorry about capitals).
Sometimes converters need a minimum  load.
You gave an "adveretised" output voltage...but from what input voltage range does this apply? The datasheet should say.

[chris_leyson]

I don't know, you've got two boost converter PCBs to play with why not read up on how boost converters work, change the input voltage change the output load and scope things up, you might learn something. Don't expect to learn anything from a half populated PCB you bought on the cheap, that's not the way it works.

[timelessbeing]

Sometimes converters need a minimum  load.

They were loaded.

You gave an "adveretised" output voltage...but from what input voltage range does this apply? The datasheet should say.

5-32V. I believe I was using 9v.

I don't know

Right. So don't post.

The diode across inductor switching node and ground

It's on Vin, not the switching node.

[ocset]

Thanks, can you locate the source sense resistor and send a scope shot of its voltage?
Mind you , maybe the sense reisstor is inside the chip, which is not going to help us.
iF YOU SCOPE IT, USE A BIT OF MODIFIED COAX AND NOT A DANLGING-GND-LEAD scope probe (sorry again about caps).
If no sense resistor, then you could always add a sense resistor in immediately upstream of the boost inductor and scope that ......as long as you power the xt6001 thru an isol transformer.

If you can look at either the inductor current waveshape,  or fet current waveshape, then that offers great help.

If you send me F(sw), THEN ILL SEND YOU A SIM of it. in free ltspice.....that can help to investigate.

[timelessbeing]

It's not a current regulated device. There is just a feedback resistor network connected to a comparator inside this chip.

I'll try to get some scope traces today.

as long as you power the xt6001 thru an isol transformer.

The XL6009? It's powered by DC.

send me F(sw),

Switching frequency? 400 KHz

[ocset]

thanks yes i appreciate its not current regulated, but i was requesting a sense resistor be placed so that we coudl scope the inductor current and / or fet current.

Yes i appreciate its DC input....what i meant was, power it with an isolated dc supply (which likely derives ultimately from an isolation transformer)  when you take the scopes.
I am sure you also know not to use two scope probes, each  with gnd  clipp connected to different nodes.

[ocset]

i HAVE ATTACHED A 15W BOOST CONVERTER IN LTSPICE. iT SWITCHES AT 400KHZ. (sorry about capitals)
I have also attached a start guide for ltspice.
It is free download from linear.com

IT IS 24V TO 35V.

yOU WILL BE ABLE TO CHANGE IT TO SUIT WHAT YOU ARE HAVING THERE. (sorry, my monitor is raised and i keep not seeing that im in caps...and i have no caps led)
I wonder if anyone knows how you can convert upper to lower case in one go?)

[ovnr]

sorry, my monitor is raised and i keep not seeing that im in caps...and i have no caps led

Use the shift key, not the caps lock key. And how to fix it? Retype the damn thing, don't be lazy.

And this is a DC/DC boost converter; the grounds are tied together. There's no need to confuse anything by blathering on about isolation transformers.


To OP:
#1: Missing components are noise filtering, as has been said. The "extra" output inductor along with the capacitors forms a pi filter, substantially reducing the switching noise on the output node. The less cheap versions of these have two electrolytics on the output; one before and one after the filter inductor.

#2: Output voltage difference may be due to component tolerances.

#3: The feedback resistors may have been tweaked to provide better loop compensation, or lowered to present a higher minimum load to the converter.

[ocset]

And this is a DC/DC boost converter; the grounds are tied together. There's no need to confuse anything by blathering on about isolation transformers.

Thanks, yes i appreciate, the thing is the OP timelessbeing, may be going to power the dc boost converter with  a dc supply that is earth referenced.....if timelessbeing then connects his scope ground connection to a high side part of the circuit (not the earth gnd) , then as you realise,  timeless is going to have  mucho problemo.........so i made the advice of just using an isolated psu to supply the dcdc during test.....just in case

...i will send more sims for timeless...a bientot.......but first, i am moving flat this weekend, to the other end of town, so i am busy right now carrying furniture.

[timelessbeing]

Yes thanks, my supply is isolated. But this tiny low power, low voltage module poses no danger.

[timelessbeing]

I stress tested both under max load (~20W):
- The extra-shitty one had nearly a volt of ripple.
- The not-so-shitty one had 70mV which is actually very close to the advertised spec. Great deal for $2

I also tested a variety of other eBay specials.
This one from shieldsfans (popular seller for electronics gadgets) had well over a volt of ripple and got very roasty. (no parts missing)

[ogden]

I wonder if anyone knows how you can convert upper to lower case in one go?)

Yes. Delete upper case text and re-type it in lower case. If you see that you made mistake in public forum posting and you don't even bother to correct it - you show disrespect to others. Please don't do it anymore.