Boost converter won't work in USB power supply

[msknight]

I've got a buck converter which works fine when attached to a bench power supply, but the moment I plug it in to a USB power supply, the supply goes blank/power ceases. Even when the buck converter is turned right down to almost the same as the input voltage.

To test this, I've had an in-line USB power meter which lights and works when plugged into the USB power charger, but the moment I then plug the USB cable in to feed the buck converter... it dies.

I've tested the cable (by using a socket to attach the bench power supply) and also used more than one USB power charging unit with plenty of amps to give... (and also tested the polarity of the cables) ... but I can't work out what's going wrong here.

Any ideas please?

[SmallCog]

Yeah I've seen this happen

Buck is putting noise onto its input (ie the output of the power bank), circuit in the USB power bank is seeing it and being unhappy

If you have access to a scope you'll (I predict) see a bunch of noise

Your bench supply any anything else it happily runs off is just not giving a damn about the noise.

You can either play with filters, or do like I did and grab a different module out the prototyping bin and it all just works.

*An EE rather than a mere tech like me will be long shortly to use bigger words and correct me in many ways*

[msknight]

Thanks for this. I guess I must have been just lucky with the first couple I used. Fortunately I have a few others around.

Is there any other way to nudge voltage up a little, say from 5 to 8 or 9 without having to use one of these chinese circuits please? A different source of more reliable units would be a start I guess 

[Nusa]

Sounds like you actually have a boost converter, even though you keep writing buck converter.

[SmallCog]

Yeah, buy a reputable brand from a reputable supplier

https://au.element14.com/w/c/power-line-protection/power-supplies/dc-dc-converters/prl/results



I buy eBay junk too, but only for messing around you just need to be aware of it’s limits

[LateLesley]

To explain what may be happening, the USB power supply might be tripping out on over current protection. When you initially connect the converter, it'll have an inrush current spike, which may be triggering the PSU protection circuitry.

USB specs is usually 500mA for a USB port, but many devices will supply 1A or 2A, sometimes more on chargers.

what you could try, is a sufficiently rated low ohm resistor in series with the supply, to limit the switch-on spike in current. You can try a 2 ohm or 1 ohm resistor. The downside of this is it will create a voltage drop across the resistor in use.

Or maybe you can use an NTC? I'm not sure that would work though in low voltage circuits. It depends how much current you circuit draws in use, because NTC's have a minimum current draw to work, and might have too much voltage drop at low currents.

I've seem some articles saying you can use a FET for inrush current limiting too, so maybe that is an angle worth investigating.

[sleemanj]

Yeah this is not a buck converter, if you call things the opposite of what they are things get very confusing.

  * Buck = Step Down from a High voltage to a Low voltage
  * Boost = Step Up from a Low voltage to a High voltage

Anway, I would...

Try without a load, try with just a resistive load, different resistive loads

Try adding a small resistance to the input to discover if it is an inrush issue.

Try adding some more bulk capacitance on the input side.

Try adding a diode on the input side. 

Try all of those things at once, or on their own.

Try paralleling the USB supply and the lab supply (probably with diode-oring) to power it and then while powered disconnect the lab supply to determine if it might be an issue on startup.

[KT88]

To tackle the problem some more information is required:
Board specs
Input voltage 5V
Output voltage: ?
Load: ?
Can you identify the chip?
A close-up photo of the board would be nice...

The board doensn't look too bad - they used a shielded inductor instead of a drum core.
The capacitor values hint a boost: 100u/50V out; 220u/35V in.

Edit: would it start without a load?

[Munyua44]

The USB rating is mostly 5V and 500mA. Most back converters devices deliver around 1A to 2A current which much higher compared to that of the USB. This stopes the USB from connecting.
thanks!

[msknight]

Thanks all. Yes, boost converter. Picture will follow. Part of the problem is that I still haven't managed to start learning properly about electronics and here I am, still trying to get projects completed and things working. I'm also trying to do this in the odd few moments between work and caring for my Mother, so I don't have a clear window to just sit down and do things... so I really appreciate your patience with me.

[KT88]

One trap for young players is the negative input imedance a switch mode converter poses to the source: while the voltage at the input is still low it draws a much higher current. You will see that the input current increases when you use your PSU and reduce the input voltage. This causes some time issues when the source has a current limitation like a USB port. Better regulators provide a soft start function for that reason.

[msknight]

The lowest I have is a 10 ohm quarter watt resistor and it's running so hot it would burn my finger if I kept it on there.

The unit does now power up but with no load is putting out less than half a volt where I would expect around 5V with what was shown in the previous pictures.

[sleemanj]

Did you reverse the input polarity accidentally?

[msknight]

Polarity is fine.

Before the resistor I'm seeing 4.8v. After the resistor it's down at 0.75v

Do I need lower value resistors at a higher wattage? What kind of things should I add to my parts stock?

[sleemanj]

If you are drawing 400mA with no load, something is seriously wrong. 

That's why I suggested polarity as that would be the first thing to check there and the sort of thing I'd expect to see if you did that (either currently, or previously and the device got killed), if you have checked and it's not incorrect currently, and the device still works normally when you have it on the power supply, I dunno.

[KT88]

The data sheet of the XL6009 states Vin from 5V - 32V. It also has a UVLO (undervoltage lock-out) funktion. Pls. use the lab PSU to find out at which voltage it starts or stops switching (w/o the resistor). The XL6009 also has an EN-pin which could be used do delay the start of the boost operation to allow for proper charge-up of the input cap. If the input cap alone still takes out the USB port, the input cap can be reduced to 100uF as an example.
However operating this boost-IC at the very edge of its rated input range is not recommended as even as it may work on the bench, it can refuse to work at higher or lower temperatures as an example...

[nali]

Polarity is fine.

Before the resistor I'm seeing 4.8v. After the resistor it's down at 0.75v

Do I need lower value resistors at a higher wattage? What kind of things should I add to my parts stock?

That could be indicative of a short on the output... what is the voltage on the actual output terminals?

That chip is a XL6009 buck/boost converter, the board probably follows the example schematic from page 3 of the datasheet https://datasheetspdf.com/pdf-file/775384/XLSEMI/XL6009/1.

By the way, do you only have the one 10 ohm resistor? You could parallel a few up to get the resistance - and dissipation - down to a more suitable value. (either more 10R or other low values)

[msknight]

From what people are saying, I believe that operating the boots at the minimal range is likely to be the problem.

I have a headphone amplifier that works with a power input between 5v and 20v. Running on 5V is not producing enough volume, hence I want to boost the input voltage to get more amplification.

What I'm probably going to do, is hook up the input to the bench power supply, where I can up the input voltage until I reach the power I need, and then get one of those DC/DC converters which is more aligned with what I need.

Sound like a plan?

[msknight]

OK - scratch everything. I attached the amplifier to the bench power supply, but increasing the volts didn't result in louder audio... probably just less amps.

Back to the drawing board... completely.