Convert Quick Charge 3.0 to high current 5V

I recently made a big mistake and built a device containing a raspberry pi, a powerbank and a few other power hungry components, only to find out that the Pi constantly crashes because the powerbank doesn't provide enough current.

I now realized that I could get up to 18 Watts out of the powerbank if I used Quick Charge, but my devices don't understand Quick Charge and only take 5V.
Switching to a different powerbank is out of question because I already spent a lot of money on the 3D printed case which is specifically designed for this powerbank.

Is there some sort of cheap module that can convert 18W of Quick Charge 3.0 to 5V @ ~3.8A?

Could I combine these two products into what I need?

Would that work?

What voltage is the "Quick Charge"? It can be from 9V to 20V depending on which Quick Charge version.

Quick Charge is more voltage, but at a lower current capacity. The Watts passing through the wires and connector doesn't change much. The USB spec allows for 5V, 2A (10 Watts). QC4 would be 20V, 0.5A, or 10 Watts.

On your second link:
I have used similar voltage converters in my Wemos projects. I use a simple 12V Wall-Wart through the voltage converter module and power all of my 5V devices from there.

I've never heard of a "decoy" module, your first link. I didn't see any power spec on the AliExpress page.

tobsun.jpg

tobsun.jpg

I'm not sure where you got all that information from, but you are kinda misinformed.

SteveMann:
It can be from 9V to 20V depending on which Quick Charge version.

QC3 alone supports 3.6-22V.

SteveMann:
Quick Charge is more voltage

Not always.

SteveMann:
The USB spec allows for 5V, 2A (10 Watts).

I don't know about "The USB spec", but we have standards like USB PD and Quick Charge which support way more.

SteveMann:
QC4 would be 20V, 0.5A, or 10 Watts.

QC 4 allows for a wide variety of voltages and currents. Up to 100W actually.

This is the powerbank. It says:

Qualcomm Certified Quick Charge 3.0 Technology:
Quick charge your QC 3.0 Smartphone to 80% within 35 minutes, providing 5V 9V 12V for different QC 3.0/ 2.0 devices. Compared to QC 2.0, QC 3.0 has 15% greater charging efficiency, charges up to 27% faster, reduces power consumption by up to 45%
The QC 3.0 Quick Charge Mode will take effect with only Qualcomm Certified QC 3.0/2.0/1.0 Smartphones, Power Adapters and Cables for Fast Charging and Recharging.

Still looking for a solution.

The moral is get a working prototype before finalizing anything...

Try this perhaps?: Bitbanging Qualcomm Charge Controllers | Hackaday

Well, I had the prototype working. It just didn't occur to me that it may not be stable.

QC3 only increases the voltage, not the current. So the QC3 output on that powerbank will happily do 3A @ 5V. Thing is, a Pi is super duper picky about the voltage. Add the fact a powerbank is not + crappy cables and you have a complaining Pi.

You might get away with adding extra capacitors. Or in addition add a 5A DCDC step up and boost the voltage to 5,2-5,3V. No benefit in going to 12V and back, this will only increase the loses.

In my head the powerbank could be set up like this:

                   |---[3.7v to 5v step up (max 3A)]---[USB port]
[batteries 3.7v]---|
                   |---[3.7v to quickcharge3.0]---[USB port]

I could totally imagine the circuit responsible for normal 5v charging only supporting up to 3A (15W), but the quickcharge circuit allowing for 18W. QC3.0 doesn't necessarily mean it will use 12v, QC3.0 by spec actually allows for 3.7v output.

Yeah, so?

Only benefit would be if it indeed has two step up converters. That way you could use the QC port for the Pi and the normal for the rest. But this may still not solve the voltage problem. Although QC3 allows for pretty small voltage steps, the powerbank will not be very precise. The idea of that is that the device under charge can increase the voltage to compensate for losses. But it's not designed to be quick or very stable. But with a QC3 dummy you may set the QC port to 5,2V or even 5,4V in the hope it will stay high enough for the pi.

I don't understand what you're talking about.
I simply want to find a cheap device that can convert QC3.0 to 5V. This way I could draw 18W instead of 15W.
There are no voltage problems and I don't need to use QC3.0 with another USB port in parallel. I just need the 18W that the powerbank can provide via QC3.0 and convert it somewhat efficiently to 5V so that I can draw around 3.6A instead of just 3A.

That 3W is NOT going to make any difference. It's the fricking unstable voltage (power bank and cable combined) you need to fight. ONLY unstable (including to less) voltage is what makes a Pi unstable.

Even if you would gain that extra 3W you will lose that again in a DCDC to drop it down to 5-ish volt. Efficiency of DCDC converters is around 80-85% leaving you with just 14,4-15,3 usable Watts.

What you say is simply not true. If the Pi is not able to draw the current it needs it will crash as well and that is my problem here. The voltage is not an issue and the cables are fine.
Maybe there is a way to get QC3.0 at 5V directly from the powerbank. In that case I wouldn't need an additional DC-DC converter.

True, but that will cause the voltage to drop. 2,4A is more then plenty for a Pi. Only thing is, you may not let the voltage drop.

felic:
Maybe there is a way to get QC3.0 at 5V directly from the powerbank. In that case I wouldn't need an additional DC-DC converter.

There is, just use that damn connector :smiley:

Have you confirmed that voltage actually drops when the Pi crashes? By how much?

Have you tried adding a capacitor?

Is there anything you could do to reduce the current draw of the other devices?

@septillion If only it would support more current on that port. :sweat_smile:

@ShermanP Well, technically the voltage may drop, but that would be a consequence of the powerbank not being able to provide enough current. I've seen that phenomenon a lot with wall chargers. Once you try to draw more current than supported, the voltage will go down. But that's merely a symptom.

I have not tried adding a capacitor. Can I use a normal electrolytic capacitor? What would you recommend in terms of µF and voltage?

Unfortunately I can't reduce the power consumption of the other devices. I guess I could undervolt the Pi, but I need all the performance so that's not an option.

I can assure you, the power bank will not limit the current. That would only add unnecessary complexity to it. Only thing QC does is change the actual voltage. So tada, when you use the QC port you're able to draw 15W, probably more. You can probably even draw 2,4A from the non-QC port at the same time.

And yeah, the voltage WILL drop. Even when you change voltage level with QC that level may drop a bit when loaded and you will definitely lose voltage over the cables.

About capacitors, yeah, electrolytic. Voltage rating is the max the capacitor can handle (aka, needs to be larger then you apply) an higher capacity is more better.

I have a hard time believing that. What you're saying is that the current is limited by the lipo batteries. I think the QC or DC-DC circuit is simply limiting the current because it's not designed for more (e.g. PCB traces are too small, wires too thin, a certain chip doesn't support more current or the device lacks the required cooling capacities etc.).

What would be the minimum µF rating that would even make this worth a try? I don't think I have very big ones.

My question about the voltage drop was to make sure that in fact the voltage is dropping when the Pi crashes. If you have no material drop when a crash occurs, then the problem is somewhere else.

The capacitor will help only if the problem is short-lived. The capacitor will give you a little extra support to withstand a quick glitch, such as when you turn something on. The voltage rating probably won't matter because this is 5V. The capacitance value can be anything, the more the better, and you can use multiple capacitors in parallel. If there's a way to add those temporarily on the 5V line that feeds the Pi, I would certainly try it. Even 100uF might make a difference.

It seems to me if voltage drops it could be the batteries in the powerbank not being able to supply enough current, or it could be the switching regulator that produces the 5V output not being up to the task, or it could be resistance somewhere along the line after the powerbank. Do you have a scope?

I don't know anything about the QC stuff, but it would help only if it delivers more watts at the higher voltage output levels, or more watts at 5V for that matter. It seems it would have to do that to have any value for charging, so I assume it does. But I don't know how you persuade it to deliver those extra watts.

All the high-current extra stuff you're using - are they powered through the Pi, or do that have parallel power connections with the Pi?

That's not what I said. I only said the limit will probably not change, just set at the input of the DCDC. That having said, a lot of power banks really don't have any current protection and will self destruct or thermal lockout. But for sure the current limit will not change if you keep 5V. Aka, the QC port will do 3A, which is plenty for ANY Pi, out of the box. QC purely relies on more voltage to get more power across and to limit power loss in transmission.

Talking about transmission, do you use a short and very good quality cable? Like I said, Pi extremely picky about the voltage. Most USB wall charger are not even compatible with a Pi.

And minimal uF, I would say bare minimum is around 1000uF if you want it to do anything.

But did you actually plug the Pi in to the QC port and the rest in the other already?