MP1584EN Shutdown / Sleep Current Not as Expected?

Hi,

I am hoping someone can help me work this problem out please!

I am using a MP1584EN DC-DC Converter Module in my project - specifically the one made by D-Sun as shown in the image below:

The reason I specifically use D-SUN is because they have the necessary components on board required to break out the EN pin so I can control it using a pin on my Atmega to shut the converter down when not being used. Other manufacturers omit these components and / or slightly re-design the board to save money but you can't tell unless you look closely.

The problem is, I had expected to see around 20uA being drawn while the MP1584EN converter is shutdown / sleeping - but instead it's taking around 94uA. Waaay more than I want.

I got the expected figure of 20uA from the datasheet here:
https://sc02.alicdn.com/kf/HTB1rWbsRVXXXXarXXXX760XFXXXt/MP1584EN-Step-down-power-module-DC-DC.png_350x350.png

This is a screen grab of the section I was looking at where it says Shutdown Supply Current Ven MAX 20uA

So am I looking at the wrong figure or am I doing something wrong?
As far as I am aware, all I should need to do is ground the EN pin to get it to fully shut down - and it does. There's no voltage coming from the output but the current draw is higher than expected for some reason.

To rule out anything else upsetting my measurements, I set up just the MP1584EN module on its own with no load. I measured around 210uA at idle and I measured 94uA when the EN pin was grounded to put the IC to sleep.

Have I read something wrong / done something wrong or is there something with the design of these modules that causes this problem and if so is there anything I can do to modify / adapt them?

If not then it's back to the drawing board because that's a massive amount of wasted energy where my project is concerned.

Any help really appreciated!

It is hard to see on picture, but there is a resistor on pin #2, try put ground = 0V directly on pin#2, if its not working you have a problem with ground or IC is not good quality, sometime tht hapents with not originals, Chinise replica.

ted:
It is hard to see on picture, but there is a resistor on pin #2, try put ground = 0V directly on pin#2, if its not working you have a problem with ground or IC is not good quality, sometime tht hapents with not originals, Chinise replica.

Hi, yes, you are correct.

Those resistors that are connected to the EN pin aren’t always there on other MP1584EN modules which is why I only use the ones made by D-SUN which do have all the components from the recommended MP1584EN circuit

And yes, you are correct again, EN is pin 2 on the MP1584EN. I have carefully soldered a fly wire on to this pin and have tried grounding it to 0V and while it does shut the converter down and there’s no voltage on the output, the current being used while the MP1584EN is sleeping is far too high.

But yes, you could be right. It might be down to cheap components.

The other week I purchased 10 more MP1584EN modules from ebay and at a glance, they looked correct but when they arrived they were not D-SUN ones and had empty pads where components should have been and also the MP1584EN IC had had all the markings washed / etched away. Needless to say I didn’t bother even trying them.

The ICs on the D-SUN MP1584EN modules are marked clearly as MP1584EN but that probably doesn’t mean much either. They’re probably reclaimed from scrap or remarked IC packages or anything these days.

Although I have to say these D-SUN ones do work well in every other way, it’s just the high current consumption at sleep. Otherwise they’re perfect for my project…

why you need sleep mode, can’t you use switch ?

ted:
why you need sleep mode, can't you use switch ?

Hi,

No, unfortunately a switch will not do in this case. I have tried relays, MOSFETS etc but all had drawbacks of one sort or another. The MOSFETS were too sensitive, the relays too power hungry.

Since the MP1584EN has a dedicated EN pin, it was decided the best option was to make use of it to put the MP1584EN to sleep at a cost of 20uA - or it should have been.

Everything in my project is either in sleep mode or switched off completely until needed to save as much power as possible to extend battery life and standby time.

The Atmega and 3V3 regulator main board I've designed only draws 6uA which is great but I need to get these MP1584EN converters down to their 20uA sleep specifications somehow...

Edited to Add:-

I have also just sent an Email to Monolithic (the makers of the MP1584EN) to see if they can offer any advice as well. I'm not sure how much they'll be able to help since the Buck converters are a third party device but it's at least worth asking. They might recognise the problem and offer some guidance. Or that's what I'm hoping.

F7204
I used long a time ago F7204, I don't remember the details, battery was not discharged wden dewice was not used forweeks. For sensitivity reduction I used pull up or pull down resistor, don't remember the value, try 10k or 1 meg.
Google - touch switch

Is it a case that EN floats, and a pullup resistor is used on the board to allow usage without any input on the EN pin? By grounding the EN pin, the chip will shut down, but at the cost of current through the pullup resistor.

I am talking about something, like this link shows.
You need only MOSFET and and resistor from that circuit, your sleep mode is now MOSFET.

http://www.seekic.com/circuit_diagram/Power_Supply_Circuit/DC_power_on___off_switch_circuit.html

tinman13kup:
Is it a case that EN floats, and a pullup resistor is used on the board to allow usage without any input on the EN pin? By grounding the EN pin, the chip will shut down, but at the cost of current through the pullup resistor.

I think you might be on to something there tinman13kup.

This is the typical application circuit taken from the Datasheet:

And this is what the datasheet says about the EN pin:

Enable Control
The MP1584 has a dedicated enable control pin
(EN). With high enough input voltage, the chip
can be enabled and disabled by EN which has
positive logic. Its falling threshold is a precision
1.2V, and its rising threshold is 1.5V (300mV
higher).
When floating, EN is pulled up to about 3.0V by
an internal 1μA current source so it is enabled.
To pull it down, 1μA current capability is
needed.
When EN is pulled down below 1.2V, the chip is
put into the lowest shutdown current mode.
When EN is higher than zero but lower than its
rising threshold, the chip is still in shutdown
mode but the shutdown current increases
slightly.

Looking at the typical application circuit, there are two resistors which I have circled in red.
R5 which is 100K Pull Up and R6 which is 56K and connected to ground (pull down?). That configuration forms a voltage divider but I'm not quite sure of it's function as a whole.

I get that the 100K pull up is there to keep the EN pin at logic high but I'm not quite sure of the role of the 56K resistor.

The extract from the datasheet I have quoted above says that if the EN pin is left to float, it is pulled up internally. And it also quite critically says that if a pull up resistor is used that the sleep current will be higher. Which is basically what you have suggested tinman.

So from that, I am assuming I should remove both the 100K and 56K from the DC converter board to leave the EN to float freely.

I've got a converter I can test this theory out on so I'm off to give it a try. I'm too impatient to wait LOL

Thanks!!!

I have de-soldered R5 and R6 from the buck converter board and grounded the EN pin and the current consumption is now 12uA!!!

And to think, I was buying the D-SUN boards because they had these resistors on them.

Well, I think that has solved that problem. Thank you very much tinman and ted for all your help / suggestions.

I may still look in to updating my design to MOSFETs but for now, I'm happy with my current design to get me by until I have the time to do more experiments with MOSFETS.

Thanks again!!!!

Naw when I see the circuit it is obvious that R5 was causing the problem, that's why I mentioned 1 meg resistor.
Good luck, problem solved.

there is something wrong with it you should have 20 not 12 uA - LOL

You are welcome. Just make sure you meet the requirements for the chip itself in re to EN. Board mfg's rarely just throw extra components on to make them look cool, unless we are talking led's. Then bright and flashy sells.

ted:
there is something wrong with it you should have 20 not 12 uA - LOL

Hi Ted, thanks for the help & sorry for the confusion.

20uA was the MAX value given by the datasheet when the IC is disabled. 12uA is typical value. When using datasheet figure estimates, I always expect to see the MAX values quoted and anything less is a bonus.

tinman13kup:
You are welcome. Just make sure you meet the requirements for the chip itself in re to EN. Board mfg's rarely just throw extra components on to make them look cool, unless we are talking led's. Then bright and flashy sells.

Hi tinman,

Many thanks again for the help.

You will probably have noticed from the image I posted of these Buck Converters in my OP that none of the manufacturers of these Buck Converters give any kind of access to the EN Pin. Which is why I have to solder a fly wire on to the EN Pin to access it.

I think what D-SUN have done with their Buck Converters is to include the Pullup Resistor as per the typical application circuit to keep the IC permanently switched on.

Whereas, with the cheaper manufacturers, they have left out these resistors and instead rely on the internal pullup and leave the pin free to float.

I can't believe I didn't see the pullup resistor or take in what was being said about a higher current when sleeping if the pull up is present and I've been over that MP1584EN datasheet dozens of times. Obviously it just didn't register when I'd been reading it.

Still, no harm done and it got there in the end. Now I'm pleased to report my entire project draws a tad under 40uA at standby - which is very close to the predicted estimates for the design.

According to my calculations, that would give me well over 2,000 days of standby time on a pair of 18650 Cells in series (7.4V @ 2600mAH).

I was looking for around 6 months to a year standby life and I've well exceeded that. The batteries would probably die of natural causes and old age before I got anywhere near 2,000 days.

Thanks again guys. I really appreciated all the help. Sometimes it's easy to overlook the obvious starring you in the face and a fresh mind and pair of eyes ALWAYS saves the day!

Have a great day!

google -battery shelf life.
Post # 11 - you overlooked -LOL

Have a look at this:
DC Stepdown MP2307DN mit Arduino steuern