High frequency noise from DC-DC Step up!

Javalin

Hello All,

Im working on a GPS project that uses a single cell LIPO (so 3.0-4.2v range) and a 5v LCD - so am using a MAX856 DC-DC Step up IC.

From the IC I get slightly higher than 5v - about 5.4v - and a high frequency noise - the LCD works ok though which is good!

On the picoScope·:-)·(www.picotech.com) I get the attached waveform.· Something is not right!

I've also attached the relevant part of the circuit· Currently the SHDN pin is held HIGH, C11 is 30uf, C7 is 20uf·and L1 is a 22UH axial Inductor.

Thanks for any help!

James

PS - also attached the MAX856 datasheet.

Paul Baker

Yup, seen that before, it's switching noise. Leading causes of this is improper layout and/or using an antiquated chip. It appears you chose an old chip (datasheet's last revision was 1996). My advice is to pick a modern switching chip which has recommend layout in the datasheet. I've had good results with Linear.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Javalin

ok thanks Paul.

J

Javalin

Oh - is there anything I can do with the current one? Change C7,C11 values, change Inductance value, Diode?

James

Ale

Assuming you have a good output capacitor and you add a 100 nF and a 10 nF in parallel... not much. Make some shots with your osci to see how those spikes look like.

Edit: Are you using an inductor designed for power switching ? Some inductors, especially the ones that look like resistors "Drosselspule" have very small cores not suitable for almost anything like this.

Javalin

Hi Ale,

Thanks for the reply! The Inductor is this one - http://uk.farnell.com/516533/passives/product.us0?sku=epcos-b82144a2223k&_requestid=465419

So maybe not - the product page text says:

INDUCTOR, AXIAL, 22UH
Inductor type:High Frequency / RF
Inductance:22µH
Tolerance, inductance:±10%
Resistance:0.35ohm
Current, DC max:1100mA
Frequency, resonant:12MHz
Case style:Axial
Q factor:50
Material, core:Ferrite
Tolerance, +:10%
Tolerance, -:10%
Climatic category:55/125/56
Diameter, body:5.2mm
Diameter, lead:0.63mm
Frequency, Q factor:2.52MHz
Frequency, inductance measurement:0.1MHz

I was thinking of the capacitance though - as my board has a design fault for that - i.e. the spec says 68uf and I have 30uf!! I'll probably add more!

James

Ale

Well, those are the inductors I said were not for that application, and the datasheet recommends a 47 uH, for light loads maybe a 22uH will work, but get yourself an inductor for power switching,

Something like that:

http://uk.farnell.com/7429878/passives/product.us0?sku=epcos-b82462a4473k

And fix the capacitor problem . Just one 68uF or 100uF will do, with the decoupling I suggested before.

Have fun

Peter Jakacki

James,

As mentioned the inductor you had is definitely not suitable as it even if the inductance is somewhere in the correct range it does not have the "storage" that a larger core has and so it saturates. A plastic food bag may hold a litre of water but it won't hold a litre of wet cement (that argument should hold water).

However, your signals show the output going to 0V which seems highly improbably as the 1N5817 schottky diode and output capacitor will prevent that from happening. I hope C7 is the right value and that you have actually wired this circuit the way it is drawn. The 4V steady level you are seeing is probably the battery voltage less a few hundred millivolts drop through the diode. So I get the feeling that you have taken the signal from before the diode rather than after and this is exactly what you would see.

*Peter*

Javalin

Thanks Peter - food for thought. I took the o'scope readings from between the inductor and diode - so that explains that.

Thinking about it my prototype was on a parallax carrier board - so would have had a large(er) decoupler on the linear voltage reg output. So definatly one to fix - cause I didn't have this problem then.

I shall look for an axial/radial package power inductor too.

James

Erik Friesen

Most of the datasheets for these step up converters are adament that you use a low esr ceramic.