Frequently Asked Questions

All our AEMs (PMICs) share the same purpose: manage the power from an harvester to a storage and/or a load.

However, their coldstart, MPPT ratio and timing might be different to be adapted to the source they should harvest from.  For more details, please see  “What is the MPPT ratio?” question.

function mpptRatio() {
const myButton = document.querySelector(‘[data-target=”#distributors-item1-5″]’);
const myDiv = document.getElementById(“distributors-item1-5”);
myButton.classList.remove(“collapsed”);
myButton.setAttribute(“aria-expanded”, “true”);
myDiv.classList.add(“show”);
myDiv.scrollIntoView({ behavior: “smooth”, block: “start”, inline: “nearest” });
}

To choose wisely your PMIC, please keep in mind it must be adapted to your harvester and its behavior.

Explore e-peas’ PMIC Selector Guide for more detailed comparison of PMICs.

When working with energy harvesting, the most important step is the power budget analysis.

Indeed, to build a viable system, enought power must be harvested to supply the target application.

Here below, we propose to study the power budget for choicing the adequate elements.

This study is composed of steps to define the parameters useful in the storage and source choice.

Those steps have no real order; and different ways for using them exist.

• On ways is starting from the load consumption to define the source and the storage element size.
• Another way is starting from the source to estimate the available left power for the load and the storage element size.

Please note that below, TS represents the time with available power at the source; and TNS represents the time without power available at the source.

If starting from load consumption :

• STEP EC : Evaluate the energy consumed EC by the load during a repeatable period (a day, a week, a month, …). This period must be chosen to include all TS and TNS time.   For example, one day is a typical period for outdoor solar application since light is coming back every day.  A week could be adapted for indoor application without light during the week-end.

The formula below includes 2 different active current peak [A1 and A2] and passive current [P].  Units below are A for current, V for voltage and s for the time.

EC = (IA1 x VA1 x TA1 + IA2 x VA2 x TA2 ) + ( IP x VP x TP )  [ J ]

Active consumption is mostly due to radio communication or processing.  Passive consumption is the sleep power or any constant consumption required by the load.

• STEP E BOOST  : Calculate the energy to be harvested EBOOST in order to supply the load.

The formula below includes converter used between the storage and the load. If using internal LDO converters, their efficiency is provided in the AEM datasheet.  For external converter, the efficiency should be given in its datasheet.

Please note that the internal leakage of the storage should be taken into account.

​​E BOOST = EC / n LDO (or n DCDC) + E leak    [ J ]

• STEP E STORE : Calculate the required energy E STORE to be stored.  Based on the autonomy and the load consumption, size of the storage must be define.

The formula below implies an autonomy egals to TNS – time with no power available at the source.

E STORE = ( TNS / TNS + TS ) x E BOOST​​​​    [ J ]

For longer/smaller autonomy, the ratio R must be adapted in order to store the energy consumed during the autonomy defined.

​E STORE = R  x E BOOST​​​​      [ J ]

If the autonomy required is 2 days, the ratio R is defined as 2*864000 / (TNS+TS) = 2 days / 1 day = 2

-> See FAQ about storage size : How to estimate the storage element size ?

function estimateElementSize() {
const myButton = document.querySelector(‘[data-target=”#distributors-item1-15″]’);
const myDiv = document.getElementById(“distributors-item1-15”);
myButton.classList.remove(“collapsed”);
myButton.setAttribute(“aria-expanded”, “true”);
myDiv.classList.add(“show”);
myDiv.scrollIntoView({ behavior: “smooth”, block: “start”, inline: “nearest” });
}

• STEP P SRC : estimate the P SRC source power required to supply the load.  Based on the time with available power at the source, the efficiency on the boost; the required power from the source is calculated.

Please note that, if longer autonomy is required, the additional energy to be stored must be harvested first.

Here below, we will define E BOOST* as the energy to boost for the autonomy required which is egal to E BOOST + E BOOST additional.

This E BOOST additional  is defined as E BOOST * Z with Z = add time / (TNS+TS).

If the additional time is one day, E BOOST* = 2 x E BOOST.

P SRC = E BOOST / (TS x n Boost )  [ W ]

For RF system, the efficiency over the RF path must be include in that step to estimate the power to be received.

P RECEIVED = E BOOST / (TS x n RF global)

This n RF global efficiency is provided in the AEM30940 datasheet for the e-peas solution.  For custom design (external rectifier with associated matching network), this efficiency must be characterized.

If starting from source power :

In that case, the first is to estimate the available power at the source.

This information is given in the harvester’s datasheet for solar, thermal or vibration harvester.

For RF source, the losses in the air must be evaluated to estimate the received power P RECEIVED based on the emitted power.

P RECEIVED = P EMITTED – Losses air

• STEP E BOOST  : Calculate the energy  harvested EBOOST from the source.

E BOOST = P RECEIVED x TS x n RF global    [ J ]

E BOOST = P SRC x TS x n Boost    [ J ]

• STEP EC : Evaluate the available energy for the load based on the energy harvested.

EC = E BOOST x nLDO – Eleak  [ J ]

The size of the battery is estimate based on E STORE = E BOOST.

The maximum input current of the AEM is 110 mA and the maximum input voltage is 5V.

P_max @ input voltage 5 V = 550mW

P_max @ input voltage 3 V = 330 mW

Yes, but some caracteristics are important for the global perfomances.

Indeed, the AEM boost efficiency depends on the the inductor associated. The one recommanded in the datasheet are the ones showing the best performances. If LBOOST inductor has to be replaced, please choose the one with the lowest Rdc resistor and the highest peak saturation current.

The actual recommanded inductors present a Rdc resistor around 300 mOhm and a saturation current around 600 mA.

We can propose those inductors from Wurth Electronik : 744 040 321 00 and 744 383 361 00.

We can propose those inductors from Wurth Electronik : 744 040 321 00 and 744 383 361 00.

We also identify those ones but we did not test them :

https://product.tdk.com/info/en/catalog/datasheets/inductor_commercial_power_vls201612hbx-1_en.pdf

https://www.digikey.com/product-detail/en/taiyo-yuden/MDKK2020T100MM/587-4093-1-ND/5035253

http://www.bourns.com/docs/Product-Datasheets/SRN2009T.pdf

Details are provided page 12 in the datasheet.

For any harvester, the power curve goes first up, achieves his maximum power point and goes down with the voltage until reaching the open-circuit voltage – Voc. Looking at this power curve, we can see that a maximum power point exists at a certain voltage called Maximum Power Point Voltage – V MPP. This MPP Voltage is a certain percent of the open-circuit voltage – Voc. This ratio varies depending on the harvester :  MPPT ratio = VMPP / Voc.

To resume, it defines the best voltage to work with in order to harvest the maximum power out of the harvester.

1. For a PV cell, the ratio varies from 70% to 90% depending on the PV cell material.
2. For a piezoelectric generator, the ratio depends on the power curve of the piezo including the rectifier.
3. For a thermoelectric generator, the ratio is most of the time around 50%.
4. For RF, using the integrated rectifier in the AEM30940, the optimal ratio is around 65%.
5. For the RF, using an external rectifier with the AEM30940, the optimal ratio of our design is around 50%.

For any harvester, it might be interesting to draw its power curve in the targeted conditions.

Note that the P-V curve might vary under different luminosity for PV cell for example, but the ratio will remain the same.

The AEM10941 performs a Voc measure each 5 s during 81 ms to adapt the VMPP voltage to work at.  The ratio is defined by the SELMPP0-1 pin-configuration (See page 12 in the datasheet).

The AEM20940 performs a Voc measure each 21 s during 330 ms to adapt the VMPP voltage to work at.  The ratio is defined by the SELMPP0-1 pin-configuration (See page 12 in the datasheet).

The AEM30940 performs a Voc measure each 0.33 s during 5 ms to adapt the VMPP voltage to work at.  The ratio is defined by the SELMPP0-1 pin-configuration (See page 12 in the datasheet).

The cold start is the very first phase of energy harvesting. This phase happens when all the internal nodes are discharged.

The AEM uses a specific architecture to cold start at 380mV with 3µW.  Coldstart can be re-worked with an additional (internal or external) circuit.

The DEEP SLEEP MODE is the state where all nodes are discharged and no energy is available at the source.

As soon as the required cold-start voltage of 380 mV and a sparse amount of power of just 3 µW becomes available on SRC , the WAKE UP MODE is activated. Vbuck node rises up to a voltage of 2.2 V.

Vbuck gets regulated around 2.2 V and Vbatt starts rising up to Vovch (starting from the storage element voltage connected to BATT).

When looking for solar or photovoltaic harvesters, please consider the different information below.

1. Is your environment suitable ?

To harvest energy from a solar cell, you need light. That light can be the outdoor light or an indoor light.

• Using outdoor light, the harvetster is often called “solar cell”.   Power from the source is available around 8 hours per day (depending on your location and the season.  Please find more information on SoDa), implying that your system has to be supplied by your storage element during 16 hours.
• Using indoor light, the harvester is often called “photovoltaic cell”.  Note that new generations are adapted to LED light or any typical indoor light.   It is necessary to estimate the time with available light in your working conditions.  Be aware that your solar cell can be shadowed by a human, a tree, a cabinet and even dust ! The angle is important for technologies using silicium, while new organic technologies do not depend on the angle.

2. Do you have mechanical constraints ? 

Depending on your mechanical integration, space allowed to solar/ photovoltaic cell can be an issue.  Different types of harvester are available on the market: with integrated protection, rigid, flexible, printable (giving you the possibility to shape your photovoltaic cell )…

Do consider the environment constraints in your choice : the level of light available (can be measured with any luxmeter), dust, angle, protection, space allowed, shape… But keep in mind there are a lot of new technology and solutions on the market every day !

3. Power budget

Choosing the harvester starts with power budget study as described in the “How to analyse the power budget “ question.

function analyzePowerBudget() {
const myButton = document.querySelector(‘[data-target=”#distributors-item1-2″]’);
const myDiv = document.getElementById(“distributors-item1-2”);
myButton.classList.remove(“collapsed”);
myButton.setAttribute(“aria-expanded”, “true”);
myDiv.classList.add(“show”);
myDiv.scrollIntoView({ behavior: “smooth”, block: “start”, inline: “nearest” });
}

When selecting the PV cell, you should verify that the open-circuit voltage is not above 5 V and the maximum current is below 110 mA. A higher voltage or current can damage the input stage of the AEM.

A common open-circuit voltage for a cell is around 0.7 V. With that voltage, 7 cells in series is the maximum because it will reach 4.9 V of open-circuit voltage.

If no other options are available, it is always possible to protect the AEM with a clamping diode at 5 V.  Select one with a small leakage current below 5V.

To harvest the maximum from the PV cell, it is important to choose wisely the MPP ratio. This ratio might be given in the datasheet or can be analyzed.  If you want to analyze it, draw the I-V curve and then the P-V curve to see where the maximum power point is located.

* Note that the number 7 announced in the AEM10941 datasheet title comes from the maximum open-circuit voltage. Indeed, a common open-circuit voltage for a cell is around 0.7 V; and 7 cells represents an open-circuit voltage of 4.9 V.

Choosing the receiver antenna will be correlated to the choice of the emitter antenna. Indeed, they will have to work together (transferring power) and the transfer can be improved with gain and directivity/ polarization.

• The emitted power depends on the emitter antenna manufacturing and type (isotrope or directive). Multiple antennas can also be used to increase the emitted power. Please refer to RF expert for antenna design.

The effective radiated power ERP or effective isotropic radiated power EIRP is one important parameter to look at in order to choose the antenna.  This power depends on the gain dBi – normally given in the datasheet of the antenna – and on the radiation pattern of the antenna – how much power is radiated in the desired direction.

• Receiver antenna also might add some gain and the intrinsic gain should be given in the datasheet [dBi].  Be careful to the setup used to measure this gain.
• The environment can influence the power received since it can contain obstacles.  Walls, window, objects can return or deflect the signal.
• Losses on the PCB depend on the square of the frequency and losses in the air depends of the frequency and the distance : Working at lower frequency reduces those losses.

Matching and rectifier required between the receiver antenna and the AEM are proposed by e-peas (under NDA).  Please note that this only proposition and you are free to re-design it your self.

e-peas also proposes design and support (under quotation) for new matching design.

For higher current peak to supply the load, we can propose to you to use an external BUCK/BOOST or LDO converter controlled by the AEM.

In that case, the internal LDOs may be turned OFF and the STATUS0 signal is used to enable the external converter.  With that solution, the battery protection is still performed by the AEM and the external converter will be disabled if the voltage battery is going below the Vovdis level.

Depending on the external converter characteristics and on the type of storage element, the protection levels might either be defined by the pre-configured modes or by the custom mode.

We also recommend to place a PMOS between the battery and the external converter in order to really disconnect the converter from the battery, as implemented inside the AEM.  This PMOS is controlled by the opposite signal of STATUS0. This way, the external converter won’t draw current from the battery even if disabled.

For the ZMPP configuration,

1. You need to set SELMPP[1] and SELMPP[0] high ( connect to Vbuck/aux)
2. Connect a resistor to ZMPP pin.

The AEM will connect this impedance to the harvester when evaluating an MPP. The voltage is read and set as working voltage for the input. The AEM will behave as the defined impedance.

To measure or define it, please draw the P[W]-R[Ω] curve for the complete harvester.  This can be done with different loads or with a configurable load.

Vprim_min is defined as the voltage of the primary battery at which it has to be considered as empty.  The AEM won’t use the primary battery as a source if its voltage goes below Vprim_min.

The primary battery is a non-rechargeable battery used as a safety margin.

Depending on your system and on its safety level, you need to choose this primary battery size. It is a backup solution if not enough power is coming from the ambient source.

An example of size could be the minimum one to send a message if the system is on the primary battery too long.  Keep in mind that your MCU will know when it is on primary battery with STATUS[1] asserted if primary battery is not discharged. If the primary battery is discharged and not enough power is coming from the source, your application will shutdown.

Help is provided in the “configuration tool” for each AEM family product on e-peas website.

Depending on the chip, the AEM will measure the MPPT every 0,33s, 5s, or 21s. When the measurement happens, STATUS[2] is asserted and the AEM is going to open the circuit of the harvester to measure the open-circuit voltage (Voc). After that measure, the AEM will optimize its boost control.

The BUCK efficiency is included in the LVOUT efficiency (voltage from internal VBUCK).

Leakage current announced integrates BUCK efficiency to supply the AEM.

When CFG[2:0] are tied to GND, the custom mode is selected and all six configuration resistors must be wired as follows:

Vovch, Vchrdy and, Vovdis are defined thanks to R1, R2, R3 and R4. The total resistors (R1 + R2 + R3 + R4) are defined as RT, R1, R2, R3 and R4, and are calculated as follows:
• 1 MΩ ≤ RT ≤ 100 MΩ
• R1=RT(1 V/Vovch)
• R2=RT(1 V/Vchrdy – 1 V/Vovch)
• R3=RT(1 V/Vovdis – 1 V/Vchrdy)
• R4=RT(1 – 1 V/Vovdis)

Vhv is defined thanks to R5 and R6. If we define the total resistor (R5 + R6) as RV, R5 and R6 are calculated as follows:

• 1 MΩ ≤ RV ≤ 40 MΩ
• R5=RV(1 V/Vhv)
• R6=RV(1 – 1 V/Vhv)

The resistors should have high values to make the additional power consumption negligible. Moreover, the following constraints must be adhered to ensure the functionality of the chip:

• Vchrdy + 0.05 V ≤ Vovch ≤ 4.5 V
• Vovdis + 0.05 V ≤ Vchrdy ≤ Vovch – 0.05 V
• 2.2 V ≤ Vovdis
• Vhv ≤ Vovdis – 0.3 V

Help is provided in the “configuration tool” for each AEM family product on e-peas website.

The estimation of minimum size differs with the storage type.  Indeed, capacitor and supercapacitor store the energy between the Vovch and Vovdis levels (best case) while battery stores the energy at constant voltage – called nominal voltage.

Find below the formulas to estimate the minimum supercapacitor size – Cmin –  and the minimum battery size – Batt Min.

Cmin = 2 x E STORE / (Vovch² – Vovdis²)   [ F ]

Batt min = E STORE / (Vnominal x 3600   [ mAh ]

Care must be taken with current peak since they can be critical for storage size definition.  Indeed, estimations above are based on average consumption.  However, some consumption profile can include high peak current and ultra-small passive current.   The storage element must be able to substain such a current peak cithout excessive voltage drop (ESR related for battery or voltage drop for a supercapacitor).

In that case, the type of converter used is of great importance.  Indeed, working with LDO implies working with current transfer (for better output stability).  Working with a DCDC switching converter implies working with energy transfer.

See belows the theoretical formulas for LDOs and switching converters with Ip the current peak in A, Tp the peak duration in second and Vp the operative voltage. The formulas below estimate the minimum size for one current peak.

Cmin LDO = Ip1 x Tp1 / (Vovch – Vovdis)   [ F ]

Cmin SW = 2 x Ip1 x TP1 x VP1 / (Vovch² – Vovdis²)   [ F ]

If storage must substain more than one current peak, formulas should be adapted.

WiFi emission is limited +20 dBm by regulations.  This power is most of the time not enough to be harvested.

Indeed, the power that RF loses in the air and over the RF path on the PCB is significant.  Please note that those losses are related to the frequency and will be smaller with lower frequencies.

In order to control the emitted power, the antenna used and the duty cycle of emission, we recommend you to use a dedicated emitter.

From the 100 mW available (emitted), 237µW are received after 20 cm.

If using internal LDO (~ 89 % efficient), around 37 µW is available to supply a load.

This power can be bufferized and used as DC but its amount can limit the application.

Yes any storage element can be connected.

However, please note that the AEM won’t harvest energy to recharge this primary battery.  This primary storage must be recharged externally.

When Vbatt drops below Vovdis and no power is available from a primary battery, the circuit enters SHUTDOWN MODE to prevent deep discharge potentially leading to damage to the storage element and instability of the LDOs.

The circuit asserts STATUS[1] to warn the system that a shutdown will occur. Both LDO regulators remain enabled.

If no primary battery is used, this allows the load, whether it is powered on LVOUT or HVOUT, to be interrupted by the low-to-high transition of STATUS[1], and to take all appropriate actions before power shutdown.

Yes you can. You just have to take care of two things :

• Discharge all the nodes before changing configuration
• Set CFG[0-2] as required

Please refer to table 1 in the datasheet. 

ZMPPT, SET_OVDIS, SET_CHRDY, SET_OVCH, FB_HV must be left floating if not used.

STATUS can be left floating.

BATT can be left floating.

SELMPP, CFG, ENxV must be defined.

/!\ If you are using an evaluation board, do not leave any floating jumper. /!\