Skip to main content

Inconspicuous Pitfalls in Datasheet Analysis

A chess grandmaster, who also happens to be an electronics hobbyist, decides to design his own chess clock. The LEDs, normally-open push-button switches, and microcontrollers are all in place, driven by a pair of 1.5V AA batteries. But the clocks only last for a few hours. So he checks online for a power management IC to extend the battery life, and finds one promising an efficiency of 94%. After purchasing the chip and ramming it onto a prototype, he becomes puzzled when the operating lifespan doesn’t change much. He then checks the wires and solders and connections again in a vain effort to find a solution.

Let us take this to the industry level. Apple is designing its 23rd Iphone. The company producing the processor for the Iphone23 is looking for a PMIC that can deliver a transient response of around 10mV with a 10A shifting load (Absurd!). They contract a semiconductor company to design the chip. After numerous tedious design reviews, the PMIC becomes ready. When the chips are tested in tandem with the processor, trouble ensues. The chip is oscillating at 8A! The processor is toggling between an on and off state. What happened here?

Certain measurement plots and point data in datasheets of ICs can mislead consumers into believing that that IC possesses a certain set of electrical characteristics that aren’t exactly 100% true due to the limitations and imperfections of the measurement procedures done in yielding those plots and data. In this article, I elaborate on those limitations and imperfections hoping to give the readers a deeper insight on the matter.

Regarding the latter example, such dilemmas occur less frequently due to the presence of field application engineers (F.A.E.s). They are there to specialize in the evaluation of the product, and if a problem occurs such as an oscillating output at 8A, they can analyse and debug on the problem.

However, in the absence of F.A.E.s, the designer faces the problem alone. With the problem being handled from a singular perspective, the underlying cause (usually a misunderstanding of the consumers required specifications) has a smaller probability of being properly addressed.

To aid the explanation of the points in question, let us do an analysis of the datasheets of a PMIC for processors of mobile applications, a motor driver, and a low power dual operational amplifier.

MC34708 Power Management IC for i.MX50/53 Families (Freescale Semiconductor)

The MC34708 is a PMIC designed for use with the Freescale i.MX50 and i.MX53 families. Typical applications include tablets, smart mobile devices, and portable navigation devices. It is lead free and available in 2 package sizes (do you still happen to stumble upon any commercial ICs that do not label their products lead free?). It offers 6 bucks, a boost, 8 LDOs, USB/UART/Audio switching for a mini-micro USB connector, a 10-bit Coulomb counter, an RTC, and of course, an SPI/I2C interface for external control. It doesn’t support battery charger management though.

Figure 1. Application block diagram of MC34708

Looking at the block diagram, we get a view of how the entire system is supposed to function. Now I’m not going to quote all the features of this PMIC (it’s available for public download anyway composed of 159 pages… Ugh) except for some plots and data.

Let us start with the thermal characteristics. Below is a table of the thermal protection thresholds of MC34708.

Figure 2. Thermal Protection Thresholds of MC34708

It says to set ADEN to 1 to enable thermal monitoring. Signals will be generated to identify whether the chip has exceeded 110, 120, 125, or 130 degrees. Upon crossing the 140 degrees threshold, the chip will shut down. Now, anyone who has had any experience with testing a chip’s thermal characteristics will tell you that doing so is not quite straightforward (this explains the minimum and maximum margins). First off is the kind of thermocouple used. Most semiconductor companies would use a type T thermocouple because the testing range usually falls in between -175 to 225 degrees. In fact, typical thermo streamers can only give around -70 to 200 degrees (at 300 degress, the heater can damage itself after prolonged exposure). Type T thermocouples have a tolerance of +/- 1 degree, which already tells a lot on the accuracy of the reading.

Another factor to consider is the soak time (the amount of time that elapsed after the chip has been exposed to the set temperature). In spite of the thermocouple indicating that the set temperature has been reached, I have experienced getting 2 different measurement readings at 2 different intervals after soaking a device-under-test (DUT), especially when it is a low voltage/low current measurement. Of course, the latter reading would be the more appropriate measurement, since higher soakage times mean a higher probability that the system has achieved steady state.

And yet another point of intricacy is the physical set-up. The chip is at 125 degrees, but are the adjacent external components such as inductors, capacitors and resistors also at that temperature? Is the thermocouple also placed directly above the DUT or a little bit above it? Did the evaluator use a thermostreamer or a hot air re-work station? Perhaps a soldering iron (in the extreme case)?

Now let us move on to the electrical specification. Fig. 3 shows a table of some of the electrical specifications of SW1, one of the buck regulators.

Figure 3. SW1A/B Electrical Specification

The turn-on time is pretty clear, from 50% of the enable signal to 90% of the end value. The switching frequency can be either 2 MHz or 4 MHz by toggling PLLX. The quiescent current consumption is at 240 uA at APS mode and 15 uA at PFM mode. This is a low current measurement, and is prone to errors by environmental factors and instrument setting (remember to set the ammeter to the lowest range). Note that at PFM mode the quiescent current is significantly lower with a 0mA load (because PFM is best used with low loads). The presentation of efficiency in this datasheet is unique, since it does so in tabular form. Usually, efficiency is expressed in curves. Given the point data, we don’t know whether the efficiency would dip between a 1 mA and 200 mA load (which most probably wouldn’t if you’ve seen other efficiency curves). Also, efficiency is more sensitive than quiescent current consumption because the measurement is dependent both on the very low off current of SW1 and the actual input voltage, output voltage, output current, and ambient temperature (yes, just a slight change in Ta can ruin the entire efficiency measurement, requiring re-measurement of the off current).

L293 Quadruple Half-H Drivers (Texas Instruments)

The L293 is designed to drive inductive loads as well as other high current and high voltage loads.

Figure 4. L293 Motor Driver Block Diagram

It features a wide supply voltage range and separate input logic supply. It also includes an internal ESD and TSD protection. It drives a typical output current of 1A per channel up to a maximum of 2A. The output clamp diodes for inductive transient suppression are visible in the block diagram.

Now let us focus on the switching characteristics. Fig. 5 shows a table of the typical delay and transition times of L293 using a PDIP package. 

Figure 5. Switching characteristics of L293 for a PDIP package.

One time there was a design engineer from XY company who received a customer complaint over a motor driver product. “Darn it, what IC did you give me? This isn’t the propagation delay time I asked for!” cried the indignant costumer. And so the design engineer humbly countered. “But isn’t this value that we agreed upon?” “Yes, it is. But that is not what I’m getting here on my end. In fact, I’m getting a much shorter delay time!” replied the costumer. After a few bouts, both parties discovered that they had different definitions of the delay time. The designer understood it as a point from 50% to 90% or 10% while the customer understood it as 50% to 50%. This resulted in the design engineer constructing a circuit with a propagation delay much longer than what the customer envisioned.

Looking at L293, one good question to ask is what reference levels were used to define that 800 ns and 400 ns low-to-high and high-to-low propagation delay times. Closer inspection of the datasheet would show that the propagation delay was measured from 50% to 50%. This is useful information that shouldn’t be ignored and that could save the consumer from the pitfall described by the scenario above.

LMx58-N Low Power, Dual-Operational Amplifiers

The LMx58 is composed of 2 operational amplifiers. It is internally frequency-compensated for unity gain with large DC voltage gain of 100dB and 1MHz bandwidth. It features a low input offset voltage of 2mV and its input bias current is temperature compensated. I think the block diagram is straightforward (2 op-amps) so I won’t be including it here.

Figure 6. Input current vs. temperature

The plot above (Fig. 6) shows a temperature sweep while simultaneously measuring the input current. It isn’t feasible to sweep the temperature over all voltages, so if a problem occurs, say at hot temperatures and at a voltage of 10V, the trouble-shooter shouldn’t exclude the possibility of the input current going awry since the characteristic isn’t actually evaluated at that condition.

Figure 7. Voltage follower pulse response.

Looking at Fig. 7, there is no definition of the rise and fall times used for the input voltage pulse. If the consumer were to use different rise times or fall times, there is a chance that the output pulse would have a different rise and fall time other than the observed 5 us. On the other hand, I don’t know the sensitivity of the circuit to a varying rise and fall time so it could be negligible.


Identifying the limitations of a datasheet saves a lot of time in troubleshooting and re-designing circuits. It is also economical due to the expenses avoided in re-manufacturing a product. At the same time, an in depth analysis of the datasheet gives those assigned in product support a better chance in identifying the source of the conflict as well as equipping them with a firm understanding of the product’s behaviour.


Popular posts from this blog

Calculator Techniques for the Casio FX-991ES and FX-991EX Unraveled

In solving engineering problems, one may not have the luxury of time. Most situations demand immediate results. The price of falling behind schedule is costly and demeaning to one's reputation. Therefore, every bit of precaution must be taken to expedite calculations. The following introduces methods to tackle these problems speedily using a Casio calculator FX-991ES and FX-991EX.

►For algebraic problems where you need to find the exact value of a dependent or independent variable, just use the CALC or [ES] Mode 5 functions or [EX] MENU A functions.

►For definite differentiation and integration problems, simply use the d/dx and integral operators in the COMP mode.

►For models that follow the differential equation: dP/dx=kt and models that follow a geometric function(i.e. A*B^x).

-Simply go to Mode 3 (STAT) (5)      e^x
-For geometric functions Mode 3 (STAT) 6 A*B^x
-(Why? Because the solution to the D.E. dP/dx=kt is an exponential function e^x.
When we know the boundary con…

Yay or Nay? A Closer Look at AnDapt’s PMIC On-Demand Technology

Innovations on making product features customizable are recently gaining popularity. Take Andapt for example, a fabless start-up that unveiled its Multi-Rail Power Platform technology for On-Demand PMIC applications a few months back. (read all about it here: Will PMIC On-Demand Replace Catalog Power Devices?) Their online platform, WebAmp, enables the consumer to configure the PMIC based on desired specifications. Fortunately, I got a hands-on experience during the trial period (without the physical board (AmP8DB1) or adaptor (AmpLink)). In my opinion, their GUI is friendly but it lacks a verification method for tuning (i.e. the entered combination of specs). How would we know if it will perform as expected or if there are contradicting indications that yield queer behavior? Also, there is not just one IP available, but many that cater to a differing number of channels and voltage requirements (each with their own price tag).
Every new emerging technology has the potential to oversh…

Common Difficulties and Mishaps in 6.004 Computation Structures (by MITx)

May 6, 2018
VLSI Project: The Beta Layout [help needed]Current Tasks: ►Complete 32-bit ALU layout [unpipelined] in a 3-metal-layer C5 process. ►Extend Excel VBA macro to generate code for sequential instructions (machine language to actual electrical signals).
Current Obstacles/Unresolved Decisions:
►Use of complementary CMOS or pass transistor logic (do both? time expensive, will depend on sched.
►Adder selection: Brent-Kung; Kogge Stone; Ladner Fischer (brent takes up most space but seems to be fastest, consider fan-out) [do all? time expensive, will depend on sched.)
►layout requirements and DRC errors

Please leave a comment on the post below for advise. Any help is highly appreciated.