RoadTest: Become a Tester of the Toshiba e-Fuse IC Evaluation Kit EVB-TCKE805NA
Author: strb
Creation date:
Evaluation Type: Evaluation Boards
Did you receive all parts the manufacturer stated would be included in the package?: True
What other parts do you consider comparable to this product?: Comparable options are Littelfuse LS0505EVD22, MPS MP5017AGD, onsemi NIS6452, ST STELPD01 and TI tps25925.
What were the biggest problems encountered?: The evaluation process went smooth, I just found a bit strange that some parts were missing on the board, most noticeably the mos for the reverse current blocking. For current consumption, care is needed to take into account output diode leakage.
Detailed Review:
No matter if you are an expert designer or the final user, you’ve likely experienced a blown fuse without an obvious cause or, conversely, a device failure where the main fuse or protection didn’t trip.
In complex systems (think, for example, of a server with many hard drives) the failure of one device should not disrupt global operation. Fast-acting, precise protection devices are essential to isolate the faulty component without affecting the rest. Standard fuses, PTC or resettable fuses aren't always up to the task. In these situations e-fuses stand out, providing fast and repeatable protection along with additional features, all integrated into a small IC.
Before starting, I want to thank Toshiba and element14 for providing the opportunity to test and evaluate the EVB-TCKE805NA. If you have any questions, feel free to ask in the comment section below, I will do my best to answer.
Unboxing
The EVB-TCKE805NA came in a small white box. The only content anticipation is given by two labels, indicating various information. The QR code on the side is probably some sort of internal code or traceability and it does not include any link to the product page. Shacking the box, you can easily perceive something moving inside.
| {gallery}The box |
|---|
|
|
|
|
Opening the box, the first thing that appears is the evaluation board instruction manual (Japanese and English versions). It is printed on a greenish-tinted sheet that makes the bright red Toshiba logo pop up.
The instruction manual is brief yet comprehensive: it includes the board overview with some PCB layout guidelines, an introduction to the TCKE8 series of e-fuses, scheme and BOM, how to set up the board and the usual long list of disclaimers. The color-printed images and tables add a nice touch.
The EVB instruction manual is also available in digital format, as well as the datasheet for the TCKE805NA that will be tested here and a nice FAQ page to get everybody not familiar with e-fuses a quick start. The product page has also available documentation like reliability data, certificates and app notes. As a design engineer, I appreciate all this effort to have good documentation because it makes designing easier.
Under the manual, layers of bubble wrap protect a small ESD bag containing the evaluation board, which measures just 46mm x 66mm. The board comes without any connector but lets you choose between standard 2.54mm pitch connectors or direct wire soldering on the big copper pads left exposed for inputs and outputs. Passive components are relatively large, presumably to facilitate component swap for experimentation. Surprisingly to me, the "FET1", used for optional reverse current protection at shutdown, is not mounted. All test points and components are nicely labeled on the silkscreen.
| {gallery}Box opening |
|---|
|
|
|
|
|
|
Why choose an e-fuse and market comparison
There are many available protection possibilities, namely standard fuses (both replaceable glass fuses or SMD ones), resettable fuses and sometimes also PTCs are used. E-fuses are the latest entry in the market, but they can provide unique advantages over their older competitors. Toshiba in its application note resumes this very well in a table:
Fuses type comparison from Toshiba app note
While a single small fuse may be appealing from the initial cost perspective, it requires replacement each time it trips, making it less suitable for long-term maintenance. Standard fuses also lack repeatability and require large margins to be reliable under normal operation. Resettable fuses can partly address some of these issues, but they are still sub-optimal in terms of intervention precision and time. None of these alternatives offer the additional features of an e-fuse IC, like inrush current limiting, overvoltage protection and so on.
Given the clear advantages that e-fuses IC have over "traditional" protection, how does the TCKE8 series compare to what it's available in the market? I looked for 5V compatible, 5A rated parts from different manufacturers and made a brief comparison. It is important to make clear that this is not a complete comparison, as there are more options available and more IC specific features for every product, but it is almost impossible to compare them all in a single table. Here, I reported what I value as "main features".
Comparison table between TCKE805NA and other e-fuses IC
Starting with Ron, the TCKE805NA is the strongest of all, slightly edging the TPS25925 and providing a fair advantage over all other competitors.
The input voltage range is a bit controversial to compare, and that's the cause for all those asteriscs. The TCKE805NA provides up to 18V on the recommended operating condition, and that matches the absolute maximum rating. Ranges reported with "*" refer to absolute maximum ratings, with a recommended operating condition much lower than that. On this topic, the Toshiba TCKE805NA is the only one that has a programmable UVLO that can accept the full input voltage range. TI e-fuse is the only other with programmable UVLO, but this input has to be limited to max 7V. This proves to be an advantage for the TCKE805NA, as for always on operation, the EN pin can be directly tied to Vin without any additional component.
All of the candidates have their fair set of protection and additional features, with a slight twist from one to another. Here I would not say "this is better than that" in a specific way, as every IC is probably optimized for a slightly different use case scenario.
As far as price goes (evaluated for 1k lot), it is only a reference obtained using an online tool. Prices can vary a lot from supplier to supplier, depending on availability and so on. Based on my research, the TCKE805NA is placed exactly in the same price range as most of the other e-fuses. The only outlier here is the Littelfuse option, appearing cheaper at the time of my search.
Last but not least the TCKE805NA shows compliance with the latest IEC62368-1 standard for ICT and AV applications. The only other candidate that mentions some kind of safety standard is the TI proposal, indicating compliance with the old 60950 standard.
Overall, the TCKE805NA places itself quite well in the market, with an aligned price and a good set of features, showing unique advantages over competitors.
If the TCKE805NA does not suit your needs, there are other options available within the same series under development, with slightly different specifications.
Options table from the EVB instruction manual
Board preparation
Before testing, I prepared the board by soldering standard 2.54mm pin strips on voltage inputs and to the EN/UVLO selector, connecting some wires at outputs and shorting out R1. This setup takes advantage of the EN pin compliance up to 18V, allowing me to have an automatic turn on of the device when an input voltage is applied.
Since FET1 is unpopulated, I mounted one mosfet to later test the reverse current blocking function.
I decided to leave unpopulated input and output sense test points.
Board testing
I had to pay attention while connecting the board because, as you may have already noticed, it has inputs on the right and outputs on the left, the opposite compared to what I'm used to. Nothing to worry about, just needed a bit of extra attention.
To start, I connected my function generator to supply the board with a 0V-10V triangular waveform and monitored it with my scope input and output voltage. Except for some noise at the turn on point (probably due to the output impedance of my signal generator) I got exactly what I expected: output is disabled until the turn on threshold is reached, then output tracks the input up to about 6V, then the OVP kicks in and regulates the output voltage regardless of the input. Same story decreasing the voltage. Comparing up and down slope we can see the hysteresis on the UVLO.
ch1: input voltage; ch2: output voltage
I thought it would be cool to have the hysteresis plotted in X-Y mode, so I tried. Unfortunately, my scope has some limitations on this aspect and the resulting curve is not great. We can certainly see that there is a hysteresis, but not much more.
Half failed attempt at plotting hysteresis curve using X-Y mode
Switching to a normal bench supply to provide power to the board, I put the jumper in the "EN" position and enabled it with an external signal. By the way, the EN pin recognizes a high value with voltages just over 1.2V, so it is compatible with 3.3V, 5V and even 1.8V logic levels.
Turn on happens about 350us after the control signal is high and it's visible that the output voltage is enabled in a controlled way, showing a nice ramp from 0V up to 5V. With the default 120pF mounted on the board, the turn on has a duration of about 300us but it can be easily programmed by changing the capacitor connected to this pin.
| {gallery}Turn on |
|---|
|
|
|
|
At turn off (without any load) we can appreciate the auto discharge function. Measuring the time it takes to reach 5% of the input voltage and knowing the output capacitance (1uF on the board itself), we can estimate an output discharge resistor in the 4kΩ range. We also get a similar behavior when the UVLO enters into action.
Output discharge function. Ch1: enable signal; ch2: output voltage
One thing to be aware of is that this discharge function does not work on the "Vout_B" with the reverse blocking mosfet, because the mos itself is shut off before the output is completely discharged.
Vout_B discharge not completed during UVLO event (no load). Ch1: output voltage
Turn on with an overload condition (1.25Ω load resistor) is no difference in terms of turn on waveform cleanliness, but the output voltage does not reach the input because the TCKE805NA is limiting the current to 3A as set by the Rilim resistor.
Turn on with excessive load. Ch1: enable signal; ch2: output voltage; ch3: output current
Leaving the TCKE805NA on in this condition for long enough, it enters overtemperature protection and cycles between on and off state until the fault is removed.
OTP. Ch2: output voltage; ch3: output current
In case of a short circuit like failure, the TCKE805NA does its job very well. I tested it using 1A load as steady state load and a TIP122 transistor controlled by a function generator to create a high current pulse. I placed the current probe only on the short side to measure the current pulse. I did not use a mosfet to create a faster, sharper current pulse because my current probe has a bandwidth limited to 800kHz, so I would not be able to display extremely fast transitions.
The TCKE805NA responded rapidly and precisely. The total pulse duration from start to when it drops back to zero is about 4.5us, which means that intervention time is even faster.
Setup to test the short circuit behavior
Short circuit pulse behavior. Ch2: output voltage; ch3: short circuit current; ch4: control signal
Directly shorting the output for a longer period makes as first step trip the fast protection and after a short period, it restarts entering in regulation until the OTP trips or the fault is removed.
Short circuit behavior with retry. Ch2: output voltage; ch3: output current
Comparing the TCKE805NA to a 3A rated fuse is not even fair. Not only the fuse allowed to circulate a higher current, but it was not able to interrupt the circuit with a fault duration up to 2ms.
| {gallery}Fuse cc |
|---|
|
|
|
|
Short circuit behavior with a 3A rated fuse. Please ignore the current oscillation on the second image, it is probably due to my power supply reaching its limit. Ch3: fuse current; ch4: control signal
On this comparison, I suggest you to have a look at the Toshiba application note, they covered this topic extremely well, showing an even larger performance gap compared to what I'm doing here.
Moving on to additional features, I tested the reverse current blocking function with the external mosfet. During normal operation, the current is allowed to flow both ways (like a normal fuse) but when shut off with the EN signal, the input and output path is interrupted. Leveraging the external mosfet, the current can be cut off regardless of polarity.
Reverse current shut off. Ch1: Vin; ch2: Vout; ch3: current; ch4: EN signal
The response is fast and the current is shut off within 2us from the EN command. In this specific case, the output voltage rises to 7V due to the current generator forcing the output node higher than the input.
Measuring Ron was straightforward: I loaded the IC with 1A and measured the voltage drop between Vin and Vout using the provided sense points. The multimeter showed 27.6mV, corresponding exactly to the 28mΩ typical specification provided in the datasheet.
Voltage drop across the TCKE805NA with 1A load
Especially for battery powered devices, current consumption can be a big concern, so it's important to keep it low. When off, the TCKE805NA does not disappoint, requiring only 35uA.
Measuring the on state power consumption initially provided some difficulties because I was consistently getting higher than expected values. Changing Rilim seemed to have an impact (a higher resistor lowered the current consumption), but still, all readings were higher than expected.
| {gallery}Initial current consumption test |
|---|
|
Rilim at 35kΩ |
|
Rilim at 100kΩ |
|
Rilim at 220kΩ |
Current consumption with Rilim at 35kΩ, 100kΩ and 220kΩ
I double and triple checked my measuring setup, but nothing seemed wrong. I tried disconnecting the external mosfet but nothing changed. It took me a bit but in the end, while writing the introduction for this RoadTest with the EVB instruction manual under my eyes, came into my mind that the EVB features a low voltage fast schottky to protect the TCKE8005NA from parasitic inductances current spikes when the protection activates. Schottky diodes usually have relatively high leakage, even under modest reverse voltages, so I removed it and repeated once again the measure.
Current consumption with Rilim 220kΩ and without the output schottky diode
This final measure showed what I expected, a current consumption of 430uA with Rilim of 220kΩ. The datasheet states 480uA with Rilim set at 120kΩ, so I'm finally aligned with the expected value!
Conclusions
In the end, the Toshiba TCKE805NA performed great, showing fast and precise protection response, confirming the lowest Ron in his category of e-fuses and completing with a nice set of additional features. It also offers compliance with IEC62368-1, which can streamline certification processes. If you’re looking for an e-fuse with solid performance, the TCKE805NA is an excellent option.
Product performed to expectation: 5/5. The TCKE805NA performed as expected on all tests, providing fast, precise and reliable protection.
Specifications were sufficient to design with: 5/5. Datasheet is comprehensive with all key informations.
Demo software was of good quality: 5/5. Not applicable for this RoadTest.
Product was easy to use: 4.5/5. The board is easy to use with multiple options to connect input and output, sense points are provided to evaluate the internal mos resistance and passives are big to make components swap easier. My only minor complaint is the lack of the external mosfet for reverse current protection.
Support materials were available: 5/5. Toshiba provides a complete set of additional material such as app notes and certificates.
The price to performance ratio was good: 4.5/5. To me the EVB is slightly pricey considering the BOM, but it’s worth it for the design time savings. The IC itself is well priced, comparably with most other competitors.
