As many OEMs, plus tier 1 and tier 2 manufacturers will attest, the multi-layer ceramic capacitor (MLCC) industry is currently experiencing a significant capacity and supply issue. The last time we saw such shortages was around the turn of the millennium. Manufacturers are putting in capacity, but this takes time to feed through and presents little comfort for manufacturers and engineers dealing with looming line stop situations. It also trickles down to supply-chain managers, who have to seek parts from non-preferred sources.
At times like this, engineers should explore new options and alternative techniques that don’t necessitate having to go through circuit or product redesigns. Polymer electrolytic capacitors such as KO-CAP offer one alternative that, given certain conditions, can help. Going to KO-CAP isn’t always trivial, but if certain things align, such capacitors can be a viable route to explore and take.
KO-CAP tantalum-based polymer electrolytic capacitors, developed by KEMET, are like any other tantalum capacitor: They comprise a slug of sintered tantalum powder that has a tantalum-pentoxide layer grown on it, with a layer of conductive polymer acting as the cathode. This conductive polymer gives the capacitor much lower equivalent series resistance (ESR) than “traditional” tantalum capacitors.
For engineers, solving problems is the norm, and as is the case with many engineering challenges, deciding to opt for a solution like KO-CAP over MLCC is a matter of managing tradeoffs. A number of parameters and factors must be considered when assessing the opportunity to switch capacitor technologies. The main, and most critical ones are capacitance, voltage, ESR, frequency, leakage current, size, and qualifications.
The flow chart in Figure 1 can serve as a guide to making decisions when considering each design parameter.
1. The flowchart shows the decision tree for replacing MLCCs with KO-CAP.
KO-CAPs tend to have more capacitance than a similarly sized ceramic capacitor. They don’t come in values smaller than 680 nF. Thus, if the total capacitance needed is less than that, KO-CAP isn’t a viable option. When it comes to capacitance, replacing a bank of MLCCs with just one or two KO-CAPs could be a key factor in terms of cost efficiency.
In any tantalum-based capacitor, the dielectric layer is very thin—a typical value is approximately 20 nm. Having such a thin dielectric gives a large amount of capacitance, but it also has the effect of limiting voltage. A “high voltage” KO-CAP would be anything more than 35 V. In general, if your operating voltage is more than 50 V, KO-CAP isn’t a suitable option.
ESR is another important parameter to consider; ceramic capacitors have lower ESR than an equivalent KO-CAP counterpart. That’s not to say very low ESR KO-CAPs aren’t available. Some go as low as 8 mΩ, but a typical cutoff of 10 mΩ is adequate.
When considering the frequency characteristics of KO-CAP, it’s the self-resonant frequency that requires close attention. Usually, you want to operate capacitors below this point. Although it isn’t always the case, if your switching frequency goes beyond 1 MHz, then you may be approaching the limits.
Bias is a simple but not to be overlooked aspect when looking into the viability of an alternative to MLCCs. KO-CAPs are polar devices; therefore, they can’t take reverse bias voltage and be placed and utilized in a location where reverse bias is possible or must be tolerated.
Having considered all of the key parameters, now let’s look at an example using Texas Instruments’ TSP54560B-Q1, a buck converter for automotive applications. The circuit diagram can be seen in Figure 2.
2. MLCCs have become unavailable for this buck application.
Assuming no availability of MLCCs, using the replacement guidelines we can come to the following conclusions:
The input-side capacitors C1, C2, C3, and C10 are 2.2-µF, 50-V, 1206 X7Rs. There isn’t a drop-in replacement for the ceramics, but it’s possible to take the total 8.8-µF capacitance and replace the four ceramics with one 10-µF, 35-V KO-CAP. It’s more than the original capacitance needed, but it’s still within the required range for this regulator. ESR, leakage, and frequency aren’t a concern on the input side, as long as the input isn’t direct battery voltage. A simulation tool such as KEMET’s K-SIM can be used to show their side-by-side comparison.
From a cost perspective, the replacement of four MLCCs with a single device will typically yield a significant cost reduction.
The output-side capacitors C6, C7, C9, and C11 are 22-µF, 10-V, X7R 1206s. In this case, there’s a drop-in replacement KO-CAP. It’s 6.3 V, but more than the output voltage range. Here, the KO-CAP ESR is higher than the ceramic equivalents, but still within the design specification limits. The switching frequency of the circuit is 300 kHz and the self-resonant frequency (SRF) of the proposed replacement device is around 1 MHz, and thus is acceptable.
In terms of cost, the situation is similar to that of the input side, with the replacement saving close to 50%.
C4, C5, and C8 are other caps that support the functionality of the device. They’re suitable candidates for substitution because of both their physical size and capacitance value. Capacitors of this type aren’t experiencing quite the same supply issues. In this example there’s little discussion of leakage current, because that’s only a concern in systems with fixed non-rechargeable batteries.
Finding a drop-in replacement for difficult or impossible to source MLCCs is possible, but it doesn’t have the same level of value proposition as replacing a bank of capacitors with a much lower quantity of KO-CAPs. Sometimes a review of a problem application will reveal that substitution won’t be feasible. Nonetheless, during times of capacity and supply issues, finding solutions through other avenues could yield a solution that not only overcomes the problem that instigated the review, but bring other benefits, too.