Lithium batteries for the Yaesu FT-897

Internal batteries in the FT-897

However, my irregular use of the NiMH batteries meant that their internal resistance was often quite high, as is often the case with NiMH batteries that are neglected. Even when new, the voltage dropped too much when transmitting at 20 W, so I generally stuck to 10 W. I even built a discharge circuit to condition the batteries now and then, but when the batteries finally died, I began investigating other options.

Here's an authentic FNB-78 battery pack cracked open. I would have liked to rebuild the pack with low-self-discharge NiMH batteries (such as the Sanyo Eneloop), which work exceptionally well in cameras, flashes, and the like, even with irregular use. They are not, however, available in the "4/3 A" format used in the FNB-78s and other similar NiMH battery packs. Other people have opted for sealed lead-acid (SLA) batteries, a.k.a. gel cells, which are cheap, and quite tolerant of any kind of use. Apparently some type of SLA fits quite snugly inside the FT-897. Their capacity, however, is significantly lower than that of the FNB-78.

Lithium batteries

I finally ordered four 6.6 Ah 7.4 V ("2S" or two cells in series) LiPos from a local RC hobby store. Two such batteries in series produce a voltage range of 16.8 V (when full) to 12 V (when empty). The high end just exceeds the radio's voltage specifications of 13.8 V +/– 15%, but not by much, and the radio survived my initial testing. The low end of the voltage range is still within spec, so the radio should work fine till the batteries are exhausted. As a bonus, the battery indicator LED will begin to blink around 12.4 V, which is a perfectly acceptable lower limit for discharging the batteries! The four 2S batteries form two identical and independent battery packs, with independent safety circuits and independent charge connectors, just like with the original FNB-78s. However, the battery capacity is almost 50% higher.

Safety circuits

For safety, lithium battery packs always include some protection circuit module (PCM) or battery management system (BMS) which monitors each cell separately and breaks the circuit, if any one of them is charged too high or discharged too low. Either case might otherwise destabilize the cell and shorten the battery pack's lifetime (at best) or cause an explosion (at worst). Often it also limits the total current to a safe level, and sometimes it contains circuitry to balance the cells after charging. The model PCM-LI14.8V10A PCM I ordered from BatterySpace.com implements all these functions. For even more safety, I added an 8 A PolySwitch overcurrent protector and a 72°C thermal fuse in the circuit, and placed 1 A picofuses on the balancer connections. The balancing function won't use that much current, but if a soldered balancer wire should crack free from the PCM, the short circuit won't destroy the battery. Some PCMs don't have the balancer function built in. These can also be used, but then the batteries should be charged with a balancing charger. There was not much room in the FT-897's battery compartment for packaging the PCMs and other safety devices, so after soldering wires to them I simply covered them in heat-shrink tube. Even so, I was just barely able to fit the mess of wires and connectors in the limited space. If only I'd had the connectors available, I could have shortened the power leads on the batteries, instead of having to fold them over each other in the cramped space.

The thermal fuses (middle, in the picture, with crimp lugs on its leads) could obviously not be safely packaged in heat-shrink (soldering can also destroy them, so I crimped the wire connections), and I wrapped them in Kapton tape instead. They were taped to the sides of their corresponding batteries in such a way that each fuse is in contact with all four cells of that circuit. Thus if any one of them begins to heat up (which should not occur at all with these currents), the fuse cuts the entire circuit. I chose a 72°C thermal fuse, because that was the lowest temperature that was available. The PolySwitch overcurrent protector (the yellow thing) is purely redundant, since the PCM contains overcurrent protection. But it does not cost much either. It should be kept away from the batteries, because if it ever does activate in a short-circuit condition, it will heat up quite hot. On the right is a picofuse.

Since my PCM contains the balancer function (which is vital with multi-cell lithium packs in the long run), I installed a pair of two-pole Molex connectors on the back of the radio as charge connectors. (If the PCM does not contain a balancer, a balancer connector would also be needed, and a balancing charger should be used instead.) The radio has male 3-pole JST VH headers with 3.96 mm pitch for the internal batteries, so you need the corresponding female connectors. Not many electronics stores seem to have them, but they may be found on eBay (search for e.g. "JST VH 3p 3.96"). If you can't obtain the connectors anywhere else, cut them from old FNB-78s (that's what I did), or improvise. Here's one 2+2 cell battery and its safety circuits under test. Interestingly, with all connections made to the PCM, the output connector did not provide any voltage, almost like the PCM were faulty. It turns out that the batteries need to be charged, if only for a second, with the PCM in place. That "woke up" the PCM, and the battery pack became operational. I don't know if this is a deliberate feature of the PCM, or just a quirk of how the overdischarge protection is implemented. It could be either one. Here is the circuit diagram. It replaces a single FNB-78 battery pack. You can build two inside the FT-897.

Charging the batteries

While lithium batteries don't self-discharge (not much, anyway), they don't like being stored long times at full charge. I wouldn't worry too much about that, but it's best to discharge them to approximately 40% charge (some 15 V or so) if they won't be used at all, say, during the winter months. The Bantam charger has a special "storage charge" function that charges or discharges the cells to a suitable voltage for long-term storage. Or the cells can first be charged full, and then discharged by leaving the radio on, until its voltage indicator shows 15 V.

The new batteries in use

The voltage of the batteries falls quite linearly with use, and provides a relatively good indication of how much operating time is left. Only at the very end does the voltage begin to fall more drastically, so when the battery LED begins to blink, it's time to quit. At that point, trying to use the last drop of charge won't give much more operating time, and the LiPos won't appreciate it either. And anyway, with a total capacity of 13.2 Ah, these batteries last long enough: about 9 hours total when working SSB at 20 watts! I am very happy with them. When transmitting at 20 W, the voltage as indicated by the radio, does fall about 0.5 V, but I'm sure this is due to the protection circuits, not the batteries themselves. The voltages of the original NiMH FNB-78s could fall by several volts, especially after a period of disuse.

The LiPos were quite expensive. Much more expensive than Yaesu's FNB-78 battery packs. But I think the investment was worth it. So how long will the batteries last, then? How many years, how many discharge cycles? I have no idea. The batteries are rated for a maximum output current of 165 A (yes, really, one hundred and sixty-five amps), whereas I will only ever draw some 5 A or so. Hopefully the light use will prolong their lifetime. In any case, battery failure usually comes in the form of gradually increasing internal resistance. In this relatively low current use, I can tolerate a bit of internal resistance. But only time will tell.

During two summers of use, I have encountered only one problem: Once, when transmitting an AM carrier (admittedly at unnecessarily high power) into a tuner after switching bands (resulting in high SWR at the transmitter), the entire battery suddenly died. The protection circuit had kicked in and disconnected the cells. There was still plenty of charge left in the cells, but the unusual RF conditions tripped the low-voltage safety circuit. Applying a charger very brifly reset the safety circuit. Unfortunately, I did not have a charger with me out in the field, so I was stuck with only one working battery. I think I will build a "dummy charger" out of two 9-volt batteries and a resistor, so I can reset the low-voltage cut-out if it happens again.

UPDATE: Fail!

One interesting alternative I've seen on the Net is making a new bottom plate for the radio, with its edges bent upwards, so there's more room for C or D -size batteries in the battery compartment. Some manufacturers (for example Ansmann Max-E; but no Sanyo Eneloop, for some reason) make low-self-discharge NiMH batteries in C-format, which provides some serious capacity. If physical size is not a primary issue, one might even go D-format. But since I eventually decided to go for an external pack, I'm now wondering if I could fit my LDG Z-100Plus Autotuner inside the empty battery compartment in some sensible way—that might be quite convenient also...

A warning, one more time

The batteries should not be dropped, banged, bent etc., they must never be charged to over 4.2 V/cell, nor should they ever be discharged below 2.5 V/cell, and their current ratings (which are different for charge and discharge) must be obeyed. There are many reported cases of Li-ion or LiPo packs exploding. And when you notice them starting to puff up, like mine did, it's best to cease using them before something bad happens.

Of course, LiPos should only be charged with a dedicated LiPo charger. Also, LiPos (or any other batteries, really) should never be charged unsupervised! Read more about lithium batteries at Battery University. And if you choose to try this yourself, you do so at your own risk. (Oh, and if you can't explain without googling, how many mAh will a 1 amp device consume in two hours, then please don't get involved in any DIY project that has rechargeable batteries in it!)

You have been warned.