Warning:
If you are going to stick with the metal servos included in Nybble kit, this reconfiguration will usually improve the servos' stability. If you are going to use plastic servos, DO NOT try it because the direct voltage from battery will burn your servos.
I've got some users who wish to make Nybble run faster and longer. In short, there’s an easy fix to unleash the full potential of Nybble. However, I'll discuss the pros/cons in design thinking before you make the decision.
Nybble is using “digital + metal gear + aluminum alloy case” servos. The manufacturer rates the lifespan as 500K cycles. It’s the best servo I could find at its price range to make Nybble affordable. You can check my previous study on servo comparison.
With the aluminum alloy case for heat dissipation, it can stand the batteries' voltage (could be 8.4V with fresh charged batteries) directly and provide maximal output. However, there will always be tradeoff for performance:
The servo’s motor brush, gear system and potentiometer will wear out much faster; It’s more likely to break the mechanics if a leg hits obstacles by accident.
Long time running with large current (3~4A) will overheat the step-down chip and powering cable.
Battery will run out faster if Nybble keeps running; Larger batteries will add burdan to the servos and won’t necessarily improve the overall user experience.
Also consider safety. The best USB power bank outputs 5V x 2A = 10W. But Nybble could consume 8V x 4A = 32W when running.
So I don’t recommend running Nybble as a racing car, but a platform for studying programming and motion planing.
NyBoard was designed as a general purpose driver board not only for Nybble, but also for other projects. Some people want to drive their cheaper plastic servos that cannot stand 8V. So the step-down circuit for servos is introduced. The chip is rated at 5A, but the momentary current fluctuation by many metal servos could accumulate and trigger the over-current protection in the chip or the batteries. In other cases poor quality batteries simply cannot sustain a high discharging rate.
Now talk about reconfiguring NyBoard to make Nybble run faster and longer. In condition that:
Your battery can supply large current and its over-current protection is not set too low.
You are using the servos included in Nybble kit, not cheaper plastic servos.
You understand the tradeoff between performance and hardware's lifespan, and have spare servos for possible replacement.
You need to upgrade the battery holder's wires with thicker ones to avoid too much heating up.
you can bypass the step down circuit as described below:

Desolder inductor L2 from pad A, B.
Then there are two paths you could take:
Path 1 directly wires BATT+ and pad A. L2 is removed from the board.
Path 2 wires BATT+ to one leg of L2, and the other leg of L2 is still soldered to pad A. Add insulation (could be a piece of paper) between pad B and L2.
Theoretically Path 2 should supply more stable current. But I don't see a big difference on the servos’s end with oscilloscope. And L2 will generate moderate heat.
I think it's better to keep L2 on the board in case you need it for other projects. So below is my modified board for making a “racing Nybble”.

You can further improve the circuit by adding an additional capacitor in parallel to the servos' power line, which is quite common for servo driver boards. NyBoard already has a 1000uF capacitor in place.
On NyBoard V0_2, the similar reconfiguration will be more convenient with a jumper selector SW3, but less convenient than a slide switch. The limit to moderate Nybble‘s running time will be done by a cool-down counter in the program.
I soldered path 1, now my nybble is working. Before I was getting frequent resets and crashes, even loss of calibration values. I am using a 1500 mAh 2S 35C 7.4 V LiPo, which is a bit too big. Might have to get a smaller one. The 14500 batteries I haven't found yet in Germany.
Hi - the 14500 unprotecteds are tough to come by in the US. I tried using a flat Adafruit 3.7 Lipo, but the servos behaved oddly & I wasn't able to properly calibrate. I've ordered a FLOUREON 2S Battery 7.4v 35C 1500mah li-ion battery: https://www.amazon.com/dp/B06Y2M2J7D/ref=cm_sw_em_r_mt_dp_U_tLlXDb5AKAKCP
It looked like it had specs similar to others folks were using. I've got Nyboard 0.2 with the jumper. I've currently configured it like the image below. Do I slide this over disconnecting batt & connecting V+ to accomodate the 7.4v rechargeable? I was otherwise going to cut the single red and black (not the group that has the white, as well), and put these in the + and - terminals on the Nyboard. Do I have this right? Thanks!
It’s better to keep the original plug of your battery pack in case the two wires get short circuit. Use two metal pins as an interface between the NyBoard’s power terminal and your battery.
@Rongzhong Li thx, as always, for your kind advice. The holes are flat shapes at a right angle, that I hadnt expected. Thought they’d be pinholes. Any thoughts on how to attach pins to these so I can use this lipo in Nyboard’s + and - terminals?
Also: do i just slide the jumper so it’s no longer connected to Batt and is instead connected to V+?
Thanks!
EBL 14500 3.7V is easy to buy at Walmart.com
Are the new servos shipped late in 2019 after your move still metal gear and safe to try this? The enclosure looks like plastic to me.
Yes, they are still metal gear. Tested on my side.
Actually for late 2019 batch the boards are already V0_2. You just need to connect the jumper between BATT and V_S.
Hello,
I made this mod and it fixed my voltage issues, now the servo works perfectly and Nybble can run without any issues, awesome!
However, I noticed my nybble was not able to balance anymore.
Does it has an effect on the MPU ?
I ran the testMPU programme and it seems that it doesn't read any value.
Here is the output from the serial port:
?nitializing I2C devices...
Testing device connections...
MPU6050 connection successful
Send any character to begin DMP programming and demo:
Initializing DMP...
Enabling DMP...
Enabling interrupt detection (Arduino external interrupt 0)...
DMP ready! Waiting for first interrupt...
Thank you very much for your help.
Well, I let it run for some time and not it started to read some values about 10min after startup.
After i reprogrammed the Nybble code, it started to work correctly again.
I don't understand, does it needs a long initialization process ?
I am sorry to waste your time, I will keep you updated if I notice something.
It shouldn't take that long to initialize. I guess some some bit flip events happened when the voltage was unstable, and re-uploading the code fixed it. If it happens again, try to redo the uploading process with WriteInstinct.ino and Nybble.ino.
Have V0_1 of the board. Had several issues with crashes and resets. After soldering path 1 all is working.
We cannot directly use the voltage battery because it is heavy as well as it can ruin our servos and burn their wires. On the other hand, rushessay.com tries to attach the battery with servos.
Small tip. If you are worried about input voltage. I often use cheap switching BEC's from e.g. Hobbyking. They will convert varying input voltage (6-26 volt) to constant 5 or 6 volt output. Also ensures servos run constant speed.
@Jakob Leander Never mind. The comment you replied to seems to be a scam generated by bot. Thanks for your suggestion!