Inline Power Bit

A module that allows you to use a second power supply for subsequent bits in a circuit.
Proof of Concept:

As you can see from the one minute movie above, the module performs as expected, at least with two number bits… The signal of the second half follows the signal of the first, but each number bit has it’s own, independent power supply. Having a separate power supply should (should!) allow a high current bit, such as a servo motor, run more smoothly.

I said should in italics because in fact, this circuit does not perform as expected at high current levels. When I add a servo motor to the end of the circuit, it works fine up to about 75% of full voltage. But when the number bits say 75 or greater, the motor goes wild, lurching from right to left, and the number bit turns on and off. I put the circuit on an ammeter, and the current of the servo side bounced between zero and 200mA, while the the other side drew a steady 20-ish mA.)

Here’s the schematic of the breadboard:

Simple/dimple, right? No wonder it doesn’t work.:frowning: I believe the main problem is due to the op-amp I used– or rather, abused. The maximum output current of the TLV2462 is 80 milliamps. The servo bit can draw 200mA or more when it starts to move. The op-amp gets over amped and checks out, then the current falls and it jerks back, and oscillates between maximum and zero milliamps.

I need to solve the current problem that @alexpikkert hinted at in this post.

1 Like

Hi Chris @chris101,
You made a major leap towards a solution !
I am currently AFL for a few days (Away From Littlebits) so I cannot experiment…
Maybe you could try the following:
I think the opamp’s only purpose in life is transmitting the SIG signal and prohibiting all feedback from the second system to the first… Could it make a difference when this opamp is powered from the first circuit P1 instead of from P2?
I suppose you did connect both GND levels from both circuits? I could not see this in the diagram or the movie… I think the whole setup needs both GNDs to be connected…
( the SIG signal will need the same GND in both systems…)
:bulb:

1 Like

Hello @chris101,

this is a great idea. The servo does draw its current from the power line, in your case P2. It does not take the current from the SOUT of the previous bit. This line draws only a very little bit of power.
However, if P2 is noise, due to the motion of the servo and the high current, this could feedback into the opamp and - in turn - generate a noise SOUT signal, which might generate jumps or whatever.

I also believe that Alex’ idea could be a solution, as P1 is quite quiet and no disturbance from the servo could feedback into the opamp. Of course same GND is required for the whole circuit. I suggest to add some capacitors, too.
best regards
7th Dwarf

2 Likes

Oh yay! You guys rock!

Thank you @alexpikkert and @SeventhDwarf! Using the first power supply to run the op-amp solves the problem, with the standard servo bit, with my ‘heavy duty’ servo (a larger servo with a larger angular range), and with both of them in the circuit together.

For clarity, here is the current circuit, the test setup (with the large servo), and the details of the connections:

Hi Chris @chris101 and SeventhDwarf @SeventhDwarf,
Inline Power Bit ready for take off!
A few things crossed my mind when I saw the diagram:
I cannot test this this week because I am on a short holiday, but rethinking about my projects until today, I think most (or maybe all) noises and disturbances were distributed via the power line VCC.
With the use of P1 and P2 these lines are 100% separated. This takes care of this.
Could this bit work without the opamp and just a solid connection between Sin and Sout?
I think all original bits can withstand feedback into the SIG line and as long as P2 is always limited to 5 Volt DC there will be no harm done…
Second thought:
What if P2 is >5 Volt DC? Will an opamp withstand this and coud the other bits be harmed?
:sweat_smile:

1 Like

Thanks Alex!

I tried the thing without the opamp, and yes it works. There are a few issues with it run like that though. First, there is a surge of noise through the signal line, the servo goes crazy for a few seconds, but then settles right down to a steady output.

Second, it’s off by one!

Here’s the detail of the proto connection:

1 Like

Hi Chris @chris101,
Off by one ?? That’s funny!
Not when the opamp is in the circuit?
So then it will be better to keep the opamp in. Then indeed possible noises on the SIG line are blocked too.
To block the connection of wrong voltages on the P2 line, the following bit design could be used (just a drawn idea) The blue bitsnap is an input bitsnap for a power bit so it will always receive 5 Volt DC…
The SIG connection in this blue bitsnap must be isolated and only the power VCC and GND are connected to the output bitsnap.

:bitstar:

1 Like

Right, I forgot to address that cause I was on my way to work when I answered. Since the second power comes in from a bitsnap, I figure it will always be regulated 5 volts. Same as regular power in the littleBits world. In my circuit, I used a wire bit cut in half to get a bitSnap connector to hook into the proto module.

If I were going to build this for real, it would lookjust like your picture:

PS, per @SeventhDwarf, I stuck a capacitor in there. Probably needs more. The opamp and cap seem kinda lonely.

1 Like

Hi Chris @chris101,
May I suggest some company for the lonely opamp?
You can add a 1 M Ohm resistor and a 0.1 uF capacitor which are used in a lot of existing bits to define the SIG input impedance.
And maybe a large capacitor (1mF ?) on the output VCC line to GND to stabelize it even more…

1 Like

Thanks Alex @alexpikkert, (and @SeventhDwarf, cause I know you are lurking :wink: ) here it is with all the littleBits regalia (except for the ubiquitous TLV, which I do not have at hand.)

I’ll try to solder one together tomorrow or the next day, then I can start playing with it in various situations.

1 Like

I have finished the inline bit:

There is an extra socket on the device in case I want to play with an opto-isolator in the future, but it is working fine with only the op-amp so far. Here are the details of construction:

Top:


2 Likes

QD
Quick Development !
:bitstar:

1 Like

Thanks Alex - I’ve had this project in mind for some time, so here it is! I’ve uploaded it as a project to the Invent page.

1 Like