Thread: *** Making Those Servo Props a Bit Easier and Cleaner ***

Tweet

Hello all, this is my first post but this is not my first time here. I am an Electrical Engineer and I have made a few props over the past year which are triggered microcontroller self running props. This year I would like to take up another level, something like Halloween Bob has, but different. I designed a few boards that should make it much easier to build, set up, and take down props. This is how it works. You take an SSC-32 and it becomes a daughter card to this main interface board. There are indicators for ATX power, SSC-32 power, xmt, rcv, and strobe.



This board can be set into a 3.5 to 5.25 bay adapter and placed inside of a PC.



You would connect your power from an ATX supply, a serial connection which I will run mine directly from the motherboard header, and a PCI serial card header to the back of the board. I plan on running 3 of these to start. If you would like to, you can run a parallel port to the back of this board with either from the motherboard LPT header, or I plan to run a cable from a PCI card back into the PC and to each interface board. What this board does is it splits out the servo outputs in different configurations along with the 4 SSC-32 inputs and 8 parallel port ULN driven outputs. These can be used with VSA or other SSC-32 software, or you can create your own.

Now for the receiver modules. I have created 4 different receiver modules. The first one can be mounted inside of a skull or on the back of your skeleton. This board is about 3.5" x 4". It takes in 3 of the RJ45 outputs from the main interface board and you can use a standard ATX power supply. You can pick up a peripheral supply which delivers 2A on 5V and 2A on 12V from eBay for dirt cheap. This board makes 6V from 12V. You can also bypass the regulators and run on 5V. I designed this board to run nod, tilt, turn, 2 axis eyes, and jaw servos. It also takes in 3 parallel port lines for RGB eyes. There is a potentiometer to adjust the fade in speed of each color. There are also 3 SSC-32 inputs. I also had holes drilled for a 60mm fan to cool the heat sinks on the 6V regulators.



The next module I plan on using for arm movements. It contains 4 servo outputs from one RJ45 connection. It also uses a standard ATX supply input and servos can use wither 6V or 5V. This was if you are demanding too much, you can always share the 12V and 5V load. 2 regulators are for higher torque needs, and 2 are for lower torque requirements. Holes were drilled to mount a 40mm fan. This board is about 2"x2"



The next module I plan on using is for fingers. It takes in 2 RJ45 connections and the ATX power and goes out to 5 servos. This board is also 2"x2".



The final board is a miscellaneous board. Again, an ATX supply and an RJ45 from the main interface board. It goes out to 2 servos, a relay, and an input. The relay can be bypassed.



All servos can use either the 6V regulator (from ATX 12V) or the ATX 5V. I also have made labels for the from 5.25 bay adapter which shows what each port can be used for. All signals are twisted with a ground to maintain a good signal over long lengths. Power is not run on RJ45 since typical current on servos can be roughly .5A under load and as much as 1.5 amps per servo if they are driven to their stops. All servo outputs use ployswitches for circuit protection.

So what do you think?

Brilliant... would love to see it in action!

I think it looks awesome. I see no power outs for power to the servos. One thing we know is it is not a good idea to provide power to servos through the CAT 5. On my 3 axis skull this year I ran speaker wire from the ssc32 board to the skull interface card. So signal came in via CAT 5 but power came in separate. Power to the ssc32 was a 6 cell nicad pack. The one thing I really do not like about the ssc32 board is the screw down connectors. I figure a main buss for power with a feed to the ssc32 is likely the way to go.

Thanks. The boards have 3 pin headers which provide signal, power, and ground. The SSC-32 does not provide the power, only the signals. The 12V that powers the SSC-32 comes from the 4 pin floppy power connector, servo power is derived from a separate power supply. On the main board, the only header pins that are connected from the SSC-32 are the input power, the RS232 RX and TX (DB 9 pin is disabled on SSC-32- jumpers are removed from the RX and TX and my interface board has a MAX232 to bypass the SSC32 RS232 to TTL converter), the 32 servo signals, the 4 input signals, the SSC-32 5V power out (used for a front panel indicator), and a handful of grounds.

With these boards, all I have to do is plug the servos into the receiver boards, an ATX power supply or peripheral power supply, and the RJ45 connectors which just provide signals (twisted with grounds).

OMG I love you...lol. A couple of us have been going through multiple ways to create an easy way to control our props but none of us really have the experience of a professional electrical engineer. We have made hobby level solutions that have worked great but these pieces look amazing, clean, and professional. I hope you plan on making these available to the community. I'm sure you would be able to sell a good amount. I know I would be in! Please keep us updated.

I agree with Jimmy.

Thanks again! Jimmy, I love the Pirate display!!

I currently had one of each board built, but I do have several PCBs, and I ordered enough parts to build a small batch (way more then what I currently need). Between myself and a friend of mine who is a certified IPC specialist (J-STD-001), we would love to make these available to others. I enjoy watching what other people have done with their props, and I would like to help out in anyway I can. First I wanted to do a good thorough test on these boards so I have full confidence that others will be very happy with them.

I think I want one

I've read over and over that running servo power down the same CAT5 cable as servo signal is not a good idea (even though servo cable itself does this). We have been doing it for 6 years now and have never experienced a problem on any of the 8 animated skulls or other props that run off our SSC-32 board. The CAT5 runs back and forth, parallel, perpendicular and otherwise all over all sorts of AC power cables, DC power cables and DMX cables. All of the servo jitter and apparent crossover issues I have witnessed have all been resolved by repairing poor ground connections or replacing old servos that were pretty much worn out. Running power and servo signal over CAT5 has been really convenient. Perhaps we've just been lucky in that we haven't had problems. Perhaps there's something in the quality of the wire we use, (Belkin stranded). I don't know why it works so consistently well for us when so many warn that it is a bad idea.

You can run power over Cat-5, but there are 2 issues. First, once you have a cable plugged in, you now have power on exposed pins. This is a big no-no. You risk shorting out your supply and this can cause a spike in voltage which could take out sensitive electronics, fuses, etc. Output power should be on sockets, not pins. The other issue is the amount of current a 24 awg wire can handle, which is just under 600ma. A single high torque servo will normally draw about that much under normal use. Even medium or low torque servos can draw almost as much. I did some extensive testing on a few different servos and I have seen an excess of 1.6A of current is drawn at 6V if the servo pegs out at a stop. Cat-5 was designed for signal transmission. For convenience reasons, I originally wanted to do this as well until I did some testing which then I decided that I would not use Cat-5 as a means of power. As for noise, the servo signals are not high speed signals which are typically susceptible to noise both generating and receiving. It also depends on how your signals are run. Twisting each servo signal with a ground will give you the best possible signal over Cat-5. Even better is if it is shielded.