Creating a Tx timed control unit for 12 boats

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

I've been busy for some time learning and creating a timed unit for our Av-a-Go fleet of radio controlled boats with which we raise money for local charities and such.

History

A group with in our Club got together and either built or bought kits of small model boats for the younger generation to Av-a-Go at radio controlled boating. We have a fleet of Graupner tugs, some scratch built fishing boats a few US river tugs all running on 27MHz or 40MHz these units being donated by club members as they have replaced them with 2.4GHz units.

As this project expanded a large 8 x 10 m inflatable pool was purchased direct from China with the auxiliary gear, a large 8 x 10 gazebo for the control and display area and a road trailer to transport all the necessary gear.

Two years ago it was realised that the battery requirements alone were getting out of order. Each Tx requires 8 AA batteries, and with a fleet nearing 18 boats the total amount of charging was getting out of hand [288 batteries]. Then there's the boat power units as well! Each unit wanted two sets to see it through a busy day (up to 7 hours) and on site power not always to hand. Indeed some events run over three consecutive days. We needed a simpler way for the Tx units.

Add to this that each skipper was donating for a set time at the helm, [£1.50 for 5 minutes] so we had to purchase timers, either digital (that fail in direct sunshine – yes we get some, and they need batteries too) or mechanical which are not very accurate. AND, would you Adam and Eve, the young skippers soon realised that a busy set-up with us ol' codgers in charge soon missed their reseting the timers or us not hearing them go off.

Requirements

We needed to simplify and control the Tx system.

1. All the Tx units could be run from a single car battery no problem, except someone needed to make the connections and carrier for this.

2. The Tx units required a timer unit, obviously external to the Tx, via which the Tx units could be time controlled and, switched on or off without buttons or switches that any small hand could find, BUT simple for us aged controllers to understand as well.

Seeking the solution.

We started with a simple timer kit from Kitronics, sold cheaply for educational establishments. The unit starts with a selectable time period (run time) followed by a buzzer sounding for a selectable time period. An LED indicates when the unit is timing, which goes out when time is up and the buzzer sounds.

From this basic unit I started out on a fairly long path into 555 timers, circuit boards, pull down resistors, etc etc till eventually arriving at a working unit I shall call Alpha. [Actually each black box is labelled Tango but whatever?] This in itself was quite a satisfying achievement to one with no knowledge and very little understanding of electronics. I'll plot the progress in my next instalment, a unit that timed and sounded the buzzer but……….

Kimosubby (electronics???)

Edited By Kimosubby Shipyards on 10/07/2015 21:02:09

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

Our plans to electronise our kid’s have a go boat units

[Dave MilbournParticipant @davemilbourn48782]

Fascinating, Kim! Looking forward to Part 2.

Dave M

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

Hi Dave, thought you'd take an interest, any suitable comments most welcomed especially when I travel the wrong creek!

The story so far, we had some time kits (3 years old and un-opened) but a starting point. A quick brain storming gave us the target.

TARGET. Power 12 Tx units from a car battery 12V supply so that each unit is individual, can be switched on without an obvious switch, have a visual indicator that a) we had power and b) the unit was timing and after the required time period the unit turned power off to the Tx and gave an audible alarm that the time had expired. This meant regulating the 12volt supply to 6V for the control unit and then getting the timer unit to switch the 12V supply onwards to the Tx themselves. Of course, a simple harness was also required to connect all the unit together.

First up was building a timer unit, and here's what they look like when made.

You can see all the bits: battery clip for a 4 x AA clamp, the red push to make start button, the power on slide switch and the buzzer (quite excruciating tonal qualities – just the job.) Also obvious are the two variable resistors, the lower one is 1M and controls the timing period in conjunction with the associated capacitor, the upper one is 100k and adjusts the buzzer on period. When the switch is slid to on, all should be off no sound or LED, but when the red push to make is engaged the timer starts and the red LED is lit. When the time is up the LED goes out and the buzzer sounds.

The simple diagram is this (and my first lesson in understanding circuits and drawings and final board layouts.)

The timer uses a single 556 chip, the diagram, at first, seemed to me to indicate two 555 chips. Off to the books and the inter-webby to learn that it is common practice to draw a 556 circuit as above, it's less messy and easier to read. The 556 is actually two 555s in one chip, each can control the other. Here's the pin layout on a 556, from it it becomes easy to relate the circuit.

I'll not show any more, but those that know understand the internal workings far better than me.

First thing noticed was the time period was not extendable to 5 minutes. I changed the time capacitor from a 100uf to a 220uf and was then able to cover 5 minutes. At this stage it was also noted that the voltage played a big part in the time period setting. I was using a battery clip with 4 x AA for 6V (actually 6.6V) but at one time used a plug-in variable voltage supply for 6V (actually 6.2V) and noted a longer time elapsed.

I had to regulate the supply anyway from 12V approximate incoming to 6V, so set about this first. Using a LM317 regulator (I had a few as I was making regulators of various voltages for my Arduino and Pollulu units and singing Gerbils). Each control box required a regulator, so at least 12 needed making, job number 1.

Having established that the unit could cover the time period wanted, I now needed to identify the output (pin 5 which lights the LED) and how to use that to switch the 12V supply on for the Tx themselves.

I made another full kit up, but this time assembled it from the circuit diagram shown above. This was used as the experimental unit for all ideas for use. It was also only later on that I realised that the circuit layout from the drawing and on the breadboard was not actually the same in terms of connections as soldered on the PCB.

It was simple enough on the bread board to introduce another LED (green) to indicate when the timer was powered up and waiting to run. The power switch on the soldered PCB is across the negative.

We had another requirement now for the unit. With demonstrating to the operator (children) how to use the Tx and ensure they were able to use it, we were using up some of their 5 minutes, we needed to be able to re-set, after all there are two re sets on the chip. Another look through the thousands of 555 timer circuits showed how to create the re-set button. The reset is on pin 4 and by introducing a 'draw down resistor' to it it is possible then to insert a switch between pin 4 and earth. This was done on the breadboard with a 10k resistor and it work first time. At any time in the time period the whole run could be re-set back to the start condition.

Oh, now I find that on the PCB things are different. I'll show you what I mean next time. Suffice to say that pins 4 and 10 are linked and not separately wired to earth as shown on the drawn circuit. Indeed they appear to be connected via the positive. What to do next?

Drilling holes and hoping. Kimosubby

[Dave MilbournParticipant @davemilbourn48782]

Kim

PM sent.

DM

[Malcolm FraryParticipant @malcolmfrary95515]

555 timers are great for short times (a few seconds) but to accurately time longer periods do need a very stable voltage supply. The main cause is the comparator looking at the rising voltage on the timing capacitor – when it gets to the switching voltage, the "ramp" is getting very shallow, meaning that a minute difference in the supply voltage results in the reference voltage shifting, giving spurious switching.

Feeding a bunch of them from a single regulator, it will help if each card gets its own decoupling capacitor (an electrolytic of a few dozen microfarads, not critical, just needs to be there to counteract wiring losses).

The reset pins, for the circuit to operate, need to be connected to the positive. They don't need much current, so can be connected via a pull-up resistor, but if the circuit designer wasn't planning to use them, a direct connection is simpler and cheaper. Just how easy or difficult it is to cut the track linking pins 4 and 10 to pin 14 and leave the power connected to 14 needs a view of the card.

This is off memory, and the last time I meddled with 555s was about 1985. FWIW, for the purpose of actually calculating timings, we tended to opt for the CMOS versions – the input impedances were so much higher that you could forget their effect on timing.

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

Hi Dave, thanks for PM,

Malcolm, welcome and thanks for your comments. The timer units were used as a starting point and have been a little modified along the way. The actual time period is not that critical so long as it is always just over the 5 minutes but not by more than 15/20 seconds we are quite happy. Tests have proved that once set at 5m 10s the drift is not great, and temperature can play a significant part too. Our customers have been known to time themselves, so we always err on the plus side.

As you've guessed, I made some track cuts along the way and drilled a few holes for extra wire connections too. I'll present the PCB next, empty of components but with my alterations. As I stated, I'm learning quite a bit with this project and not all my "solutions" worked.

When I've typed up the next part I'll post, thanks for your interest, aye Kim

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

Hello again.

We now had a timer board that seemed usable for our purposes, we could get the time period required and a buzzer sounding when the time period had expired. On the breadboard the re-set was easy to add and also a power on LED as well to show the board is live.

There were a few alterations now made to the PCB as received. Here's the board and marked on are the current changes/additions.

This image was created by overlaying photos of both sides so that the track can be seen below the component layout. I isolated pin 4 from pin 10 by cutting (small drill) the track, and also cut the track 3 times between pin 4 and the +ve (just in case my new holes encountered the track and a short might be created).

As pin 10 was originally connected to the +ve via pin 4 I added a wire connection (blue) from the +ve back to the nearest direct solder station of pin 10. For the re-set on pin 4 I inserted a 10k resistor between the +ve and pin 4 (blue), and then a connection via hole 2 to pin 4 for the re-set switch (another magnetic reed) the other end of which went to ground via hole 1. A green LED was introduced again via a 150R resistor through hole 4 to the +ve and the earth via hole 5 (the start switch is across the -ve).

We had thought to use red for power on and green for timer running, but after operating trials in is much easier to see the red LED in bright sunlight. Some want a flashing amber when the buzzer sounded, but we argued away from flashing lights.

Powering up, the green LED illuminated, the timer started with the start reed switch (red LED on) and when the re-set switch was activated the timer stopped (red LED off) and the buzzer sounded once. [The buzzer period was kept short at this point, it's very annoying!] Note that everything is operating in constable mode, each 555 runs once and awaits a start signal. [It is possible to set up for astable operation, where the cycle is continuous till stopped, I mention this as I encountered this action a short while later, much to my chagrin!]

All seems well. Now to measure up for a box for all this, referred to as the control unit from now on. Most of my electronics materials come from Rapid, so an appropriate plastic enclosure was found to take this board and the regulator unit as well. The control box in use would display the green and red LEDs and have the buzzer central. The two magnetic switches would be positioned diagonally opposed tight beneath the lid, top right for a magnet sweep to start and bottom left for a re-set magnet sweep. Currently we are using neodymium magnets, but weaker ones will do the switching but require a slower accurate pass over the reed. The on/off would be on the upper end, and a power in jack socket alongside it. Also out the top end would come the power out to connect to the Tx unit.

Also found was a clip system that fitted this box that enabled it to be attached to the operatives console (a plank about 3m long on which we attached 6 Tx units each. ) The clip is attached to the enclosure via two nylon bolts, which, as luck has it, could then be used with nylon spacers to mount the PCB inside with a slight alteration to the hole position in the PCB (hole 7).

Now, where to get an output to switch on the 12V supply for the Tx and what to use as the switch? Suggestions included CMOS chips and regulators, transistors, and relays both sold state and solenoid.

The output on the 556(1) is pin 5. The voltage measured here was 4.2 volts, and it was responsible for the lighting of the red LED amongst other things, like providing a high voltage to pin 8 (the trigger for 555 2) and signal to pin 10 etc. I drilled hole 6 and took a line out from the inlet side of R3.

Not being electronic savvy I acquired some sold state state relays to experiment with (and some 5V solenoid relays too). The solid state relays appeared to work fine, when using a continuity mode on the meter, but when wired direct to switch the 12V they failed, stayed closed. I then tried the solenoid relays, having read about diode protection for bounce and emf these were included in that circuit. They switched ok BUT the buzzer function now went into astable mode, so I now had a pulsing buzzer and worse, the power out to the Tx did not shut off (because of the astable operation) and only by using the re-set could we get back to a start position.

Still, at least I was now switching 12V to the Tx unit via the control box, but was at a loss as to why the system was cycling. Our first public run was with 6 units in this mode, it was quite a success, we had achieved most of our aim, using one car battery to power 6 units, the units having integral timing with secret squirrel switching for start and reset. Four hours and no failures.

So, now to cure the astable consequences. It seems I would have to switch the relay with a transistor after all – this is getting me deeper into electronics, I had to do some quick learning now find out firstly how to wire a transistor to the circuit and what type to use, there are hundreds out there……………….. AND of course, I should have taken into consideration the power required for the switching – I needed to know that now in order to get a transistor that could do the job.

Next time – the solution is found (and why I changed the on/off switch to +ve on the incoming power line), and some pictures of the units and some of the unholy mess I've created within – but it works,,,,,,,,,,,,

Kimmosubby (aka sparky)

Edited By Kimosubby Shipyards on 20/07/2015 20:54:36

[ashley needhamParticipant @ashleyneedham69188]

Kimmo. A few electronic kitchen timers not good enough for you then ??

Ashley : unable to muster the concentration needed to assimilate the previous posts…..

[Trevor HollowayParticipant @trevorholloway99134]

I am assuming there will be no problems when the boats receivers loose the transmitter signal.

[Dave MilbournParticipant @davemilbourn48782]

Hey, you naysayers – leave the guy alone! This isn't M*yh*m, you know.

This is very impressive stuff for a bloke who professed to know the square root of sweet Fanny Adams about electronics to start with (well, so he says………)
Ashley
Kitchen timers would not have given the warning buzzer facility and would have been open to tampering/cheating by the users. Don't assume children are all sweetness and light (remember your own childhood?).
Trevor
Practically all electronic speed controllers will stop the motor if they don't see a valid signal from the receiver. A formal failsafe system in the radio itself isn't necessary.
Kim
More power to your elbow, m'duck.

Dave M

[Malcolm FraryParticipant @malcolmfrary95515]

Since the radios are all semi-redundant 27 and 40 MHz sets, the loss of signal will just result in the receiver not giving an output. If 2.4GHz gear was being used, the results of loss of signal would be anybody's guess.

Well done, Kimo, welcome to the brotherhood of burnt fingers and squinty eyes.

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

Ashley and Trevor,

Sorry Ash, I did explain about the goals in post 1, kitchen timers are excellent, at about £6 an item not bad, multiply by 12 plus a couple of spares – £84 and then add in some spare batteries. Now experience tells us, and we have been running these boats for about 7 years now, is that exposure to bright sunlight a) heats the timers up and they fail and b) the digital display is destroyed. So each year we have to get some replacements.

Trevor, with using the 27 and 40 sets we have no problems at all with switching on and off of the Tx units. As with any sets there is a slight delay as comms are established, but no problems. We have used new crystal pairs too as older ones had some frequency bleed into neighbouring bands. With 2.4 gear, there could be. It's one of my popular hang ups about failsafes. On the cheap 2.4 sets the failsafe is fixed in the receiver, no way to turn it off etc, so we will have to ensure that such sets are altered internally so that we do not use the THR channel (normally 3 if memory serves) else they have been proved to work quite nicely – in fact better as there is very little interference with them. Our main interference is the marquee metal poles forming a large ariel effect about the 27 and 40 Tx sets and their ariels. I'm seeking opinions on 'chokes' and where to fit them in the sets to lessen stray emissions.

The major component in the system that can be a problem is the ESC [speed controller] type used. [We are seriously considering not using one and reverting back to micro switches and one servo, as almost all users of the boats are full stick either direction.] The ESC we use is the 15A Mtroniks older type not the plug and play and they are superb! Why? The plug and play type on switch on and power up requires the first throttle stick movement be for ahead (forward) movement of the vessel. The older 15A unit is set-up once and remembers the settings even when switched off – the button on the ESC has to be pressed as the two lights are flashing in order to re-set.

Thanks for your kind words DM – I oft wonder if you remember that unit I sent back for repair, one of my first attempts at soldering a board – you kindly advised some improvements to technique and equipment all of which I took to heart. And you fixed the item of course – MOFSETS as I recall.

I now have 14 control units operated exactly as required. I'll continue their development on the next post – just creating some images as they better show the end product.

Happy sailing, we have a two day event this weekend, 8 hours each day, so as good a test we could get for the new gear. Just think, thats near 450 AA batteries that do not require charging! It will also confirm our calculations re the car battery and duration. I'll try and make some measurements during the day on power consumption etc with all units in operation.

Aye, Kim

[Dave MilbournParticipant @davemilbourn48782]

Kim

That must have been one of my better days; I wasn't usually that patient. These days I just raise my eyebrows, shrug my shoulders, fix the problem – and bank the cheque!
DM

[Trevor HollowayParticipant @trevorholloway99134]

Certainly wasn't meant as a "dig", just that other posts have warned about loss of signal. In a small mobile pond and slow boats there would be no danger maybe just some boats wandering around aimlessly (zombie boats ?).

As usual foresight is switched on.

[Kimosubby ShipyardsParticipant @kimosubbyshipyards]

Hi Trevor,

sorry, I didn't take it as a dig, it is a genuine problem with 2.4 systems if the operator is unaware of the default situation with some cheaper sets. I remember in a yacht race loosing command at about 60 yards away, so I took the back off the Attack set and stuffed in a set of alkaline batteries – got enough extra signal to regain control, continue to the mark and return (near the back of the fleet I must add.)

At most, 14 boats on a 8 x 10 metre pool makes for quite a lot of barging about – hence the kids love it. And yes, we do restrict the speed by restricting the stick travel on the Tx units using hex screws into the casing.

I suppose I could now expand and make a zombie boat by introducing a Pollulu chip to take command if the boat receives no signal for 5 minutes – could do away with skippers then……….

Aye, Kim

[Fred Smith 1Participant @fredsmith1]

If you have a common power supply for all of the transmitters you have the perfect situation to also use a single timer function to control all of the TXs.

Get a raspberry PI and some geek who understands programming. Its a simple enough job to create a simple system to monitor each power feed and time it out. One PI should be able to pole and control at least 20 feeds.

Since the control is now remote from each TX its safe from little fingers. The person controlling the stand can sit and monitor a display board. Using such a system will not be accurate to microseconds but I should expect an accuracy of better than 1min ( actually better than one second). The PI can turn a relay on and off for each TX.

If you are worried by the idea of uncontrolled boats then you can instead use a relay in the TX and cut throttle modding the TX so when the relay trips it puts out a low throttle or stop signal.

[Author]

Posts