Credit: John Rice
It has been known for some time that marine invertebrates such as corals, sponges and anemones do better in an aquariums outfitted with some form of wavemaker or surge device. Even fish do better in this environment as dead food products appear to come to life and they are forced to regain balance in the changing currents.
From my own experience I can attest to the effectiveness of such devices. The natural reef enviro- nment sways in the ‘breeze’ caused by surface waves and under-tows and the fish and invertebrates that evolved there depend on such action. It behooves the hobbyist to create this same water motion.
For the captive system, water motion is important for moving food around and in preventing local dead spots that can build-up toxins. Many hobbyist use their main return with submersible pumps running continuously to provide this water motion. However, pulsed wave motion is more efficient and provides added benefits for inverts and fish.
Corals, sponges, anemones and other marine invertebrates have two requirements related to water motion:
1) They need to capture food from the water column.
2) They need to excise waste products.
Unfortunately these two requirements need different types of water motion. The first requires relatively slow or even still water motion while the second requires rather brisk, turbulent water flow. The only way to satisfy both requirements is through some mechanism that will vary the rate of water movement past the animals. For more information about wave motion and wavemakers as it relates to our aquariums see the following link .
In the reefkeeping hobby there are two ways to create wave motion: wavemakers and surge devices. This page will briefly discuss surge devices in general and will present my design of a DIY wavemaker I call ‘The WavePulse’ that as been in operation on my (and others) tank for several years. Finally the use of this wavemaker will be described along with some helpful hints. I hope this page will help you to build one if you are so inclined.
For more information on wave energy use by corals see the following link.
The purpose of a surge device is to introduce into the aquarium a measured amount of water for a relatively short period of time; after which there is no flow for another (usually longer) period of time. The action is in bursts or pulses of high flow rate water into the aquarium and this produces a wave- like effect.
The device itself is a water holding tank located some distance above the aquarium water line that is supplied by a pump pushing water up from either the aquarium or a sump. The greater the height of the surge tank above the display tank the more energy there is available to the surge flow (and the larger the pump needs to be for the same fill flow rate). A mechanism is used to cause the start of water flow out of the tank and into the aquarium and is usually based on water level in the surge tank. When all or nearly all of the water has left the surge tank this same mechanism will shut flow off and the tank will begin to fill again. This cycle continues over and over and is the basis for creating waves.
As the water enters the aquarium from the surge tank it tends to be very turbulent throughout and this turbulence is much appreciated by our inverts and fish. Often there are two surges, one on either side of the aquarium and this has the effect of creating wave energy that runs back and forth across the length of the aquarium.
There are two basic surge designs: the dump bucket and the syphon surge. The dump bucket system has been used in conjunction with turf scrubbers as well as tank wave motion generators. One is described by Delbeek and Sprung in ‘The Reef Aquarium’, Vol 1, page 148. The device can have problems, most of which involve maintenance and splashing. They are generally impractical for small home aquariums but have been used with success by large public aquariums where they can be accounted for in the system design and maintenance schedules. ‘The Reef Aquarium’, Vol 1, page 160 describes some of the problems with dump buckets.
The syphon surge system is more popular and practical for the home aquarium. The Carlson Surge device is a good example. A smaller version of the device may be easily built by most hobbyist. How- ever, they do have a problem with adding bubbles to the tank during surge initiation. While Carlson explains a way of reducing this effect, bubbles will form nonetheless and may be objectionable to some. In addition the syphon break at the end of the surge can be noisy.
Another related version of the syphon surge is the ‘toilet flapper’ system. This device uses a standard toilet flapper and ballast to create surges. Instead of a manual handle the device uses a ballast to lift the flapper at the end of the fill cycle. Often a second smaller ballast is needed, attached near the flapper to hold it up during dumping. Dr. Eric Borneman explains such a system in the FAMA May 1998 issue. In the end, this device may be the most useful for the home aquarium since it can be made bubble AND noise free.
Another way to create wave motion is through the use of a wavemaker. These are created using an electronic controller or timer and a set of submersible pumps located at either end of the tank. The pumps are alternately turned on and off and are run in tandem – first the pumps on one side are turned on then, after a period of time, these are turned off and the pumps on the other side are turned on. This continues over and over and is the basis for creating waves.
Wavemakers produce a fundamentally different wave pattern than surge devices. The surge device produces waves by creating differential water pressure from one side of the tank to the other. As water is introduced on one side of the tank, a build up occurs and water will flow from this side to the other. Since wavemakers use submersible pumps, they produce a flow pattern that is circular about the pump – this is especially true if the pumps are located near the top of the aquarium. Water is drawn from near the bottom of the aquarium and propelled away at the top. This creates a water pressure differential between the top and bottom (as opposed to left and right as in the surge system) and the flow will tend to be circular from top to bottom. This creates a backflow along the substrate towards the pump intake that is relatively laminar while the water flow near the top of the tank will be fairly turbulent. The surge devices produce fairly uniform turbulent flow going across the tank and this is considered more natural. Also, since wavemakers produce this localized circular pattern, it is necessary to have pumps located at both ends of the tank; while only one surge device is needed to achieve satisfactory water movement.
Wavemakers do enjoy a number of advantages over surge devices: they do not cause splashing or bubbles, are relatively maintenance free and do not interfere with sump water levels and evaporation monitoring. In addition, strategic wave placement is possible by using a number of pumps located at several points inside the tank. Finally, surge devices, due to their dumping of external water into the aquarium, require careful design of the overflow box if a sump is used. It may be necessary to over- design the box. In addition, with a surge device it is usually necessary to leave some headroom between the water surface and the top of the display tank. Instantaneous water flow rates from even a small surge device can be tremendous, causing a large surface wave to travel down the length of the display tank of several inches height! If not careful, the hobbyist may find that the surge will cause an overflow from the aquarium! Wavemakers do not have these concerns since all water flow is local to the aquarium (i.e. now additional water volume is added during surge). On the other hand, a properly designed surge device in an aquarium specified to handle the surge level is really something to see!.
In the end, wavemakers are easier and more flexible to work with for the average hobbyist but surge systems are better for wave generation.
Wavemaker Design Requirements
There are a number of wavemaker controllers on the market and as DIY projects. Virtually all of them allow at least two sets of pumps to operate in tandem. Wavemaker controllers should have the following minimum requirements:
A large on-time control range.
To accommodate a wide range of aquarium sizes and rock work densities a range of 2 to 60 seconds is desirable.
It should be able to drive several amps of AC power.
This allows hobbyist to use any number of submersible pumps (powerheads) for good water flow in even the largest tanks.
It should have a variable dead-time control.
Dead-time is a period of time when all pumps are off. I have found that this allows greater flexibility in wave motion set-up.
The input/outputs should be protected from shorts and transient conditions.
As more research is done, transient protection begins to look very desirable for use in electronic equipment to increase reliability. Recently a number of very good protection devices (transorbs) have been developed for this purpose.
The controller packaging should have some level of water tightness.
Virtually every hobbyist has splash a bit of salt water on and around their equipment. While the controller need not be water proof, it should be able to survive a light spray of saltwater without burning up.
In addition, the controller might incorporate an electronic night sense system to reduce the rate or intensity of wave action when the lights go out. Some systems have more than two timed outputs for additional flexibility. A ‘feeding’ button can be added that, when pushed, turns the controller off for a few minutes and allows the hobbyist time to feed the fish.
Survey of wavemaker designs
There are a number of DIY and commercial wavemaker designs available. Blue Line Products makes a commercial unit they call the Tsunami (the word means ‘the tidal wave’). It features a third output that is not linked to the timing of the first two and has a night sense system. There are lots of bells and whistles on this unit!
The typical DIY wavemaker uses a 555 timer and photo coupled TRIAC to drive the pumps. The photo couplers in this example use ‘opposing’ drive from the same signal (i.e. one is on when the other is off) to generate the mutually exclusive drive. This however, does not permit dead-time control.
The WavePulse Design
Figure 1 shows the schematic diagram of the wavemaker I designed a few years ago. I call it the ‘WavePulse’ since it creates waves by pulsed water motion using a set of submersible pumps on either side of the tank. It features variable on-time AND dead-time control. The following is a rather technical description of the circuit operation. Those not interested may want to skip ahead.
[All of the component part numbers that end with ‘-ND’ were purchased from Digi-Key Corporation and can be purchased from them on-line.]
Figure 1 Schematic Diagram of the ‘WavePulse’ Wavemaker.
The controller works off the 120 VAC, 60 Hz, power brought in through E1 and E2. E1 should be connected to the black AC line (the high side) and E2 should be connected to the white or neutral line. No ground (green) line is provided since the controller electronics and box are isolated and the sub- mersible pumps we use don’t require the ground.
A 3 amp fuse is provided to insure that a shorted pump or other high current conditions do not cause the box to ‘burn up’. Z1 is a bi-directional transorb and is used to limit input voltage transients above 200V and protects the controller electronics, especially BR1, from excessive voltage during lightning storms and other transient conditions. Z1 is located after the fuse so that in the event that it shorts it will not pop the house breaker (it will blow F1 instead). Z1 can short with a near direct hit from lightning (I had this happen once).
C1 is a substitute for a transformer which are large and expensive. However, C1 will produce an apparent power of about 9.8 watts due to power factor issues. This power is not dissipated but the power company thinks you are using it all the same.
Power then is transferred from the power lines to half-wave rectifier BR1. One diode is connected from the right side of C1 to R1 (anode to cathode respectively). During the positive half cycle of the input power, C2 will be charge through R1 and C1 to a minimum of 10 VDC set by zener diode, CR1. C1 acts as an impedance (1.47 kohm at 60 Hz) limiting the amount of current that can flow through C2 (115 mA peak at 120 VAC).
However, during transients, the current in C1 can become quite large and this needs to be limited by R1. The 27 ohm resistor will allow a maximum current of 14 amps for a maximum instantaneous input transient of 390 Vp (maximum swing allowed by Z1) [Note: this is for very short transients that may occur due to lightning etc. and virtually all of it will go through C2, C1, BR1 and R1. R1 just keeps this short duration current limited thus protecting BR1, R1 plays no significant role during normal operation].
Since C2 is much bigger than C1, it takes about 15 seconds to pump up the voltage on C2 to 8 volts after which time, U1 will begin to operate.
The high side of the line voltage is fed directly to the output solid state relays S1 and S2 (pin 2). These are zero-volt turn-on devices that are photo coupled TRIAC AC switches rated for 2 amps at 240 VAC and are made by Omron. They are a small SIP package designed for PC mounting. S1 and S2 act as simple switches between the high side of the line and the pumps which are connected to E7-8 and E9-10. The other side of the pumps are connected to local ground which is the low side of the AC line. So when S1 or S2 are active, a pump will see the full line voltage across it. Z2 and Z3 are transorbs used to protect S1 and S2 in case a pump is unplugged while active (this causes a large back EMF due to the inductance of the pump – this energy will be absorbed by the transorbs protecting S1 and S2) and are more reliable than snubber circuits.
Control for S1 and S2 is provided by a current through pins 3 and 4. There is an internal LED with series resistor that requires 4-6 VDC at about 10 mA for switch turn-on. This current flows through the device, an indicator LED (CR2 and CR3) and into the output transistor of U1 (pins 12 and 13). R5 and R6 are used to shunt excess current around the indicator LED’s (since they take less current than S1 and S2). When U1 outputs are inactive, the voltage at pins 12 and 13 is the same as the supply (10 VDC) and no current flows through the control pins of S1 and S2 and the indicator LED is also off. For troubleshooting and modification, it is important to remember that a logic 1 output of U1 turns the pumps (i.e. negative logic is used).
At the heart of the design is U1, a Pulse Width Modulation (PWM) controller designed specifically for push-pull forward switch-mode power supplies (a fancy name for a topology used in the power supply industry). I used this device since it is readily available (and thus low cost) and has facilities for the addition of dead-time control. The device used is an LM3524 made by National Semiconductor Corp. It was originally made by Unitrode (and still is) with the same base number UC3524. These are exactly the same die with the same characteristics – I used the National part because it was available from Digi-Key. It has everything needed to provide reliable wavemaker control at a low cost. Its only disadvantage is that (due to it being an old design) it uses a lot supply of current. You may be able to find a substitute that is functionally the same but draws much less current.
U1 contains an internal sawtooth generator the timing of which is set by C5. The series combination of R4 and VR2 is used to program a current mirror internal to U1 and the output of this mirror charges C5 with a constant dv/dt. An internal comparator senses when the voltage on C5 reaches 3 Vdc and will cause a discharge circuit to kick in. This discharge current is derived internally and will cause C5 to discharge about 20 times faster than it charges. The same comparator will sense when C5 has discharged to 1.5 Vdc (due to hysteresis) and will the circuit will then revert to the charge mode. This makes a sawtooth generator where the frequency is adjusted by VR2 (RATE) control. With the combination of parts shown, the maximum frequency is 0.5 Hz and the minimum is 0.017 Hz.
Internally, this sawtooth is applied to another comparator where the threshold is set by the voltage divider formed by R2, VR1 and R3. R2 and R3 limit the range of VR1 to just within the sawtooth voltage span. VR1 allows the user to set a voltage level trip-point along the sawtooth. The on-time of the outputs begin as the sawtooth finishes discharging and ends when the sawtooth has charged to the level set by VR1. This allows a very precise pump on-time. In addition, internally, U1 has gating that does not allow both outputs to be on at the same time – each output is active every other sawtooth cycle (this is needed for push-pull topologies for which the IC was designed). So each pump will be on for the time (specified by VR1) once in every two cycles (the other pump is active for the same amount of time but on the other cycle).
The time between when the sawtooth has reach VR1 level and the discharge time can be defined as the system ‘dead-time’ when no pump is running. So by varying VR1 and VR2 one can set both the absolute pump on-time and the ratio of on-time to dead-time.
[4 Jan 00] It should be noted that the display LED’s, CR2 and CR3 are of a special type. They have an internal resistor that allows them to be used with 5 volts directly. So they can not be substituted with a general LED type (these only drop 1.2 volts). R5 and R6 are used to increase the bias current through S1 and S2 internal LED’s.
ITEM QTY PART NUMBER DESCRIPTION SRC PRICE TOTAL
1 1 EF2185-ND CAP, 1.8UF/250V MET POLY DK 1.24 1.24
2 1 P6232-ND CAP, 2200UF/16V ELECTRO DK 1.01 1.01
3 1 P6216-ND CAP, 330UF/10V ELECTRO DK 0.18 0.18
4 1 1N4740ACT-ND DIODE, Zener, 10V, 1W DK 0.99 0.99
5 1 LM3524-ND IC, PWM CONTROLLER DK 3.00 3.00
6 1 DB104-ND BRIDGE, 1A, 600V DK 0.76 0.75
7 3 1.5KE220CAGICT-ND SUPPRESSOR, 1.5KE, 200V DK 1.28 3.84
8 2 Z909-ND SS RELAY, 2A, 250VAC DK 7.72 15.44
9 2 MR305QT-ND LED, RED, 5V DK 0.45 0.90
10 1 CT6X103-ND POT, 10K, SINGLE TURN DK 0.99 0.99
11 1 CT6X104-ND POT, 100K, SINGLE TURN DK 0.99 0.99
12 1 F329-ND FUSE, 3A/250V, 3AG DK 0.61 0.61
13 2 F040-ND FUSE HOLDER CLIP, 3AG DK 0.12 0.24
14 1 Q114-ND LINE CORD, 18 GAUGE DK 1.31 1.31
15 1 PC11-ND BRD, COPER 1OZ, 6X9 [1/16] DK 0.99 0.99
16 1 HM105-ND BOX, 5.9 X 3.1 X 1.8 DK 5.25 5.25
17 8 A0914-ND TERM., RING #4 DK 0.11 0.88
18 4 J241-ND STAN-OFF, 1/2′, #4 DK 0.06 0.12
19 4 SJ5508-O-ND PAD, STAND-OFF, BLACK DK 0.06 0.24
20 9 Screw, 4-40, 1/2′ HS 0.01 0.09
21 14 NUT, 4-40 HS 0.01 0.14
22 23 WASHER, 4-40 HS 0.01 0.23
23 14 LOCK WASHER, 4-40 HS 0.01 0.14
24 3 GROUMET, 5/16′ HS 0.10 0.30
25 2 WIRE, #18 LAMP CORD, 3′ HS 0.30 0.60
25 2 AC SOCKET, 120V HS 0.75 1.50
26 1 LABELING LP 5.00 5.00
Table 1 Hardware parts list for the WavePulse.
DK and HS in the Source column denote Digi-Key and Local Hardware Store respectively. LP is for a local printing company that made the labels to my specification.
Here the final hardware cost is $46.97 and as would be expected, the most expensive item is the solid state relay. However, both the box and labels take their toll. I am sure with adequete search I could find a less expensive box. The labeling cost is for a nice looking box presentation and most DIY folks can just use masking tape and marker :-).
Finally, resistors and 0.1 uF capacitors were from my own stock but would not have changed the final cost much anyway. You can get these at your local Radio Shack (Digi-Key tends to sell bulk for these items).
There are a number of ways to build this wavemaker. Since I planned on selling a number of them, I decided to make a number of etched circuit boards. However the DIY hobbyist can just use a perf- board with 1/10 inch hole spacing and wire wrap it up. No ground plane is needed since the transient currents are very low and all components can be placed on a 1/10 inch grid.
Circuit Card Etching
Figure 2 shows the copper etch layout for the circuit board. The board is 2 x 3 inches and 0.062 inch thick with 1 oz copper (bulk stock available through Digi-Key).
Figure 2 Etch Layout of the ‘WavePulse’ Wavemaker.
The first step is to cut the circuit board to 2 x 3 inches. I used a table shear to do this. Then I put up a 1/10 inch grid pattern to show where the holes go. I did this by overlaying a piece of 1/10 inch grid perf-board material that is a bit bigger than the circuit board to use as a hole guide. Place the perf- board over the copper side. The assembly is held in place by drilling holes in the perf-board at several points around where the periphery of the circuit card will be and putting in 4-40 hardware to hold the circuit board in place. I then mark the holes (using Figure 2 as a guide) to be drilled on the perf- board using a crayon and headed for the drill press. I drill with the perf-board facing up and use the marked holes as a guide. I found that I could drill four circuit boards at once using this technique (any more and the drill bit will begin to wander). Remember, the board material is glass-epoxy, so use a good carbide tip drill and you may need several if you are doing a large number of boards. Since the copper was facing up when drilled, there will be no need to prepare its surface. The back side may need some steel wool to debur.
The next step is to put some etch resistance material on the boards using the pattern shown in Figure 2 (hatched areas). There are a lot of different products made for this, but I just went down to Radio Shack and got a circuit card etching kit that included ‘stick-on’ symbols for traces and hole eye’s.
After putting on the resist material you will need to etch the exposed copper away. This is done by using Ferric Chloride (got this at Radio Shack as well) put into some kind of pan. I used a Tupper- ware tub but you can use any glass or plastic container. Do not use the container for any other purpose after etching (do not store food in it etc). Pour out enough Ferric Chloride to cover the board (about 1/4 inch). Put the board in, copper side up, and rock the container back and forth to cause a wave action over the copper. Keep this up and note how much copper is dissolving. It should take about 10-15 minutes to do 1 oz. copper and can be sped up by heating the solution. When all the exposed copper is gone take the board out and rinse well using tap water. You can pour the spent solution down the drain.
Next you will have to remove the resist material with steel wool. Then again rinse with tap water. Finally you will need to protect the exposed copper traces. This can be done with a purchased solution, but I chose to just heat the traces up with a soldering pencil and wipe with solder (since there is not much exposed copper this was easiest).
An alternative approach is to use an iron-on etching material. This method was suggested by Christopher Lorne Arko and is a good approach. This is what he writes:
It’s a printable edition of the PC etch board for the Wavemaker. More or less, I used a new (to me) technique outlined inCopymask.Essentially, it’s a way of masking the copper with a photocopier, overhead transparencies and a laser printer. You just print out the design, photocopy onto a transparency, and iron on to the copper. It transfers the toner and masks the areas — in short, it’s great! Saves time and effort — see a more detailed explanation at the site above.
I’ve attached the ‘mask’ edition of the PC board. Just resize to two inch wide, print and photocopy, then iron on. There are two, and they’re flipped. This is because most times you can never tell which side of the plastic the toner will appear on, and it sometimes ends up a ‘mirror image’ on the transparency. (In short, the pictures (figure 3) I’ve attached will work no matter which way you screw it up!). Again, the link above gives a better explanation. I was even able to skip a step by printing directly to an overhead using the laser printer I have access to.
Figure 3 Iron on image. Top and bottom images are mirrored.
Now you are ready to stuff the board with components. Figure 3 is the assembly drawing for the wave- maker. Since I was planning on selling these, I used five minute epoxy to hold the parts down as I inserted them into the board. This gave the card plenty of resistance to shock and vibration. Insert the parts from the non-copper side and solder.
Figure 4 Assembly Drawing for the ‘WavePulse’ Wavemaker.
Make sure that all parts are level and not tilted or set off the board surface too far. As the parts are put in, bend the leads at a 45 degree angle to help hold them in place (if you are not using epoxy). Turn the board over and apply a liquid solder resin to the pads and wires to be soldered. Then solder each component into place – make sure you make good solder connections.
Add wires between E3 and E4 to an LED using about four inches of wire. Do the same for E5 and E6. Use some shrink tubing over the LED leads to insulate. The LED’S will be glued into holes in the front panel later. [For high levels of production I would use little connectors here along with all the external wiring to allow easy replacement of the board – but for your purposes you will want to use the cheapest method]. Also note polarity of LED’s.
Note that Z1 is not on the board but is attached to E2 using a #4 size eye connection (available at your local hardware store). The other side of Z1 is connected to top of F1 (as F1 is shown in figure 2). This is a change from the original design and so is not shown in any of the figures. You will need to drill a hole for this Z1 lead such that it can be soldered to the copper that connects F1, C1, Z2, Z3, S1 and S2 together. Then squeeze the part onto the board – you can just have it run around the top of F1 to a hole drilled near the bottom of Z2.
I used 4-40 hardware and eye connectors for all input/output connections. This would be for E1, E2, E7-9. This was done by using a screw with washer inserted from the non-component side and attached with washer, lock washer and nut on the component side. As final assembly continued, the eye connections would be slipped over the screw protruding beyond the first nut and another nut put over that (with lock washer).
4-40 hardware was also used to attach the board to the plastic box with appropriately sized stand- offs (from Digi-Key) at the four corners of the board
Picture: 4/27/99, D500L, 150 mm, +2 Tiffen, Full Flash.
The above picture shows the assembled circuit board (less Z1). It shows where all the parts go and their orientation. Note that the 4-40 hardware has also been installed. The green and brown wires run from E-points to the LED’s.
[4 Jan 00] There are several wires that need to be installed at the bottom (non-component side)
1) Wire between ‘+’ terminal of C2 to the junction of R1, S1 and S2 (i.e. the trace that connects them.
2) Wire from U1 pin 12 to R5 (left side of R5 as shown in fig 4).
3) Wire from U1 pin 13 to R6 (bottom of R6 as shown in fig 4).
4) Wire from U1 pin 6 to R4 (rignt side of R4 as shown in fig 4).
5) Wire from U1 pin 16 to R2 (top of R2 as shown in fig 4).
I chose to use a small six-sided box made entirely of plastic to hold the WavePulse hardware. That way I did not have to worry about grounding issues. I made a series of mechanical drawings to show were the various holes and labeling needed to be. Figure 5 show the front view drawing with a 0.2 inch grid. Holes needed to be drilled for the two LED’s and two time controls. In addition rub-on labels (decals) were to be made and placed as shown.
Figure 5 Front View of the WavePulse’ Wavemaker.
Figure 6 shows the top down view drawing showing the placement of the 2 x 3 inch circuit card. Some precision in placing the four holes for mounting the circuit board to the bottom is needed so that good registration occurs with the two potentiometers, VR1 and VR2, and the holes in the front wall.
Figure 6 Top View of the WavePulse’ Wavemaker.
Figure 7 shows the back view drawing. Here three holes are needed to bring AC power in and two pump drive power cables out. I planned on using 18 gauge ‘lamp cord’ for these and a rubber grommet for each as a strain relief. I bought a standard six foot power cord with plug from Digi-Key for the input power and used locally purchased ‘lamp cord’ and AC sockets for the pump output power. The holes were drilled for 5/16 inch ID rubber grommets.
Figure 7 Top View of the WavePulse Wavemaker.
Picture: 4/27/99, D500L, 50 mm, +2 Tiffen, Full Flash.
The above picture shows a top view of the finished controller. The LED’s have been glued into the holes and the circuit board mounted to the bottom of the plastic box. All wiring was brought in through the back wall and attached to the appropriate E-points using eye connectors and 4-40 hard- ware. Note that in this picture Z1 is attach to the AC input at the left of the board using eye con- nectors. This has been changed as described above (this configuration shown in this picture will cause the house breaker to pop if Z1 shorts.
Using the WavePulse
The following link is to the WavePulse technical manual I prepare after first designing the device:
WavePulse Technical Manual
It will give you an overview of the device, specifications, set-up and use and some cautions and warnings.
After you have built the WavePulse you may want to add a number of modifications. The first of interest is the ability to run the device on 240 VAC power – this is of interest for those in Europe etc. Another thing you may want to look at is getting rid of the nearly ten watts of apparent power. This involves replacing C1 with a transformer.
In addition you may want to add some features. These include a feed button and night sense shut- down circuit.
240 VAC Input Power
A number of people have emailed me from overseas to ask about running the WavePulse on 240 VAC power. In order to do this you will need to change the values of a few components.
F1 needs to be rated for 2 Amp at 250 VAC.
C1 needs to be 1 uF at 500V.
S1 and S2 need to be rated at 600 VAC.
Z1-Z3 will need to be rated at 400 volts.
Remember that this will output 240 VAC to the pumps so size them correctly as well.
C1 can be replaced with a 120 VAC 60 Hz step down transformer to provide the 10 VDC needed for the controller electronics. Something like an 8 VAC secondary at 2.5 VA shown in the Digi-Key on- line transformer page (item 10425-ND for example). An alternative is to replace C1 with a wall trans- former capable of delivering 12 VDC at 20 mA. Digi-Key sells such a transformer. (While the original design uses 10 VDC local power, 12 VDC will work just fine). To do this you will need to still provide 120 VAC at E1 and E1 and connect the transformer primary to the right side of F1 to ground then remove C1 and connect the secondary to BRA input and ground.
This can be done by shorting out C4 with a switch. This effectively shuts off the outputs to the pumps. You will need turn the switch off to allow the system to operate after feeding.
A more sophisticated technique would allow a momentary contact switch to trigger a timer to allow say five minutes of shut down then automatically restart the wavemaker.
For night mode operation the wavemaker output is reduced to a less energetic state with pump on-times considerably shortened. Normally a photoswitch is used to make the transition between day and night mode. This circuitry can be developed by using a photosensitive switch with series riostat inserted between pins 2 and 16 of U1. The switch should be negative logic (i.e. open when light is present, closed when light is missing). When the lights are on the switch will be open and the circuit operates in normal day mode with the on-time set by VR1. At night, the switch will be closed and a new operating point will be established by the riostat that will pull pin 2 if U1 up a bit thus reducing the on-time of the pumps. Here the riostat can be set once and left alone (or after setting, replace with a fix resistor). A hole is drilled into one of the sides or top plate and the photoswitch mounted for sensing ambient light conditions.