Speed 400 Speed Control

Construction

Photo of finished unit, 60K JPEG image.

Schematic diagram, 8K GIF image.

PCB layout and loading diagram, 22K GIF image or 30K GIF image. It is also available as a compressed Autocad LT .DWG file, which I think is equivalent to Autocad R12 or so.

The hole sizes on the PC board are as follows:

• Drill 0.025 (#72) for all small-lead components except for the LED D1; this needs 0.031 holes. This includes all resistors, capacitors, Q1, Q2 and U2. (If you want, 0.031 will work fine; it's just easier to solder with more pad to work with.)
• Drill 0.031 (#68) for the servo wires and for the holes for W1. NOTE - you may want to drill it out a little larger - 0.036 (#64). However, keep in mind that the pads are 0.06 diameter.
• TO-220 devices. There are two possibilities:
  1. Drill 0.046 (#56). This will cause the device to go into the board until the wide part of the lead is reached. This causes the device to set up about 0.15 from the board.
  2. Drill 1/16". The device seats down flat. This may set up excessive stresses in the leads, and makes it harder to get a good solder joint. But, it is smaller, and it seems more rugged to me.
• Power wiring. Here, there are two reasonable choices.
  1. Drill 1/16" and use #18 stranded wire from Radio Shack.
  2. First, drill 1/16 at the centers shown. Next, using a Dremel tool with a cutoff wheel, cut a short slot in from the edge of the board into each hole. Widen the slot until it is 1/16 wide. You can now use the flexible 1.5 sq. mm (about 16 ga.) wire sold for electric flight use. Slide it into the slot, and solder it down normally. This approach looks like it would be too weak; however, it seems to hold up. It eliminates the problem of threading many fine strands through a hole.

  Neither of these is perfect. Think of something better.

There is a faux pas on the artwork. The leads on the TO-220 devices are offset in the wrong direction; they should be further away from the mounting tab than the center line, rather than closer to it. This means two things:

• Q3 and Q4 are closer to the edge of the board. This does no harm at all; it even leaves room for an IC socket and an erasable processor.
• U1 is too close to R3. This is a slight problem. but it can be fudged as follows:
  • If option 1 for drilling the TO220 holes is used, just mount R3 before mounting U1. The resistor will be slightly under the regulator.
  • If option 2 for drilling the TO220 holes is used, fudge the holes toward the outside edge of the board, and mount the regulator before mounting R3. R3 can then be mounted by bending the leads very slightly sideways.

Note that although the positions are slightly off, the orientation shown on the loading diagram is correct.

On the LED D1, polarity is indicated by the fact that one lead is shorter than the other. Put the short lead next to the asterisk on the loading diagram. Otherwise, your LED will be green for stop, and red for go.

If you are using the optional RF suppression capacitor, install it while installing U2. The picture shows which pads are involved and shows the relative size of the part. When soldering the processor pins, use a minimum amount of solder on pins 7 and 8, but use enough to cover the pads. Lay the chip between the pins with the solder-covered ends touching the solder on the pads. Use the pointed of a soldering tool to press the chip to the board while soldering. If this is not done, the surface tension is likely to stand the capacitor up on its end; the result is called "tombstoning". While keeping pressure on the capacitor, heat both pads alternately and quickly. When both are hot enough, the capacitor will settle into the reflowed solder. Hold it down until the joint cools.

I have included a spacer, screw and nut that are used to join Q3 and Q4 thermally. This worked out well, and I recommend it. Since both transistors do not dissipate maximum power at the same time, they can act as a heat sink for the other. The photo shows how to assemble them. Tighten the screw before soldering the devices; otherwise, you may overstress a solder joint.

The board shown in the photo was made using the Press-n-Peel PnP-Blue toner transfer system from Techniks, Inc. in Ringoes, NJ. The sheets are available from All Electronics . If you don't feel up to creating a PC board, this design can be built with point to point wiring on a prepunched board. Just be certain it is a fiberglass board; some friends have used the phenolic boards from Radio Shack, and they break too easily. If you do go the point to point route, keep in mind the following:

• Keep C3 close to pins 1 and 8 of U2, and connect it with short wires.
• Use at least 22 gauge wire for the main power path to the motor. This includes all of the paths with heavy traces on the PC artwork.
• Use at least 26 gauge for the power path to the receiver. If you can't figure this out, include all ground leads, all leads attached to the regulator input, and all leads attached to the regulator output.

I tried gluing the TO-220 devices to the board with epoxy. It came out through the lead holes, and made a mess of the solder joints. Don't do that.

Programming the PIC Processor - see the accompanying software document. If you lack confidence in your ability to program the processor right the first time, you can use a machined pin IC socket. Do not use a cheap socket - a friend of mine is still trying to find a PIC which fell out when he landed a little hard. Mount the socket before mounting the surrounding parts. Some of the parts' leads will have to be bent; the layout makes no allowance for a socket. If you don't want to program a PIC processor, see my offer near the end of the parts list..

-- Initial Test --

Do not plug the servo plug into the receiver. Get a volt-ohmmeter and measure the resistance across the +BAT and -BAT terminals with the switch off or W1 removed. This should be a high value. If it is not, reverse the ohmmeter leads and measure again. If both measurements are less than one megohm, find the problem before proceeding. If this test passes, re-connect W1 (if removed) and connect the battery. If you are using a switch, turn it on. Feel of U1. If it gets hot, you either have a short somewhere or have a component installed incorrectly. If it stays cool, measure the voltage between the + and - servo wires; this should be between 4.85 and 5.15 volts. If the voltage is right, unplug the battery and plug the servo connector into the proper receiver channel. Use a servo to verify that you are using the proper channel; the instructions for the receiver sometimes lie. Turn everything on and see if it works. Note: you can use an 1157 automotive light bulb in place of the motor for testing. It won't tell you if the brake works, but it's safer than a whirring propeller if troubleshooting is needed.

Copyright © 1998 by Michael J. Norton.

May be copied as desired and modified as needed so long as this copyright notice is preserved.

Back to Mike Norton's Hobbies