Inline graphics on this page: 92K
Questions and answers
From email conversations with readers and builders
- Mains adapter
- Recognizing choke coils
- Capacitor values
- Resistor values
- Charger voltage
- Soldering errors
- Flux and soldering iron
- Life expectancy
Mains adapter> I can't find the Mains adapter 12V DC 200mA, the closest thing > is a 12V DC 1200mA but I'm afraid to use it, don't know if 1200mA > is too much!!! Your adapter will do just fine. 1200mA is the largest current it's able to deliver. Dizzy's charger will demand much less. However, do check your work carefully before going through the special procedures for bringing the little guy to `life', as more current will be available to `let the smoke out' when there is a misplaced component or wrong connection somewhere.
Recognizing choke coils> How can I recognize the choke coils? They look like resistors, but are much larger. I put a red circle around one of them in the photo below. You might check with an ohmmeter; their DC resistance is very low.
> Is there a difference between the leads of a resistor? Resistors and coils have no polarity. Neither have most capacitors. But take care with the diodes, electrolytic capacitors, transistors and ICs. Be patient and work carefully. Try to get assistance from someone with a little more experience in electronics.
Capacitor values> I've nearly finished the PCB. Nearly, because I can't figure out > how to deal wwith some of the components - the ones which are > marked with two little bars, in the circuit diagram. Those are capacitors. The yellow circle on the left photo below marks a capacitor of 10N. The value will usually be written on it in microfarad: "0.01" A blue circle on the next photo marks a capacitor of a different kind. Its value - 47N - will be written on it in picofarad, using a three-digit code ending with the number of zeroes: "473" The red circle marks a capacitor which has been selected to be within 2% of 100N. Its value will usually be written as "0.1"
> I know the little bars only as the symbol for capacitors, but > in the diagram, those are depicted as little beakers. The beakers are electrolytic capacitors, with plus and minus connections. > I've also studied the photo on the building instructions page > (for a long time), but it didn't help me very much, because I > received two rectangular components and a blue one. The photo > shows a red one and two blue ones. The colour of the larger (polyester) capacitors can be red or blue, that varies. The smaller (multilayer) caps are usually blue. Use them as shown; their properties differ. > If you can tell me how to measure these components, than I can > do the rest on my own. You'll need a capacitance meter, or a multi-meter which can measure capacitance as well as volts, amps and ohms.
Resistor values> I'm working on my Dizzy, but I don't know how to read the > resistor colour codes... Try to get assistance from someone with a little more experience in electronics. That will make successful completion of your project much more likely. About resistors: The package usually contains resistors which show their values in a four or five band colour code. The numbers from 0 to 9: 0 - black 1 - brown 2 - red 3 - orange 4 - yellow 5 - green 6 - blue 7 - purple 8 - grey 9 - white When there are four bands: First identify the gold coloured band (indicating 5 percent tolerance; silver means 10 percent). Start at the other end and read the numbers indicated by the other three bands in sequence. The last one (closest to the tolerance band) tells you the number of zeroes. So the combination red, purple, red stands for 2700, usually written as 2K7. Green, blue, gold means 56 and minus one zero (* 0.1), so we write 5E6 (5.6 ohms). Brown, black, yellow indicates 10 and four zeroes, written as 100K. In a five-band code, the tolerance woud be brown (1 percent) or red (2 percent). The tolerance band will usually be clearly separated from the group of four which make up the value. Again start at the other end and read the numbers in sequence. Brown, black, black, orange means 100,000, written as 100K. Other codes are possible but unlikely. Our packager will usually make things easier by leaving all resistors with a certain value in a single strip; so eight resistors connected by paper strips will no doubt have 100K as their value, and a group of four indicates 1K. Of course it never hurts to check with an ohmmeter. Here is a nice page about resistors
Charger voltage> My question: What voltage should I read when measuring at the > lunchbox output plates, when Dizzy is not eating? When the current source is not doing any work, it will merely conduct, allowing the voltage to rise to whatever your mains adapter provides. When Dizzy has connected himself to the `dinner table', his eating will drop the voltage to around 5V.
Soldering errors> A week ago, I received the Dizzy package and built him. But when > I switch him on, he goes around and around like crazy (even > without the AT90S2313). When a motor runs while the AT90S2313 is not in its socket, you've made a soldering error. Because without the controller, the two motor driver transistors won't get any base current, as the schematic diagram shows. > There were sparks between ground and the point where the yellow > wire is connected to the PCB (below the battery). That also sounds like bad soldering. > With the AT90S2313 he turns around first (but that is as it > should be) and then he makes a complaining sound every 5 seconds. > When I switch him off, he makes strange crackling sounds. Perhaps > that is normal, caused by a capacitor losing charge. No, that too sounds like bad soldering or a wrong connection. > While in the special mode for adjusting the beacon (which worked > only the first time), he makes a siren-like noise and goes > in circles for ever. This might be normal, but I thought I'd > better mention it. This indicates a short-circuited feeler, or a soldering error in that part of the circuit. > I can't think of a way to give the battery its first charge now. > Do you recognize this problem and can it be solved, or is > something wrong with my Dizzy? The latter. I suspect a short between two or more soldering points. It's best to disconnect the piezo speaker and the battery. Then thoroughly clean the PCB and carefully check all connections and component positions, before you put the robot back together for another test.
Flux and soldering iron> Yesterday I started work on Dizzy. After making the cap for the > IR-receiver and soldering the wire bridges, I did the first > resistors: 5K6, 47K and 1K were soldered in place. Then I did the > same with the 12K resistor (coded brown, red, orange). But when I > checked the value with an ohmmeter, I read only about 8K, > indicating a parallel resistance of about 24K. I desoldered the > resistor and measured between the solder pads: Indeed 24K, > instead of infinite as it should be. I checked the entire circuit > board, but could find nothing wrong. > > Can you tell me what is causing this `ghost' resistance? De circuit boards are supplied in clean condition, so the resistance between separate copper traces is close to infinite, unless components or flux residue cause a reduction. Something to contemplate: In our waiting room sits a Dizzy which was soldered together with the aid of a somewhat acid and conducting flux, causing signal shorts all over its circuit board. > After reading your reply to the email I sent yesterday, I cleaned > the circuit board thoroughly with water. Following the advice of a > friend, I had bought soldering grease and used it on the first > connections. I wasn't happy with the result and had already > cleaned the board, but apparently not sufficiently. Because after > the second cleaning operation, the ghost resistance was gone! > So the soldering grease was to blame. > > Thank you very much for the tip! My pleasure. Use resin-cored solder for the electronic components; That soldering grease (or some other flux) will come in handy for the steel wire and the brass pipes. Just make sure you grease only the steel and brass, and do clean your work thoroughly afterwards. Use preferrably a temperature-stabilised soldering iron of about 50W with a fine tip, or else a 16W iron for the elctronics, and another one of about 30W for the steel and brass parts.
Repairs> I built Dizzy, and he looks very nice. > But he doesn't work entirely as he should... ---8<---
When email doesn't solve a problem, you can return Dizzy to Pitronics for repair. Make sure you package him really well. It's best to tie the drive and wheel shafts to a sheet of strong cardboard, which tightly fits in a good quality box. Also use sufficient filling material. If whatever is wrong is our mistake, we'll repair Dizzy and return him free of charge. If the error is yours, you'll receive email with a cost estimate before we start.
So far, seven Dizzies have been repaired. The conversation above continued after inspection of the first one:There was quite a lot wrong with your Dizzy: 1) All four electrolytic capacitors are misplaced. 2) The same is true for five other capacitors. 3) A diode is soldered in the wrong way around. 4) There is a 330 ohms resistor where there should be one of 10K. 5) 5K6 has been swapped with 12K. 6) Both 330 ohms resistors are used on the circuit boards, and we found resistors of 5K6 and 120K which aren't part of the package. 7) One of the whiskers has been replaced by an improvisation which can't do the job. The brass pipe of the other one is loosely attached by a lump of solder which didn't properly flow. 9) Both castor legs are unusable. 10) Bad soldering is causing several shorts and unreliable contacts. Several components on the charger board didn't survive the soldering and/or subsequent power-up: The 74HC00, the BC337 and the BC559. ---8<--- Work done: All solder was removed, misplaced components were relocated, new castor legs and a new whisker were made; and then all solder work was redone. Result: Your Dizzy is happily exploring our `terrarium'. > I think it's really supercool that you found the time to repair > my Dizzy!
A repair job like the one above results in a final price which is of course much to high. Please don't order a package unless you understand the building instructions as published on these pages, and have sufficient knowledge and skill for this project.
The second case was much less dramatic. The critical error was an exchange of two components: The 100N cap which should be part of Dizzy's voltage measurement circuit, and a 10N cap which ought to be in his charger. No doubt the colour of these caps got more attention than their values. When in doubt, read the bit about capacitor values.
The soldering was alright except for the brass and steel parts. It looked like the soldering iron had not been hot enough. Use either a 50W temperature-stabilised iron, or one of about 30W; in the latter case, a second, 16W iron would be preferrable for the electronics.
Contact glue had been used to fit the motors - clear epoxy gives a far more pleasant result.
After returning number two, his owner reported:> Dizzy arrived today (Saturday 02-02) in perfect condition. That's good to hear. > He is exploring my room now. Thank you for repairing him. > You provide a much better service than many other companies! We like to see as many Dizzies as possible `living' happily ever after :) > Dizzy still finds it a bit difficult to make contact with his > charger. I'll try sanding the edges of his the dinner plate. Bevelled edges are necessary, otherwise the castor legs will tend to get stuck behind them. > And when one of his whiskers touched the vertical plate, > his leg is often not making contact with the horizontal plate, or > the other way around. I've increased the distance between dinner > plate and wall, which improves things somewhat. Your Dizzy had no trouble making contact with our dinner table. I noticed that you had not (yet) modified the IR LED, as shown on Linkname: LED modification URL: http://www.xs4all.nl/~sbolt/images/led.gif That really is useful. And you might experiment with the LED's direction. Aiming it almost parallel to the long wall of the `terrarium' works best here. > In all other respects, Dizzy is working perfectly. :)
The oldest Dizzy has been continuously active for ten months. Every few days, dust is brushed off his wheels and drive shafts; so far, he has needed no other maintenance. The whiskers and castor legs clean themselves as long as Dizzy is `alive'. If you switch him off for long periods, corrosion will cause bad whisker contacts. In that case, the steel wires will have to be cleaned.
The NiMH battery is still doing fine. It is likely to last at least a year. A new one will set you back about US$12.
previous (message from a builder)