Project Logs

Collapse

• In operation

  lukasz.iwaszkiewicz • 10/25/2016 at 11:47 • 0 comments

  Project finished, 3 boards reflowed so far (2 for a client). Oven rewired to 3,6kW.

• 4°C per minute

  lukasz.iwaszkiewicz • 02/07/2015 at 20:24 • 0 comments

  Yay! I just rewired the oven in parallel, so now it has 3200W. It heats up to 250°C in about one minute (4°C/s)! That's more than enough! Now the software has to do it's job and it has quite big reserve.

• The oven

  lukasz.iwaszkiewicz • 01/31/2015 at 22:35 • 0 comments

  First tests with the oven (9l, 800W, the cheapest I could find which heats to 250°C) shows that :

  1. It heats up to slow (I suspected that it may happen :d).
  2. It smokes and stinks as hell, I can't turn it on at home, at least not when my wife is around :)
  3. Even after insulating it with something which my father in law gave me (I'm afraid it may contain asbestos) it heats up only to 230°C in about 7-8 minutes.

  PS. I bought 2 heater springs as a replacement, 1kW each (and about $2 each. The oven cost about $30 or less). I hope not to blow this contraption up.

• 9

  lukasz.iwaszkiewicz • 01/29/2015 at 09:05 • 0 comments

  Just a quick update : made the PCB. Only 1 unrouted wire left (as always), but I'll fix that in no time this time :D New PC arrived (I'ts a laptop, I've never had a laptop!) and I'm excited because it is even faster than my old overclocked i7. Also the oven has been sent by the seller yesterday, so I'll expect it to arrive today or tomorrow. I have two syringes (extra small ones, about 2 grams [fraction of ounce - how much is the ounce btw?], only for tests) of soldering paste, so I'm ready. I bought both leaded and unleaded, because I don't know if the oven will manage to heat to 250°C although it is advertised as so (unleaded paste requires higher temp).

• WIP

  lukasz.iwaszkiewicz • 01/28/2015 at 08:21 • 0 comments

  Day 8: Now, that the prototype, and firmware works, I did the schematic and started to route the board (that was the night before). Tonight I'll try to finish the board. Since I do not do PCB by myself (mostly because I have 3yrs kid running around, and I don't want that he drank iron chloride, whether accidentally or on purpose), I'll have them fabricated by some local fab. Oh, and seems that the hackaday.io profile is invisible. I know that this kind of oven stuff may be boring, but I thought there are more bored people / hobbyists outside.

• Day 6, 7

  lukasz.iwaszkiewicz • 01/27/2015 at 08:45 • 0 comments

  Had a few problems, but firmware is more or less complete. It is as simple as it only could be. Got the casing and it seems to be to small to house the small PSU which would power the relay (relay is opto-coupled from the logic, and logic is powered from the USB) thus I'll go with external wall-wart, oh well. Tonight (when kid go asleep :D) I'll try to sketch the PCB.
  

• On-off

  lukasz.iwaszkiewicz • 01/26/2015 at 10:01 • 0 comments

  After few hours spend on reading about control theory and PID controllers in particular I came to the conclusion, that PID controller is not suitable for my oven project mostly because it's on/off device. Although you can set temperature on most ovens, there is no simple way for me to control the knob on the panel - this is obvious. So instead SSR (a relay) is used, transforming the oven into on/off device. In a PID controller I is used to speed up the process increasing the overall value (lets name it u, where u = P + I + D). So lets assume, that setpoint (SP) equals 100°C, and PV (actual temperature) equals 20°C. Thus the error (e) is 80, and hence P > 0, the u becomes 1, oven starts to heat up. I increases every Dt (delta t), because it accumulates the past errors, but heating cannot be speed up : you cannot turn it on more. Similar consideration goes to D.
  
  So for now I think, that the best solution will be to implement a on/of (bang-bang) controller with deadband (deadband dependent on thermal capacity of the oven used).

  Fixed some issues with thermocouple.

• Video, day 5

  lukasz.iwaszkiewicz • 01/22/2015 at 23:44 • 1 comment

  Day 5: Implemented P.I. controller (rudimentary). Successfully achieved 100°C without noticeable oscillations, but this is due to the fact, that I am using small toaster which has very low thermal capacity, and heats up almost instantly (and cools down fast).

• 1AM & 100W blinky

  lukasz.iwaszkiewicz • 01/21/2015 at 00:04 • 0 comments

  Day 4: USB control requests are working as expected. I like them because they maintain data integrity i.e. each request has its own name (number, and 2 additional words) and it is apparent what it's doing. In my case there is only one request "GET_TEMP_REQUEST" (check GitHub) which sends back 32 bits of thermocouple data to the host. The highest measured temperature of light bulb in my desktop lamp is : temp = 274.75°C. It's 1AM, and I must wake up at 6AM o'clock. Oh well...

  Day 5: Added the relay. Lesson learned : beefy solid state relays won't work with little load. My works when load has at least 24V and 50mA (or more ?). So I quit supplying 5V and other funny voltages, and went with 230V AC. Little scary, but at the end it would operate with high voltage, so what the heck. So tonight I basically made 100W blinky :D Also, tested the thermocouple with toaster, and it peeked at about 200°C.

• First steps

  lukasz.iwaszkiewicz • 01/19/2015 at 06:37 • 0 comments

  Day 1 : MAX31855 testing with TM4C123 launchpad if I remember correctly. It was a while ago. Data acquired via USB.

  Day 2: MSP430F5529 firmware (basics). Added URL.

  Day 3: µC and thermocouple ADC see each other. Raw data (32 bits) is being sent to the host via USB CDC class connection. Tomorrow I'll try to switch to clean and tidy USB control requests, and maybe make some piece of host software to visualize the temp. readings. There's not much time left, in week or 2 all pieces of CPU-gauge will arrive and I want to solder the thingy in the oven (don't have the oven itself even!).

View all 10 project logs