Différences

Ci-dessous, les différences entre deux révisions de la page.

--- wiki:projets:what_we_ve_done [2016/04/23 19:13]
royce [What we've done]
+++ wiki:projets:what_we_ve_done [2020/10/05 14:39] (Version actuelle)
@@ Ligne 9: / Ligne 9: @@
   * Since:
-**E** =ΔV/d   where **E** is the electric field
+  **E** =ΔV/d   where **E** is the electric field
   * We know that:
@@ Ligne 27: / Ligne 27: @@
 This means that kinetic energy of the particle: **Ec**= qV
-**Ec** must be higher than that of the ionization energy **Ei** in order to ionize. This energy must also be high enough to incite at least 2 ionizations in order to trigger a "townsend cascade". We have looked at the ionisation energies of multiple elements and decided that we would like most to study the ionisation of nitrogen as it is inert and relatively safe. The down sides are that atomic N does not exist in nature and we will have to procede with N2. This choice also allows us to build the machine using nothing but "air" and the change to pure N2 will actually be an improvement on the system as Oxygen (21% of the air) requires more energy to ionize. Logistically, we also have relatively easy access to liquid nitrogen of which we hope to create a nitrogen rich environment.
+  **Ec** must be higher than that of the ionization energy **Ei** in order to ionize. This energy must also be high enough to incite at least 2 ionizations in order to trigger a "townsend cascade". We have looked at the ionisation energies of multiple elements and decided that we would like most to study the ionisation of nitrogen as it is inert and relatively safe. The down sides are that atomic N does not exist in nature and we will have to procede with N2. This choice also allows us to build the machine using nothing but "air" and the change to pure N2 will actually be an improvement on the system as Oxygen (21% of the air) requires more energy to ionize. Logistically, we also have relatively easy access to liquid nitrogen of which we hope to create a nitrogen rich environment.
 {{:wiki:projets:ionizeng.gif?400 |}} This graph shows us the relatively high ionisation energy of Nitrogen.
@@ Ligne 33: / Ligne 33: @@
 More accurately the **Ei/mol** of nitrogen are:
-st: 1402.3 kJ/mol
+  * 1st: 1402.3 kJ/mol
-nd: 2856.0 kJ/mol
+  * 2nd: 2856.0 kJ/mol
-rd: 4578.1 kJ/mol
+  * 3rd: 4578.1 kJ/mol
-rd: 7475.O kJ/mol
+  * 4rd: 7475.O kJ/mol
--Calculate the voltage (energy) required to ionise 1 partcile
+-Calculate the voltage (energy) required to ionise 1 particle
 We now have two different generators:
@@ Ligne 69: / Ligne 69: @@
 === DC ===
 To create our DC High-Voltage generator we used a relatively simple [[http://www.ti.com/lit/ds/symlink/lm555.pdf|NE555 timer ]] and applied a astable circuit (page 10 of datasheet). The output of this pin was then linked to gate of Power-Mosfet [[http://html.alldatasheet.fr/html-pdf/37004/SAMSUNG/IRF220/245/1/IRF220.html|IRF220]]. The drain and source of the mos were then in series between a 60W (5Amp max) lab generator. The Drain and Source of the Mosfet are then placed in parallel with the leads of a recycled FlyBack transformer salavaged from an old TV screen.
@@ Ligne 77: / Ligne 78: @@
 This circuit is fairly straight forward to make and will require:
-Breadboard or CIF
+  * Breadboard or CIF
-NE555 timer
-Capacitors (0.01uf and C2 (explained below))
+  * 1 NE555 timer
-Resistors (30ohm for Ra, 1Kohm for Rb & 10ohm to place in series before the Mos gate)
-IRF220 PowerMos
+  * 2 Capacitors (0.01uf and C2 (explained below))
-Fly-back Transformer
+  * 3 Resistors (30ohm for Ra, 1Kohm for Rb & 10ohm to place in series before the Mos gate)
+  * 1 IRF220 PowerMos
+  * 1 Fly-back Transformer
 Since the signal frequency (square wave) can be determine using the formula:
-**f**= 1.44/(Ra+2*Rb)*C2
+  **f**= 1.44/(Ra+2*Rb)*C2
 Since: (Ra+2*Rb)=2030
-C2=1mF **=>** **f**=1 Hertz
+  * C2=1mF **=>** **f**=1 Hertz;
-C2=100uF **=>** **f**f=10 Hertz
+  * C2=100uF **=>** **f**f=10 Hertz;
-C2=10uF **=>** **f**f=100 Hertz
+  * C2=10uF **=>** **f**f=100 Hertz;
 Since the Flyback transformer's datasheet indicated that it's peak operating frequency is that of 50Khz
@@ Ligne 100: / Ligne 106: @@
 We must keep in mind that the duty-cycle of the output signal is:
-**D**=Rb/(Ra+2*Rb)
+  **D**=Rb/(Ra+2*Rb)
 {{ :wiki:projets:09ed74b4394.gif?300 |}}
 By choosing the resistances that we did, we can have a our desired 50Khz signal while at the same time keeping a near 50% duty-cycle. Another final note is that we can keep the same duty cycle and frequency by increasing the Resistors by a factor **X** as long as we decrease **C2** by a factor **X**. This maybe desirable to protect our NE555 from higher amps.
-We will improve upon our current circuit by replacing Ra with a variable-resistance. This will allow us to adjust the frequency while at the same time minimally impact our duty-cycle.
+This photo demonstrates the relation between increasing **Rb** from for a circuit with a fixed C2 and Ra. We can see that increasing **Rb** will increase the duty-cycle as the frequency decreases. This is not desirable in our set-up as we are generally after high frequencies and a decreased **Rb** imposes a decreasing duty-cycle.
-This photo on the left demonstrates the relation
+{{ :wiki:projets:img_1178.jpg?400 |}}
+This next image demonstrates the relation between increasing **Ra** for a circuit with a fixed C2 and Rb. As we decrease **Ra** our frequency **&** duty-cylce increase.
+{{ :wiki:projets:img_1179.jpg?400 |}}
+We will improve upon our current circuit by replacing **Ra** and **Rb** with a variable-resistance. In order to do this we will not only need to replace the resistors but also determine a means to recover the actively changing values of **Ra** and **Rb** as their values will have a direct impact on the power of the circuit.
+  **CIF BONUS**
+We chose to make our circuit using the CIF technodrill on PCB as we will be submitting this circuit to high Amps and we know that the breadboards tend to melt after just 1-2amps.
+ Our circuit was drawn on KiCad and we made sure to include 1mm traces everywhere and 2mm traces for the High Amp section of the circuit. I recommend using this [[http://www.engineeringtoolbox.com/wire-gauges-d_419.html|amp/diameter chart]] to determine trace thickness for a given amperage.
+{{ :wiki:projets:capture_d_ecran_4_.png?300 |}}
+{{ :wiki:projets:capture_d_ecran_58_.png?400 |}}
+ Since we are operating at "high frequency" we propose that the electrical charge will follow the surface of the trace as it would a cylindrical wire due to electrical skin effect. Because of this we prose that:
+ Trace width (**T**) : **T**= 3.141*D where **D** is the wire diameter for the given Amperage required
+In our case **T**=3.141*D(5amps) = 3.141*0.64mm = 2mm
+When designing a circuit on KiCad make the track as large as you can (upto 2mm, unless high amperage requires more). The CIF is capable of 0.1mm trace widths, however this level of precision requires tedious calibration that will only add time to the work. It is also recommended to increase the pad size as large as possible.
+To better understand what "as large as possible" means, we must consider the tool that we will be engraving with. The standard CIF PCB engraving tool is a 3mm 10degree engraving wedge. Pictured below:
+{{ :wiki:projets:cif.jpg?400 |}}
+Therefore our the width (**W**= of material that the tool will cut on either side of a trace will be:
+  **W**=2*D*Tan(Θ/2)  : where D = depth that the engraver will plunge into the material and Θ = the engraver tip angle (in our case 10 degrees)
+The thickness of the copper on PCBs can be found [[http://www.sunstone.com/pcb-capabilities/pcb-manufacturing-capabilities/copper-weights|here]]. Standard thickness is 0.35mm. Therefore, D>0.35
+This implies that **W**>= 0.06mm = 60 microns
+Gallaad (the CIF operating system) will automatically off-set the tip of the engraver a distance of **W** from
+the track. We will see however, that while fixing the PCB to the CIF, the distance from the engraver tip to the "XY plane" (**ΔZ**) will vary around 0.05mm at very best. Since the machine can not compensate for this, the variation in the **Z** will result in the narrowing/enlarging of the track width itself.
+The width that is lost from the track (**W'**) is equal to 2*ΔZ*Tan(Θ/2)
+In our case, it is only 0.008mm. This however, is the BEST CASE SCENARIO. Until a CIF user is confident in their abilities and methodology, a **minimum distance** of 0.2mm between traces should be respected. This value will increase if the relative flatness of the "XY plane" is poor & if the angle Θ increases.
+To best decrease the variation **ΔZ** apply the double sided tape evenly over the entire back surface of the PCB.
+Using a multi-meter set to "ohms with beep indicator" attach the cathode to the engraver with the spindle off. Attach the anode to the copper plated PCB. Slowly descend the engraver on all four corners & center of the PCB and note the **ΔZ**. Apply the above forumals to your setup, and verify that your track routing will support the variation with compromising the final circuit.
+{{ :wiki:projets:img_1174.jpg?400 |}}
+//Note// There are minor changes from the above KiCad photo to the final circuit
+Our circuit was finally routed with relatively large insulating routes around the track.
+The machine spindle was set to 1/2 speed and the advancement was set to 8mm/second.
 === AC ===
@@ Ligne 168: / Ligne 222: @@
 -add margin of error-
 ==== The Vaccum ====

FablabSU

Outils pour utilisateurs

Outils du site

Différences

Outils de la page