44 lines
2.4 KiB
Plaintext
44 lines
2.4 KiB
Plaintext
|
Name Chirag
|
||
|
|
|
||
|
|
1. Question text:
|
||
|
|
Design a system using this Honeywell ammonia sensor (https://drive.google.com/file/d/164WvPD7GauvseOfH-K3v8iZ4YYmQFkq-/view?usp=sharing ) to detect when ammonia concentration in ambient atmosphere exceeds 100 ppm an alert an operator in a control room about 100m away. This device will be mounted in various sections of an industrial processing plant. The complete plant will need a fairly large number (~20) of sensors so unit costs needs to be kept under control. Also, once mounted it will be very difficult to remove/service/repair the sensing devices.
|
||
|
|
|
||
|
|
|
||
|
|
2. Design summary:
|
||
|
|
Summary is ok. The current change is not fixed - you need to calibrate it - you have overlooked that.
|
||
|
|
On the receiver side, how do you check which sensor(s) have triggered and/or by how much?
|
||
|
|
Also, how do you handle the return lines for each sensor? You are powering all the sensors using separate isolated supplies!
|
||
|
|
|
||
|
|
3. Block diagram
|
||
|
|
System design is ok.
|
||
|
|
Lots of issues have been glossed over - what if one sensor (or its PSU) fails?
|
||
|
|
The current change is *extremely* small per ppm. What is the final current change and how do keep it immune from noise, drift, errors?
|
||
|
|
We discussed so many issues in class but your circuit has not even a mention of them?
|
||
|
|
|
||
|
|
4. Power supply design
|
||
|
|
Why use the same transofmrer for both 5V and 15V? Won't you have immense losses using a 15V transformer and 7805?
|
||
|
|
And even for 15V, your 24V rms secondary will give you far higher voltage than you need - headroom for 7815 can be nominally
|
||
|
|
5-7V even on a worst case basis to account for supply dip.
|
||
|
|
|
||
|
|
5. Industrial Design:
|
||
|
|
What is "pull" marked on the sensor box?
|
||
|
|
|
||
|
|
6. KICAD Design
|
||
|
|
Schematic:
|
||
|
|
Your part 1 has only the power supply. Where is the connector for the sensor?
|
||
|
|
Part 2 needed to be scalble - since all your processing is being done on this board you will have to make a 20 opamp version
|
||
|
|
irrespective of how many are actially deployed.
|
||
|
|
Why have you not used the correct symbol for OmAmp?
|
||
|
|
|
||
|
|
PCB:
|
||
|
|
Layut is good overall.
|
||
|
|
You have a 6pin SMD footprint for the line-frequency transformer - this shows a very cavalier approach to this design.
|
||
|
|
DRC fails on both designs, no outline.
|
||
|
|
Why is the LED pointing inwards? How will it be seen?
|
||
|
|
Your industrial design diagram has multiple "LEDs" but only one LED in the design.
|
||
|
|
No power supply indication? Power On LEDs??
|
||
|
|
|
||
|
|
|
||
|
|
7. Pricing / Assembly
|
||
|
|
There is no pricing information in the BOM and therefore no concrete justification for unit cost.
|