Upgrading Old Semiconductor Equipment??

Thread Starter

ESSO

Joined Feb 14, 2015
30
Hi Everyone,

I need you all feedback/opinions about what could be the best way to update old Semiconductor Equipment. Currently our Equipment runs using the OEM SW base on windows and has a determined amount of digital/analogs inputs that we can monitor (Essentially we have used them all) and everything is integrated as part of the SW and the automation systems that we used in our area - However we have run into the need to monitor more signals (Motor Rotation, Object detection, Chemicals Flows, Sensor (On/Off), etc). Since the OEM does not support this equipment anymore. There is no possibility to ask them for a SW upgrade with our needs, in addition the price will be tremendously high.

I have thought about the possibility of installing a PLC (Automation Direct or Allen Bradley) with I/Os cards/modules and a separate Buzzer/Alarm system per machine(maybe an HMI), with in house PLC programming and support. This system will not be integrated into our automated system, however will allow us to monitor and have visibility of the actions/tasks that we want to monitor. I feel this to be the cheapest and still good professional option.

However - I am not sure if there are other cheaper/simpler options. I was looking into Turck( Modular I/O - CODESYS) but I dont know, if I will have this capability.

Thanks for reading and hope to hear your opinions

ESSO
 

MrSalts

Joined Apr 2, 2020
2,767
This is just monitoring and recording, right? No control or feedback loops, right?

will you require xp enclosures (vapors or dust)? What are you making and what process steps?
 

Ya’akov

Joined Jan 27, 2019
9,071
I concur with @MrChips. As a consultant (not in industrial automation but in information technology) I was often called in when the client had gone as far as possible with a poorly crafted “solution“. This is too late.

I have turned down several lucrative contracts because “fixing their solution” would not have solved the problem they set out to address. I have also been able to convince some clients, who understood sunk costs and my explanation of why the “solution” wasn’t one. to do the hard thing of abandoning the mistake and starting properly.

Since you seem willing to build and maintain this in-house, now is the right time to bring in someone who knows more than you to help with the design. This will ensure that you are starting out right, headed for a real solution.

Also, plan and budget for them to return at intervals and keep things on track. That way if things are drifting they can be set back on the path before it is impossible without a lot of pain.

Who knows? You might find that a consulting engineer would have a way to do some kind of integration with the older system that would improve utility. I don’t have a lot of information, but it strikes me, for example, the old system could be input to a newer one so you could have integrated monitoring instead of two silos. But that’s speculation.
 

Thread Starter

ESSO

Joined Feb 14, 2015
30
This is just monitoring and recording, right? No control or feedback loops, right?

will you require xp enclosures (vapors or dust)? What are you making and what process steps?
Thanks Mr Salts,

Yes - Only monitoring and recording,
Some sensors can be in a chemical(Slurries) wet environment- others would be in a dry environment monitoring flow
PLC will be in a secure/clean environement
To make chip.

Sensors wont be in a furnace or extremely high temperature.

Currently we can monitor these tasks through human glue(Technicians performing daily checks) but we know from experience that this can fail. by installing this upgrade - we can eliminate this human glue and have the system Alarm- This will get the attention necessary to investigate further.
 

MrSalts

Joined Apr 2, 2020
2,767
Depending how robust you want to make this, yes, you can move everything to PLC and start from scratch but, with programming, wiring, sensors and licensed/proprietary control terminals, thst can be a 100k to multimillion dollar effort.

Since you are only monitoring and I'd assume your operators can continue to validate any electronic monitoring with a visual comparison on a regular basis so the world doesn't come to an end with a minor failure, you can get something as simple as
- a sensor and old-school paper chart recorder with multi-range amplifier on-board.
- a raspberry pi (I think they have industrial versions or industrial housings available) with Any possible interface board you can imagine being available as "shields" or readily designed. Python libraries are commonly available for most hardware to write your own code in Python. Data can be collected and displayed in real time or simply stored for future quality checks. The pi can be an HTML server (full Linux-like operating system) so you can monitor via a browser on another PC or phone connected to your local network - or you can create a local network with the pi to keep an air gap from any hacking. The pi generally runs on an SDcsrd for operating system so you'd be best off writing some code to offload data to a remote system for safekeeping.
- labView Data acquisition boards and monitor/control software and interface with many analog and digital signals to run on standard PC (and possibly raspberry pi).
- ...
- full PLC solution.

it is hard to judge what's best without knowing budget, urgency, criticality and end value of the collected data. But those are some options.

Note: i wasn't joking about Simple cart recorders. I still see chart recorders monitoring humidity and/or temperature in storage areas and labs in Fortune 100 manufacturing companies.
 
Last edited:

Thread Starter

ESSO

Joined Feb 14, 2015
30
It sounds like you need to consult a professional systems engineer to assess your situation.
Thanks for the feedback - definitely that is an option, however I think the project can get stall very quickly in upper management because, then the question as to why not used the OEM to perform the upgrade will come up and then we will go back and forth many times with no progress.
 

Thread Starter

ESSO

Joined Feb 14, 2015
30
I concur with @MrChips. As a consultant (not in industrial automation but in information technology) I was often called in when the client had gone as far as possible with a poorly crafted “solution“. This is too late.

I have turned down several lucrative contracts because “fixing their solution” would not have solved the problem they set out to address. I have also been able to convince some clients, who understood sunk costs and my explanation of why the “solution” wasn’t one. to do the hard thing of abandoning the mistake and starting properly.

Since you seem willing to build and maintain this in-house, now is the right time to bring in someone who knows more than you to help with the design. This will ensure that you are starting out right, headed for a real solution.

Also, plan and budget for them to return at intervals and keep things on track. That way if things are drifting they can be set back on the path before it is impossible without a lot of pain.

Who knows? You might find that a consulting engineer would have a way to do some kind of integration with the older system that would improve utility. I don’t have a lot of information, but it strikes me, for example, the old system could be input to a newer one so you could have integrated monitoring instead of two silos. But that’s speculation.
Thanks for the feedback Yaakov! - Definitely, bringing someone else will be good, but I think could be more expensive then what my module could be interested in spending.

Even,

if it is possible to bring a third party - Do you all have some ideas of the system or technology that can apply to implement this solution regardless- definitely I would like to be more knowledgeable and understand it better
 

MrChips

Joined Oct 2, 2009
30,714
Before you choose the PLC you need to look at which fieldbus to select.
Commonly used buses are MODBUS, CANopen, Ethernet.
Choose the fieldbus first and then select the PLC.
 

Thread Starter

ESSO

Joined Feb 14, 2015
30
Depending how robust you want to make this, yes, you can move everything to PLC and start from scratch but, with programming, wiring, sensors and licensed/proprietary control terminals, thst can be a 100k to multimillion dollar effort.

Since you are only monitoring and I'd assume your operators can continue to validate any electronic monitoring with a visual comparison on a regular basis so the world doesn't come to an end with a minor failure, you can get something as simple as
- a sensor and old-school paper chart recorder with multi-range amplifier on-board.
- a raspberry pi (I think they have industrial versions or industrial housings available) with Any possible interface board you can imagine being available as "shields" or readily designed. Python libraries are commonly available for most hardware to write your own code in Python. Data can be collected and displayed in real time or simply stored for future quality checks. The pi can be an HTML server (full Linux-like operating system) so you can monitor via a browser on another PC or phone connected to your local network - or you can create a local network with the pi to keep an air gap from any hacking. The pi generally runs on an SDcsrd for operating system so you'd be best off writing some code to offload data to a remote system for safekeeping.
- labView Data acquisition boards and monitor/control software and interface with many analog and digital signals to run on standard PC (and possibly raspberry pi).
- ...
- full PLC solution.

it is hard to judge what's best without knowing budget, urgency, criticality and end value of the collected data. But those are some options.

Note: i wasn't joking about Simple cart recorders. I still see chart recorders monitoring humidity and/or temperature in storage areas and labs in Fortune 100 manufacturing companies.
Thanks - I did not thought about the Raspberry pi - But that is another option that I should consider.
Previous upgrades (not related to this one in particular) used PLC(Automation Direct) to monitor valves(Open/Close) and activate solenoids
 

MrChips

Joined Oct 2, 2009
30,714
I have been seriously thinking about your situation.
Without knowing the function, size and layout of your plant, here is what I would implement in your situation.

1) Select DIN rail mountable input/output modules
2) Select MODBUS compatible modules
3) Run RS-485 twisted pair cabling
4) Use USB to RS-485 adapter at the PC
5) Use Labview

Hope this helps.
 
Top