BanSci Hello! I am from China. It is my honor to brush your video, haha! The Mass Flow Controllers you buy are basically for semiconductor devices. I'm a semiconductor equipment maintenance technician. I look forward to more interesting videos from you. Thanks for sharing!
Thanks for the insight! I thought they were for vacuum systems from the VCR fittings, but I didn't know the application. I assumed some deposition process, definitely could be semiconductors! They were bought from a private seller in the UK.
Very interesting video. Since you were talking about GCMS a lot, i have a quick question. I know the GC usually works off of a carrier gas (mobile phase), which means it needs positive pressure to push it through. Then the MS works by ionizing the pgas samples and sending the charged particles through a vacuum. Question is - how does the GC continously pass the samples off to the MS if one works off of positive pressure and the other works off of a vacuum?
Hi, so the MS is in a vacuum which is "leaky". The sample gas "leaks" into it continuously, and a vacuum system constantly removes this sample gas to maintain the vacuum. To introduce a sample gas, you just place it near the leak. All of the sample gas doesn't all need to go into the MS, so you flow the sample gas past the leak. The leak being a tiny hole stops too much gas entering (removing the vacuum). So with a GCMS, you have a high pressure at the start of the column, that drops down to more-or-less room pressure by the time it exits the column to the detector. If the detector is a MS, then it flows past a sample port for the MS before being released into the lab or into an extraction system. Hopefully this helps understand the pressures!
Our current design uses >10 different sensors as we want as diverse a range of response to different gases as possible. Some examples are metal oxide sensors like Figaro's TGS range (i.e. TGS6812-00 and the TGS821-A00), as well as optical sensors like Sensirion's SCD30 CO2 detector, with built in humidity and temperature sensing. A big part of our research is modelling the sensors to predict the optimal set for a particular application, as there's a lot of different sensors out there we could choose.
Glad you liked it! After I did this video I found this library github.com/numat/mfc. Unfortunately for me, my MFCs are set in direct valve control mode, and the flow limit set in the web interface restricts the values input, so you can't actually control the MFC over it's full range, or even get PID control. Watch out for this, there's no way to debug this from the interface, you just notice that you can't set the full flow rate, and that flow rate varies with input pressure! I found this was true for all my MFCs unfortunately. I think the factory menu allows configuring this, but I did not have the password for that.
@@bansci the only difference I noted with my interface was a radio button option on the set point value to be digital or analogue control. I wonder if that has anything to do with being in direct valve control mode ? Appreciate the heads up and repo nonetheless.
The video is informative. How can I find out the default login IP for first time login? My MFC don't have any display. Its a mks GM50A series MFC. No default IP address found in the manual.
From the manual "The MKS IP Address format is 192.168.2.X, so the Host IP Address must be of the same format. A recommended Host IP address is 192.168.1.Y with a Subnet Mask of 255.255.252.0, where Y is some number between 5 and 254" This means you will need to scan the subnet 192.168.2.X for any http servers. You can do this using tools like NMap, or just try pinging all X=1 to X=254 values using a script. You could also try a broadcast ping, by pinging 192.168.2.255, this should cause all devices in the subnet to respond. Only the port scan is guaranteed to work though, some devices ignore pings.
It depends on what state the valve is in, as that's the big current draw device in the MFC. I would guess around 100mA for the positive voltage rail (which seems to carry the valve), but much less for the negative, which I would assume is only used for op-amps. I built a MFC controller for those older types, (oshwlab.com/m.bannerman/banflow), and use ITX1215S which only supply 200mA for pairs of these MFCs without trouble, but they are extremely expensive for this application. I would design a switch-mode power supply these days to do the same job.
@@bansci Excellent information great description for these Mass flow Controllers. I purchased two older units which appear to be a consistent manufacturing design of similar input and output functions. Only difference if the lack of digital display or interface to a computer via usb or any other typical connections. Your videos are great and keep up the Great work and explanations!!!
I have done a PhD, but it wasn't on enoses or electronics. I'm always amazed by how much specialised knowledge you can find on TH-cam, people like The Signal Path, Huygens Optics, Applied Science etc do such good jobs explaining what they know that I decided I wanted to try and be like them. I'm sure you've got some deep expertise in something, make the world a tiny bit better by sharing that knowledge!
@@bansci thanks for your reply.... I could sense it with the clarity of knowledge you were explaining... Post PhD experiences nowadays people are preferring to do mostly in different domains. Indeed you are doing a great job on spreading the knowledge you are gaining... All the very best in this journey...
Hello sir.. I’m using MFC Model no: GESDA0045025BV020 argon gas flow controller. More variation coming during machine running. Can suggest what’s the issue in this Card. I have also checked ICs and resistors no any variation found their readings.
Sir @@banscithanks for support. I serviced the valve no any seal found in damaged condition then all cleaning done and fitted again also checked the line pressure found ok & other valves are working on same pressure. Now I found only fixed value 240.43 during startup. Can suggest how to resolve this issue. when I operate the valve by hmi no any gas flow increase shown on display. sir can I connect with you if possible on mail id or WhatsApp number.
@@rajeshsain3591 Hi, I'm sorry I'm not an expert in repairing these, and I can't offer support. It may be with the oscillation that you have bad PID settings, but if the flow rate is stuck now that's a new problem. Good luck!
So the sensor is fast/cheap enough to reach that rate (or current prototype can be run continuously with about ~5 seconds of time between samples, and is around £100 in parts, so we can run several in parallel very cheaply); however, I'm not sure if it can detect TCA at the ppt level. You'd probably be relying on several co-incident markers that are associated with the presence of TCA (assuming there are any). It would be an interesting experiment to check, can you provide some "corked" corks and some control corks to see if there's sensitivity?
@@bansci I am very happy for your interest, I am happy to send samples of cork stoppers sorted by amount of Tca. if your sensor works, can you create a prototype? I am willing to pay for this service. what is the address to send the samples?
We already have a working prototype (we're even on version 5), which is why I'm happy to try to detect the TCA in the corks. Please send me an email to enose@bansci.com and I'll share contact information. I can't promise success but we will give it a try.
A friend of mine is selling MFCs cheap, if you need more. Cheap means around 30..40 Euro. I can help you to get them from Poland. They should be working.
Thanks! I've got quite a few now, more than I need, but until I've checked them out fully I may need more. If you know what model and what flow rates your friend has I can ask a colleague who's also looking for some? The lower flow rates are harder to find on eBay...
@@nkdm87 Pretty big flowrates unfortunately for our applications (microflow reactors, the nose, etc), thanks for the offer but I'll have to decline. If I need extremely small concentrations I might come back, that 10000 sccm MFC is a beast!
![#1 [Repair/RF] HP 8591A Spectrum analyzer](http://i.ytimg.com/vi/WFnWgubNcEY/mqdefault.jpg)
BanSci
Hello! I am from China. It is my honor to brush your video, haha! The Mass Flow Controllers you buy are basically for semiconductor devices. I'm a semiconductor equipment maintenance technician. I look forward to more interesting videos from you. Thanks for sharing!
Thanks for the insight! I thought they were for vacuum systems from the VCR fittings, but I didn't know the application. I assumed some deposition process, definitely could be semiconductors! They were bought from a private seller in the UK.
Very interesting video. Since you were talking about GCMS a lot, i have a quick question.
I know the GC usually works off of a carrier gas (mobile phase), which means it needs positive pressure to push it through. Then the MS works by ionizing the pgas samples and sending the charged particles through a vacuum.
Question is - how does the GC continously pass the samples off to the MS if one works off of positive pressure and the other works off of a vacuum?
Hi, so the MS is in a vacuum which is "leaky". The sample gas "leaks" into it continuously, and a vacuum system constantly removes this sample gas to maintain the vacuum. To introduce a sample gas, you just place it near the leak. All of the sample gas doesn't all need to go into the MS, so you flow the sample gas past the leak. The leak being a tiny hole stops too much gas entering (removing the vacuum). So with a GCMS, you have a high pressure at the start of the column, that drops down to more-or-less room pressure by the time it exits the column to the detector. If the detector is a MS, then it flows past a sample port for the MS before being released into the lab or into an extraction system. Hopefully this helps understand the pressures!
Hi BanSci, really interesting project, what king of sensor elements do you will use before de ADCs?
Our current design uses >10 different sensors as we want as diverse a range of response to different gases as possible. Some examples are metal oxide sensors like Figaro's TGS range (i.e. TGS6812-00 and the TGS821-A00), as well as optical sensors like Sensirion's SCD30 CO2 detector, with built in humidity and temperature sensing. A big part of our research is modelling the sensors to predict the optimal set for a particular application, as there's a lot of different sensors out there we could choose.
Amazing vid! About to write my own script to set flow rate, but thought I'd check to see if there wasn't a github repo already posted. Cheers!
Glad you liked it! After I did this video I found this library github.com/numat/mfc.
Unfortunately for me, my MFCs are set in direct valve control mode, and the flow limit set in the web interface restricts the values input, so you can't actually control the MFC over it's full range, or even get PID control. Watch out for this, there's no way to debug this from the interface, you just notice that you can't set the full flow rate, and that flow rate varies with input pressure! I found this was true for all my MFCs unfortunately. I think the factory menu allows configuring this, but I did not have the password for that.
@@bansci the only difference I noted with my interface was a radio button option on the set point value to be digital or analogue control. I wonder if that has anything to do with being in direct valve control mode ? Appreciate the heads up and repo nonetheless.
@@scotty3837 nah that radio button is only for the 15 pin variant which has an analogue input. You should be fine with that variant too
The most common unsolved problem of MAS is when the sensor going bad, also capilary clodged, heater and temperature sensor problem. Then Trashed.
The video is informative.
How can I find out the default login IP for first time login? My MFC don't have any display. Its a mks GM50A series MFC. No default IP address found in the manual.
From the manual "The MKS IP Address format is 192.168.2.X, so the Host IP Address must be of the same format.
A recommended Host IP address is 192.168.1.Y with a Subnet Mask of 255.255.252.0, where Y
is some number between 5 and 254"
This means you will need to scan the subnet 192.168.2.X for any http servers. You can do this using tools like NMap, or just try pinging all X=1 to X=254 values using a script. You could also try a broadcast ping, by pinging 192.168.2.255, this should cause all devices in the subnet to respond. Only the port scan is guaranteed to work though, some devices ignore pings.
How much current is required for one of those mks devices with the split 15V voltage rating?
It depends on what state the valve is in, as that's the big current draw device in the MFC. I would guess around 100mA for the positive voltage rail (which seems to carry the valve), but much less for the negative, which I would assume is only used for op-amps.
I built a MFC controller for those older types, (oshwlab.com/m.bannerman/banflow), and use ITX1215S which only supply 200mA for pairs of these MFCs without trouble, but they are extremely expensive for this application. I would design a switch-mode power supply these days to do the same job.
@@bansci Excellent information great description for these Mass flow Controllers. I purchased two older units which appear to be a consistent manufacturing design of similar input and output functions. Only difference if the lack of digital display or interface to a computer via usb or any other typical connections. Your videos are great and keep up the Great work and explanations!!!
Thanks for posting... Incredible knowledge... Are you PhD student or completed PhD?
I have done a PhD, but it wasn't on enoses or electronics. I'm always amazed by how much specialised knowledge you can find on TH-cam, people like The Signal Path, Huygens Optics, Applied Science etc do such good jobs explaining what they know that I decided I wanted to try and be like them. I'm sure you've got some deep expertise in something, make the world a tiny bit better by sharing that knowledge!
@@bansci thanks for your reply.... I could sense it with the clarity of knowledge you were explaining... Post PhD experiences nowadays people are preferring to do mostly in different domains.
Indeed you are doing a great job on spreading the knowledge you are gaining... All the very best in this journey...
Hello sir.. I’m using MFC Model no: GESDA0045025BV020 argon gas flow controller. More variation coming during machine running. Can suggest what’s the issue in this Card. I have also checked ICs and resistors no any variation found their readings.
Your valve might need servicing, also check your upstream pressure, perhaps you have other things switching on the line affecting the pressure?
Sir @@banscithanks for support. I serviced the valve no any seal found in damaged condition then all cleaning done and fitted again also checked the line pressure found ok & other valves are working on same pressure. Now I found only fixed value 240.43 during startup. Can suggest how to resolve this issue. when I operate the valve by hmi no any gas flow increase shown on display. sir can I connect with you if possible on mail id or WhatsApp number.
@@rajeshsain3591 Hi, I'm sorry I'm not an expert in repairing these, and I can't offer support. It may be with the oscillation that you have bad PID settings, but if the flow rate is stuck now that's a new problem. Good luck!
Please tell when will you post the design of e nose vedio
It's an active research project for a PhD student, so it might be a while, but I'll ask him what he's happy to have shared
The new video is out now, let me know what you're interested in particular
Hello I need to detect TCA in natural
Cork stoppers, can you help us?
We need to control 1.500 natural cork stoppers per hours?
So the sensor is fast/cheap enough to reach that rate (or current prototype can be run continuously with about ~5 seconds of time between samples, and is around £100 in parts, so we can run several in parallel very cheaply); however, I'm not sure if it can detect TCA at the ppt level. You'd probably be relying on several co-incident markers that are associated with the presence of TCA (assuming there are any). It would be an interesting experiment to check, can you provide some "corked" corks and some control corks to see if there's sensitivity?
@@bansci I am very happy for your interest, I am happy to send samples of cork stoppers sorted by amount of Tca.
if your sensor works, can you create a prototype? I am willing to pay for this service.
what is the address to send the samples?
We already have a working prototype (we're even on version 5), which is why I'm happy to try to detect the TCA in the corks. Please send me an email to enose@bansci.com and I'll share contact information. I can't promise success but we will give it a try.
A friend of mine is selling MFCs cheap, if you need more. Cheap means around 30..40 Euro. I can help you to get them from Poland. They should be working.
Thanks! I've got quite a few now, more than I need, but until I've checked them out fully I may need more. If you know what model and what flow rates your friend has I can ask a colleague who's also looking for some? The lower flow rates are harder to find on eBay...
@@bansci MKS 1179BX52CM14V - with Profibus and analog control, with 500, 1500, 5000 i 10000 sccm for N2 :)
@@nkdm87 Pretty big flowrates unfortunately for our applications (microflow reactors, the nose, etc), thanks for the offer but I'll have to decline. If I need extremely small concentrations I might come back, that 10000 sccm MFC is a beast!
@@bansci yes it is 😅 I got some 500 sccm for my magnetron sputtering project.
I need couple. Can I have your email ID?