Budget Digital Vacuum Sensor, read a 60$ pirani gauge through your PC

diy
guide

#1

In the hunt for a reasonable budget vacuum gauge @JacobsLadder was kind enough to mention the MKS 901p sensor and provide some information to help me get it working properly.

This gauge can be easily purchased on ebay for quite cheap, around 55$ shipped.

The transducer easily communicates over serial port to a PC. This requires 2 wires for serial communication, and a separate power supply that ranges from 9vdc to 30vdc. You want to choose a power supply with a very low power output. I searched through saved walwarts I had and found a 12v with only 300ma output. Some individuals might choose to power the tranducer with a battery. I used pin jumpers to easily connect the wires for testing but I plan on cutting up a serial cable I have to make a more durable version.

This is what my wiring setup for testing looks like:

With labels:

Pin Diagram:

I used an older PC I had with a serial port but you can easily get serial to usb adapters.

The transducer has a standard KF16 fitting. You want to mount it with the connector upright. I would read some of the manual which ideal vac is nice enough to post: https://www.idealvac.com/files/manuals/901manual.pdf

The best serial program I’ve found is the opensource ScriptCommunicator, https://sourceforge.net/projects/scriptcommunicator/

First go into settings and use the default settings for the 901p sensor, you can also connect on this screen, your com port might be different but SC usually finds it:
connect

Go to Scripts, add the vacuumpressure.js:

When you start the script it will first ask the transducer what unit it is set to. When you click the Start Polling button the script will ask the transducer for the pressure every 1 sec and post it to the large window, add it to the graph, and add the local timestamp and pressure to a table (which can be copied and pasted to excel). When you click the button again it will stop polling. If you use the comboboxes to change the unit or gas type the script will send the command to the transducer to make the change.

The code is stable and works well to take readings but is still a work in progress and I’m in the process of making the plot range values auto adjust to the readings over time, as well as adding commands to the menu structure such as changing/reading some of the other transducer settings and working with multiple transducers.

Here is the script running on my welch 2090b:

And when I turn it off you can see it return to ATM:

Here is the script:
http://s000.tinyupload.com/?file_id=61892485709490171897

This is the thread where we first talked about it and @JacobsLadder mentions his arduino powered version: Vacuum gauges

My plan is to put a monitor near my setup and use a wireless mouse to control the script. I had the font a lot larger but changed it when debugging but I will update the thread when I’ve finished making stable changes. At first I was going to build the arduino setup and then decided I could do more with the PC version. I would like to figure out a good way to read thermocouple temp values and combine that with the vacuum readings in a table format…

I was super happy to get it working and to confirm that my 2090b is truly a beast, pulling 25 microns (it is rated @ 33), which is really damn good for a chemical duty diaphragm pump. Here is a picture of the badboy:

If you want to know how your vacuum is sucking and you don’t have thousands of dollars to spend doing it I highly recommend the 901p transducer.


#2

I’m so glad you got this going. I’m almost done with my diy write up for anyone that wants a smaller lab box that they can put together themselves. I’ll supply the code and all the schematics for anyone that has some basic wiring skills. Great job!


#3

Exelent job gentelman :pray::pray::ok_hand:


#4

Do You think iT s possible to run on a rasberry computer ?


#5

What about a Bluetooth to serial adapter? And a cellphone…

https://www.ebay.com/p/Bluetooth-Rs232-Rs-232-Adapter-Serial-External-Dipole-Antenna-100m-to-USA-Only/20025212424

@Bill?
@Doc?


#6

It is possible but arduino is easier. Here are some guys who have detailed several arduino projects that might give some clues:

http://www.fusor.net/board/viewtopic.php?f=10&t=11730&p=77957&sid=644a4003abb4870dc0d74e7c975a2637

http://www.fusor.net/board/viewtopic.php?f=10&t=11533


#7

How big is your head ?
Darnit that would rock vac on the phone WiFi conect long distance monitoring :grinning:


#8

Yes, you could use a raspberry pi. It uses a ttl serial connection same as the arduino. So, you’d need a rs232 to ttl converter as well. Completely doable though.


#9

The only issue i see with a bluetooth to serial adapter would be the power requirements of the sensor. It needs a 9v-30v power supply. I have a lot of android/bluetooth development and could probably put something together pretty easy. I’ll have to see what kind of bluetooth adapters i have on hand. The android app would be a piece of cake. In fact, i think i have something that would almost work out of the box that i developed for another prototype.


#10

Which is where the 9v battery comes in…every other digital vac gauge uses a battery, so it’s not a huge problem.


#11

Hmm… Not a problem at all. I do have a few serial to bluetooth adapters on hand. I’ll need to see if they’re using ttl logic or not. If they don’t it’ll add a bit of complexity to the equation. Now that i think of it. They’re almost certainly ttl logic since they rely on a 5v power supply, but i might be wrong. Ill need to investigate. Someone should pay me to return to being a software/hardware engineer so i can ditch the extraction business:) Any takers?


#12

Well the power supply does not have to be connected or a part of the interface. My power supply and sensor are about 20’ away from my computer. Only the 2 signal wires travel from the sensor to the PC. Serial to bluetooth could be really challenging or really easy depending on the software interface of the adapter. If it is easy to get at the bits and send bits through it then it would be pretty easy to use.

I’m doing a little research to see what an android development would look like. I think if I had an rs-232 to bluetooth I could figure it out pretty easily, but it doesn’t really make sense for me to get one. (most of us Mainers are pretty poor). Google has a lot of examples posted and there are a few bluetooth libraries that would make it pretty easy. The idea of having it go to a tablet seems pretty cool and would open up some neat functionality. The transducers do have built in relays that could be used and it could be easier to work with multiple transducers over bluetooth I think.


#13

The 2090 is rated for 25 Torr (not micron) and 33mBar. I had to double check because I’ve never heard of a diaphragm pump pulling that low. Probably just a simple typo. Great thread and post. Very cool to see where this develops.
For the script is that something written by yourself?


#14

Ah I figured the mbar and torr were fairly interchangeable but I guess the deeper you get there is a pretty big difference.

Yeah I wrote the script and plan on improving it some before I change around my vacuum lines. I used the example worker script that came with ScriptCommunicator as a skeleton template. One thing that threw me off some was that I wrote it on my desktop with a usb to serial adapter which sends the packets whole, while using it in my lab with a serial port the packets come in pieces so you have to save the pieces and then check for packets. ScriptCommunicator implements the QTscript language which is a special version of javascript and includes some handy widgets such as the plotwidget that I use to make the graph.

I really like that vacuum pump and its such a great value. In the short time I ran my rotary vane pump I spent about 80$ on oil, after my next run the 2090 will have paid for itself just in vacuum oil, let alone the other challenges that go with the rotary vane.


#15

The smaller version on that wob-l will actually get down to 5 torr. I think both the 2581 and 2585 both go lower than the 2090. I was looking at all these in November.

And when I mean smaller I mean a tad bit smaller cause those pumps are pretty large.


#16

Those ones aren’t chemical duty whereas the 2090 is in the chemical duty category. I know a lot of people have done ok exposing the wob-l series to some vapors but I like being able to not worry about it altogether, but yeah, it is a big pump.


#17

Thanks for clarifying that, and thanks for correcting that information. Don’t want to lead people down the wrong path.


#18

have you gotten any further with the bluetooth research? how about a simple usb to mini usb for plugging into a cheap android tablet then we can get an app made to read the signals? id love to wire up a pirani and use the many old tablets i have kicking around as a vacuum gauge.


#19

lmk if i can buy or send you any parts to get this accomplished? id love to have an app that reads these 901p sensors!


#20

Meh, i just more time in the day. I have a simple version working right now, but it requires an arduino for it all to work. I don’t think a standalone 9v battery powered setup is a real possibility right now. Of course its an achievable objective, but with my workload right now i don’t see it happening. I was thinking about adding it to the vacuum gauge boxes that i’m putting together for people. It’ll just take me some time to put it through its paces. That wont happen for some time though. The vacuum gauge boxes will be ready for purchase after i get a few into the wild and make sure they’ll work in a lab environment. I use my prototype box in my lab all the time, but I need to get more feedback from other users. I’ll add bluetooth to them after I get the dual gauge version locked down.