Vacuum feedback control / PID

[drewbot]

Wow that's quite a bit of info to digest. Hijack the thread all you want TConnolly as I'll probably be using your method to control vacuum, sweep gas etc. I'll be tuning this process up for some months.

So as I understand that 1 phase output small motor VFDs exist, but are only for permanent split capacitor motors. Apparently the main reason VFD doesn't work on the single phase motors is due to the starting capacitor.
edit: looks like there is no shortage of small 3 phase vac pumps. I think that's the route I'm headed.

I need to be as precise as possible as I expect to be fractionally distilling if at all possible. I'll be considering the bleeder/butterfly valve control. Perhaps both motor speed control and gas valve control. I did notice that the new line of scroll pumps by edwards (xds6i) have internal drives and are single phase, but they're pretty overpriced.

 

[NetNathan]

What are you evaporating?
What is the vacuum range you need to control at to maintain evaporation?
I assume the higher setting of the vacuum range is the vapor pressure of the liquid to be distilled and you want to stay below that pressure.
What will be exhausted from the pump?

You said in an earlier post that it was 10-4 Torr to atmosphere. This is fairly large range.

 

[drewbot]

What are you evaporating?
What is the vacuum range you need to control at to maintain evaporation?
I assume the higher setting of the vacuum range is the vapor pressure of the liquid to be distilled and you want to stay below that pressure.
What will be exhausted from the pump?

You said in an earlier post that it was 10-4 Torr to atmosphere. This is fairly large range.

Yes you are correct that this is a very large range. I'm doing a pilot scale project which will separate each component of some interesting saps and rubber latex, paraffin wax if possible.

I expect that 3x10-2 torr is the lowest vacuum I need to achieve. The lower rating is for ballast gas and a little room for error, if I'm not mistaken. It might be a bit more than I need but I also get to build one of these for my own personal silliness. As a chemist I'm new to resins and higher vacuum conditions. The pump just adds a vacuum to the condensing column and so it won't be transferring anything aside from ballast gas. I didn't expect to stumble upon a forum where there are people who even know what vac distilling even is, so this is a pleasant surprise. Thanks again.

 

[TConnolly]

What kind of sensor are you using right now? I'm asking because most vacuum pressure sensors are kinda slow and they are highly sensitive to the molecular weight of the gasses. Gauges are usually calibrated to nitrogen, so if you start introducing organic gasses into the chamber atmosphere you'll find that you can't control the pressure because you'll really have no idea what the real pressure is. I have successfully used the MKS baratrons with low concentrations of paraffins and acrylates (much lower than you are talking about). However T/C, convectron, and piranni type gauges are not going to read correctly with higher molecular weights - and they aren't off by just a little bit either. As an example, one gauge that indicates atmospheric pressure with N2 will read about 15 torr with Ar, and thats with atomic masses of 28 and 40, imagine how bad heavy organics will be. The capacitance manometer is not sensitive to the molecular weight of the gas however they aren't very accurate at 10^-2 torr and lower pressures. You may very well need an assortment of gauges - capacitance manometer for distillation and convectron and ion gauges for evacuating the system to 10^-4 torr.

Another thing that slows down vacuum measurement is that in most systems somewhere between 1 and 10^-2 torr the mean free molecular path becomes longer than the largest dimension of your vacuum system, you no longer can entrain flow and the system becomes governed by random molecular flow - or in other words, that itty bitty connection port to the vacuum gauge body becomes an issue, which manifests itself as slow gauge response.

Also, do you also have a diffusion pump or turbo-molecular pump? Without one of those you are not going to get any real system into the 10^-4 torr range, and with one you may as well shoot for 10^-5 or even 10^-6 torr. You didn't mention one so I thought I would bring it up so you don't have unrealistic expectations.

Do a forum search for vacuum, there have been a lot of discussions on the topic.

 

[drewbot]

What kind of sensor are you using right now? I'm asking because most vacuum pressure sensors are kinda slow and they are highly sensitive to the molecular weight of the gasses. Gauges are usually calibrated to nitrogen, so if you start introducing organic gasses into the chamber atmosphere you'll find that you can't control the pressure because you'll really have no idea what the real pressure is. I have successfully used the MKS baratrons with low concentrations of paraffins and acrylates (much lower than you are talking about). However T/C, convectron, and piranni type gauges are not going to read correctly with higher molecular weights - and they aren't off by just a little bit either. As an example, one gauge that indicates atmospheric pressure with N2 will read about 15 torr with Ar, and thats with atomic masses of 28 and 40, imagine how bad heavy organics will be. The capacitance manometer is not sensitive to the molecular weight of the gas however they aren't very accurate at 10^-2 torr and lower pressures. You may very well need an assortment of gauges - capacitance manometer for distillation and convectron and ion gauges for evacuating the system to 10^-4 torr.

Another thing that slows down vacuum measurement is that in most systems somewhere between 1 and 10^-2 torr the mean free molecular path becomes longer than the largest dimension of your vacuum system, you no longer can entrain flow and the system becomes governed by random molecular flow - or in other words, that itty bitty connection port to the vacuum gauge body becomes an issue, which manifests itself as slow gauge response.

Also, do you also have a diffusion pump or turbo-molecular pump? Without one of those you are not going to get any real system into the 10^-4 torr range, and with one you may as well shoot for 10^-5 or even 10^-6 torr. You didn't mention one so I thought I would bring it up so you don't have unrealistic expectations.

Do a forum search for vacuum, there have been a lot of discussions on the topic.

Thanks once again. I was considering a turbo pump but I didn't think that the high vacuum was necessary. On the other hand, I had a hard time figuring out why there are many pumps claiming ultimate vacuums of 10^-4 range yet no bench scale distillation systems offering anything under 1 torr.

In my case I wanted to strip the overpriced vac system and put in my own. You have just outlined the complications of doing so, obviously.

On the other hand the vacuum measurements would be taken at the top of the condensing column, apart from the large organic molecule process stream. I don't know exactly what to expect in terms of the latent response time to the sensors so I can say that I have a little bit of researching left to do. Thanks for outlining some of the problems I'll be running into.

If I decide to go with a turbo-pump I will be able to achieve a higher vacuum level but if I have a hard time controlling it with a scroll pump how will this be solved by going over to a turbo?