Digihelic controller - might suit some ones motor control needs

[jsa]

EDIT by JSA Ignore this post, it is misleading

Not low cost, won't replace PTS DAQ, but might provide a means of motor or valve pressure control.

[Tony]

[jsa]

Tony,

It has a 4-20ma output, which at first glance of the web pages seemed suitable for setting the speed of a speed controller or the opening of a valve directly.

On reading of the operating manual, I thank you for catching my misinformation.

As you are probably aware the manual reveals the 4-20ma lacks sufficient setting parameters to achieve use as a 'decent workable' pressure controller.

I'll go and write 100 times, RTFM and ignore the sales hype.

[Tony]

Not to worry, the devil is usually in the really fine print.

At first glance it did look good, but unfortunately it is not what we flow benchaholics here are really after.

[49-1183904562]

John,

Not to take this as a bullet over the bow but an opportunity to promote the power behind the PTS-DM. It should be noted that the PTS-DM is in its infancy! But depending on the ADC used internally would determine whether a closed loop Depression control could be implemented. This would commonly incorporate some type of PID circuit but may be able to be done through software. Not to open an argument based on Auto vs. Manual depression control but more as to get member feedback.

Is auto depression control something of real value?

What if it could be turned on and off?


Bruce why are you looking at me like that? ???


Rick

[bruce]

[Tony]

Here are a few things to think about.

Firstly, the test pressure MEASUREMENT SYSTEM should never be used to control the test pressure. It is a very bad idea.
The pressure CONTROL SYSTEM should be kept entirely separate.
Try to imagine what happens if the pressure transducer goes funny or starts to drift. The test pressure could then wander all over the place, but the DAQ wiill tell you it is holding solid at the set 28.00 inches.
If the measurement system is completely independent to the control system, and it is telling you the test pressure is holding steady at the correct set pressure, then it is. And you can be entirely confident that the whole system is stable and behaving itself. The aerospace and biomedical people do this all the time with mission critical or life support control systems. It a the golden rule never to be broken.

Next we need a PID control system that must be tunable to suit the blower motor characteristics. Three possible ways to do it. You can use the main computer which requires an extra DAQ pressure input channel and some type of analog, PWM, or serial data output to control the blower motor power. Or buy a commercial PID module which will not be cheap. Or lastly use two resistors, a capacitor and a one dollat chip to build your own PID system which will work every bit as well, if not better than a software based PID system.

Last thing we need is the power control module to be able to control ALL the motors together. That may be 30 to 100 Amps ?? All the motors need to be controlled because the control system needs to adjust the flow from absolutely flat out, right down to completely blocked flow, and still hold constant set test pressure. No motor switching allowed, it must be continuous smooth control all the way. That is a heck of a lot of power to control.

Only practical way to do it is by using the usual mains phase control system that everyone else uses. That is simple, although the high power switching devices and the required heat sink will be fairly large and expensive however it is done. The really big issue here is that this power switching module will be directly connected to the mains supply and all of it will be electrically alive, and potentially lethal if touched.

That worries me greatly, if this is to be a home built Forum project. Buying a commercial router speed control sealed entirely in a plastic box with a plastic knob is one thing. Constructing something vastly larger with exposed metal parts that are alive, for Forum members to construct themselves could be rather dangerous.

It is a bit like guns, explosives, and very dangerous chemicals. Each of us might feel entirely confident doing it themselves.
But suggesting that someone else (you do not really know well) handle these same very dangerous items is extremely unsettling.

Supplying a PTS 100 Amp power module all built up and ready to go will be expensive, because the parts to build it will be very expensive, way beyond what most here would considder being a viable budget DIY home project.

The last problem is this electrically alive power module needs to be connected to the computer output port, so it can be controlled.
That also requires full electrical isolation which needs to be done properly, because your life depends on it being done properly. It all adds to the cost.

The last hurdle to overcome is that mains phase control needs circuit timing changes if operating at either 50 Hz or 60 Hz. It absolutely must be accurately synchronised to every mains cycle, so the phase control system must be built into the power control module. A computer has absolutely no idea what the mains supply is doing at any instant, so that part of it cannot really be done in software without providing another mains reference input port. It all add to the cost and compleity.

O/k sorry for all the doom and gloom, but this ain't at all easy. I have been thinking about all this now, for a very long time.

My current best approach for this would be to provide a completely built up independent test pressure control module. No computer required (sorry Rick !)

It would be completely analog, with no software at all. That cuts down the cost and complexity significantly. It might be just a large plastic box that connects into the power cable going to all the motors to control the power.

It would have a plastic barbed fitting to connect an air pressure sensing line to the bench to sense the test pressure. That overcomes all the difficult mains isolation problems, and ensures absolute electrical safety in a very simple way. There would be an insulated plastic knob on this box to set, and then automatically maintain that test pressure constant by varying to power to the motors.

This approach is the best I have come up with so far, that combines electrical safety, a control system completely independent of the measurement system, simplicity, and lowest cost. This may not be at all what people want or expect, but I am wide open to any other ideas or suggestions.

[49-1183904562]

[bruce]

I'll second what Rick said! Left hand adjust the valve setting, right hand adjust the pot . . .

Once you learn your bench it's not that hard to know what is required to set the depression.

[Tony]

That is how I do it too.

But some guys here wish to have automatic depression control, which is entirely possible.
It is just not something that can be done very simply or at very low cost.

[bruce]

[Tony]

Yes indeed, the PTS-DM is a huge leap forward at an affordable price for flow bench enthusiasts all over the world.

[jsa]

[quote="1960FL"][/quote]
Rick & Tony,

What you are doing with PTS-DM is great, and the addition of some form of PTS pressure control would be a good thing.

I fairly well agree with the issues Tony has raised, but I'll put some thoughts up for discussion.

If we are relying on manual pressure setting presently, we don't know if PTS-DM's displayed test pressure has drifted or not, because we don't have a comparison instrument.

My line of thinking is that if we currently blindly accept one instuments reading as correct to manually adjust our pressure, then why not blindly allow it to adjust pressure automatically.

Certainly two instruments can provide a level of comparison checking, but if both instruments are based on the same components, designed by the same group, and of similar design, both could drift in the same manner. The result being two inaccurate instruments agreeing with each other.

Speed control of a VFD on a single blower is well served by using a VFD with internal PID and a pressure transducer. Cheaper non PID VFD's could be an option if a PTS-DM pressure controller was built.

The PTS-DM could have an output scaled to the pressure reading for input to a PID VFD, eliminating a seperate pressure transducer.

As I understand it the blokes with vac motors and router controllers are only speed controlling a fraction of the total motors with the balance being on or off. That would not 'have' to change. Auto control of one or two motors with a small speed control module/s could avoid big custom power electronics. IIRC a couple examples of modules have been posted a little while back.

Of course PTS-DM could just drive a valve actuator.

We certainly don't want to be getting involved in bespoke power control devices, or encouraging unsuitably experienced people to play with electricity.

I don't imagine anything beyond a PID and or scaled output such as 0-10V or 4-20ma.

I've had a flow performance box for some time with a PWM out for pressure control. The implementation is fine for R/C servo control but less than perfect for motor speed control. As I'm off on the LFE tangent, PTS-DM with or without speed control doesn't quite fit for me either. This is not criticism, just an observation on the different output types.

[Tony]

Some really interesting ideas there John.

As far as drift and stability go, it is assumed that the original design is adequate, and any obscure drift problem that develops would be due to a faulty component of some kind.

The thing about that, is using two completely separate systems, one to control, and another to monitor. Any fault becomes immediately apparent. That is the golden rule for safety critical systems, that it either shuts down to a safe condition, or instantly reveals a problem.

With shared parts, a fault may go completely unnoticed, because the system can be lying to you. This is called common mode failure. Telling you everything is working fine, when in fact something can be desperately wrong. People in the nuclear industry have nightmares about these types of issues.

The difficulty is, the fault may not be sufficiently bad to alert you that there is a fault. Your flow readings may randomly vary by a few percent. Just enough to drive you nuts trying to do some serious flow development.

Your idea of using a closed loop system to just fine tune one motor to hold constant depression is excellent. But it requires all the complexity of full motor control, while still requiring manual motor switching to get into the ballpark.

It is certainly well worth thinking about though, and an excellent idea, because test pressure can change as motors heat up, and the mains voltage moves around as other people switch things on and off.

A commercial VFD with built in PID controller would be a ready off the shelf solution for anyone right now. It would work either with a single three phase motor, or a bunch of single phase vacuum motors split into three equal groups. In other words, three, six, nine or twelve motors.

Most VFDs have very sophisticated load monitoring, and if the load is not almost identical on all three phases, the VFD will spit the dummy and shut down knowing something is wrong.

A VFD would not be a low cost solution though, but it would do everything required, including computer control, because modern VFDs usually have a computer serial port provided for this exact purpose, remote programming of a motor.

Add a pressure sensor and the VFD with its internal software PID controller could do everything. Unfortunately the cost will be at least one, up to to two thousand dollars. Nice, but a fairly expensive solution.

But for anyone really serious it does offer a fairly straightforward solution to automatic test pressure control.

[pablo]

I think a depression control is good for production type flowbench work. You really need a automatic valve actuator also to make it all come together. Most commercial depression controls are a seperate expensive module.
Bruce's Flowbench is a simple design. No rotary orifice plate etc. Less parts to make, less parts to break. Perfect concept for small shop, lowbuck racer, etc.
I personally would rather see a swirl meter addition for the DM.