Depression Controller. - does anyone know how to make one?

[Greg]

I need to pick peoples brains, I am using 10-12 vacuum motors(depends how much space they take up) and would like to set up a controller that would up the motors to maintain a set depression throughout a test. I see performance trends sell one but surely it cant be too complicated to make? or is it?

[Shawn L]

I haven't been able to finish my flow bench yet, hence have a question.
I'm aware that you have to re-adjust test pressure when making changes like valve lifts, different tests, ect. but
If you setup the test pressure, does it change over time as the air heats up from the vacuum motors?
Ie do you have to constantly adjust this over time even if no test changes are made?

[84-1074663779]

A lot depends on the characteristics of the blower you use.

For instance a large diameter single stage centrifugal blower will have a very flat pressure/flow curve, so once you adjust the blower RPM, you can vary flow up and down a lot, without it having much effect at all on the blower outlet pressure.

On the other hand, a multi stage centrifugal blower, like vacuum cleaner motors have a very steep pressure/flow characteristic. At zero flow you can often easily get a hundred inches or more of pressure developed. At high flows the pressure dies away very quickly to almost nothing. So every small change in flow means a fiddly readjustment of test pressure.

While automatic adjustment of blower speed is possible to maintain a fixed test pressure, it may or may not be easy to put into practice. The problem is going to be instability in the feedback loop. If the blower speed over corrects they might start to surge. A lot depends on how stable the blowers are and the tuning constants of your feedback loop.

It could work really well, or your choice of blower, and the the air volumes in your flowbench might make for a very tricky feedback loop tuning problem. If you end up with surge or flutter in your control device, fixing it might not be all that easy.

If your bench is very sensitive and "hairy" to adjust test pressure, give some thought to the characteristics of the air blower you are using. I do not like vacuum cleaner motors for this very reason, but they are a low cost way to get started.

[Mouse]

The FP1 has a mode where you can set a depression "window", say, a window of 9.95" to 10.05"H20 depression. Then, only when that particular depression window is met or "crossed" do you get a readout. This may only happen within a few milliseconds, but it will catch it if it occurs. When I use a shop vac, just holding my hand near the exhaust can create a "window crossing", or simply toggeling the on/off switch will work too.

I have a bleed valve prototype that really works well at controlling the depression, but not enough memory in the current processor chip to handle it. When the new processor chip is available, it will be a feature.

[Shawn L]

I was refering to physically static conditions. Ie after the head(or whatever test device) is setup on the flow bench. You open a valve at a constant height, then set the test pressure.
Now leave the thing sit there for a 20 minutes. Will the test pressure change? or should it stay constant when nothing os being modified from the original setup?
The only thing that I can see will change is the vacuum motors will heat up a bit, but will this change the test pressure substantially?

[84-1074663779]

In my experience, test pressure can drift due to mains voltage changes, and motor heating. It depends on how hairy the blower(s) are, and the type of motor used.

A large three phase induction motor is not going to be in the least bit sensitive to mains voltage or load, and will run at an almost constant speed. Brush type universal ac/dc motors can be highly sensitive to voltage, mechanical load, and motor winding temperature.

Motor winding temperature is an interesting one. It varies with motor current (load) and also the cooling airflow rate through the windings. As the same airflow rate through the bench also cools the motors, and also changes the motor load, it is difficult to guess at what the combined effect might be.

[98-1074649673]

My question would be why would u let your setup set there for 20 minutes running without making changes? I make all my tests at the same flow depression to remain constant in my testing procedure. Yes I have to "tweak" my depression from time to time but thats just the cost of performing a consistent test.

My bench uses a flatslide gate valve to control the flow through the bench. U can see it in the pics of the back side of my flowbench.

[Mouse]

Using a shop vac, just someone turning on a light in the house will cause a dip in the depression.

But if you are blowing air, and it is getting hotter, you will also experience changes due to everything downstream of the motor expanding from the heat. Orifices, pitot tubes, valves, your test ports, all expanding from the heat.

[84-1074663779]

I agree, it can be a frustrating exercise. You change the flow, set the test pressure, take a flow reading. You then look back to find the test pressure has moved from where it was.

So you go back and do it all over again, to discover the readings have changed yet again.

Oh the joys of the flowbench.

[Shawn L]

Bruce, I was using the 20 mins to explain "a long period" of no physical changes to the setup.
But I can see this happening if one is playing with the effect of port shapes on flow

Most of the time I would say one is simplying going to do a test withing a few minutes of setup. I hope I don't have to re-adjust the test pressure within that period of time or it is going to be a "fun" exercise.

[98-1074649673]

I'm sorry but, I just have to tweak my flow and take reading go on to the next test point tweak the depression take a reading etc etc . . . I don't see a big change? I perform a test 3 times once up the scale, once down the scale and once back up this gives me a good feel if my "numbers are correct" I don't see much if any change at all?

[84-1074663779]

Repeatability is really probably the most important single feature of any flowbench. If you can measure a standard test fixture and get the same readings to repeat, it is something to be really pleased about.

But if you take ten readings in half an hour, and they are all over the place, this definitely needs to be looked into and fixed.

There seem to be two root causes of unstable readings.

The first is turbulence upstream of an orifice plate. This can be reduced by having a large plenum volume and reducing airspeed as much as possible immediately before the measurement orifice. And also using as much depression as possible across the actual measurement orifice. An orifice design pressure of two inches is going to be more vulnerable to upstream turbulence than an orifice design pressure of twenty inches for example.

The second factor is blower stability. There are all sorts of factors at work here, but the biggest is probably the pressure/flow curve versus motor RPM and supply voltage. A good steady airflow is needed, not one that can rapidly vary up and down with every slight change in operating conditions.

I am now on my third orifice flowbench, and am just starting to realise the importance of some of these things. My main interest is in orifice style benches.

Bruce and Mouse have expert knowledge on pitot benches, which are quite a different beast altogether.

[Shawn L]

Sounds like I should be ok then, I was just trying to rule out a possible problem that I should be addressing in the flowbench design before it is build and find out it is too late and i have to rip it all apart again.
i must say I'm taking quite while ot finish this beast due to other priorities. Gives me time to think on my design options.