Page 1 of 2

Hot bench CFM rise

Posted: Thu Jul 24, 2014 8:01 pm
by tt911er
Hi everybody,

I started new topic, could't find this from previous ones.
Has anyone have CFM rise/ creep out of calibration after longer runs ( I thought because getting hot).
Mine will start to show about 4% higer figures after about two hours of testing. I do have sound mat at " engine room" so that might get heating effect worse.

Juhani

Re: Hot bench CFM rise

Posted: Thu Jul 24, 2014 8:10 pm
by Old Grey
What type of bench SF, PTS, MSD?

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 3:11 am
by tt911er
PTS, ( of cource) ;) with 8 surplus motors.

Juhani

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 3:45 am
by Old Grey
It's hard to get a PTS wrong because it's basically a box with a divider, so it would be weird if it was the cabinet.

I could do a leak test when it's hot, I mean masking tape a test orifice closed and put it on the top while running exh, and feel for leaks on the cabinet.

You can flow a test orifice (PAP 100cfm) with std cd when hot.

If it reads high by equal amounts on both int and exh, the leak is in the top section of the cabinet.

If it reads high on int and low on exh by the same amount it's leaking on the "+" tube of the incline, and vica versa for the "-" tube.

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 5:21 am
by tt911er
Thank you veru much! That's locigal, I'll test that and will report findigns after tested.

Juhani

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 8:40 am
by tt911er
After pitot probing six single heads I couldn't dublicate the porblem but membered that at least once I have noticed one motor to fail after longer period of running ;) All motors are controlled by Bruces hardware which has been working superbly!

When/if the the problem will gome again and it's not from failed motor I'll do the leak test as adviced.

Juhani

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 9:57 am
by Old Grey
Motor problems only reduce the capacity of the bench without affecting the calibration.

You are comparing the pressure between the 2 chambers while always using the same dep, so if there is less motor it doesn't matter because it gets adjusted to the same dep again.

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 4:16 pm
by tt911er
When you think about it, of course. After been able to get the greep again I'll start leak testing.
Thanks for your update!

Juhani

Re: Hot bench CFM rise

Posted: Fri Jul 25, 2014 5:26 pm
by Old Grey
I supposed you have to think about what could change when it's hot.

With intake testing most of the bench will stay cold except for the panel between the motors and the orifice chambers. Maybe a gap could open up between them with heat.

Exh could be anywhere as everything gets hot except the the inlet side of the motor box.

Re: Hot bench CFM rise

Posted: Sun Aug 31, 2014 5:34 pm
by Tony
Basically the motors get quite hot, and the copper wire increases in resistance with temperature.
Even if the supply voltage stays constant, motor current falls as the motors heat up.
Less current, less motor torque, and the rpm slowly creeps down.
Pressure is proportional to rpm squared, so even a very slight drop off in motor rpm causes the pressure to fall away.

Unless you have some kind of automatic constant pressure correction, there is not much you can really do about it.

Even if the bench has been running continuously for hours, changing the flow conditions changes motor load and motor cooling, which changes motor temperature (up or down) and it takes time for the motor temperatures to stabilise at the new operating point.
So pressure can sometimes slowly creep up as well as down.

One advantage of running a big single blower bench from a VFD is that blower speed depends only on the frequency coming out of the VFD which once set remains constant.
Motor temperature and mains voltage variations can have no effect at all on set motor speed.
This type of big blower bench will have a much more stable test pressure than a bench using vacuum motors.

If you run vacuum motors as most here do, the only real solution to this problem is to fit automatic test pressure correction.