Problem with Transducer on Hydraulic Pump

We manufacture a small press line, we have made and tested 6 of these lines. We tested them in our facilities many times (100+ for each unit) with out any problems. When we shipped it to the costumers site all they had to do is run 480v AC to the main cabinet. 2 if the 6 presses the transducer stops working after 10 – 30 cycles. If they replace the transducer with a new one it will work for 10 – 30 cycles before they have to replace it again. We have gone through 4 transducers on one pump before giving up. We were not able to test all the hydraulic pumps at our facilities.



I have had two electricians from an out side company and 4 controls guys from my company check it out with out seeing any thing electrically wrong. The customer supplied the pumps and the people they got the pumps from will not look at the pump because they are saying it is no way it is them. I had the company that made the transducer come out and they are not sure what it could be. They say electrically there is no way we could fry it (based on the set-up). They suggested some things we could do to lessen the peek surges that hit the transducer, but said the transducer should not have a problem with this.



We currently use a 3000 lbs transducer but also tried a 9000.



Does any one else have any ideas on what I can check or try?

I used to work for a API pipe plant, and we test pipes over there up to 10000 psi every 40 seconds or less. We used pressure transducers for customer data and at the same time we use murphy switches.
In order to avoid surge problems we used upstream the transducer and the murphy switch an element from murphy i don´t recall its name but is was something designed to alleviate surges on the line so that the murphy ans the pressure transducer do not soffer.
http://www.fwmurphy.co.uk/products/pressure_vacuum/pressure_vacuum_engine_gauge.htm

attached you will find murphy web site address.

hope this will help.

william

Your biggest challenge it to find if it an electrical or mechanical problem. May be worth posting on the ifps.org board, though it was down when I just checked.

You have to get seriously analytical and logical, tell everyone to get lost whilst you are at the machine. Nothing worse than someone looking over your shoulder giving suggestions and telling you that they had exactly the same problem on a 57 Ford except that it was slightly different. You have 4 machines working OK, so what is different about them, where are they, where does the supply come from, is the pipe different, is the process different. I am guessing that you are using 4-20mA, if so try adding a loop isolator. Also make sure that everything is earthed and bonded. Make sure that the control supply goes through an EMC filter.

One comfort is that most of us have been through something similar and there is always a reason and an answer. Well most of the time....

Bryan

Brandon, exactly what kind of a transducer are we talking about here?
Is this a pressure transducer?
Or is it a load cell?
Or is this a swash plate angle transducer?
Or is it a power transducer?

Or something else?

What is the transducer being used for?
Pressure control?
Or displacement/velocity control?
Or pump input horsepower limiting?
Or force control?
Or something else?

How is the transducer powered?
What is it connected to?
What kind of signal (4-20mA, +/-10V, etc)?

I'm just wondereing because your post title indicates it a transducer on the pump, not a transducer on the press - yet there are many different kinds of pump transducers.

See below in red

Alaric said:

Brandon, exactly what kind of a transducer are we talking about here?
Is this a pressure transducer? Yes, a pressure transducer
Or is it a load cell?
Or is this a swash plate angle transducer?
Or is it a power transducer?

Or something else?

What is the transducer being used for? We read the curent presure and when it gets to a set presure we release.
Pressure control?
Or displacement/velocity control?
Or pump input horsepower limiting?
Or force control?
Or something else?

How is the transducer powered?
What is it connected to?
What kind of signal (4-20mA, +/-10V, etc)? 4-20ma

I'm just wondereing because your post title indicates it a transducer on the pump, not a transducer on the press - yet there are many different kinds of pump transducers.

Click to expand...

Check the impedence of your 4-20mA input. It should be 250 to 500 ohms. If it is too high then the transducer has to work too hard to push the current through to the input (ie, it has to ouput a much higher voltage) and this is what could be burning it out.

i had a similar problem years ago.


turned out it was back feed voltage / current from the plc backplane through the analog card.

you may want to look at all power sources for secondary power.
also look at the wiring for nicks.

regards,
james

Not sure what type of controler or I/O you are using but some analog I/O lets you set whether it is for voltage or current input. Voltage inputs are high impedance while current are low impedance (see previous post explaining a little more about this).

If your's is selectable you need to make sure you have it selected properly. One type of AB analog input has a selection in software as well as a DIP switch setting that is hard to see and both need to be set to current to make everything work properly.

The Loop isolator should help here if the transducer is being damaged electrically.

Also, the transducer manufacturer should be able to give you a report on why the unit failed. Sometimes they charge a fee for the investigation but at this point, you need to determine the cause.

You should never install a transducer directly in the product line, but always use a dampener of some sort like a pigtail syphon to minimize the over pressure when the pump turns ON.

ndzied1 said:

One type of AB analog input has a selection in software as well as a DIP switch setting that is hard to see and both need to be set to current to make everything work properly.

Click to expand...

Thats right.

Brandon, what is the pressure transducer connected to?

Lets not play 20 questions. The more information we get the better the answer we can provide. The problem solving capability of the combined board is formidable, but its only as good as the input. If you can you post an excerpt from the wiring diagram (even a screen shot will suffice), and provide details on the transducer and the instrument it is connected to it will be a huge help. You say two of the six machines are having the problem, so this should be an anomoly we can help you track down.

You should never install a transducer directly in the product line, but always use a dampener of some sort like a pigtail syphon to minimize the over pressure when the pump turns ON.

Click to expand...

I don't know what a pigtail syphon is but a dampener is the first thing that came to mind. Put a fitting in-line with the transducer that has a small hole (0.030") this will keep the pressure spikes from taking out the transducer. I have seen 6000psi transducers go out on a 3000psi system because of this.

He says they work for a few cycles then die. I think if it was wired wrong he would not relate the failure to cycles he would relate it to time. But you never know.

Has to be something different from the other 4, find that and you find the problem.

I agree that it is probably a power quality issue.

More Info

Sorry, below is more detail.

Pump info:
Max pressure: 4000 lbs (But we have it limit set to 2500)
Application: Pressing metal nail plates into lumber.

Application info:
We run the press until it hits 2000 lbs of pressure they release pressure. We repeat this process over and over about every 8 minutes. We have the peruse transducer measuring the pressure from the manifold to the press.

Transducer:
Name Brand: Hydac
Max PSI: 3000 measured and 4500 before breaking. Also tried a 9000 psi
Signal: 2 – 40 ma

Electrical:
PLC: Koyo



Testing that has been done:

Electrically: Had two different electrical companies. Check the cabinet. They checked power going in, power to transducer, and verified the cabinets were identical.

Controls: I had out control engineers check as much in the cabinet as they could think.

Hydac Rep: Attached a transducer to a monitoring device. They did not find any thing wrong and they took back a transducer to see what is wrong. Still waiting on a report, I have been calling daily but they are really busy.

Pump Manufacture: Nothing, they say it has to be the electrical. Despite the fact that the cabinet worked on anther pump before shipment.



What I am trying today:

I am putting a gauge with a max pressure indicator on it, to see what the max on that port is. After I see what it comes up with I will install a snuber. I am also trying to find some one local that stocks a EMC filter (per a recommendation from some one on the forum).

I'm guessing

" Signal: 2 – 40 ma"

is a typo?

Yes, 4-20ma, my bad.