TS – Core

Some systems claim to run “up to” 7200 samples per second (SPS), but also state that it is optimal to run the system at a fraction of that number. Why would that be? When you have more than enough quantity for your application, improving the quality of your samples becomes paramount. Running 7200 samples a second yields unnecessary noise that actually ends up reducing the quality of those samples.

Imagine it this way: Every sample (hertz/cycle) is like a water hose (voltage) filling up thousands of measuring buckets (bits). The more times you want to fill up the buckets over the course of a second, the messier things get. Not only do you have to fill up the buckets at an appropriate speed to prevent you from splashing into the next bucket, but you have to empty those same buckets completely before you fill them again. Also contributing to the mess is a small voltage spike each time you turn on the hose. All of this adds up to a large quantity of inaccurate data.

Although our indicator is capable of running 1000/SPS, we set it to run at 600/SPS to reduce noise (drift/interference) from line power which can run at either 50 or 60 Hz, thus increasing the quality of our samples. So in conclusion, would your rather have $600 US dollars or 7200 Vietnamese dongs? If you’re confused right now, one of them is a worthless currency.

Understanding the Application

Imagine you’re watching a football game and your job is to most accurately record ball movement over the course of 60 minutes. You can log this data up to 7200 times (sampling frequency) over the course of an hour and you have 4095 spots (bit depth of a 12-bit system) on the field that can be used as specific measurement points. This would mean some pretty accurate measurements for a game of football.

Now step back and imagine the entire football game was condensed into 1 second, but instead of 100 yards, the field is now 40000 yards.

Now that the game has changed, being able to measure 7200 times in a single second is impractically redundant. At the same time, determining the exact location of the ball is far less accurate because you simply don’t have the required resolution (40000/4095 = every 9.76 yards), you’d have to bring out the chains every time!

Working with a higher bit depth is like measuring with a ruler that has finer increments: you get a more precise measurement. When the values are in finer increments, the converter doesn’t have to quantize as much to get to the nearest measuring increment.
For example:

*Resolution determined by load cell capacity (10000 lbs) x Max Lever Length (48”/12”)

As you can see by the chart above, a 12-bit system has data acquisition resolution for every 20 lbs. of torque. Whereas Alltorque’ 24-bit system has a data acquisition resolution of close to 2 grams torque.

However, this is not a perfect world and with 24-bit resolution and 600 SPS our system runs with 0.002% noise + THD (Total Harmonic Distortion). This means we are recording closer to 0.5 lbs. of accuracy.

A sample every 0.0016 seconds with 0.5 lbs. accuracy seems like it would translate into a lot of data to look at and to log. This assumption would be correct. So every 16.67ms we take 10 unique data points (taken every 1.6ms) and average them out. So when it’s all said and done, you can analyze 60 incredibly accurate data points per second on your graph. This graph is produced real-time and is not constructed after the data has been “analyzed”.

Another difference you’ll notice when analyzing our graphs is the data is extremely precise, giving you the ability to accurately evaluate the potential occurrence of:

• Dirt between threads
• Decanted running compound
• High friction running compound
• Running compound contamination
• Pipe thread imperfections
• Pipe friction from improper alignment
• Dump valve problems
• Load Cell problems
• Thread interference
• Cross–threading
• Threads with burs or razor edges
• Jaw dies with poor grip
• Dirty or worn tong, slips, or spider dies
• Spider movement
• Tong operator Error

Our graphs have a tendency to look more like seismograph readings than dead penny stock charts. This is because we are showing far more data points at greater accuracy than anyone in the industry. We use the latest in data compression code and powerful processors in our hardware that allow us to produce highly detailed graphs with ease.

Dump valves are limited to a response time of 10 – 35ms; these limitations can vary depending on the solenoid being used. To achieve faster response times, solenoids must be overpowered. A solenoid size must be selected which will produce a force at the start of the required stroke several times as large as would be needed under normal speed operation. Stroke should be as short as possible to keep plunger travel time at minimum. Solenoids will also respond faster when they are energized for a short period of time (normally off) vs (normally on) spring. We have tested both normally off and normally on solenoids and registered identical response times in both cases.

Instead of using the computer to process the data, we use our digital circuitry to process and display load cell data while directly mounted on the tongs or derrick leg. The digital circuitry uses the raw torque data from the load cell to directly engage the dump valve.

To break down the steps: First, the A/D compares each sample to the set limits of the torque specification and will open the dump valve on the first sample to reach that threshold (optimal torque). After the dump valve has been opened and during the delay of the physical release of hydraulic pressure the DAC continues to monitor the sensor output until the peak reading of the joint has been reached (actual highest torque value, not the torque value the dump valve was opened at). The results are stored and averaged on the DAC hardware until the Computerized Control Unit calls for a data sample at the rate set by the program 10-60 SPS ensuring that vital information is accurate, maintained and allowing for a high speed control of the torque joint reducing the chances of operator error with respect to correct application of torque.

Using computer algorithms to try and predict dump initiation based on the torque curve or using the value the dump valve was triggered at as the peak value is just bad practice. It can result in over/under torqued joints and completely false graph data. With the Alltorque’s unique dumping system, what you are seeing is what you are getting. No post processing or algorithms needed to compensate for poor design. Everything we are doing is for a reason, and that reason is to give you the fastest, most candid and precise dumping and graphing system in the industry.

Control Unit – Core

• Control Unit/Computer
• 10’ Power Cable
• 50’ Main Cable
• Junction Box
• TC10K – Tension Load Cell
• 6’ Load Cell Cable x2
• Rotation Counter Module (Magnetic Gear Tooth Counter)
• 5’ Rotation Counter Cable
• Dump Valve Assembly (Or components to plumb tong)
• 15’ Dump Valve Cable (4’ if Tongs are plumbed with valve)
• Spare Cartridge (For Dump Valve)
• Spare Coil (For Dump Valve Cable)
• Remote Display (With magnetic mount)
• 50’ Remote Display Cable
• 50′ Power Extension Cable
• 400W Power Inverter (For running computer from truck power)
• 720 Case (Large) x2