Menu

Articlets

Recorders FAQ

Recorders FAQ

Last updated, May 2008 by Daniel Burk, IST Applications Engineering

General Information: All Models 
EDR3-series recorders including the MSR-1 and MSR3C 
EDR4-series recorders 
Snapshock Plus



1. General Information: All Models

1.1 How often should I replace the batteries in my recorder?
1.2 What kind of batteries should I buy for my recorder?
1.3 What types of accelerometers do you use in your systems?
1.4 How do I mount this recorder to my test?
1.5 What kinds of tips do you have to ensure that I get good data?
1.6 I inherited this recorder from another office. How do I make it work?
1.7 I've had this recorder for several years. Do you have any updated software?
1.8 I can't get this recorder to communicate with my computer! What is wrong?
1.9 What is an anti-aliasing filter?
1.10 My company is working on ISO9000 certification. What is your recommended calibration interval?
1.11 How does sample rate affect my recording time?
1.12 My recorder has a lithium battery inside. What are the transportation restrictions?

 

1.1 How often should I replace the batteries in my recorder?
The answer can be a very complicated, however, we recommend a very simple one. You should replace the batteries in your recorder before each and every field test. The way we look at it, it is the cheapest insurance policy around.
1.2 What kind of batteries should I buy for my recorder?
Nearly all IST models are designed to use common batteries such as 9-volt, C-cell, or D-cell batteries. In all cases you should use the highest quality alkaline battery you can find. If your system uses a 9-volt battery, you should look for a non-metal jacketed lithium 9-volt such as the Ultralife UV9L. The Lithium has nearly double the energy capacity of the Alkaline, and nearly four times the capacity of a Zinc-Carbon. Triaxial Snapshock Plus systems should always use the Lithium battery. An Alkaline such as ProCell, Duracell, or energizer will work in a pinch, but offers reduced operational time.
1.3 What types of accelerometers do you use in your systems?
Most of our recorders feature an internal triaxial accelerometer cluster. The cluster consists of three piezoresistive bridges mounted on silicon micromachined beams. These accelerometers measure DC acceleration including Earth Gravity. This yields a measurement bandwidth starting at DC to upwards of more than 1600 Hz, depending on the measurement range.
Most of our systems also offer the ability to use a special low-bias current piezoelectric accelerometer by PCB. These Voltage-Mode accelerometers do not have the limitations associated with conventional charge mode P.E. transducers. The bandwidth of our P.E. acceleromter may range from as low as 1 Hz to well over 15 Khz. The P.E. accelerometer is handy in situations where mounting of the recorder is a problem, such as mass loading or limited space considerations.
1.4 How do I mount this recorder to my test?
Mounting considerations should ensure that the recorder is affixed to your structure as firmly as possible. IST offers the MMB-110 magnet set for attachment to ferrous surfaces that experience low levels of vibration. Other methods such as 3M double-sided tape work well in low-frequency low amplitude acceleration. Rivet-Nuts from McMaster-Carr work well for reuseable installations in sheet-metal or fiberglass. Unusually shaped structures may require custom mounting fixtures. Contact us to discuss your mounting considerations.
1.5 What kinds of tips do you have to ensure that I get good data?
First, always use new, quality alkaline batteries. Next, be sure to use the recorder in the office several times to get used to the behavior of the system. This is an advanced piece of instrumentation with a lot of electronic controls: There is a learning curve associated with any piece of advanced electronics. Test your recorder setup a few times before actually commiting it to a test. Lastly, gather as much contextual data about the test as you can. Take pictures of the installation and surrounding environment. This ancillary data may help answer questions about unusual or unexpected behavior that you find in the field.
1.6 I inherited this recorder from another office. How do I make it work?
If you find yourself with an old IST data recorder, give us a call. We probably can supply you with copies of the user manuals, and any applications software that you need to run it. In addition, we have records of the system that show the calibration and service history. In many cases, the old hardware can yield great results when used with updated software.
1.7 I've had this recorder for several years. Do you have any updated software?
For some products, Yes. If you have any questions about the software, contact us. Quite often, we can send you the latest version electronically as an email attachment.
1.8 I can't get this recorder to communicate with my computer! What is wrong?
Most communications problems are at the computer end, such as hardware conflicts with a mouse or IRDAdriver. Sometimes it is as simple as a dead battery in the data recorder. Communications problems are not always easy to diagnose, but IST will help you troubleshoot the cause.
1.9 What is an anti-aliasing filter?
The anti-aliasing filter (AAF) is essentially an insurance policy. If you run a finite sample rate when digitizing data, you have to be on guard against signals of appreciable amplitude that have frequency content of 1/2 the sample rate of greater. Since it takes a minimum of two data points to reconstruct the frequency content of a waveform, signals with frequency content greater than half the sample rate cannot be reconstructed. These "aliased" signals fold back to form erroneous signals with lower frequency attributes. Once aliased data enters the digitzed data stream, there is no way to remove it, or tell it from real data. Thus, filters are used to restrict the bandwidth of the data acquisition front-end. This reduces the possibility of aliased data. Most anti-aliasing filters are chosen at about 1/3 the intended sample rate, as a minimum.
1.10 My company is working on ISO9000 certification. What is your recommended calibration interval?
ISO9000 certification requires that you set at least some kind of calibration interval. It is up to you to determine the length of the interval. IST reccomends a 12-month calibration interval for all IST equipment. If you can establish a history of calibration for the recorder and show that the system has experienced very little drift with regard to the calibration, you can extend this interval to a longer time frame and still be within ISO9000 requirements.
1.11 How does sample rate affect my recording time?
Since the recorder has a fixed amount of digital memory, it can store a fixed number of acceleration samples. How you choose to use this capacity is up to you: The faster you sample the environment per unit of time, the less overall time-history you can store in terms of number of seconds of digitized time-history. For instance, a 4MByte equipped EDR3C can digitize about 33 minutes of vibration when running at 512 samples per second. If you double your sample rate, you cut your available storage time in half. The number of digitized data points remains constant.
1.12 My recorder has a lithium battery inside. What are the transportation restrictions?
None! All recorders that use a lithium battery that meets all transportation specifications for unrestricted access. For more information, refer to the May 2000 hardware tech tips.

2. EDR3-series recorders including the MSR-1 and MSR3C

2.1  What is the difference between an EDR3 and EDR3C?
2.2  What is “autozero correction circuitry”?
2.3  What is a “trigger source”?
2.4  What is a “recording engine”?
2.5  What is a “channel set”?
2.6  Can I use the internal accelerometer and the external accelerometer inputs at the same time?
2.7  How can I use the single-ended inputs on my recorder?
2.8  I set my acceleration threshold to 5.0 g and it came back as 4.92 g. Why did it change?
2.9  My acceleration readings look like little steps. They never read zero g either: It’s always on one side or the other of zero. Why?
2.10 My system didn’t record any acceleration. How do I know if it was even recording at all?
2.11 How do I calculate the maximum amount of digitized time-history data that I can store?
2.12 What is the difference between “Overwrite” and “Sliding Window Overwrite”?
2.13 How do I calculate the maximum number of events that I can store in my recorder?
2.14 I want to use the quick-start menus, but also want to configure my environmental settings. How do I do this?
2.15 I have no idea on where to start choosing my setup parameters. Got any suggestions?
2.16 My recorder has nonsensical humidity readings. Temperature too! What is going on?
2.17 Will IP alcohol affect my humidity measurements?
2.18 What is the lock switch? How is it different between the EDR3 and EDR3C?
2.19 Why should I NOT use the lock switch?
2.20 How do I use the “Delayed Start” function?
2.21 Explain to me how the Sliding Window Overwrite process works with regard to the EDR3-series recorder
2.21.1 With regard to the EDR3-series recording engine
2.21.2 With regard to the EDR3C-series recording engine

2.1  What is an EDR3 and how is it different than the EDR3C?
The original EDR3 used a Hitachi-based microprocessor with a built-in EPROM. When the recorder was factory-calibrated, the calibration numbers were burned into the EPROM. Subsequent calibrations were burned into new microprocessor chips which were then installed in the recalibrated unit. Since the chip is no longer available, we have devised an alternate method for storing calibration numbers in RAM. With the Hitachi chip no longer available, the recorder was redesigned to incorporate a Motorola chip. The EDR3C is the result, and it sports more features than the EDR3 such as Sliding Window Overwrite, and faster download speeds. Calibration numbers are stored in EEPROM, so we no longer need to burn a new chip each time the recorder is calibrated.
2.2  What is “autozero correction circuitry”?
The Autozero Correction Circuitry (AZC) is a method used to deal with offsets that result from the use of an accelerometer with DC response. With DC response, the accelerometer can measure earth's gravity. This shows up as a constant 1 g of acceleration. Subsequent orientation changes of the recorder with respect to earth gravity show up as acceleration change. The AZC helps "trend" this acceleration away. AZC also eliminates any offset changes associated with temperature drift. As the piezoresistive bridge heats or cools, it changes bias due to thermal expansion. AZC is NOT a low-frequency filter. You can't rate it as a function of decibles per octave: It is a function of percentage of full scale signal per second.
2.3  What is a “trigger source”?
The recorder is always measuring acceleration, and storing acceleration in the circular buffer (used for pre-trigger data). It will not store the acceleration in RAM until a valid event condition becomes true. The valid event condition may be set true from any number of trigger sources: The accelerometers themselves, an external device (such as a TTL-level signal), or periodic measurement of temperature. Thus, acceleration events are recorded in memory whenever any of the trigger sources are turned on, and considered true.
2.4  What is a “recording engine”?
The recording engine is the actual electronic board located inside the housing of the recorder. Some recorders such as the MSR series, or the EDR3D may contain multiple recording engines. Each of these recording engines record three high-speed channels such as acceleration, and two periodic "slow-track" channels such as temperature and humidity. Thus, to construct a nine-channel MSR3C, we use three recording engines that work in tandem.
2.5  What is a “channel set”?
A channel set is a grouping of some sort of channels. In the conventional EDR3C-series recording engine, there are two channel sets: The high-speed channel set, and the slow-track endironmental data channel set. Once data has been collected from an IST recorder, DynaMax uses these channel sets as the basis for analysis. The 32-bit application, DynaMax Suite, can actually synthesize new channel sets and build "virtual recorders" based upon your specialized needs.
2.6  Can I use the internal accelerometer and the external accelerometer inputs at the same time?
A single recording engine can only access three high speed channels at a time. Thus, EDR3 and EDR3C recorders can access either internal (differential) accelerometers, or external (single-ended) accelerometers, but not both at the same time. Recorders with multiple recording engines such as the MSR-3C or the EDR3D can access both single-ended and differential acceleromter channels, but only in groups of three. You cannot "mix & match" single-ended and differential channels within a single channel set.
2.7  How can I use the single-ended inputs on my recorder?
The single-ended inputs on each recorder are designed primarily as a means to measure from voltage mode piezoelectric accelerometers. These accelerometers require a bias current to operate. Thus, the IST recorder provides a constant 400 microamp current excitation on each single-ended channel. If you wish to use only one of the three single-ended channels, you must first provide some sort of current path on the other two channels via two 6.8 Kohm shunt resistors. IST sells these resistors in a handy 10-32 cap. Ask for the SHNT-1 channel shunt. These single-ended inputs are frequently used with other transducers such as rate gyros, strain gages, or load cells. For other interface requirements, call IST Technical Support, or check for the latest hardware users manual.
2.8  I set my acceleration threshold to 5.0 g and it came back as 4.92 g. Why did it change?
This is a phenomenon called "quantization error". The calibration of the recorder is stored in a digital number. A digital number is by nature quantized: There are only a finite number of values that can be stored which is dependent on the number of bits used to form the digital word. In this case, 5.0 g falls between two quantized values and thus is rounded to the nearest value.
2.9  My acceleration readings look like little steps. They never read zero g either: It’s always on one side or the other of zero. Why?
When a recorder is used to measure small-level signals, the data begins to resemble small steps. This is the result of quantization due to the digitizing process. If your data is typically stepped, it may be that the range of the recorder is too great, and you are under-utilizing the abilities of the Analog-to-digital converter. The EDR3 and EDR3C recording engine (which is also used by the MSR-series) uses a 10-bit A/D converter. Thus, signals below 10% of full scale will be represented by only about fifty discrete digital steps.
Because the data is bi-polar (positive and negative), there is no bit specifically  assigned to represent zero. Instead, there are two bits representing +- 1/2 Least-Significant-Bit that the recorder bounces between. No measureable acceleration will thus show up as a "random" waveform flipping from one state to another.
2.10 My system didn’t record any acceleration. How do I know if it was even recording at all?
Look at the start/stop times or the temperature sample data. There should be at least two temperature samples, and at least a start time. If there is no temperature sample data, then the system was not activated. There is a method that ensures that the recorder records at least something: Turn on the periodic (slow track) temperature sampling as a trigger source. The recorder will now store an acceleration event each time the temperature is recorded. Even though the acceleration levels never exceed your acceleration trigger level, you will at least gather some record of acceleration.
2.11 How do I calculate the maximum amount of digitized time-history data that I can store?
The maximum amount of digitized time-history is entirely dependent of the amount of RAM that is resident in the recorder, and the sample rate. The faster you sample in a given second, the less number of seconds' worth of data will be able to be digitized. This is because the total number of samples available in memory is fixed.
2.12 What is the difference between “Overwrite” and “Sliding Window Overwrite”?
Overwrite is a subset of Sliding Window Overwrite where the number of time-based windows is set to one. the length of the window is as great as the test length. Therefore the system will not "slide" to the next window. For more information, refer to the hardware users guide for your recorder.
2.13 How do I calculate the maximum number of events that I can store in my recorder?
The maximum number of events is a function of the memory capacity of the recorder, the number of samples that are dedicated to a single event, the number of events allocated to a given window, and the numver of windows used in a given test. For more information, refer to the hardware users guide for your recorder.
2.14 I want to use the quick-start menus, but also want to configure my environmental settings. How do I do this?
First, configure your recorder using the quick-start menu. Next, re-enter the "set RCPs" menu and choose "Advanced RCPs". The communcations program EDR3Ccom or EDR3com will extract the parameters chosen by the previous quick-start from the recorder and display them for editing. Leave the sample rate and event configuration alone, and modify the environmental settings to your specifications. Re-send the newly modified RCPs to the recorder.
2.15 I have no idea on where to start choosing my setup parameters. Got any suggestions?
Sure! You can start by using the "quick-start" control parameters to get in the ballpark for your application. Once you have gotten some ideas about the nature of your environment, try the following tips:
Choosing your RCPs require you to balance sample rate with regard to system resources and recorder bandwidth. For instance, if you have a system with 60 Hz anti-aliasing filters (low pass filters placed on the accelerometer inputs) it is probably pointless to run the sample rate at 3200 samples per second. Usually a 5:1 to 20:1 sample rate ratio is chosen with regard to the measurement bandwidth of the recorder. Assuming your system has 110 Hz filters, this means a sample rate between 512 s/sec and 2200 s/sec. You probably can get by with less than 1000 samples per second; probably even 512 samples/second, if suspect that your packaging is going to dampen any high-frequency vibration.
Since the recorder has a fixed amount of digital memory, it can store a fixed number of acceleration samples. How you choose to use this capacity is up to you: The faster you sample the environment per unit of time, the less overall time-history you can store in terms of number of seconds of digitized time-history. For instance, a 4MByte equipped EDR3C can digitize about 33 minutes of vibration when running at 512 samples per second. If you double your sample rate, you cut your available storage time in half. The number of digitized data points remains constant.
You will probably use the "Sliding Window Overwrite" feature of the recorder. This means that each digitized portion of time-history, called an event, will feature a fixed length of "x" number of samples. A portion of this event will consist of pre-trigger information. This pre-trigger information are samples that are stored immediately before an event trigger condition goes true. A trigger could be caused by level sensitive triggering, such as the X-axis exceeding the X-axis acceleration threshold, or by time-triggered recording which coincides with the temperature sample interval. For initial tests, I would suggest that you make the length of the event equivilant to about two seconds of time-history. Thus, for 512 s/sec, the event length would be 1024 samples.  Pre-trigger is generally low-amplitude information, and should thus only be a small percentage of the overall event length: No more than 30% of the total event length. Thus, with a 1024 sample event, the pre-trigger might consist of 256 or288 samples. Pre-trigger serves an important role of providing "context" for a given impact.
The acceleration trigger threshold should be set low enough such that as the package reaches its final destination, the very last empty event space in memory is filled as it hits the engineer's desk. This is, of course, impractical. It is hard to predict what the environment is going to provide as far as impacts go, and this is why the overwrite feature is so handy. Read about it in the users guides. I suggest you start out with a trigger threshold of 0.5 g, and a dead time of 100 milliseconds. The non-zero dead time is essential for allowing the system to re-bias after orientation changes. Non-zero dead time (100 msec is the minimum non-zero value) is required for the "Zero after event" function to work.Use all three of the zero correction options: Initial zero, zero after event, and zero correction during event. This will reset the system in case of large temperature changes or orientation changes.
One thing I like to do is keep my event length and sample rate set to some integer power of two. This is because I like to do spectral analysis of my data to analyze vibration. Sample window lengths in integer powers of two are required for the Fast-Fourier-Transforms used to process data. If the event length is not an integer power of two, DynaMax "zero pads" the data to make it such. If you have the luxury of choosing an integer power of two, then your data remains undistorted in the frequency domain. Zero padding adds frequency resolution thus "adding" information to the waveform. The effect is subtle, and zero-padding is an accepted industry practice so you don't necessarily have to follow my preference.
I suggest that you review the manuals on the theory of sliding window overwrite and the other materials for more suggestions.
2.16 My recorder has nonsensical humidity readings. Temperature too! What is going on?
Either you do not have a humidity sensor installed, or you have an un-matched humidity probe. The remote RH-2 temperature/humidity probe is factory matched to the recording engine. In addition, the RH-2 uses the temperature compensation element that normally accompanies e accelerometer block. If your recorder does not have an internal accelerometer block, as is the case in the MSR-series, the sensor is moved to the end of the RH-2 probe. Thus, to correctly access the temperature of this sensor you must choose the "INTERNAL" or "ACCEL TEMP" sensor, NOT the External temperature sensor! The TMP-1 External temperature sensor is yet another probe of different construction and type.
2.17 Will IP alcohol affect my humidity measurements?
No, but the water that is dissolved in the Isopropyl alcohol might. Tests show that a wipe of 99% IP alcohol directly on the sensor made humidity vary be only about 5%. If you are using IP alcohol with a greater water content, you are likely to see a higher deviation.
2.18 What is the lock switch? How is it different between the EDR3 and EDR3C?
The Lock switch serves several roles in the recorder: The first application was to "lock" the recorder into the current operating mode. This mode was useful to help reduce aborted tests due to unintentional tampering by curious personnel. The second role is that of entering manual commands in the field. You can use the lock switch in conjunction with the select button to enter simple commands. The difference between the EDR3 and EDR3C are two-fold: The sequence of commands, and the behavior of the system when in the "active-standby" mode.
The lock switch operates by holding the recorder in the current mode, no matter what happens on the RS232 port or the Select button. Once locked, the recrder stays in the current mode, be it Active, Standby, or Communicate mode. There is one additonal mode, Active-Standby, that the recorder resides in once a start command has been issued. Once a recorder receives a start command, it switches immediately to active-standby, where it compares the start time to the current time. It stays in this mode until the start time elapses. At this time, the recorder will shift from active-standby to active mode. There is a difference in how the system operates this transition when the lock switch is used however: The original EDR3C will successfully shift from ACT-STB into Active regardless of the lock switch selection. The EDR3C recording engine however, will not. Instead, the EDR3C will attempt to shift to active, but fail to activate. Instead, it will simply shut off. Therefore, we suggest you look at the next question:
2.19 Why should I NOT use the lock switch?
The lock switch was useful in the original EDR1 where the controls were located under a simple plastic plug. When the EDR3-series recorders were developed, the controls were located under a billet T-6061 Aluminum plate cover which is bolted to the EDR3 housing. Thus, the controls are not easily accessed by non-authorized personnel. therefore, the lock switch is rendered redundant by the top cover. Because of the differences stated in section 2.18, IST recommends that you NOT use the lock switch in everyday operations. Several users have experienced data loss due to the improper use of the lock switch. Dissimilar behavior with regard to the lock switch function between the EDR3 and EDR3C recording engines further complicates the operation. Since the top cover of the recorder serves as a deterrent to unauthorized tampering, the lock switch is probably best left only for the entering of manual commands. Keep the lock switch UNLOCKED during your test.
2.20 How do I use the “Delayed Start” function?
The Delayed Start function is handy for when you find it hard to be on-site for the installation of the recorder and subsequent manual activation. There are a few things to be aware of: The use of a delayed start requires the system to be placed in the "Active-Standby" mode. In addition, the delayed start time must be in the future with regard to your computer system time. Be sure the computer has the correct time zone: All times are converted to GMT before being placed in the recorder. Finally, once you have established your setup, check-mark the box "Start recorder immediately after sending setup". This will cause the computer to send a start command to the recorder after the setup file has been uploaded. Please note that future versions of the DynaMax Suite will probably change this text box to read "Place unit in Active-Standby after sending setup". The EDR will now enter active-standby and wait for the start time to elapse. Be sure to see section 2.18 about the use of the lock switch when in active standby!
2.21 Explain to me how the Sliding Window Overwrite process works with regard to the EDR3-series recorder
Sliding Window Overwrite (SWO) is a patented method for gathering dynamic environmental data. Essentially, SWO ensures that your recorder continues recording until the end of the test, no matter how many times conditions exceed the triggering criteria. The "window" is a window of time. Each window of time is allowed a finite number of time-history segments called events. Each event has an equal number of digital samples. Generally, the recorder is configured such that the total number of samples required by all of the events in all of the windows does not exceed the memory capacity of the recorder. There are several other papers that are located elsewhere in this web site that describe the sliding window overwrite process further.
2.21.1 With regard to the EDR3 recording engine
The original EDR3 recording engine only has the ability to hold one window. This window spans from the start time to the end time and thus beackets the whole test. Events with large impulse sums at the very end of the test can cause early events with small impulse sums to be discarded.
2.21.2 With regard to the EDR3C recording engine
The EDR3C recording engine has the ability to hold multiple windows. When an event is logged in a given window, the impulse sum of that event is compared only to events recorded during that window. Other events from a previous time interval (window) are not included in the overwrite process, and thus are preserved. Once the last window is reached, any additional time (or event space) stays resident in the last window.

3. EDR4-series recorders

3.1 When I press the select button I see three LEDs flash at me. What does this mean?
3.1.1 One short flash
3.1.2 One long flash
3.2 How do I orient the batteries in this recorder?
3.3 What operating systems are supported by DynaMax?
3.4 What is the latest version of DynaMax?
3.5 What will happen after year 2000?
3.6 When I retrieve my setup, some of the parameters have changed. What happened?
3.6.1 Sample Rate
3.6.2 Low-pass (anti-aliasing) filter setting
3.6.3 Trigger Level
3.7  What full-scale range can I use with my EDR4-200?
3.71 What is the minimum resolution of my EDR4? Can I reduce it?
3.8 How fast can my recorder download data?
3.9 How do I communicate with my recorder?
3.10 There's a lot of stuff that comes up on my screen when sending the setup. What does it all mean?
3.11 How should I store my recorder when it is not in use?
3.12 What does the "time constant" represent in the setup?
3.13 I have an EDR4M6 model: How do I set up both channels?
3.14 How are the single-ended channels different than the EDR3-series?
3.15 What are the voltage requirements of the single-ended channels in the EDR4?
3.16 Explain to me how the sliding-window-overwrite process works with regard to the EDR4.
3.17 When we set 'Low Pass Filter Frequency' to 2000Hz,and then send the value to EDR-4, we find the value(2000Hz) changed to 1704Hz.Why it is?
3.18 What's the relationship of the parameters 'Low Pass Filter Frequency' and 'Sample Frequency'?
3.19 We want to record vibration to 2000Hz. How do we set up the recorder?
3.20 The EDR-4 have 108M memory. How long can we record, if sampling at 5 Ksps?

3.1 When I press the select button I see three LEDs flash at me. What does this mean?
  The flash of three LEDs indicates some kind of power failure.
3.1.1 One short flash
This is an indicator that the charge on the D-cell batteries is insufficient to start the recorder. Replace the batteries and try again. The recorder should either start, or should begin flashing one long flash of the three LEDs. If the replacement of the D-cell batteries does not help, and the system continues to show one short flash (or no flash at all) it indicates that the internal battery is discharged. The system cannot start without a charged backup battery. Call IST for details.
3.1.2 One long flash
It means that the memory has been cleared and the Programmable Logic Array has lost the programming. The EDR4 uses DRAM in conjunction with a programmable logic array (PLA) that is loaded with a portion of the firmware. Both of these components are volatile and require current for memory refreshing. Whenever the power is interrupted to the system, the data memory and the PLA get cleared out. Normally there is an internal battery that keeps the memory and PLA refreshed while you change your system batteries. If this internal battery is good, your system will refresh the memory for eight minutes without the D-cell batteries. After eight minutes, the system shuts off, and flushes the PLA and memory. This keeps the non-rechargeable internal battery from being excessively discharged.  When you attempt to communicate with your recorder after a memory flush, DynaMax will prompt you with a warning that the memory has been flushed. DynaMax will then re-load the PLA, find the latest calibraition file for the EDR4 from your hard drive, and load the calibration. If you have had your EDR4 for a long time, and have lost the calibration file from the hard drive, contact IST. We probably have the most recent cal file and can send you an electronic copy over the internet. It is a binary file that is unreadable, except by the EDR4.
3.2 How do I orient the batteries in this recorder?
The EDR4 has a battery clip arrangement designed to keep spring contact with the D-cell battery, even under high rms vibration conditions. Since the system operates at 5-volts, and the alkaline batteries only provide 6 volts, there is not much room for the voltage drops of a reverse-protection diode. Therefore it is important that you do not reverse the batteries in the clip! Later model clips have had a "+" and "-" etched into the top of the clip. Earlier models do not have this marking. An attempt was made to make it impossible to insert the batteries backwards, but it turns out that you can get them in if you try hard. For the record: Each clip has two features: One terminal is recessed into the battery clip and one is not. The end of the battery that is supposed to sit in the recess is the negative side of the battery. the positive side of the battery simply pops into a little hole in which a spring is resident. Unfortunately, there is also a little hole with a spring inside at the bottom of the recess as well. That is for the NEGATIVE side of the battery! Call IST if you have any questions whatsoever. If the batteries are reversed, it will blow a reverse protection fuse located inside the battery clip.
3.3 What operating systems are supported by DynaMax?
DynaMax for the EDR4 runs on DOS, Windows 3.1, and Windows 95. Windows 98 has a spotty record of operation. Windows NT will not work. Windows NT divorces the software from directly accessing hardware, except through a device driver. The EDR4 will require additional hardware development to support communications under Windows NT.
3.4 What is the latest version of DynaMax?
DynaMax Version 4.29 is the latest version. It is still DOS based. Currently, there is no release date available for any upgrades to Windows-based software. If you do no have DynaMax version 4.29, call IST. We can send you an electronic copy via email. Previous versions had several printing errors, as well as setup errors when using sample rates less than 4096 samples per second.
3.5 What will happen after year 2000?
The EDR4 will function merrily beyond the year 2000. There is a display issue with the processed data in DynaMax however: Version 4.29 (and all other previous versions) displays the year "00" as "100". This will not change the way the data is analyzed, sorted, or collected. It only effects the displaying of the data.
3.6 When I retrieve my setup, some of the parameters have changed. What happened?
Because the EDR4 is a digital device, there may not be an analog equivilant to the setting requirement that you choose. Some settings have only a finite number of different settings. You can check your settings against the recorder by issuing a "get RCPs" and "View RCPs". the most obvious changes involve:
3.6.1 Sample Rate
The sample rate choices are 16, 32, 64, 128, and 256 samples per second, then increase in intervals of 128 samples per second, up to 14976 samples per second. If a sample rate is requested that falls in-between an integer multiple of 128, DynaMax rounds it DOWN to the next sample rate. Please note that low sample rates under 1024 s/sec. will reduce the rate at which the recorder can auto-correct DC offsets from the transducer. The time constant for the autozero-correction circuit will be so large at 16 s/sec, for instance, that the autozero correction will be effectively disabled, allowing for DC offsets in the data.
3.6.2 Low-pass (anti-aliasing) filter setting
The anti-alias filtering uses an active filter that employs a digital potentiometer. The ratios and ranges of filters is not linear with respect to the potentiometer settings. There are some fine graduations at lower frequencies, but once the anti-aliasing filter values move into the kilohertz region, the graduations become rather coarse. This is a design tradeoff with regard to the power consumption and board space on the EDR4. The minimum filter setting is 20 Hz, and the maximum is 5 Khz.
3.6.3 Trigger Level
Trigger level is determined by the chosen full-scale range of the recorder. There are 2048 discrete digital levels between zero and full-scale. Therefore, the trigger level must fall on one of these discrete levels. If the trigger level is between two levels (known as "counts") it is rounded to the nearest count on the A/D converter.
3.7  What full-scale range can I use with my EDR4-200?
Your model number tells the story: The EDR4-200 system is equipped with 200g internal trixial block. Therefore, the highest possible g-range (at unity gain) is 200 g.

The EDR4 does however, have a variable gain amplifier that lets you turn down the full-scale range by turning up the gain. This lets you improve the resolution of the system for applications where there is little chance of experiencing 200g. There are practical limitations to this scaling though, due to the "noise floor" of the system, which I will address later.
3.71 What is the minimum resolution of my EDR4? Can I reduce it?
Your resolution is governed by the 12-bit A/D converter as well as the noise floor. When the system is configured as a 200g system, the noise floor should be 1 bit or less. With a 12-bit A/D converter, there are 4096 discrete counts between +full scale and -full-scale. Thus, there are 2048 counts between zero and full-scale. The resolution is therefore 1/2048 of 200g, or .098 g.
98 milli-g's is the resolution when full-scale range is set to 200g's. What if we turn the full-scale down to 100 g (by increasing gain of the amplifier)? The resolution is reduced to 49 milli-g's. Can we go even lower? In a perfect world, yes. Unfortunately, we run into limitations with regard to noise.

The noise floor of the data recorder is a fixed value. This value changes somewhat from system to system, but is usually less than one count on the A/D converter (at unity gain). When we increase the gain on the amplifier, we also amplify some of the noise. Therefore, noise rises up and begins "consuming bits".

The law of diminishing returns thus begins to take effect: Increases in gain allow for more accurate measurements at low signal levels, but noise increases begin masking the very measurements we want to measure! Any further gain increases simply offer no net gain.

If you wish to find your lowest practical gain setting, run your system with a progressively lower full-scale range, and a trigger level of zero g's. Vibration isolate the system as much as possible. I sometimes set systems on a piece of soft open-cell foam that sits on the unpowered table of our electrodynamic shaker. You should eventually see a band of random vibration of so-many-milli'gs that never changes with lower full-scale ranges. Once you identify your noise floor, you can more accurately choose the minimum resolution, and choose the most practical full-scale measurement range, which will be 2048 times that resolution.

Your system measurement accuracy is therefore 5% of the measured signal, 1/2048 of full-scale, or the noise floor, whichever is greater.

3.8 How fast can my recorder download data?
This depends on the rate at which your computer can process information through the parallel port. Pentium class machines running straight-DOS can see some very fast download speeds: Up to 1 Megabit per second. Machines running Windows 95, and lots of embedded applications may go as slow as 100 Kbits / second. It is mostly a function of your specific computer setup.
3.9 How do I communicate with my recorder?
Through the parallel port. This offers greater throughput than conventional RS232 links.
3.10 There's a lot of stuff that comes up on my screen when sending the setup. What does it all mean?
The descriptive information that comes up when sending the recording control parameters to the EDR4 relates to the complex task of setting the proper gain, anti-aliasing filters, adjustable autozero correction circuitry, and correct bias for the accelerometers. This process takes a while because the system makes several passes to correctly calibrate the circuitry. Overall, the process takes about a minute to complete. Once it is finished, the system will be ready to make accurate acceleration measurements.
3.11 How should I store my recorder when it is not in use?
Always store your system with a good set of D-cell batteries installed. The current from the D-cells will be used to keep the PLA alive, and the memory refreshed. Otherwise, the system will draw from the internal battery for eight minutes, then shut down. At this time, the memory, PLA, and calibration records will be flushed.
3.12 What does the "time constant" represent in the setup?
The time constant represents the number of seconds required for the autozero correction circuitry to correct off 1/2 of the full-scale value of the recorder's measurement range. Thus, if you have a 10 g recorder, and your time constant is set to 12 seconds, the system will correct out 5 g of constant acceleration over the course of 12 seconds. This correlates to an autozero rate of 4.2% of full scale per second.
3.13 I have an EDR4M6 model: How do I set up both channels?
An EDR4M6 has two recording engines resident in the housing. Each recording engine has a unique serial number that you use when performing a system setup. All EDR4-series recorders have a nine-digit serial number, of which the last five digits correspond to the serial number of the recording engine(otherwise known as "board"). The first digit of the five-digit serial number is always 4. The second digit is either a zero or a one, depending on whether the recording engine is the primary or secondary recording engine. The remaining three digits corespond to the build sequence of the recorder. Thus, if the EDR4 serial number is 119740062, the board serial numbers will be: 40062 for the primary recording engine, and 41062 for the secondary recording engine. Once you have established the serial number of the recording engine that you wish to set up, you must choose the correct communications port for the unit. The primary board is accessed through the 9-pin connector located under the top cover, and the cesondary board is accessed through the 19-pin KPTO-series connector on the front of the recorder.
3.14 How are the single-ended channels different than the EDR3-series?
The EDR4 has the ability to shut off the current excitation to the single-ended channels. This allows you to use other types of transducers and signal inputs without worrying about the loading difficulties associated with a shunt resistor.
3.15 What are the voltage requirements of the single-ended channels in the EDR4?
The EDR4 does not have a fixed gain such as the EDR3. The EDR3 is limited to 500 millivolts as the full-scale range of an external transducer. The EDR4, however, can measure up to 1500 mv from the external transducer. With adjustable gain, your full-scale range is adjustable from approximately 50 millivolts to 1500 millivolts. The practical bottom-end (full-scale ranges less than 100 millivolts)  is very dependent on the transducer characteristics and the electronics of your particular recorder.
3.16 Explain to me how the sliding-window-overwrite process works with regard to the EDR4.
The Sliding-Window-Overwrite (SWO) process is essentially the same as the EDR3C, with one exception: The EDR4 uses two different impulse sum buffers: High-speed events are recorded in memory whenever a triggerable condition occurs, such as acceleration exceeding a predetermined threshold. The two impulse sum buffers compare events based on the source of the trigger: Acceleration-triggered events are compared only to acceleration-triggered events, and time-triggered events (triggered from the temperature sample interval) are compared only to time-based events. This allows you to perform a mixed-mode test, where the overwrite process gathers the "worst-case" events from acceleration threshold triggering, and "typical" events that are triggered by time, independent of the level of acceleration. Both statistical representations are necessary to adequately gage the envelope and severity of an environment. Previously, users had to perform two tests, or use two recorders to gather both the "statistical average" and the upper bound of a shock & vibration environment.
3.17 When we set 'Low Pass Filter Frequency' to 2000Hz,and then send the value to EDR-4, we find the value(2000Hz) changed to 1704Hz.Why it is?
The filter settings are non-linear, and there are only a finite number of settings. There are only a few low-pass filter settings above 1 Khz. In fact, there are only two or three above 1.7 Khz. I think they are 2.4 Khz, and 5 Khz.
3.18 What's the relationship of the parameters 'Low Pass Filter Frequency' and 'Sample Frequency'?
DynaMax does not let you set the filter frequency to anything  higher than half the sample rate in order to avoid aliasing. If, for instance, you set the sample rate to 5,000 samples persecond, the AAF filter frequency will only go to 2.4 Khz. You can set it lower, if you wish.
3.19 We want to record vibration to 2000Hz. How do we set up the recorder?
Set your sample rate to 5000 samples per second, and your filter to 2.5 Khz. DynaMax will "round down" to the next lowest filter setting which is, I believe, about 2420 Hz. This is the closest you can get to 2 Khz given the existing hardware topology.
3.20 The EDR-4 have 108M memory. How long can we record, if sampling at 5 Ksps?
108 MB of memory means that you can store about 24 million samples in memory. 24 million divided by 5000 samples per second means 4800 seconds of triggered time-history data can be stored. DynaMax can only download 5000 events, so your event length should be at least 5000 samples (1 second) per event. These 5000 samples are divided between the pre-trigger and the post-trigger. Max event length should also be 5000.
4800 seconds is 80 minutes of continuous time history. Remember though: The recorder can trigger from acceleration. Thus, in periods where there is no acceleration to measure, one need not store an event because no signal will be present on the accelerometer. Thus, the recording time may actually be longer than 80 minutes. In addition, the overwrite criteria will ensure that the 4800 seconds represents the highest level of vibration that occurred during the test.

4. Snapshock Plus

4.1 What type of battery should I use in my SnapShock Plus?
4.1.1 What type of battery life might I expect from my SnapShock Plus?
4.2 How do I cycle my system through the various modes?
4.3 Is this recorder waterproof?
4.4 Why can't I get the system to activate?
4.5 I didn't record any impact data. What happened?
4.6 My data shows a "LOW BAT" marker. What does this mean?
4.7 How do I use a delayed start in this recorder?
4.8 What is the "Maximum Event Length"?
4.9 I keep getting the warning "Recorder time precedes computer time". What does this mean?
4.10 I have received the message "Heap/Stack pointer not set correctly - Do not cycle to Active!" What does this mean?
4.11 I received the error message "check-sum failed!" when setting up my recorder. What does this mean?
4.12 My SSP recorded some events greater than the advertised full-scale range. Is there something wrong?
4.13 The SSP didnt clear its memory, and it wont start when I try to activate it!

4.1 What type of battery should I use in my SnapShock Plus?
Your Snapshock Plus will work best with an Ultralife Lithium 9-volt battery. The lithium battery offers 1200 mAH of charge capacity versus the 600 mAH of a typical alkaline battery. A Zinc carbon 9-volt will provide 400 mAH or less. Low-temperatures will exacerbate the problem to the extent that systems will not activate below freezing temperatures.
4.1.1 What type of battery life might I expect from my SnapShock Plus?
Most single-axis SSP recorders typically draw 1.8 milliamps when active. The triaxial version typically draws 3.4 milliamps. To estimate battery life, take the capacity of the battery and divide by the current draw. This will estimate the number of hours of operation from a brand new battery. Other factors to consider include the temperature of the system in the operating environment, and any current consumed while the system is in COM mode. Never leave a SSP in COM mode for the long-term: IrDA mode draws 3.6 ma of current, and hard-wired COM draws 24 milliamps, whenever the cable is plugged into your computer!
4.2 How do I cycle my system through the various modes?
This is the mode sequence: The system spends exactly two seconds in each operating mode. The first second of the operating mode is spent with the corresponding status LED turned off, followed by the last second spent with the LED turned on. Therefore, whenever you see the LED representing your desired mode, you have less than one second left before the system reverts to the next mode in the sequence. Here is the sequence:
Mode sequence 0: Standby. Designated by two LEDs: ACT and COM
Mode sequence 1: Hard-wired RS232 COM. Designated by the COM LED. (First flash)
Mode sequence 2: IrDA COM. Designated again by the COM LED. (Second flash)
Mode Sequence 3: ACTIVE. Designated by the ACT LED. This is a green LED in later systems. Earlier systems have a red LED in both positions.
4.3 Is this recorder waterproof?
No. The system is not waterproof. The system has been filled with a silicone potting compound to make it resistant to high-g shocks, as well as offer weather resistance. The system can still admit water in around the mode switch. Therefore, avoid immersion of the system. Place good quality vinyl tape over the mode switch area if you expect to expose the system to water or salt spray.
4.4 Why can't I get the system to activate?
There are several possibilities:
1) The power fail flag is set. clear it by re-sending the Recording Control Parameters to the system.
2) The battery is too weak to provide the initial inrush current required to start up the analog circuitry. Substitute a new,  quality lithium battery. Note: This is a common problem. Used Lithium batteries will show an increased internal resistance which will cause this failure. Recorders have been returned for repair in which the problem was simply a weak battery!
3) The battery temperature is too cold. If in inclement weather, stick the SSP in your pocket and warm it up. Freezing temperatures slow down the chemical changes necessary to release electrical current in the battery.
4) The mode switch is being held too long. Remember, if you see the active LED, you are more than half-way through the 2-second mode sequence.
5) Too much offset required to bias the accelerometer. This is evidenced by the active light staying on for more than 30 seconds, followed by the system reverting to standby. The active LED will be illuminated for a steady half-minute or more.
4.5 I didn't record any impact data. What happened?
Either the system was never activated, or the RCPs are too aggressive for the environment. If the recorder has start marker that coincides with the beginning of the test, then it is most likely a case of aggressive RCPs with too-high of an acceleration ot velocity threshold. If "old data" from a previous test appears, it indicates that the system never was activated in the first place. (For questions on "old data", See question 4.13)
4.6 My data shows a "LOW BAT" marker. What does this mean?
The "LOW BAT" marker is created by ShockView 32 software and is not stored inside the recorder. It indicates that no stop marker was found. The next record found in the recorder was from a previous test. The most likely cause is a low battery, and a subsequent shutdown of the processor by the power-fail circuit. ShockView 32 software will time-code the LOW Battery condition with the time of the last known good event. What this time means, is that something happened AFTER this time to cause the system to shut down. It could have occurred 1 second after the last valid event, or 1 month after the last valid event.
4.7 How do I use a delayed start in this recorder?
Just check-mark the start and stop times in the setup menu. Cycle the system to standby after you are done. When the start time elapses, the system will turn itself on and begin measuring.
4.8 What is the "Maximum Event Length"?
This is the amount of time required before a sustained acceleration event is terminated, and tagged as a "rollover". The recorder will enter the event in memory with a zero duration time. After event termination, the recorder will re-bias the accelerometers for the next dynamic event.
4.9 I keep getting the warning "Recorder time precedes computer time". What does this mean?
This warning simply means that the clock in the SSP was reset due to a power failure, such as when changing batteries.
4.10 I have received the message "Heap/Stack pointer not set correctly - Do not cycle to Active!" What does this mean?
This error indicates that the pointers in the processor are not in sequence. There are many reasons why this error occurs: It means that if the recorder is re-activated, loss of data or calibration information might occur. Be sure to download all of the data, and re-send the RCPs to the recorder before activating. This will re-synchronize the heap and stack pointers.
4.11 I received the error message "check-sum failed!" when setting up my recorder. What does this mean?
The check sum is used to validate the information sent to the SSP. If the checksum fails, re-send your RCPs. If you are experiencing frequenct checksum failures, try repositioning the IrDA interface or place a piece of paper over the IrDA interface and the SSP to shield out interference. Keep the IrDA a minimum of six inches from the SSP, but not more than twelve inches away.
4.12 My SSP recorded some events greater than the advertised full-scale range. Is there something wrong?
Not at all. Every system is capable of measuring beyond the advertised full-scale range. The advertised range is the ABSOLUTE range, which is guaranteed to be a valid measurement. the remaining range area is reserved for digital offset correction. Thus, a signal beyond the absolute full-scale range might be valid, or it might be clipped at the A/D converter. Because the SSP does not store the digitized waveform, there is no way of knowing. The chance of a signal being clipped rises the farther you go beyond the absolute full-scale range. I have a PDF that explains it further: SSP Digital Offset explained!
4.13 The SSP didnt clear its memory, and it wont start when I try to activate it!
It sounds like your system is having troubles with the battery. Try a new alkaline or new lithium battery, then reset the RCPs and try starting the recorder.  Here's why I think it's the battery: If the battery has a high internal resistance, it will "brown-out" upon initial startup, causing the power fail circuit to trip. Thus,  the system never sets a start marker in memory because the test never got started.
The reason you can see the old test data is because it is never truly erased; It is only overwritten. Since the Snapshock Plus uses FLASH memory, we do not needlessly write to it. The memory has about 10,000 write cycles available. Therefore, we never really erase the old data, we just write over top with new data.
Thus, if the battery has a high internal resistance, the large inrush current needed to power up the system will cause a very brief voltage drop which is still long enough in duration to sometimes cause the power fail circuit to trip. When the power fail circuit trips, it cuts power to the microprocessor, which then never gets a chance to write anything to memory. The system shuts down, and the old test data remains intact. (See also question 4.4)



Comments are closed.

Showing 0 Comment