Chapter 2 - GETTING STARTED

2.1 Upgrade notice

If you are upgrading from earlier versions of the LTP program, you will have write new *.pro protocol files. The newer LTP24 program will not load protocol files you made using a previous LTP program for reliability reasons.

Note that you can analyze the same ADsweep files with LTP24 that were obtained with all earlier versions of the LTP program.

2.2 Installing the LTP Program

2.2.1 Go to the Explorer (in Windows 95/98) or File Manager (in Windows 3.x) and create the folder \LTP24 on the appropriate drive (I will assume the C: drive for this discussion). Then download the self-extracting zip file LTP24Z.EXE from the 'LTP' program web site, www.ltp-program.com.

Then double click on the LTP24Z.EXE file to extract the following files which should now be in your \LTP24 folder:

2.2.2. For Windows 95 or 98:

Because LTP24 uses a newer DOS4GW.EXE 32-bit DOS extender that can access 64 MB of RAM memory (version 2.01a,  April 3, 1996), and replaces the older DOS4GW.EXE supplied with LTP114J which can access only 32 MB of RAM memory (version 1.97, Feb. 29, 1996), delete the older DOS4GW.EXE in the \LTP114J folder.  Also, remove C:\LTP114J; from the PATH statement in the autoexec.bat.

To implement the changes made to the autoexec.bat file you must first reboot the computer by clicking on 'Start', then 'Shut Down', and then 'Restart the computer'.

2.2.3. For MS-DOS 6.0/6.2 (with Windows 3.0/3.1/3.11):

Because LTP24 uses a newer DOS4GW.EXE 32-bit DOS extender that can access 64 MB of RAM memory (version 2.01a,  April 3, 1996), and replaces the older DOS4GW.EXE supplied with LTP114J which can access only 32 MB of RAM memory (version 1.97, Feb. 29, 1996), delete the older DOS4GW.EXE in the \LTP114J folder.  Also, remove C:\LTP114J; from the PATH statement in the autoexec.bat.

To implement the changes made to the config.sys and autoexec.bat files you must first reboot the computer by pressing CTL-ALT-DEL, pressing the reboot button, or by turning OFF then turning ON the power again.

2.2.4. For Windows ME:

Same as for Windows 95/98.

2.2.5. For Windows NT, 2000 or XP:

Same as for Windows 95/98.

2.2.6. Check For Correct Memory Manager, HIMEM.SYS

If you are using DOS 6.x / Windows 3.x, you should have HIMEM.SYS is in your AUTOEXEC.BAT and not another memory manager like QEMM.SYS or 386MAX.SYS. The normal installation of DOS 6.x / Windows 3.x automatically installs HIMEM.SYS. Similarly, if you are using Windows 95/98 HIMEM.SYS does not have to be in your AUTOEXEC.BAT - it automatically used. However, you should make sure that another memory manager like QEMM.SYS or 386MAX.SYS is not in your AUTOEXEC.BAT. (It is possible that other memory managers such has QEMM.SYS, 386MAX.SYS may work fine, but I have not tried them and therefore will not recommend them. Only use them if you are 100% sure they are working correctly.)

The reason for being careful about what memory manager to use is the following. The LTP24 program uses a DPMI (DOS Protected Mode Interface) host which helps the LTP program talk to the computer hardware. If the HIMEM.SYS memory manager is used with the LTP program, then the DPMI host is the Watcom C / Tenberry's DOS/4GW DPMI host. This is a well defined DPMI host which works well in real-time, and has been well tested by me.

DPMI hosts can be either 'real' or 'virtual'. When the DPMI host is 'real', the DOS program has complete control of the machine. When HIMEM.SYS and Watcom C / Tenberry's DOS/4GW are used, the DPMI host is 'real', and when the LTP24 program blocks keyboard, disk drive, or network card interrupts during data acquisition and stimulation, they are really blocked and there are no delays!!!!

Some DPMI hosts can be 'virtual' DPMI hosts meaning that the DOS program does not have complete control of the machine. With a 'virtual' DPMI host, the keyboard and disk drive interrupts appear blocked to the DOS LTP24 program but they are not actually blocked in the computer - and therefore substantial delays of 1 msec or longer can occur during acquisition/stimulation and your synaptic potential would appear much shorter.

Therefore, be sure you are using HIMEM.SYS, and test for any delays during the data acquisition / stimulation sweep by using a triangular waveform input (as described in Section 2.9.1). Also test the sample-to-sample jitter by measuring the width of a train S0 pulse using an oscilloscope (see Section 2.9.2).

2.3 Installing the Data Acquisition Board

The LTP24 program presently runs on four data acquisition boards: the Axon Instruments Digidata 1200 board, the Scientific Solutions Labmaster board, the Pico Technology 42-ADC board, and Measurement Computing's CIO-DAS08/JR-AO.  

2.3.1 Axon Instruments Digidata 1200 Board

The Digidata 1200 board is one of the most popular data acquisition boards for neurophysiology, almost solely due to its use by the Axon Instruments program pClamp. However, Axon Instruments no longer sells the Digidata 1200.

The Digidata 1200 board comes with its own interface. LTP24 uses two analog input channel AnalogIn0 and AnalogIn1, two analog output channels AnalogOut0 and AnalogOut1, and four digital output channels DigitalOut0, DigitalOut1, DigitalOut2, and DigitalOut3.  Note that there is only up to 1 bit AD channel crosstalk for a -10 to +10 volt input into the other AD channel.  AD channel crosstalk is not a problem with the Digidata 1200 board.

Presently, the Digidata 1200 board is programmed using the slower Software controlled acquisition and stimulation, and therefore does not run on some of the slowest computers (e.g. a Tandon 16 MHz 386) but does work on faster computers (e.g. a Tandon 25 MHz 486 and above). With Software controlled acquisition there is also some Analog input and Analog/Digital output sample-to-sample jitter on the order of 2-7 usec. For timing errors in sample-to-sample jitter and sweep-to-sweep jitter with the Digidata 1200 board, see APPENDIX B.4 and APPENDIX B.6, respectively.

2.3.2 Scientific Solutions Labmaster Board

The Scientific Solutions Labmaster data acquisition board in the past been one of the most commonly used AD boards in neurophysiology, again due in part to its use by the Axon Instruments program pClamp. The Scientific Solutions Labmaster board was also previously marketed by Axon Instruments, and Axon Instruments also previously sold their own interface unit for the Labmaster board, the Axon Instruments TL-1 interface. Scientific Solutions still sells the Labmaster board, although Axon Instruments does not.

Note that the Labmaster board has 2-3 bits AD channel crosstalk.  In general, I think that the Labmaster board is very good for 1 AD channel recording (AD channel crosstalk is not an issue).  However, the board is just acceptable for 2 AD channel recording, because of the 2-3 bit AD channel crosstalk, although I think this crosstalk is low enough to live with.  However, if you are doing 2 AD channel recording, it is best to use a Digidata 1200 board because there is only up to 1 bit AD channel crosstalk.

2.3.2.1  Program Startup and Connections for the Labmaster Board

The electrode <-> interface connections for the LTP program are based on this TL-1 interface. LTP24 uses two analog input channels (physical AD channel AD15 for logical AD channel AD0, physical channel AD14 for logical channel AD1), two analog output channels DA0 and DA1, and four digital output channels C4, C5, C6 and C7.  The input channels (AD15 and 14), and the output channels (DA0, DA1, Ports C4, C5, C6 and C7) are those commonly used by the Axon Instruments program pClamp 6, and the LTP24 program will automatically work on a computer set up to use pClamp 6.

Fig. 2.3.2. Labmaster Electrode Connections Help Box.

For timing errors in Analog input and Analog/Digital output sample-to-sample jitter (on the order of 2-4 usec) and sweep-to-sweep jitter with the Labmaster board, see APPENDIX B.4 and APPENDIX B.6, respectively.

2.3.2.2  Measurable AD Channel Crosstalk in the Labmaster board

The Labmaster board is a small, but measurable, amount of crosstalk between the 2 AD channels.  Note, however, that AD channel crosstalk is NOT a problem when using just 1 AD channel.

If -10 to +10 volts is input into one AD channel, then a 2-3 bit negative shift occurs in the other AD channel  Because the 12 bit AD converter in the Labmaster has 4096 different levels or 'bits' over the -10 to +10 volt range, the AD channel crosstalk is ~ 3/4000 or ~ 0.075%.  Furthermore, the amount of crosstalk is roughly proportional to the negative input voltage into one AD channel, such that -5 volts input into one AD channel causes only a half the bit shift in the other AD channel.  

The AD crosstalk is up to 1 bit in the Digidata 1200 board, and 1-4 bits in the DAS08/JR board. 

However, Labmaster boards can be slightly different, so be sure to measure the amount of crosstalk between the two AD channels for your Labmaster  board (see Section 2.9.3 and also APPENDIX B11).

If you are amplifying well above the 'bit' level, this should not be a problem.  In other words, if the EPSP or other signal you are measuring is say 400 levels or bits in amplitude (approximately 1/10th of the full scale of 4096 bits), then a 3 bit shift (caused by the other AD channel going from -10 to +10 volts) would cause a 3/400 or 0.75% error in amplitude measurement.  For example, when measuring extracellular EPSPs, if your AD0 Gain is x1000 and you are recording 4096 different levels over the -10 to +10 volt input range, this means you effectively have ~ 4000 levels or bits in a 20 mV range.  Therefore, a signal of 400 levels or bits would be 1/10th of 20mV, or 2 mV.  Therefore, for 0.75% error of the EPSP peak measurement, the EPSP amplitude should be at least 2 mV.

If you think this amount of crosstalk is too large, either increase the gain of the AD channel you are measuring from, or use a different  board such as the Digidata 1200.

Feel free to contact support@winltp.com if you do not understand this and want to use 2 AD channels on this board.

2.3.3 Pico Technology's ADC-42 Board

I wanted to have the LTP program also work with a low cost data acquisition board. At present, the cost an AD board can be twice the cost of a completely adequate Pentium computer. The Pico ADC-42 board is a very low cost (ca. £100 / $150) 12-bit AD 'board'.  

However, in LTP24 I installed another inexpensive board, Measurement Computing's CIO-DAS08/JR-AO that fits into an ISA slot, uses 2 AD channels and has two analog outputs.  Although the Pico board is very good (especially with the added Software Calibration), the DAS08/JR board can be easier to install than the Pico board (sorting out the parallel ports can be a problem), it has two analog outputs (versus none on the Pico board), it is very good for 1 AD channel recording, and it is acceptable for 2 AD channel recording (as long as AD channel crosstalk is considered).  Therefore, in general I think that the DAS08/JR board is better than the Pico board.  

Nevertheless, the Pico board continues to be useful because it has been found to acquire data on at least some recent PCI only bus desktop computers (such as the Dell 8200 1.8GHz P4 computer).  The Pico board has an accuracy of +1% which is adequate for extracellular experiments, but in LTP24 I added Software Calibration making it accurate enough for intracellular and patch-clamp recording.

2.3.3.1  Program startup for the Pico board

The Pico ADC-42 'board' plugs into a parallel printer port and acquires data on one AD channel. Digital output can either come from the same printer port or a second printer port (which can be separately purchased for ca. £10). Sometimes the second printer port is necessary because they often have greater digital current output which may be necessary for triggering some SIUs. 

If you are using the Pico board, it is good to put these commands into a batch file (say LTP.BAT) in the C:\LTP24 directory, and always call the LTP24 program from the batch file.

2.3.3.2  Connections for the Pico board and Printer Port

Fig. 2.3.3.2.1  Pico42 Electrode Connections Help Box (using one printer port).

Use Pin 18 (chassis Ground) if you want to ground coaxial shielding. The wire connections are shown schematically in Fig. 2.3.3.2.2, and a picture of the Pico-42 ADC board and digital output using the LPT1 parallel port on a laptop computer is shown in Fig. 2.3.3.2.3.

Fig. 2.3.3.2.3. Photo of the Pico ADC-42 connected to the parallel port of a laptop computer by a ribbon cable and two DB-25 plugs. The parallel port also produces four digital outputs (Right to Left - Green/Right plug, Wire 1, DigitalOut0; Yellow plug, Wire 2, S0; Red plug, Wire 6, S1; White plug, Wire 8, IC) and a ground to the parallel port (Black/Left plug, Wire 10, GND).

For an LTP experiment with the Pico ADC-42 board connected to LPT1 and digital output through LPT2, connect the electrodes as shown in the Pico42 / Two Port Electrode Connections Help Box (Fig. 2.3.3.4).

Use Pin 18 (chassis Ground) if you want to ground coaxial shielding. The wire connections are shown schematically in Fig. 2.3.3.5 for the connections of the digital output from the second of two parallel ports.

There are two potential problems with the Pico ADC-42 board. First, is your computer fast enough? When the LTP program is started it will automatically determine if the computer is too slow for the Pico board, (for example by running ltp24 pico42_dio_1 as described above). If the computer is too slow, the program will terminate (see APPENDIX B.10). You can even test this without the Pico board being installed. This is because the speed of the acquisition is determined by how fast the data can be read from the parallel port (one bit at a time!), and the Pico board does not actually have to be installed to do this. Pico board AD conversion is always complete within 100 usec.

The second potential problem with using the Pico ADC-42 board is: Are the parallel port digital outputs capable of triggering a SIU? We've found that LPT1 has difficulty triggering our Neurolog SIUs, but LPT2 output will trigger them. Again you can test this digital output capability without having to have the Pico board installed.

If you want to use the Pico board with the LPT1 parallel port and it does not have enough current output to trigger your SIUs, you could interface to a TTL chip for amplification. Alternatively, you could use four 4.7 Kohm pull-up resistors connected to a 5 volt source - read the Web document "Interfacing the Parallel Port" at www.senet.com.au/~cpeacock/parallel.htm#2 - although like they say, you do so at your own risk.

Finally, according to the Pico Technology, the accuracy of the Pico ADC-42 board is about +1%. Because of this we presently use the Pico ADC-42 board only for extracellular LTP experiments, although, if you calibrated your system and you were sure the board was within your accuracy requirements, you could use the board for intracellular recording and patch clamping.

Also note that the input voltage range on the Pico board is -5 to +5 volts (in contrast to the -10 to 10 volts on the Labmaster and Digidata 1200 board). For timing errors in Analog input and Digital output sample-to-sample jitter (on the order of 2-5 usec) and sweep-to-sweep jitter with the Pico board and printer port, see APPENDIX B.4 and APPENDIX B.6, respectively.

Finally, note that the fastest the Pico board can do repeat ADsweep (LTD) stimulation is 1.0 Hz (see APPENDIX B.5).

2.3.3.3  Software Calibration of analog input to the Pico board

According to the Pico Technology specifications on the ADC-42 board, the inaccuracy +1%.  In LTP24 I have added software offset and gain calibration that is applied to the sweep data just after the sweep is obtained.  For the acquired data, this adjustment is made at the 16 bit level even though the board has only 12 bits accuracy, so this ensures monotonicity.

WARNING: This software calibration is specific to each Pico board.  If a new Pico board is installed it must be calibrated anew.

Enabling the software calibration

When the LTP24 program starts up, the Pico offset and gain adjustment fields cannot be altered (for protection against inadvertent changing).  However, in the Options menu that appears when a Pico board is being used, picking Enable Pico42 offset/gain adjust (Fig. 2.3.3.3.1) enables offset and gain adjustment fields for the Pico board to appear in the Miscellaneous page (Fig. 2.3.3.3.2).  Thus the AD0 Offset Adjust field and the AD0 Gain Adjust field can now be changed.  

Software Calibration Proceedure

Note: It is important to perform offset adjustment before gain adjustment. 

2.3.4 Measurement Computing's CIO-DAS08/JR-AO and CIO-DAS08/JR Boards

Because of the advent of 2 channel acquisition in LTP222A, I wanted to have the LTP Program work with an inexpensive 2 channel AD board (rather than just relying on the single channel Pico ADC-42 board).  After trying several boards, I settled on Measurement Computing's CIO-DAS08/JR-A0 (Fig. 2.3.4.1).  It fits into an ISA slot like the Labmaster and Digidata 1200.  

NOTE:  In general, I think that the DAS08/JR board is very good for 1 AD channel recording (AD channel crosstalk is not an issue), and is better than the Pico board for this because it is easier to install than the Pico board (sorting out the parallel ports can be a problem), and it has two analog outputs (versus none on the Pico board).  However, the board is only acceptable for 2 AD channel recording, because there is measurable 1-4 bit AD channel crosstalk, although I think this crosstalk is low enough to live with (particularly when considering the low cost of the board).  However, if you are doing 2 AD channel recording, it is best to use a Digidata 1200 board because there is only up to 1 bit AD channel crosstalk.

The DAS08/JR-AO  is a simple board that essentially contains an Analog Devices AD574A acquisition chip, sample-and-hold and multiplexer chips, an Analog Devices AD7237 analog output chip, and a TTL output port chip (74LS273N).  The only difference between the DAS08/JR-A0 and the DAS08/JR is that the DAS08/JR does not contain any analog output chip and relies on the TTL output port chip for all output.  The DAS08/JR-A0 costs ca. £180/$200 and the DAS08/JR costs ca. £140/$150.

2.3.4.1  Program startup for the DAS08/JR-AO and DAS08/JR boards

Startup for the DAS08/JR-AO and the DAS08/JR is the same.

2.3.4.2  Measurable AD Channel Crosstalk in the DAS08/JR boards

The 'weakest link' of the DAS08/JR board is a small, but measurable, amount of crosstalk between the 2 AD channels.  Note, however, that AD channel crosstalk is NOT a problem when using just 1 AD channel.

If -5 to +5 volts is input into one AD channel, then a 1-4 bit positive shift (usually 1-3 bits) occurs in the other AD channel (see Fig. 2.9.3)  Because the 12 bit AD converter in the DAS08/JR has 4096 different levels or 'bits' over the -5 to +5 volt range, the AD channel crosstalk is ~ 4/4000 or ~ 0.1%.  However, the crosstalk shift in one AD channel is usually much larger for negative going than positive going input voltages into the other AD channel, so the over the negative input voltage range, the AD channel crosstalk is closer to ~ 0.2%.  Furthermore, the amount of crosstalk is roughly proportional to the negative input voltage into one AD channel, such that -2.5 volts input into one AD channel causes only a half the bit shift in the other AD channel.  

The AD crosstalk is up to 1 bit in the Digidata 1200 board, and 2-3 bits in the Labmaster board. 

However, DAS08/JR boards can be slightly different, so be sure to measure the amount of crosstalk between the two AD channels for your DAS08/JR board (see Section 2.9.3 and also APPENDIX B11).

If you are amplifying well above the 'bit' level, this should not be a problem.  In other words, if the EPSP or other signal you are measuring is say 400 levels or bits in amplitude (approximately 1/10th of the full scale of 4096 bits), then a 4 bit shift (caused by the other AD channel going from -5 to +5 volts) would cause a 4/400 or 1% error in amplitude measurement.  For example, when measuring extracellular EPSPs, if your AD0 Gain is x1000 and you are recording 4096 different levels over the -5 to +5 volt input range, this means you effectively have ~ 4000 levels or bits in a 10 mV range.  Therefore, a signal of 400 levels or bits would be 1/10th of 10mV, or 1 mV.  Therefore, for 1% or less error of the EPSP peak measurement, the EPSP amplitude should be at least 1 mV.

While the presence of this AD channel crosstalk is not ideal, I think this is a limitation that can be lived with, especially when considering the low cost of the board.  However, if you think this amount of crosstalk is too large, either increase the gain of the AD channel you are measuring from, or use a different  board such as the Digidata 1200.

For safety, make sure channels AD2, AD3, AD4, AD5, AD6 and AD7 are grounded (see Fig. 2.3.4.3.2 and Fig 2.3.4.3.3).

Feel free to contact support@winltp.com if you do not understand this and want to use 2 AD channels on this board.

2.3.4.3  Connections for the DAS08/JR-AO and DAS08/JR boards

Fig. 2.3.4.3.1  DAS08/JR-AO and DAS08/JR Electrode Connections Help Box.  For the DAS08/JR-AO (having the analog output chip), use DA0 (Analog Out 0) or DigOut5 for S0 stimulation.  On the DAS08/JR (with no analog output chip), use only DigOut5 for S0 stimulation.  For the DAS08/JR-AO, use DA1 (Analog Out 1) or DigOut4 for IC stimulation, but on the DAS08/JR use only DigOut4 for IC stimulation.

To test the board, the simplest way is to buy a 37 pin shielded cable with female DB-37 connectors (C37FFS-5) and a screw terminal board (CIO-MINI37) from Measurement Computing.

This simplest way of making the connector to the DAS08/JR-AO or DAS08/JR board is to directly solder coaxial cables to the DB-37 socket and put BNC connectors at the other end of the cables (Fig. 2.3.4.4).  An important step is then to secure the wires by gluing them in with Epoxy (Araldite).

Also note that the input voltage range on the DAS08/JR-AO and DAS08/JR boards (as with the Pico board) is -5 to +5 volts (in contrast to the -10 to +10 volts on the Labmaster and Digidata 1200 board). For timing errors in Analog input and Digital output sample-to-sample jitter (on the order of 2-5 usec) and sweep-to-sweep jitter with the DAS08/JR-AO and DAS08/JR boards, see the values for the very similar Pico ADC-42 board in APPENDIX B.4 and APPENDIX B.6, respectively.

Finally, note that the fastest the DAS08/JR-AO and DAS08/JR boards can do repeat ADsweep (LTD) stimulation is 1.0 Hz (see APPENDIX B.5).

2.3.4.4  Software Calibration of analog inputs and analog outputs

According to the Computing Measurement and Analog Devices specifications on the AD574A chip, the accuracy of the DAS08/JR-AO and DAS08/JR  board is a maximum +5 LSBs offset and maximum 0.25% gain error at full scale (which includes the offset error).  This inaccuracy is because there are no additional on-board potentiometer offset and gain adjustments.  However I have installed software offset and gain calibration that is applied to the sweep data just after the sweep is obtained.  For the acquired data, this adjustment is made at the 16 bit level even though the board has only 12 bits accuracy, so this ensures monotonicity.

According to the Computing Measurement and Analog Devices specifications on the AD7237 chip, the accuracy of the DAS08/JR-AO board is about +5 LSBs offset and  +5 LSBs gain at full scale (which includes the offset error).  This indicates that the DAS08/JR-AO has an accuracy of +5 LSBs at full scale, and  +5/2047 or  ~ +0.25% inaccuracy at full scale.  This inaccuracy is because there are no additional on-board potentiometer offset and gain adjustments.  However I have installed software offset and gain calibration that is applied to the stimulation data before the sweep is run.  For the output this adjustment is made at the 12 bit level.  If the gain is adjusted, this would not produce monotonicity for a triangular wave output (which LTP24 doesn't do anyways), but it means that square pulse amplitudes are slightly more accurate.

WARNING: This software calibration is specific to each DAS08 board.  If a new DAS08 board is installed it must be calibrated anew.

Enabling the software calibration

When the LTP24 program starts up, the DAS08/JR-AO and DAS08/JR offset and gain adjustment fields cannot be altered (for protection against inadvertent changing).  However, in the Options menu that appears when a DAS08 board is being used, picking Enable DAS08JR offset/gain adjust (Fig. 2.3.4.4.1) enables offset and gain adjustment fields for the DAS08 boards to appear in the Miscellaneous page (Fig. 2.3.4.4.2).  Thus the AD0 Offset Adjust field, the AD0 Gain Adjust field, the AD1 Offset Adjust field and the AD1 Gain Adjust field can now be changed.  In addition, the DA0/S0 Offset Adjust field,  DA0/S0 Gain Adjust field, the DA1/IC  Offset Adjust field DA1/IC Gain Adjust field can also be modified.

(Note that the DA1/IC Offset Adjust, and the DA1/IC Gain Adjust fields are actually mistakenly labeled DA1/S1 Offset Adjust, and the DA1/S1 Gain Adjust fields, a cosmetic error that I didn't correct because the 2.30D version was well into beta test by the time I wrote this.)

Software Calibration Proceedure

Note: It is important to perform offset adjustment before gain adjustment. 

2.4 Starting the LTP Program

The starting of the LTP24 program has drastically changed from the earlier LTP114J version primarily due to automatic creation of data folders during program startup.  You no longer have to: 1) manually create a data folder, 2) move to that data folder, and 3) start the LTP program from that data folder, as you had to with LTP114J.

2.4.1  Windows 95/98 Acquisition

Starting Acquisition in the Windows 95/98 MS-DOS Mode

The LTP program will run on-line in only the MS-DOS Mode of Windows 95 or 98.  LTP24 is started by first going into the MS-DOS Mode and then starting LTP24 from the DOS command line.

To put the computer into the MS-DOS Mode (DOS 7.0), click the 'Start' button, then click the 'Shut Down' button and click 'Restart the computer in MS-DOS mode'. Do not click on the MS-DOS icon in Windows 95/98 because this will merely go to the DOS Compatibility Box. The MS-DOS Mode in Windows 95/98 is not equivalent to the DOS in the DOS Compatibility Box in Windows 95/98.

Alternatively during the boot-up process on computers using Windows 95/98, just after the 'Starting Windows 95' or 'Starting Windows 98' line comes up and the computer beeps, press F8 to go directly into the Windows 95/98 Startup Menu, then enter 6 for the 'Command prompt only'. This bypasses the need to go into the Windows 95/98 GUI and then down into the MS-DOS Mode. (After exiting the LTP program type C:>\ win to enter the Windows 95/98 GUI.  If this  seems a bit like Windows 3.x and "deja vu all over again", it is.)

The LTP24 (and LTP222A) program can now be started from within ANY folder, in contrast to LTP114J (and it doesn’t have to be the root folder used here).

When the LTP program starts up, it automatically determines if the computer is too slow for the chosen data acquisition board, in which case it terminates. However, after this initial 'speedtest' the LTP program does not again check if the computer is fast enough for the data acquisition board. This means that if there was a microprocessor slowdown (which could occur if the microprocessor when into a power saving mode on a laptop computer, or if the 'Turbo' speed was inadvertently turned off), a 100 msec sweep would actually take twice as long (200 msec) in real time. This would mean that a 10 msec duration EPSP would actually appear having a 5 msec duration on the ADsweep graph, and the slope would appear twice as large. If the synaptic potential duration or slope immediately and inexplicably changes by a factor of two, check that the synaptic potential duration is right by independently measuring it with an oscilloscope. Alternatively, if an oscilloscope is not available, restart the LTP program to test the data acquisition board speed again.

All single AD waveform files (*.P0, *. BP0, *.FP0, *.AP0, *.LP0 and *.IP0 etc), all Amplitude/Calculation files (*.AMP), and all LaserJet print files (*.LJ) will be stored in this C:\YOURDATA\010704 start-up folder. All initialization (*.INI) and protocol (*.PRO) files will be stored in the program folder C:\LTP24.

2.4.2  Windows 95/98  Reanalysis

You can start LTP24 program reanalysis by:

1) Go to the Windows Explorer and click on the ltp24.exe file to run the program in an DOS box.

2) Click the MS-DOS prompt to go to the DOS box command line, and from any folder type in:

C:\> ltp24

3) Click the supplied desktop reanalysis icon shown below.

Due to a fault in the Windows 95/98 DOS extender, the LTP program doesn’t know how much RAM memory is available, only if it is too little in which case the program exits.  Therefore, in reanalysis mode in the DOS box of Window 95/98, the number of samples/sweep has been limited to 50,000.  If you need more samples/sweep during reanalysis, run the LTP program in the MS-DOS mode, and use:

C:\LTP24 reanalysis_msdosmode

if your computer has a Digidata 1200 board installed (which would otherwise be automatically detected and will send the LTP program into acquisition mode).

2.4.3  MS-DOS 6.x  / Windows 3.xx  Acquisition

The LTP program will acquire data only when started in the 'Real' Command Line DOS mode.   To enter this mode, just start the computer and don't go into Windows (e.g. do not type C:>win).  If you are already in Windows 3.x, exit Windows completely (File->Exit) and return back to 'Real' Command Line DOS.  Don't click on the MS-DOS icon in Windows 3.x because this will merely go to the DOS Compatibility Box. 'Real' Command Line DOS is not equivalent to the DOS in the DOS Compatibility Box in Windows.

The LTP24 program can now be started from within ANY folder, in contrast to LTP114J (and it doesn’t have to be the root folder used here).  

When the LTP program starts up, it automatically determines if the computer is too slow for the chosen data acquisition board, in which case it terminates. However, after this initial 'speedtest' the LTP program does not again check if the computer is fast enough for the data acquisition board. This means that if there was a microprocessor slowdown (which could occur if the microprocessor when into a power saving mode on a laptop computer, or if the 'Turbo' speed was inadvertently turned off), a 100 msec sweep would actually take twice as long (200 msec) in real time. This would mean that a 10 msec duration EPSP would actually appear having a 5 msec duration on the ADsweep graph, and the slope would appear twice as large. If the synaptic potential duration or slope immediately and inexplicably changes by a factor of two, check that the synaptic potential duration is right by independently measuring it with an oscilloscope. Alternatively, if an oscilloscope is not available, restart the LTP program to test the data acquisition board speed again.

All single AD waveform files (*.P0, *. BP0, *.FP0, *.AP0, *.LP0 and *.IP0 etc), all Amplitude/Calculation files (*.AMP), and all LaserJet print files (*.LJ) will be stored in this C:\YOURDATA\010704 start-up folder. All initialization (*.INI) and protocol (*.PRO) files will be stored in the program folder C:\LTP24.

2.4.4  MS-DOS 6.x  / Windows 3.x  Reanalysis

Just go to the File Manager and click on the ltp24.exe file to run the program in an DOS box.

Alternatively, click the MS-DOS prompt to go to the DOS box command line, and from any folder type in:

            C:\> ltp24

Due to a fault in the Windows 3.xx DOS extender, the LTP program doesn’t know how much RAM memory is available, only if it is too little in which case the program exits.  Therefore, in reanalysis mode in the DOS box of Window 95/98, the number of samples/sweep has been limited to 20,000.  If you need more samples/sweep during reanalysis, run the LTP program from the command line in Real DOS and use:

C:\LTP24 reanalysis_msdosmode

if your computer has a Digidata 1200 board installed (which would otherwise be automatically detected and will send the LTP program into acquisition mode).

2.4.5  Windows ME Acquisition

LTP24 can only be used in acquisition mode if the computer is booted up into DOS 6.x or DOS 7.0.  Windows ME is similar to Window 95/98 with the exception that it has no MS-DOS (real DOS) mode (which is too bad because that is what LTP24 uses in the acquisition mode).

Remember that using LTP24 for acquisition also requires the Windows ME computer to have an ISA bus (although the Pico ADC-42 board may work without it).

However, a computer using Windows ME can be booted up into DOS 7.0.  The simplest thing to do is to format a 1.44 MB floppy disk in a Windows 98 computer (C:\format a: /s from the DOS command line).  Put the floppy disk into the A: drive of the Windows ME computer, and start the computer.  For these computers, if the computer starts up with a disk in floppy drive A: that has been formatted with DOS 7.0 (e.g. in a  Windows 98 computer), then the operating system used to format that floppy (DOS 7.0 of Windows 98) will be the operating system that the computer with Windows ME installed actually will run in. In other words, putting in a floppy disk formatted in a DOS 7.0 (Windows98) computer causes the Windows ME computer to become a DOS 7.0 (Windows 98) computer.

Alternatively, you can use a program such as Partition Magic to install a dual boot system on the Windows ME computer and boot up into a DOS 6.x or 7.0  partition 

2.4.6  Windows ME Reanalysis

You can start LTP24 program reanalysis by:

1) Go to the Windows Explorer and click on the ltp24.exe file to run the program in an DOS box.

2) Click the MS-DOS prompt to go to the DOS box command line, and from any folder type in:

C:\> ltp24

3) Click the supplied desktop reanalysis icon (for Windows 98) shown below.

Due to a fault in the Windows ME DOS extender, the LTP program doesn’t know how much RAM memory is available, only if it is too little in which case the program exits.  Therefore, in reanalysis mode in the DOS box of Window 95/98, the number of samples/sweep has been limited to 50,000.  If you need more samples/sweep during reanalysis, restart the computer in DOS 7.0 (MS-DOS mode), run the LTP program in the MS-DOS mode, and use:

C:\LTP24 reanalysis_msdosmode

if your computer has a Digidata 1200 board installed (which would otherwise be automatically detected and will send the LTP program into acquisition mode).

2.4.7  Windows NT, 2000 or XP Acquisition

LTP24 can only be used in acquisition mode if the computer is booted up into DOS 6.x or 7.0.  And because DOS 6.x requires FAT and DOS 7.0 requires FAT or FAT32 file systems, at least one hard disk partition in the Window NT, 2000 or XP computer requires a FAT or FAT32 file system.  If NTFS is the only file system being used, convert it to FAT or (better yet) FAT32 using a program such as Partition Magic 6.   

Remember that using LTP24 for acquisition also requires the Windows NT/2000/XP computer to have an ISA bus (although the Pico ADC-42 board can  work without it).  The Pico board, using the parallel port, will work on at least some of the computers having only a PCI bus (including the Dell 8200 1.8GHz Pentium 4 computer with a thin screen monitor). 

Then the simplest thing to do is to format a 1.44 MB floppy disk in a Windows 98 computer (C:\format a: /s from the DOS command line).  Put the floppy disk into the A: drive of the Windows NT, 2000 or XP computer, and start the computer.  For these computers, if the computer starts up with a disk in floppy drive A: that has been formatted with DOS 7.0 (e.g. in a  Windows 98 computer), then the operating system used to format that floppy (DOS 7.0 of Windows 98) will be the operating system that the computer with Windows NT, 2000 or XP installed actually will run in.  In other words, putting in a floppy disk formatted in a DOS 7.0 (Windows98) computer causes the Windows NT/2000 computer to become a DOS 7.0 (Windows 98) computer.

Alternatively, you can use a program such as Partition Magic to install a dual boot system on the Windows NT/2000/XP computer and boot up into a DOS 6.x or 7.0  partition 

2.4.8  Windows NT, 2000 or XP Reanalysis

You can start LTP24 program reanalysis by:

1) Go to the Windows Explorer and click on the ltp24.exe file to run the program in an DOS box.

2) Click the MS-DOS prompt to go to the DOS box command line, and from any folder type in:

C:\> ltp24

3) Click the supplied desktop reanalysis icon shown below.

Due to a fault in the Windows 2000 DOS extender, the LTP program doesn’t know how much RAM memory is available, only if it is too little in which case the program exits.  Therefore, in reanalysis mode in the DOS box of Window 95/98, the number of samples/sweep has been limited to 50,000.  If you need more samples/sweep during reanalysis, restart the computer in DOS 7.0 (MS-DOS mode), run the LTP program in the MS-DOS mode, and use:

C:\LTP24 reanalysis_msdosmode

if your computer has a Digidata 1200 board installed (which would otherwise be automatically detected and will send the LTP program into acquisition mode).

2.5 Automatic data folder creation at startup, AutoCreate a new data folder, and Change data folder

2.5.1  Automatic Data Folder creation at startup

When LTP24 starts up, the first screen shown below is a ‘splash screen’ (Fig. 2.5.1, top panel) that shows whether the program is in Reanalysis Mode or in Acquisition Mode (by showing which data acquisition board is in use and at which I/0 port location or LPT port number).  Fig. 2.5.1 shows that here the LTP program is running in acquisition mode and that it is using the Labmaster board at I/O location hexidecimal 710.  A dialog box (“Set Data Folders”) is presented in the ‘splash screen’ which allows you to either “Create a New Data Folder”, which you would do if you were to Run an Experiment and to put your data for the day  into a new data folder.  Alternatively, you could choose to “Use an Existing Data Folder” if you were reanalysing data.

If you decided to “Run an Experiment” the “Running Experiment …” dialog box would be presented (next dialog box down) which requests that you to fill in a data root drive and folder.  The data root folder is the folder off of which all subsequent data folders will be created.  In Fig. 2.5.1, the data drive folder is C: and the data root folder is the folder containing the LTP program, \LTP22XX (just think of the LTP program being LTP22XX instead of LTP24 for the purposes of Fig. 2.5.1).  The LTP program folder C:\LTP22XX (e.g. C:\LTP24) is the default data root folder that is presented upon initial startup.  Therefore if today was July 4, 2001, then the data folder into which the *.P0, *.AMP and *.LJ files would be written would be C:\LTP22XX\010704, and this data folder would be automatically created.  The first two characters of the data folder denote decade and year, the second two characters denote month and the third two characters denote day of the month.  With this method, folders are automatically sorted year/month/day when viewed with the Windows Explorer. 

Alternatively, if you have two people using one computer to run experiments, the data root drive and folder could be changed to a non-default data root folder such as C:\BILLDATA, and the data folder automatically created on July 4, 2001 would be C:\BILLDATA\010704.  The other person would just have to change the data root folder on LTP program startup to their data root folder.

Alternatively, if you are reanalysing data, choose the “Use Existing Data Folder” option in the “Set Data Folders” dialog box and the “Reanalyzing Data” dialog box appears (see Section 6.1).

2.5.2  AutoCreate a new Data Folder (using CTL-F) while running an experiment

While running an experiment that saves several thousand ADsweep files, we have found that the speed at which files are saved decreases when over a thousand files are saved to a folder.  To fix this problem I added the ability to AutoCreate a new data folder while running an experiment.  This can be accomplished by choosing “AutoCreate new data Folder” in the File menu (Fig. 2.5.2.1).  Alternatively, you can quickly AutoCreate a new data folder ‘on-the-fly’ (e.g. while saving ADsweeps) just by pressing the CTL-F hotkey.

When this is done, the data folder is changed from C:\ltp22XX\010704 to the newly created data folder C:\ltp22XX\010704B (Fig. 2.5.2.2).  Up to 26 data folders (e.g. up to C:\ltp22XX\010704Z) can be AutoCreated in this manner.

When saving ADsweeps to a hard disk having a FAT32 file system, the AutoCreate a new data folder is unmeasurably fast, but when saving to a hard disk having the older FAT (16) file system, the AutoCreation can take several seconds.  It is therefore better to use a FAT32 file system if speed is needed.

2.5.3  Change to a different data folder during acquisition

If you wish to change to a different data folder during either acquisition or reanalysis you can choose the “Change data folder…” selection in the File menu (Fig. 2.5.2.1) to bring up the dialog box in Fig. 2.5.3.1.   

Fig. 2.5.3.1.  Change Folder Dialog Box.

Choosing “Using Same Read/Write Data Folder” allows you to set the Read/Write data drive and data folder by presenting  the “Using Same Read/Write Data Folder” dialog box in Fig 2.5.3.2.  Changing this dialog box can set the Read/Write data folder say back to the original C:\LTP22XX\010704 data folder.

2.6 Set the AD Gain, DataType, Sample Interval and Other Parameters

2.6.1 Set the AD Gain, Data Type, and Sample Interval in the Miscellaneous Window

First, set the channel AD0 DataType to "mV".

Second, change the AD0 gain to be equal to the total amplification gain from the electrode to the AD board connection. For example, if an Axon Instruments AxoClamp is used in current clamp mode with an internal x10 output gain, which then goes into an external x100 amplifier and then into channel AD0, the AD0 Gain would be 1000. 

Alternatively iif an Axon Instruments AxoPatch 1D or 200B is used in whole cell voltage clamp mode (with a CV-04 1/100 headstage where beta = 1), with an output gain of alpha = 10, and with the output going channel AD0, the channel AD0 Gain would be the following:

(Yes that's 10 0's!) If the output gain is changed to = 20, the AD0 Gain increases to 20000000000. Then set the channel AD0 DataType to "pA".

In LTP24 I have added the capability to change the gain by alternatively entering it in AD0 Units/V (Fig. 2.6).  While maybe it is clearer to just type an AD Gain of 1000 for an extracellular or intracellular experiment in "mV", with a patch clamp experiment in "pA" with enormous gain, a AD Units/V of 100 pA/V is much clearer than an AD Gain of 10000000000.

Third, set the PulseSweep ADinterval to 100 usec.

2.6.2 Set the Digital-to-Analog Output

There is no conversion or gain controlling the DA output voltage. The PulseVolt value you enter (in volts; see Fig. 3.1.2) is the output voltage you get (in volts) from the Labmaster or Digidata 1200 Digital-to-Analog output channel.

2.6.3 Set the PulseSweep P0 Stimulation Values in the PulseSweep P0 Stimulation Window

Set Pulse stimulation values (see Section 4.5).

2.7 Analog Filtering of the Signal Before Digitization

The waveform signal data should be filtered before being digitized with an analog filter set to half or less of the digitization frequency. For example, if you are acquiring an AD sample at 100 usec intervals (e.g. at 10 KHz sampling frequency), the analog filter should be set to at maximum 5 kHz (1/2 the sampling frequency), or preferably to 2 kHz or lower (except if doing an Rs exponential fit, then keep it at 5 KHz). If the analog filter is set to higher than half the sampling frequency (for example to 10 kHz in the above example), mysterious things can begin to happen due to anti-aliasing such as loosing the capacitative transients (e.g. the Rs measurement) that occur during the voltage clamp pulse when patch clamping.

The analog filter can either be part of the biological amplifier such as the one on the AxoPatch 1D patch-clamp amplifier, or it can be a separate analog filter between the amplifier and the data acquisition board input. Generally, the best analog filters for electrophysiology are Bessel filters because they cause no overshoots, only phase lags.

2.8 Pressing a Function Key to Stimulate and Acquire Data

For the simplest situation of outputting one stimulus pulse (which can be measured with an oscilloscope) and simultaneously acquiring one sweep of data (an ADsweep), press function key F5.

Examine Train output on an oscilloscope or a chart recorder by pressing CTL-F8.

2.9 Check that the Data Acquisition and Stimulation are Working Correctly

The Data Acquisition / Stimulation functions should be carefully examined by:

2.9.1  Input a triangular waveform

Input a known (calibrated) signal using voltage and time calibrated stimulator and see if the correct values are plotted on the AD sweep graphs. Be sure to measure waveform acquisition timing by using a TRIANGULAR WAVEFORM input to make sure no delays occur during the data acquisition/ stimulation sweep. If there is a timing delay, the ramp will not be smooth, but will appear to instantaneously jump to a higher voltage and then continue (Fig. 2.9.1).  Ramps can also be used to test if the slope calculation is working correctly (see Section 4.7.3)

2.9.2.  Measure Sample-to-Sample Jitter with an Oscilloscope

Use a calibrated oscilloscope to measure the voltage and timing of Analog or Digital output pulses that are generated during Pulse and TrainSweeps. This will test for sample-to-sample jitter because both Analog input and Analog/Digital output occur at about the same time because an AD conversion always starts when an Analog/Digital output occurs. For example, the sample-to-sample jitter can be tested by measuring the width of an S0 pulse during a train (Fig. 2.9.2).

2.9.3.  Check AD Channel Crosstalk in the Measurement Computing's CIO-DAS08/JR-AO and Labmaster boards

The 'weakest link' of the DAS08/JR board is a small, but measurable, amount of crosstalk between 2 AD channels (see Section 2.3.4.2 and also APPENDIX B11).  Note, however, that AD channel crosstalk is NOT a problem with the DAS08/JR board when using just 1 AD channel.  

If -5 to +5 volts is input into one AD channel, then a 1-4 bit positive shift (usually 1-3 bits) occurs in the other AD channel.  In contrast, the Digidata 1200 has only up to 1 bit shift.

Just remember that if you are amplifying well above the 'bit' level, this should not be a problem.  In other words, if the signal you are measuring is say 400 bits in amplitude (approximately 1/10th of the full scale of 4096 bits), then a 4 bit shift (caused by the other AD channel going from -5 to +5 volts) would be a 1% error.

Futhermore, the Labmaster board also has a small, but measurable, amount of crosstalk between 2 AD channels (see Section 2.3.2.2 and alsoAPPENDIX B11).  AD channel crosstalk is NOT a problem with the Labmaster board when using just 1 AD channel.  If -10 to +10 volts is input into one AD channel, then a 2-3 bit shift occurs in the other AD channel.

There are two basic ways to measure the AD channel crosstalk, directly testing the DAS08/JR or Labmaster board in the computer, and testing the DAS08/JR or Labmaster board at your setup before an experiment.

To directly testing the DAS08/JR or Labmaster board in the computer, connect the IC analog output to one AD channel (e.g. AD0) and ground the other AD channel (e.g. AD1).  Put a -5 to +5 (or -10 to +10) volt signal of substantial duration (say 50 msec per pulse) out of  IC.  Change the AD0/AD1 DataTypes to "V", change the AD0/AD1 Gains to x1, and put the AD0 sensitivity to a low value (ca 10V (or 20V) peak-to-peak), and the AD sensitivity to a very high value (ca. 0.02V peak-to-peak).  Then measure the shift in the grounded AD channel, AD1 (Fig. 2.9.3).  Use AvgAmp (and therefore also Baseline) to record the shift in voltage by setting the Baseline cursors during the other AD0 channel's -5 (or -10) volt pulse, and the AvgAmp cursors during AD0 channel's +5 (or +10) volt pulse.  Because analysis and reanalysis is done at the 16 bit level with the DAS08/JR boards, the values will be to within 1/16th bit.  Analysis and reanalysis is still done at 12 bits on the Labmaster board.  Also try different IC output voltage regemes.  

Then repeat the procedure by connecting the IC analog output to  AD1, and ground AD0.

Alternatively, particularly if you are using the DAS08/JR board (without the analog outputs), use the -5 to + 5 volt square wave input from an external waveform generator to test AD channel crosstalk.

The shift in AD1 voltage for the DAS08/JR board should be up to 4 bits positive when AD0 goes from -5 to +5 volts.  The shift in AD0 voltage should be up to 4 bits positive when AD1 goes from -5 to +5 volts.  The shift in AD1 voltage for the Labmaster board should be up to 3 bits when AD0 goes from -10 to +10 volts.  The shift in AD0 voltage should be up to 3 bits when AD1 goes from -10 to +10 volts.  Any substantial increase in these values indicates a problem, and it is questionable whether this DAS08/JR or Labmaster board should be used for 2 AD channel recordings.  Choose a Digidata 1200 board instead.

In the measurements taken in Fig. 2.9.3 for the DAS08/JR board, going from -5 to +5 volts in AD0 caused about a 0.8 bit positive shift in AD1.  Note how the -5 volt pulse was much more effective in causing a bit shift than the +5 volt pulse.

To test the DAS08/JR or Labmaster board at your setup before an experiment, say if you are measuring an EPSP in AD1, turn down the amplification to AD1, and put out a fairly long (ca. 50 msec) pulse into AD0 by say using an electrode test measuring pulse or a voltage clamp pulse such that the pulse would go from 0 to -5 (or -10) volts, and 0 to +5 (or +10) volts at the AD0 input.  You could even try grounding out the input to AD1.  Measure the shift in AD1 by using AvgAmp.

The shift in AD1 voltage for the DAS08/JR should be up to 4 bits positive when AD0 goes from -5 to +5 volts.  The shift in AD1 voltage for the Labmaster should be up to 3 bits positive when AD0 goes from -10 to +10 volts.  Any substantial increase in these values indicates a problem, and it is questionable whether this DAS08/JR or Labmaster board should be used for 2 AD channel recordings.  Choose a Digidata 1200 board instead.

Feel free to contact support@winltp.com if you do not understand this and want to use 2 AD channels on this board.

Fig. 2.9.3.  Measuring AD channel crosstalk on a DAS08/JR-AO board using AvgAmp (and its Baseline) measurements.  A -5 to +5 volt was input into AD0 from the IC analog output.  The AD1 input was grounded.  A 0.8 bit positive shift occurred in AD1.

2.10 Choosing the Printer

2.10.1. Hewlett-Packard LaserJet and Compatible Printers

The LTP program runs with Hewlett-Packard LaserJet II compatible printers (II, III, 4, 5 and 6 etc.) using PCL4 graphics commands. Moderately high resolution (300 dots/inch) plots of ADsweep and Amplitude/Slope graphs are output to the LaserJet printer.

and then from the upper group of the dialog box choose either 'None' or 'HP LaserJet II, III, 4, 5, 6' (Fig. 2.10.1).

If you are using a printer, connect the parallel printer port on the printer to parallel printer port LPT1, LPT2 or LPT3 on the computer, then from the lower group of the dialog box choose either 'LPT1', 'LPT2' or 'LPT3'. If you are already using a Pico board with digital output on LTP1 or LTP2, make sure this does not conflict with printer usage.

These values determining hardware specifications are automatically saved in the hardware.ini initialization file every time a new value is entered.

2.10.2. Hewlett-Packard DeskJet and Compatible Printers

The LTP program does not specifically support the popular Hewlett-Packard DeskJet and similar inkjet printers, but we have found through trial and error that the LTP program can plot ADsweep and Amplitude/Slope graphs to the Hewlett-Packard DeskJet 870cxi, and may well do so for other deskjet/inkjet printers. Just choose 'HP LaserJet II, III, 4, 5, 6' from the upper group of the Printers dialog box (see Fig. 2.10.1). You just have to try your DeskJet/inkjet printer to find out if this will work.

Although plotting the Amplitude/Slope calculation graphs worked fine, we did notice that only one ADsweep graph could be plotted per page (say out of four). This problem has not been corrected.

2.10.3. Other Printers/Plotters

The LTP24 program does not support Epson FX80 compatible printers, Postscript language printers or Hewlett-Packard 7475 compatible plotters, and will  not do so in the future.

2.11 Printing an AD Waveform Graph to the Printer

If the program has performed the above tasks, it is ready to be set-up in more detail for your particular experiment (Chapter 4).

Go to LTP Manual - Chapter 3

Return to the LTP Manual (Table of Contents)

Return to the LTP Program Home Page