IMP version 2.05

This is a no-cost (free) update for owners of IMP version 2.00 through 2.03. 

To get the latest version ---- 

Download the IMP205 zip file here 

and unzip its contents to a floppy disk, then copy the contents to your C:\IMP directory.  



  This will probably be useful to you only if you have a means of burning an EPROM and can make the circuitry for an IMP MLS add-on board.  Note that we are NOT able to provide these boards or EEPROMS anymore.  The file is given here for reference only!   

download the IMP MLS ROM file

Notes about changes and additions in IMP v2.05:

It has been several years since the last release of IMP/M. IMP's memory size is nearly at the limit that can be handled within its program constraints -- this and the hardware design of the IMP module limits the addition of further high leve features. To change from "real mode" to "protected mode" (allowing for use of memory size greater than about 512K) would require some fundamental changes in the programming code and strategy

The previous version has proven very robust and elegant in operation, as well, leading us to be very hesitant to tamper with its code -- seldom are changes made in a complex program without introducing new bugs (particularly when memory
limits are being approached).This release addresses some of the bugs and some of the feature requests we have received. Note that any feature additions are NOT covered in the program's Help files. Relatively few feature additions have been included in this release. Also, some requested features or behavior, while easy to describe, are not so easy to program -- sorry if your "favorite" has not been included!

1) Fixed a program crash which occurred if a Time Domain Overlay is attempted with the Memory trace not included.

2) Display of floating point values now round correctly (for example, "9.999" in 4 characters is now shown as "10.0" rather than as "9.99"). NOTE: as before, any value displayed as "ooooo" or "-oooo" (series of lower-case "o"s) indicates that the number cannot be displayed in the spaces provided -- read this a as an "overflow" or a very large (and unlikely to be accurate) value.

3) The Window selections for "autoMeasure" now work for Electrical (probe) frequency response measurements as well as for Mic measurements. NOTE that the window type for autoMeasure is taken ONLY from the [* autoMeasure Setup] menu, NOT from F7! And that it will NOT be displayed at the top menu line; and the F7 window type will be set to "None" after all frequency response autoMeasure operations. The window you select in [* autoMeasure Setup] IS applied, but its result will be seen only in the FFT result, not on the top-of-screen indicators!

4) The trace outside the selected region of a time-domain graph in the lower window is now displayed in yellow, to highlight the region (in red) that is between the time domain markers.

5) There are now a selection of printer formats for HP (LaserJet or inkJet) type printers, in various sizes and orientations.

6) There is now an option to print to a Windows-compatible Bitmap file, rather than to a printer. The file created will be placed in your C:\IMP directory (assuming you are installed on drive C) and can be either a color or a black-and-white .BMP file. This file can then be imported into Windows wordprocessors or graphic editor programs for further editing, inclusion in reports, or printing to printer types other than the natively supported "IBM/Epson" or "HP" types.

7) An [* autoMeasure imPulse] function is now included for using MLS for "live" cycling impulse response monitoring. This is handy for doing time domain alignment of multi-driver arrays.

8) The code has been recompiled using a modified Borland "CRT" Unit, to avoid a crash that could happen when using very fast (Pentium Pro 200 and up) processors.

9) Shift-F8 now does not transport the user to other menu locations. Users should note, though, that the F? buttons are not always independent of the menuing (and vice-versa). Some function keys are shortcuts to menu positions; others are "interruptions", which return to the position evident before the key was used. Also, some function keys will give a different result depending on program conditions: for example, [Shift-F8] will change "ohms per division" if an impedance plot was last active, or "dB per division" if a frequency response plot was last active.

10) Users who are uncomfortable with the "t=0" part of a waterfall showing non-empty response are directed toward readings on what the cumulative spectral decay plot means! (A short article on the subject is available on our web site at In particular, it is ***NOT*** (!!!!) "frequency response versus time"!

11) Note that smoothing does NOT work on merged curves (and is not meant to). Smoothing can be applied only after the data has been converted back to frequency response.

12) Overlay plots, because they contain multiple curves, cannot indicate
the type of smoothing used or whether the data is cal'd. Each curve can
be different.

13) There are no markers for Cal, Mic Data, or other response displays than 1: time domain, 2: frequency response, 3: impedance. These three marker sets are independent and separately maintained, appearing according to the current display mode.