A (short) History of the
Model 2000™ Control/Data Acquisition Instrument

It is because of the enthusiasm of dedicated physicists, scientists and engineers all around the world that QCM Research has been able (and willing) to put considerable capital into new product development. We have made it our goal to have the finest products possible at the lowest feasible prices.

While pursuing this goal it was learned that it would be desirable to have a better control and data acquisition system. The Model 1700^TM (M1700) controller, an analog device developed in the mid 70s, had evolved into the Model 1900^TM (M1900) by 1983. This controller had served well (and still serves well, with better than 230 units in the field) for many years, but a digital version was widely requested.

The M1900 controller worked well with the Mark 9 QCM^TM (MK9) but when the TQCMs (the Mark 10^TM and Mark 14^TM particularly) went into production, a new controller was called for. The MK9 (and that generation of sensor) had a PRT used both for heating and temperature sensing. The new TQCM still used a PRT for temperature sensing, but used a Peltier (a heat pump, using as much as 1.25 Amp, worst case) for both heating and cooling. This called for a new controller, the Model 1800^TM (M1800), which later became the M1800A. This new controller was used in conjunction with the M1900 to control the temperature, and read the frequency of the TQCMs. These units had external frequency and DC outputs on the rear panels to port the collected information to a storage device, such as a chart recorder.

This approach worked well for many years, but was expensive and awkward because if you had six TQCMs, you needed six sets of controllers. Also, with the advent of the personal computer being used as a lab computer, came the expressed desire to be able to gather data and control temperature remotely (via RS232 interface).

At about this time, the Mark 16 CQCM^TM (MK16) was developed. It has a Silicon Diode temperature sensor and 2.5 watt wire wound heater. None of the extant controllers were going to be able to be easily converted to accompany this new sensor, so in 1988 we developed a combination M1800/M1900, appropriately called the Model 1819^TM (M1819) controller.

The design requirements for the M1819 were for a controller which would be able to (1) stand alone but also to send data to and receive commands from a remote computer, and (2) independently control up to eight MK16s at a time. These goals were ultimately accomplished, but the M1819 still did not meet the request of many customers for a unit that would be able to concurrently run even more QCMs of many different varieties (quickly and accurately) yet completely independent of each other.

This demanded an entirely new kind of controller, the Model 2000 Control/Data Acquisition Unit^TM (M2000). It was to be a low cost unit for interfacing the various mass sensors we made to a general purpose host computer. The unit would have a minimal front panel (power switch, lamp, status indicator) and communicate with the host computer over an RS232 serial link.

The completed M2000 design accommodates both Si-diode and PRT temperature sensors and will drive wire wound heaters, PRT heaters, and both single and dual Peltier heat pumps. Additionally, it is able to power an external heater, such as is used for an effusion cell in our Vacuum Outgassing Deposition Kenitics Apparatus^TM (VODKA).

While autonomously controlling the temperature, both temperature and frequency data for each channel are sent to the host over the serial link. The base unit can handle four channels, expandable to twelve channels, in steps of four. The data is sent to the host at a rate of one reading per second per channel (ie. twelve channels in one second).

The temperature data sent to the host is in degrees Kelvin, with the least significant digit representing 1/100 of a Kelvin (0.01 K). The accuracy of the reading is affected by five elements: sensor accuracy, curve fitting accuracy, electrical noise, quantitization error and thermal drift. The electronics add no more than 1/4 of a Kelvin (< 0.25 K) error to the accuracy of the sensor for absolute accuracy. For relative accuracy, which eliminates thermal drift, the added error is less then 1/10 of a Kelvin (< 0.1 K).

Frequency data sent to the host is in Hertz. The relative accuracy is better than one part per million (> 1 ppm). For all frequencies (1 KHz to 600 KHz), seven digits with a floating decimal point is sent. Thus, for a frequency of 100 KHz, the least significant digit represents 1/10 Hz (0.1 Hz).

We initially intended the software to serve as a "bare-bones" example, which the customer could then use to write their own finished product. However, it soon became apparent that a turn-key system was desired by the majority. The software now included is simple to install and run. There is an integral programming language so that simple command sequences may be given if you wish (like a script file). For example: Go to 100 C; Stay there for an hour; Go to -200 C at 2.5 C/Min; Stay there until further notice.

Several customers have requested the source code to the software and have opted to write their own as we originally intended but by far the majority of customers use the included software "as is". We now include the source code as part of the software package.

A specialized version of the M2000 was designed for the VODKA (as was previously mentioned) necessitating a dedicated software design. After everyone had had their say, we could see that it would be difficult, if not impossible, to satisfy all of the requests without a major rewrite of the software so we decided that this would be a good time to go to a different operating system, OS/2. The new VODKA Software is now a huge success and makes the work of running the VODKA largely intuitive (the VODKA comes with thousands of pages of documentation).

In the works is a much requested rewrite of the regular M2000 Software using Windows and incorporating the ability to graph the results of tests in real time.

The M2000 has been on the market for more than two and a half years so there is now a large installed customer base.

The M2000:

1. Can control up to twelve (12) QCMs at once;
2. Will control any combination of QCMs;
3. Sends data at a rate of one reading per second;
4. Has provision for the control of an effusion cell;
5. Communicates using industry standard RS232 interface and
6. Starts at a lower price than the M1900/M1800A system.