|
3. NOTES FOR PRIOR L-17C USERS |
| |
| A Brief Description of Major Differences between the L-17C and the L-17D |
|
| |
Since many prior L-17C users are now switching to the L-17D, we provide here a brief summary of the major differences they will encounter.
In general, the available bandwidth is greater, the noise at a given bandwidth is reduced, and the recovery time from large signals is faster. Also, the signal transit time through the IC is now greatly reduced. The most important differences and improvements are described in more detail below.
Performance:
The performance of the L-17D has been upgraded in several important areas: |
| |
- The available bandwidth of the IC has been increased substantially. We have run the L-17D at over 60 MHz. It can, however, be run at much lower bandwidths if speed is not an issue, but rather the noise needs to be minimized. Both the transit time and the transit time spread are much lower with the L-17D at high bandwidth. It is also very easy to get an excellent output pulse shape.
- The noise has been reduced from 1.5nV/√Hz to 1.2nV/√Hz. Since the current noise is negligible for most applications, this should give about 1dB improvement in TSS for input devices with an output impedance of the order of 50 ohms (e.g. tunnel diodes), and less for higher output impedance input devices.
- The logging range has been extended. The basic unit now contains 8, rather than 7, log stages. The linear extension now has 3 stages, as opposed to the single stage on the L-17C. This enables the range to be extended substantially if the positive input is used. A unit has been built which gives 60dbm range with a single L-17D using a Schottky diode. An extra log adjust has been added so that L1, L6, and L7 can now be adjusted. The log linearity also appears to be a little better than that of the L-17C over the L-17C’s dynamic range.
- The power supply rejection ratio has been improved dramatically. It is no longer necessary to use circuitry that ensures good dual-tracking to get optimum performance. The decoupling caps on the power traces can also be smaller since the improved PSRR now extends into the MHz range. The IC can now be run at rail voltages from +/- 5V to +/- 9V. However, at higher voltages, it is necessary to ensure that adequate thermal contact is made to the board or a heat sink is used, as the required power dissipation is increased. At a given rail voltage, the L-17D uses slightly less current than its predecessor.
- The behavior over temperature has been improved in two areas. First, the output slope is now temperature independent, so it is not necessary to compensate the output feedback resistor with a sensistor net. Second, a great deal of care was taken in the design to ensure that the voltage cutoffs of A1 and A2 are independent of temperature. This makes it a lot simpler to utilize recovery nets.
- The A1 output is no longer internally connected to A2 input, and similarly A2’s output is not internally connected to the A3 input. The reason for doing this is to enable people who work with short pulses, and do not need to run DC coupled, to couple capacitively in order to avoid the need for temperature trimming. Limiters can also be inserted between A1 and A2 where optimal recovery time is required.
|
| |
11 February 20, 2008 |
|
|