Direct Digital Decoupler

Front Panel

Rear Panel

The Direct Digital Decoupler is an independent transmitter channel for MR systems that would typically be used to allow proton decoupling during acquisition of Phosphorous spectra. It can also be used for NOE.

It was designed to interface to the MR system through a single TTL signal, so that no complicated wiring or special hardware is needed to get up and running quickly.

General

• rack mountable or benchtop enclosure 3.5 in. x 15 in. x 17 in.
• TTL gate output for gated power amplifiers
• TTL input from scanner to enable decouple output (Vin > 2.4V selects decouple output, Vin < 0.7V selects NOE, 47k pullup to +5V)
• switch to select NOE output, decouple output, or both
• main power switch on rear
• WALTZ-16 sequence pulse widths of 0.16 ms, 0.32 ms, 0.64 ms, 1.28 ms, and 2.56ms
• LED: red during decouple output, green during NOE output, off if no rf is being generated

Frequency

• 0 - 120 MHz frequency range
• 1 Hertz frequency resolution
• +/- 0.28 ppm stability
• frequency set by front panel thumbwheel switches

Amplitude

• separate NOE and decouple power levels
• 0 - 150 mV p-p (-12.5 dBm)
• 256 steps chosen by ten-turn pots on front panel

Options

• higher stability ocsillator: +/- 0.03 ppm (extra cost and lead time)
• internal power amplifier: 5W or 10W RF output (extra cost)
• single sideband mixer input for generation of higher output frequencies (extra cost)
• second TTL control input

Circuit Description

This instrument is based on a direct digital synthesizer (DDS), under control of 2 microcontrollers. A stable 20 MHz reference clocks the microcontrollers and the DDS. The DDS uses a 15x PLL multiplier to generate an on-chip 300 MHz source.

The first microcontroller handles front panel inputs from the user (such as the conversion of the front panel frequency into the correct binary data scaled for the DDS). The second microcontroller takes control information from the first microcontroller, and TTL signals from the rear panel, and directly controls the DDS (generating the WALTZ-16 sequence, setting the amplitude, etc.).

This design eliminates the need for an external synthesizer and provides a basis for incorporation of different modulation sequences in the future, without the need for hardware changes.

User Notes

Please check that the following points have been met:

1. A TTL signal (both 5V and 3.3V systems work) is supplied from the MR system to the decoupler via coax cable with a BNC connector (and no 50 ohm termination). This signal should be high for the period when a decoupler radio frequency output is desired.
2. The rf output is connected to an appropriate rf power amplifier input. In most systems, the rf amplifier output will go through a bandpass filter, and then out to the probe.
3. If the power amplifier has a gating capability, connect the TTL Gate Out to the power amplifier's gate input.
4. The "DCPL Level" is set to produce a 90 degree pulse at the pulse width selected with the 5 position switch. Note that the bandwidth of the decoupler output is approximately the inverse of the pulse width.
5. If NOE is in use, set the "NOE Level" to the appropriate level. Some users set the NOE level to 10% of the DCPL level.
6. The NOE/BOTH/DECOUPLE switch is in the correct position.
7. The rf power output is at safe levels during both the NOE and DCPL periods.

Frequently Asked Questions

If the above conditions seem to be in order, please check the following FAQs before contacting us at decoupler@curdes.com:

Q1. The coax cable from the scanner is connected, but the LED isn't switching?

A1. First be sure that the NOE/BOTH/DCPL switch is in the correct position. If the input is always low, for instance, and the switch is in the DCPL only position, the LED will be off. It is probably simplest, for diagnostic purposes, if the switch is in the 'both' position until the triggering is verified (but be sure to leave it in the correct position before starting the actual study).

A simple way to check that the decoupler logic is switching correctly is to connect a 50 ohm terminator to the TTL A input, which makes that input low. But don't use the termination that's on the TTL B input, because removing that one will make the B input high, which will confuse the logic. Putting the termination on the A input places the unit in the NOE period, so the green LED should be on if either NOE or BOTH are selected. If DCPL only is selected by the front panel switch, then the LED should be off. Removing the termination from the A input should switch the unit into DCPL (because the input is made to float high). So you should see a green LED at that point (with the front panel switch in either DCPL or BOTH).

If the unit switches correctly with the terminator test, but not when the logic signal from the scanner is connected, it will be necessary to confirm that the logic signal from the scanner is switching as it should. Do this with an oscilloscope connected to the TTL A input, with a BNC Tee, so that the effect of loading by the decoupler input is verified at the same time. Confirm that signal switches to a low value below 0.7V and to a high value above 2.4V. Contact us for help with this is it is unclear.

Q2. My power amplifier's output level just changed even though I didn't touch the level potentiometers; what happened?

A2. Did you change the sweep width, number of points, or anything else at the MR console?

It is common for the logic signal duration to be set equal to the scanner's acquisition period, which is the number of points acquired times the dwell time between points (with, possibly, short gating delays). If the acquisition period has changed significantly, the power meter's response time may no longer be fast enough to give a true power reading.