Choosing a suitable Hardware Component Stack

Hi Scott. Thanks for the detailed reply.

I think the TDC1000-TDC7200EVM will allow me to achieve what I want, but I have an issue with choosing a suitable transducer.

With a 1 Mhz excitation frequency and a measured sound velocity of around 4500 m/s (which is about middle of the range for most Spruce woods), then we end up with an evaluated wavelength of about 4.5 mm.

The 'problem' with that number is that it is of the same order of magnitude as typical ring spacings for these trees (1mm - 5mm makes a good instrument candidate).

Under these experimental conditions there is going to be interaction between the excitation waves and the rings, which of course we want to avoid. So, a different excitation frequency is needed.

I think that 1 or 2 orders of magnitude difference in wavelength are really what's called for to avoid such interaction.

I have seen commercial ultrasonic transducers on the market rated at 60 kHz activation frequency.

Under such conditions the wave length is about 20 times longer (according to my calculations as shown in the attached image) such that 1 wave will encompass about 15 rings.

Can you recommend an alternative Steminc (or other) transducer pair that is 'TDC-ready' ?

Here are my calculations comparing 1 MHz and 60 kHz:

Scott, just started reading the FAQ ...

in brief,
Min. distance = 0.2 m (200 mm)
Max. distance = 1.0 m (1000 mm)
Typical distance ~0.4m (400 mm)

For the most part, these are blanks that instrument makers purchase.

I am looking primarily at already (longitudinally) cut samples (new/aged etc).
And, I am primarily interested in longitudinal (not transverse) measurement. Any transverse measurement would be at the lower end in terms of sample length.

Hi Scott.

Thanks for your latest suggestions.
What you say about measurements along the grain makes sense. The grain should not be a factor longitudinally, so the 1 MHz transducer should do the job.
In working out the min and max distances, using the info from the FAQ

Max distance = speed of sound in medium (m/s) * 0.008 seconds / 2

then 4500 * 0.008 / 2 = 18m ... so no problem there (unless we start looking at long logs !). Good.

Cross-grain minimum length measurements are the issue as you identify (at the lower frequency).
I am trying to validate this with a calculation of min times and distances according to the FAQ.
For the min. distance calculation,

Min distance = speed of sound in medium (m/s) * min time / 2
and
Min time (theoretical) = (number of TX pulses (at least 3)) * (1 / transducer frequency) + ringdown time
and
Min blanking period = 8 * (1 / system clock frequency)
In practice, the minimum time is further limited by the blanking time of the TDC1000.
The maximum system clock frequency is 16Mhz, which equates to a minimum blanking time of 500ns.

So, does this mean that the actual minimum time = blanking_period (at least 500 ns) + Min time (theoretical) ?
That is how I understand it.

The only 'unknown' to me in this situation is the ringdown_time.
Where do I get that information for any of the transducers ?

Thanks once again.
Julian

Hi Julian,

The GASEVM does not include any transducers to my knowledge. The board will also require modification to work with higher frequency transducers. Since it is set up with a bandpass filter from around 300-500khz.

1. Yes, this can be battery powered.
2. The MSP430F5528 does not have USB OTG support as far as I know. I do not know if any of the MSP430 family has that support. Please ask in the MSP430 E2E forum for the best information.
3. The MSP430 can be debugged over USB. Personally I have only used CCS for this. This could be another question to ask on the MSP430 E2E forum.

If you mean as far as incorporating the EVM in producing and selling your product please review the licensing terms for EVMs: www.ti.com/.../sszz027n.pdf

The schematic and layout of the EVM are available for reference, and I encourage you to copy as much of the schematic as works for your application since it is a known good system.

Regards,