Just finished reading through the AM25T manual....
I will be connecting different sensor types and will write my own loop to scan the channels. It doesn't seem hard, although a CRBASIC example would be a good addition to the manual.
The minimum pulse for switching channels is 50us ON + 60us OFF. So I should be ok with:
PulsePort (2, 100)
PulsePort (2, 100)
Which would give me a 400uS delay between measurements as well as a reasonable safety margin. Why would the CR10 examples use 10 time as long a pulse duration?
With the AM25T located 25m from the CR1000, do I need to extend the pulse duration to avoid glitches? How much?
What does the AM25T instruction use? What scan rate does it achieve? (I know it depends on the chosen integration time...)
Am I right in thinking that the first pulse disconnects the current channel and the second pulse connects the next channel?
Anton,
>> although a CRBASIC example would be a good addition to the manual.
I'll agree that the manual needs a bit of revising.
>> Why would the CR10 examples use 10 time as long a pulse duration?
Using EdLog and the P20 command, option 3 provided the smallest pulse duration, i.e. 1ms duration. It was the smallest pulse duration available.
>> do I need to extend the pulse duration
As the cable length increase, the impedance of the cable increases. Therefore, I would think your pulse duration would need to increase also.
The relays are advanced on the falling edge of the clock pulse (transition from >3.5 V to <1.5 V; 7 V max.) So you need to provide enough time to allow the CLK line to rise above 3.5V and fall below 1.5V.
>> Am I right in thinking that the first pulse disconnects the current channel
From page 4 of the manual:
The first CLK pulse advances the relays to the reference temperature excitation channel. The second CLK pulse advances the relays and connects HI and LO to 1H and 1L on the multiplexer.
Two clock pulses are required to advance to the next adjacent
sensor input channel on the AM25T.
The fourth CLK pulse advances the relays and connects HI and LO to 2H and 2L. The sixth CLK pulse advance the relays and connects HI and LO to 3H and 3L. This sequence is continued for the remaining input channels.
Hi Sam,
Thanks for your responses. Your details on the pulse voltage levels will let me check for trouble before it occurs.
With my question about the first and second pulse I was actually thinking of the first and second pulse in the pair of pulses that are required to switch from one channel to the next adjacent channel. I was trying to think of a logical reason for requiring two pulses rather than one.
I am still curious to know how long the pulses are from the AM25T instruction...
Anton
>>> I was trying to think of a logical reason for requiring two pulses rather than one.
Each pulse has a purpose:
The first pulse, of the pair, advances the relays to the reference temperature excitation channel. The second pulse advances the relays and connects HI and LO to 1H and 1L on the multiplexer. Third - excitation, fourth - 2H/2L, fifth - excitation, sixth - 3H/3L.
Why it is this specific way? I don't know. Generally, you only read the excitation once per scan. However by reading the RTD excitation, we can compensate measurements for long lead length effects. And the design of the multiplexer requires two pulses to advance 1H/1L, 2H/2L ... 25H/25L.
I have a query into engineering regarding the AM25T instruction pulse details.
Anton,
From one of our operating system engineers:
"The AM25T [instruction] pulse width is ~35uS, this gives 70uS per channel advanced (35 low and 35 high). There is some additional overhead for the reset line at the beginning and the end. This is about 100uS."
Anton,
Can you explain in detail what you are trying to do? Another engineer monitoring this thread stopped by my desk and cautioned using the AM25T and without using the AM25T instruction. There can be some snags related to measurement and processing timing when running Pipeline mode. I encourage you to detail your application here or to call in (435-753-2342) and ask for someone in the Industrial group. They have the most experience using the AM25T and will be very familiar with the do's and don'ts.
Sam
Hi Sam,
I finally have the equipment to work with.
I was initially planning to pause at each channel and take several readings in burst mode before moving on to the next channel. However according to the CRBasic compiler, using the minimum settling and integration time the AM25T instruction can read all 25 channels in just 16.1 msec. I won't push this to the limit, but it means I can easily put the AM25T instruction in a subscan to do a parallel burst on a few channels. Then after the subscan I can use the AM25T instruction again to pick up the channels where a single reading suffices.
I'll give you a call if I run into any more snags.
Anton