AMC 1210 and ADS 1203

Hi Maurizio Da Ros,

The processor on the AMC1210MB-EVM is not able to accomodate the Manchester Decoding Mode of the ADS1203.  The clocking scheme mentioned in the datasheet is not possible to implement with the Stellaris process found on the board.  I'll see if we can get the AMC1210MB-EVM Users Guide updated to reflect the fact that the Manchester mode is not supported.

Dear Tom,

Many thanks for your quick and clear answer but having to decide to use the combination ADS1203 + AMC1210 on a critical application we need to understand the real limits of the AMC1210 and the validity to use it.
In my opinion, all the work of decoding the Manchester encoded signal is assigned to AMC1210. The Stellaris is a simple bridge between AMC1210 -> USB -> PC. The confirmation of this behavior is the fact that in Modulator Mode = "00" the data and clock from ADS1203 are regularly processed in AMC1210 by the chain Control -> Filter-> Time Unit-> Data Register. In the configuration Sinc3 + SOSR = 127 Stellaris receives parallel datas with a speed of approximately 50 kBytes / s.
As stated on page 17 of SBAS372D, in Modulator Mode = "10" all the Manchester's operations of clock extraction, data recognition and calibration are performed by AMC1210. All the data transfers to Stellaris are performed in the same manner as in Modulator Mode = "00".
We try to use the Manchester encoding to isolate datas from ADS1203 to AMC1210 via a planar transformer with an available bandwith from 4 to 10 MHz avoiding, in a harsh EMC environment, the use of a complex Trigger - R-S Flip-Flop arrangement like the one used inside the AMC1203

We think to interface AMC1210 to our process with a C674x DSP.

Are we wasting our time?

Regards,

Maurizio

Hi Maurizio,

You are correct in that the AMC1210 can process the the Manchester stream and decode it's clock/data directly, but the AMC1210MB-EVM can't provide a 60MHz clock to the 'CLK' pin of the AMC1210.  The data stream from the ADS1203 is coming out of the part at 10MHz, so the AMC1210 needs a (min) 60MHz system clock to be able to perform the decoding.  Getting the proper system clock from your TMS320C674x device should not be an issue though. 

I'll look around through the lab to see if I still have one of the old style AMC1210EVM's which might be easier for you to work with.

Hi Tom

As stated on page 17 of SBAS372D is right that the ratio between AMC1210 and Manchester frequency must be more than 6 times. 0412.IQD-CFPS73.pdf

Do not worry about the old AMC1210 mother-board: we have in stock some SMD 64 MHz oscillators in sizes and characteristics equivalent to the 32 MHz installed on MB.

If there are no problems between AMC1210 at 64 MHz and Stellaris, we can easy replace the Y2 oscillator with the 64 MHz model.

Thanks in advance for your confirmation.
Regards

Maurizio

Hi Maurizio,

The 64MHz won't be an issue for the AMC1210 (it can run up to 90MHz) and the colck is not used by the processor at all, so changing it out is not a problem.

Hi Tom

Today I was able to change the oscillator mounting the 64 MHz model.

In this way the Manchester demodulator works correctly.

Thanks for your help.

Regards

Maurizio

Hi Tom-san,

You discussed about the solution with AMC1210 and ADS1203 for isolated Manchester encoding solution in this thread. But now our customer look for the isolated Manchester encoding solution with one device. They evaluate AMC1210 but it can be used for Manchester encoding. Do you have other devices to achieve it?

Your comment will be appreciated.

Best Regards,
Sonoki / Japan Disty

Hi Sonoki,
I'm not sure that I understand what you are asking. The ADS1203 has a Manchester encoding mode and the AMC1210 has the ability to decode the Manchester stream. Are you looking for an isolated modulator that has a Manchester encoded output mode?

Hi Tom-san,

Yes, I'm looking for an isolated modulator that has a Manchester encoded output mode. Do you have such solution?

Best Regards,
Sonoki

Yes, we do have something in the design pipline. I'll contact you off-line to provide some detail...