The good news: I've got reliable connections to the bus and am getting data. The MSO2204EA will collect 10 Mpts at 100,200,500 Ms/S & 1 Gs/S.
The bad news: I have no clue what I'm doing. I've never used the LA features before other than a brief attempt which made clear I needed more reliable connections to the bus. So I have several things to learn. The scope LA , the AWG and the SPI bus.. I know what the SPI bus is and its strengths and weaknesses, but I have no practical experience working with it. I've also never used a LA before.
FWIW I was reading the JDS thread today. The maker is very active in the forum helping owners and explaining the history of the very similar units. Apparently two engineers designed all the units. The JDS is the most recent version. I think a major issue for FeelTech is the lack of an engineer with English language skills. I urge everyone to ask for technical assistance from FeelTech. JDS is providing the bus command protocol. It's in Chinese so google translate is required for most, but they will help with problems.
Please!!!! Start a thread in the repair section of EEVblog so people can find things like bus pin assignments, etc. Trolling through this thread to find a snippet of information is a huge waste of time.
OK, the's take them in importance order last is bigger:
1) FeelTech and technical assistance: Only in China you can became an electronics engineer without proficient English, all over the world you're screwed if you try, I wish they would have exported just the textile industry there
.
Asking for technical assistance: I have asked FeelTech really politely for very minor things, I've got exactly ZERO answers, if nobody knows English there it's understandable, if somebody ask for me of something in Swahily I won't be able to answer them, or just confirm that I've received an inquiry, that make sense. Ah, who am I kidding, they are the typical Asian company, eager to copy and profit from everything and considering a weakness and stupidity to share anything back, at least with non-Asians. I would so much LOVE to be proven wrong, but in 15 years or so while dealing with them I never had a cool story, and I haven't heard too many as well. So JDS has a full specification of the control protocol from the FP to the Signal Board, all the registers and settings being explained, and compatible with FY6600 huh ? Willing to share it, but in Chinese ?
I would say bring it on, I've never seen a real unicorn yet
, and being a Mandarin Unicorn will not make it less beautiful
.
Also a link to the JDS thread will help.
2) About actually doing stuff, I don't know if your scope has an individual LA or just has a protocol decoder using the analogue scope inputs.
<Beginner explain mode>
The commands are send via SPI protocol, the protocol uses 3 to 4 wires and it's a synchronous protocol, that means the data it's send and received being aligned with a common clock shared both by the sender and the receiver.
So you have to grab that clock before everything else, in our situation it's on
pin 6 of the connector, the SPI protocol it's also a raycist
protocol, where there is a master and a number of slaves, they are selected with an individual selection pin, in our case there seem to be two slaves, but for basic stuff we can ignore it for the moment. Only master talks and generates the clock, and talks to the slaves using the MOSI signal (Master Out, Slave In),
pin 8 and listen to what are they saying using the MISO pin (Master In, Slave Out),
pin 7.
There is also a Slave Select or Chip Select pin, that goes to 0 usually, to wake up the slaves and tell them that the master is gonna talk and they better be listening and answering or else... (very raycist
This is our
pin 5, usually if you have a LA you use this pin on falling edge to trigger acquisition, if you have a two channel scope with separate sync input you put it there and use it as trigger, if you have a 4CH scope you can still put it on the sync input but also on one of the channels, and if you have just a 2 channel scope without separate sync, you ignore it and use the clock signal to trigger, but it's a bit on the edge.
So, with a two inputs scope you can see one direction only M->S or S->M, with a 4 inputs scope you can see and decode fully the SPI interface, same with a proper LA.
Now the clock and transfer size varies, in our case the clock it's 18MHz and the transfer size or word size it's 16 bits, an LA should be used in sampling mode, with a sampling frequency of at least 100MHz to get reliable results. Scope ca. 50ns/div, depends on what they recommend for SPI protocol decoding in regard with the clock frequency.
On the Signal Board side, in the FPGA, there are a number of individual addressable entities called
registers, a register is written or read using 3 x 16bit transfers, one for address and two for the data to be written or read, you can see a bit behind some of those registers read and explained, my LA has reached it's limits, so I'm waiting to get a proper one.
If your protocol decoder works, you should be able to get the registers values while doing different settings (compare first with the ones already discovered to make sure that everything it's set OK ).
In the evening I will make a new post in the Repairs forum, summarizing the details spread around here and I can assist anyone trying to get the information or trying to translate this mystical JDS document
.
I have to run for work now, cheers.
</Beginner explain mode>