request to help for ADE7753 design

[arup]

reading data from ade7753

Hello,

Below are few design issue i would like to discuss here ..

1) is there any issue with spi comm. of 7753 chip if CS pin is not controlled by MCU.Because same routine works with cs pin controlled by mcu .But gives corrupt data if CS pin is tied to GND(low).
a) My SPI routine is bit bang oriented.
b) There is only one device, ADE7753 on the bus.
c) MCU is pic
d) MCU freq : 20 Mz

attached is the wave form zip file with and without CS.This wave form depicts the read operation of IRQEN default value i.e 0x40.
Green : CS
Orange : Dout from 7753
Blue : Din to 7753
pink : Clk to 7753

Dout data i get from 7753 W/O CS is not same in every read operation.

2)Another issue is reading vrms . if read vrms register every 10ms (with timer interrupt) for a constant AC voltage i/p , i see avg result of such 10 samples is varying too much.What could be the problem?is it that i am not reaching to read vrms exactly at every 10ms??

3) i am trying to measure active energy . i am reading RAENERGY register at every 100 ms at unity p.f and accumulating value read in another register.But what i see that after 30 min i end up with reading error of -10%. This register is SIGNED type.But am not considered that. is it the issue? How to take into account of SIGNED type.

expecting some help.
Regards,

arp

 

[bobcat1]

ade7753 & waveform

Hi

I am also working on a project with the ADE7753 - we can share some info.

Your picture with the CS active (BMP) looks OK

In order to test your SPI comm. you need to write & read from a register that dose not change it context (like interrupt mask register)

You probebly have timing problem with the SPI bus - the edge of SPI CS signal cut the data stream

Some hints:

*Look that all the SPI signals lay under the CS signal all the time
*check the timing diagram against the DS to see whether you are not working to fast

Also the ADE7753 MUST NE CLIBRETED in order to read accurate result

All the best

Bobi

 

[arup]

aeenergy

hi,
it will be nice to share info regarding this project...

In order to test your SPI comm. you need to write & read from a register that dose not change it context (like interrupt mask register)

i did exactly that. waveform attached in my previous post was reading constant default data i.e 0x40 from IRQEN (interrupt enable register).

In my previous post W_O CS means CS pin is permanently gnded instead of micro controlling that pin. data sheet says that cs can be tied low permanently.
moreover for serial communication cs pin should be kept low.

timing is same for both the routine i.e with cs controlled by mcu and cs permanently gnded .And timing is more than the minimum time specified in datasheet
so i wonder why then data gets corrupted in 2nd case.

regds,
arup

 

[arup]

ade7753 project

Hello!,

Here is an observation i have taken.If CS is gnded and then chip is powered up then data reading get corrupted.Instead if i let the chip gets power up first and within a second if i gnded CS pin thru a resistance data reading is OK.

what could be it's solution now?

rgds,
arup

 

[bobcat1]

ade7753 & interrupt

Hi

It is Not a good idea to GND CS line specially in a noisy environment

Try to avoid this idea when ever you can unless you are short with IO line.

All the best

Bobi

 

[arup]

ade7753 how pf

Hi!,

Ok. Now i m controlling CS using I/O line and constant data read thru spi is ok.But the issue now is ...

if i read vrms register i get different value(which differ by 5000/10000 or even more ) for same i/p voltage ....i read it inside a timer interrupt which occurs at every 10ms.

What could be the problem here ??

For active energy reading is it sufficient to read unsigned value of AENERGY register and multiply it by wh/lsb constant and accumulate them in another user defined register....
I think we require signed value of AEENERGY register to detect reverse current or export energy....correct me if i m wrong...


regards,
ARP