My Neptune Apex web interface compatible DIY reef controller

[d0ughb0y]

btw, since I was updating the program on the controller yesterday, I figure it was a good time to add the third temp sensor. It runs and works fine. this is actually the first time I tried this many temp sensors. the main program loop execution time remains the same even with the added temp sensor. The initialization did take just a little longer. anyone tried more than 3 temp sensors? It's probably not practical to do more, but I think the circuit and the software can still handle it in case anyone needs more.

 

[Moomin1967]

The non EZO pH was 384ms, the data sheet is now called legacy..... its the top right hand side of the documentation section.

Looking forward to IVR cable arriving, and network uploading.

Ok to upload and I will post a screen shot.

M

 

[SorsCode]

d0ughb0y - Did you commit to github your Outlet changes? I just looked, and didn't see it.

For me, I updated mine for the LCD fix. I was not affected by the NTP as I use my own personal NTP server, and point to it. I'm currently waiting on my ENC28J60 Ethernet Shield so I can move to IPv6 native.

 

[d0ughb0y]

d0ughb0y - Did you commit to github your Outlet changes? I just looked, and didn't see it.

For me, I updated mine for the LCD fix. I was not affected by the NTP as I use my own personal NTP server, and point to it. I'm currently waiting on my ENC28J60 Ethernet Shield so I can move to IPv6 native.

no I have not committed it. I'll do it tonight.

the current code will use pool.ntp.org by default and will only use the ntp ip address if dns lookup fails. So it will not use your ntp by default now.
I suppose if the controller is not given external internet access, it will still use your ntp ip.

if you use enc ethernet shield, the bootloader will not work, as it uses direct w5100 register access so the bootloader is specific to w5100 ethernet shield. you can still use the bootloader, but can upload via usb only, as the bootloader detects if w5100 is present.

 

[d0ughb0y]

The non EZO pH was 384ms, the data sheet is now called legacy..... its the top right hand side of the documentation section.

Looking forward to IVR cable arriving, and network uploading.

Ok to upload and I will post a screen shot.

M

I see that now.

even the conductivity kit price went up. it was like $150 before now it is $198 with tax. there must be some hyper inflation going on, as normally these things go down in price on each iteration, even with more features.

I guess they can do this as there is virtually no competition.

 

[SorsCode]

Well that sucks about the Bootloader. Guess I'll have to play with that. I wish the W5100 supported IPv6 but it doesn't

Not to worried on the NTP piece, as I won't carry those changes over.

 

[Moomin1967]

I see that now.

I guess they can do this as there is virtually no competition.

I saw some probes offered from China on Ebay for liek $30 ish dollars, I'm thinking stamps from Atlas and try an import.

Atlas say their stamps work with other probes, so figure it is worth a go!

I'm thinking I may need a few other temp probes, here is the thinking.

Water - , Ambient - trending , Sump, ATO- is temp?, Main LEDs- fan and over temp dimming, Sump LEDs - fan and overtemp dimming.

Also thinking may run an experiment to recover heat from the Main Leds to contribute/heating the tank, during sub summer daylight periods.

This would save thousands of KWH's per year.

 

[rrbigdog]

I picked up a cheap ph probe for 8 dollars from china and seems to be working fine on my atlas stamp.

 

[Moomin1967]

ORP playing me up , and cond still will not init despite 1200ms.

have noticed init fails, but still outputs a couple pf times (leds flash)

 

[Moomin1967]

attached this time!

Non of this is in the tank at the moment, Temp probes are at ambient isn, pH is in 7 solution, Cond is a glass of fresh saline at ~35

 

[Moomin1967]

I picked up a cheap ph probe for 8 dollars from china and seems to be working fine on my atlas stamp.

Thats good to know, thanks

 

[Moomin1967]

With extra temp sensors, and seems all good!

M

 

[d0ughb0y]

ok, I'll check why the salinity is showing as 0.

the $8 ph probe works, but the cable is like 3 feet long. I had to get a bnc extension for another $7.5. I'm keeping it as backup for now. I ended up getting a pinpoint brand for around $30. The last probe I used (lab grade) lasted 5 years. You know when it needs replacing if you are getting random readings.

 

[Moomin1967]

Ok things are a little different to how I reported them earlier.

I have shortened the cond init string, as below. delay is 1200.

LED is now flashing red, err in the data! strange

/////////////////////////////////////////////////////////
// Atlas Sensors Functions
/////////////////////////////////////////////////////////
boolean initAtlas(AtlasSensorDef_t &data)
{
char responsechars[15];
responsechars[0]=0;
data.saddr.begin(38400);
for (int i=0;i<255,strlen(responsechars)==0;i++)
{
if (data.type==_cond)//set conductivity stamp to return SG only, K=1.0
data.saddr.print("response,0\rc,0\r");
else
{
#ifdef _PH_EZO
if (data.type==_ph) {
data.saddr.print("response,0\rc,0\rr\r");
}
else
#endif
data.saddr.print("e\rr\r");
}
delay(1200);
data.average=0;
getresponse(data, responsechars);
}

Also was the original plan to display salinity large and SG in the subscript, tink it would look pretty good!

 

[d0ughb0y]

you are missing at r\r at the end of your cond init. you need the r command in order for the stamp to return 1 reading.

I can't find any problem in index.htm nor in arduino code. once the stamp starts returning salinity values, it should all display.

 

[Moomin1967]

Changed the init to this same situation

/////////////////////////////////////////////////////////
// Atlas Sensors Functions
/////////////////////////////////////////////////////////
boolean initAtlas(AtlasSensorDef_t &data)
{
char responsechars[15];
responsechars[0]=0;
data.saddr.begin(38400);
for (int i=0;i<255,strlen(responsechars)==0;i++)
{
if (data.type==_cond)//set conductivity stamp to return SG only, K=1.0
data.saddr.print("c,0\ro,sg,0\ro,s,1\ro,tds,0\rr\r");
else
{
#ifdef _PH_EZO
if (data.type==_ph) {
data.saddr.print("response,0\rc,0\rr\r");
}
else
#endif
data.saddr.print("e\rr\r");
}
delay(1000);
data.average=0;
getresponse(data, responsechars);
}

Am curious how pH returns ok then goes on to test Cond, when if I am reading the init string corretly Cond is tested first, what am I missing?

M

 

[d0ughb0y]

cond is called last in initSensors function.

try using the atlas test sketch or modify it to send that same string.
double check the baud rate and connections. 38400 is the factory default if you did not change it.


I found this on ebay
http://www.ebay.com/itm/E201WM-Cond...244?pt=LH_DefaultDomain_0&hash=item1e8ebf207c

if you try to get non atlas probe, make sure it has BNC connector. Most conductivity probes do not use BNC.

 

[Moomin1967]

And one really odd thing.

As I plug and unplug the USB from computer or plug it in the Cond Leds Flash a cpl of times, alebit red

If I swap serial 2 and three around in the conf.h file would that definitivly prove dis prove any cross talk between serial ports 0 and 3?

 

[Moomin1967]

loaded Atlas test code, stamp output all good from power up.

 

[d0ughb0y]

if you are running the test code on the same mega circuit, then there is no reason why it will work differently using chauvet program.

I just checked an noticed now.

The old ph stamp, with LED off, only consumes 2.3ma active.
old orp with led off is 7ma.
The EZO ph and orp stamps with LED off consumes 13.8ma!!

The EZO cond with LED off consumes 17.5ma!
the legacy cond consumes only 4.2ma.

The new stamps are sure using up a lot of power.

I suggest you calculate the total current usage and see if the arduino can still power it or you may need to power the atlas stamps with a separate 5v supply.
The ethernet shield takes up quite a bit of power from the arduino board. The 2.3ma from old ph stamp is quite minimal draw from the arduino.