emonTxV3 not counting pulses

Hardware emonTx
,
1

Just bought a emonTxV3 from the store and i set up a pi2 with reciever and emonSD
My house is connected to a IT powersystem (Isolated Terra) and i want to log pulses from my powermeter and use 3 Ct clamps

The ct clamps are not connected yet but i hooked up a pulse counter from the store to the emonTxV3
the pulse counter blinks when the meter blinks and the PI seams to recieve data.
but the counter in my feed in not increasing.
when looking at the emonhub log it seams to recieve the input from the emontx but the pulse counter in not increasing

2017-10-26 07:26:07,037 DEBUG RFM2Pi Discarding RX frame ‘unreliable content’? 7 118 161 132 69 127 33 254 254 178 185 215 69 84 246 7 56 19 219 5 68 (-86)
2017-10-26 07:26:07,366 DEBUG RFM2Pi Discarding RX frame ‘unreliable content’? 0 229 69 187 83 31 134 43 225 214 66 254 255 224 43 46 107 214 133 143 1 (-86)
2017-10-26 07:26:13,297 DEBUG RFM2Pi 53 NEW FRAME : OK 8 0 0 0 0 0 0 0 0 63 100 184 11 184 11 184 11 184 11 184 11 184 11 2 0 0 0 (-28)
2017-10-26 07:26:13,302 DEBUG RFM2Pi 53 Timestamp : 1509002773.3
2017-10-26 07:26:13,303 DEBUG RFM2Pi 53 From Node : 8
2017-10-26 07:26:13,305 DEBUG RFM2Pi 53 Values : [0, 0, 0, 0, 256.63, 300, 300, 300, 300, 300, 300, 2]
2017-10-26 07:26:13,306 DEBUG RFM2Pi 53 RSSI : -28
2017-10-26 07:26:13,308 DEBUG RFM2Pi 53 Sent to channel(start)’ : ToEmonCMS
2017-10-26 07:26:13,309 DEBUG RFM2Pi 53 Sent to channel(end)’ : ToEmonCMS
2017-10-26 07:26:13,590 DEBUG MQTT Publishing: emon/emontx3/power1 0
2017-10-26 07:26:13,593 DEBUG MQTT Publishing: emon/emontx3/power2 0
2017-10-26 07:26:13,595 DEBUG MQTT Publishing: emon/emontx3/power3 0
2017-10-26 07:26:13,598 DEBUG MQTT Publishing: emon/emontx3/power4 0
2017-10-26 07:26:13,600 DEBUG MQTT Publishing: emon/emontx3/vrms 256.63
2017-10-26 07:26:13,602 DEBUG MQTT Publishing: emon/emontx3/temp1 300
2017-10-26 07:26:13,604 DEBUG MQTT Publishing: emon/emontx3/temp2 300
2017-10-26 07:26:13,606 DEBUG MQTT Publishing: emon/emontx3/temp3 300
2017-10-26 07:26:13,608 DEBUG MQTT Publishing: emon/emontx3/temp4 300
2017-10-26 07:26:13,611 DEBUG MQTT Publishing: emon/emontx3/temp5 300
2017-10-26 07:26:13,613 DEBUG MQTT Publishing: emon/emontx3/temp6 300
2017-10-26 07:26:13,616 DEBUG MQTT Publishing: emon/emontx3/pulse 2
2017-10-26 07:26:13,618 INFO MQTT Publishing: emon/emontx3/rssi -28
2017-10-26 07:26:13,620 INFO MQTT Publishing: emonhub/rx/8/values 0,0,0,0,256.63,300,300,300,300,300,300,2
2017-10-26 07:26:13,622 INFO MQTT Publishing: emonhub/rx/8/rssi -28

What am I doing wrong ?

2

You have included there only one message from the emonTx. (And yes, it is receiving data - ‘unreliable content’ could be other apparatus some distance away using the same radio frequency.) Are you seeing a similar series of logged events every 10s, and is the display in emonCMS refreshing?

Do you have a bad connection either in the RJ45 plug or the socket? Does it look correct? Is there any damage visible?

If my memory is correct, we have seen a case where the pulse sensor flashes, but pulses were not being seen by the emonTx. I cannot remember now what happened in that case.

3

when I tail the log, it seems to sucessfully recieve data every 10 seconds. the unreliable content is probably our wireless smokealarms. I can see the vrms number change on every message and the emoncms is refreshing. I will take a look at the socket and plug when i get home from work.

4

I hooked up my scope to the IRQ pin on the rj45 socket, but saw nothing so I used my multimeter to test the rj45 plug and wire on the sensor but that also looked ok.
one thing that i noticed is that R7 is missing on the sensor. Is it supposed to be that way ?

5

On the optical pulse sensor, or on the emonTx? On the emonTx circuit diagram, two burden resistors are shown. One is SMT and is fitted, the other is to place holes for a replacement wire-ended resistor if a different burden is needed.

Having just received an optical pulse sensor, I have to say that the description is misleading. It does not output a pulse at TTL level - it actually gives a voltage that is proportional to the incident light, 0.7 V in darkness and 3.28 V in bright light. The LED on the back of the sensor only lights within a few mV of 3.28 V at the output. Therefore the output “pulse” is formed by the flash of the LED on your meter, not by any processing in the optical sensor.

The pulse input reacts at a much lower voltage than the sensor’s LED, so the emonTx will detect pulses that do NOT flash the sensor’s LED. Which raises the question - is it going dark enough to NOT register an input?

I suggest you take the sensor off the meter and measure the voltages in darkness and in light. If you don’t see approximately the same as I do, I think you have a faulty sensor. If you do, then there’s something wrong with your emonTx - does it count pulses if you short-circuit the interrupt input (pin 4 on the screw terminals) to 3.3 V (pin 2)?

6

@Robert.Wall - when I read this I did go looking for (but failed to find) the post I wrote many moons ago about the fact I found depending on careful positioning I could persuade the optical pulse sensor to either register a pulse without flashing the led OR flash the led without registering a pulse to order by varying the position.

I don’t doubt your findings above at all but I felt I should mention my findings again as a flash but no pulse might not specifically mean a faulty unit if not positioned correctly or if it is in high ambient light to “hide” the led flashes being monitored. build tolerances or changes in production might have changed this position, but my findings were 100% at the time, hence the post back then as it seemed odd, I even asked at the time if there were 2 sensors (one of the pulse and one for the led) because as unlikely as that sounded, that was a logical explanation for my experience at the time.

Is the sensor you have not a sealed unit? All the ones I have seen have been sealed.

7

Mine is sealed, and R7 is the burden for CT1 - so the request is for clarification as it didn’t seem to be relevant. But if the SMT R7 is missing, it indicates a manufacturing fault and Glyn needs to know.

From what I measured, I think you’re describing the situation where the output voltage rises above the threshold for detecting a logic high at the interrupt input, but not high enough to turn on the LED on the sensor, and alternatively when the output voltage fails to fall below the threshold for detecting a logic low at the interrupt input, so the pulse is never “undetected” so it cannot detect a subsequent rise in voltage that eventually again lights the LED on the sensor.

I think we’re agreeing that we’ve seen the same thing, and I think what I saw gives a plausible explanation for what you saw earlier.

8

Totally!

I think that same logic explains well what I was trying to say about ambient light too. That the ambient light could prevent the voltage dropping enough to be able to register the next pulse.

9

I have not been able to look more at the problem yet ( I have toddlers:))
but to clearify my previous post: I was not able to detect any variance in voltage from the optical sensor when shining a bright light into it in a semi dark room. with a led flashlight (measured from the screw terminals on the emontx with a dso on 200mv trigger). (measured from the screw terminals on the emontx)

but I will try to hook up the sensor to a psu and measure directly from the rj45 with scope and Multimeter to be sure

The r7 im talking about is inside the optical sensor. I had to open it to verify that the wires was wired correctly
I think we have different definision of a “sealed unit”
its just a plastic enclosure. semi easily opened

the pulsecounter did increase when I just touch the input with voltage from the 3v pin on the screw terminals

10

That I think excludes the emonTx.

I plugged the optical pulse sensor into an emonTx using the RJ45 connector, and measured the voltage at the screw terminals - pins 3 & 4. “Dark” was a finger over the sensor’s lens.

11

I did some measurements today and have to correct my previous statment. the emontx did increase the pulsecounter when manualy pulsing the irq pin 4 with gnd not 3v
the irq pin 4 on the screw block is “high” with the sensor connected or not
there is no change in voltage on the pulse line from the sensor when dark or when i shine a light on it
measured on the sensor pcb and on the emontx

12

That’s as it should be. There is an “input pullup” in the software that pulls the input high. It is necessary because without a sensor connected, the ‘floating’ input would count spurious pulses.

I think then that you have a faulty sensor. Email the shop, and mention this thread.