20+ hours banging head off wall

Thread Starter

bowlingo

Joined Jun 29, 2011
162
Thanks...

The ph and conductivity transmitter is a dual output unit each of the 2 wires on the outputs goes to an individual sms controller connected between AGD and 1 (analogue inputs)

Only when I try to connect both sms controllers up at once do I get the problem..on there own they are working perfect

When I connect a multimeter across the PH output and the CF output (using 2 multimeters) the readings are perfect but when I put them into the sms controllers (remember there is 2 inside the system I have built) the readings go all over the place which must be to do with the 2 sms controllers sharing a common point..possibly negatives although ive tried 2 x power supplies but to really be sure I need to strip the entire system down of which could take me many hours as there could be a relay or something sharing a negative.

Last night I got the CF probe to work as it should with the ph wired up also by changing the base and scale settings within the sms controllers software..although a bit rough it worked well. Then I tried to calibrate the PH side by putting the probe into PH 7 buffer solution (it was in PH 4 buffer solution when I was calibrating the Cf side) what happened when I put the PH probe into PH 7 is the CF went from 0 (free air as it should be) to 2.43! This means dependant on what one probe is doing effects the other of which must be to do with the shared point between the 2 sms controllers...

I have just found these..

http://uk.rs-online.com/web/p/produ...552267573743D343636323231342677633D4E4F4E4526

http://docs-europe.electrocomponents.com/webdocs/03b8/0900766b803b8ed4.pdf

These will isolate the outputs from the transmitter so they will be floating as I see it.

Do you think this will work?..Although the sms controllers are still joined in some way..I cant picture in my head or by drawing if a circuit is being made up?

Due to the cost of these isolators I am hoping I can get away with using one then send the second one back as a return (RS are good for returns)

Thanks
Alex
 

praondevou

Joined Jul 9, 2011
2,942
I abused a bit eblc1388's schematic.

Question: Did you connect the SMS controllers in a way like in the attached diagram?
2 separate power supplies for the sms controllers and one for the CDTX, none of the GND signals of each unit connected to the others?

Btw, the isolators u found will probably work, but if you build a circuit yourself it will get cheaper I guess. I used the HCNR201 for voltage sensing and they work quite well...
 

Attachments

Last edited:

Thread Starter

bowlingo

Joined Jun 29, 2011
162
Hello..

I dont get where the second connection between the transmitter and controller should go or is now connected as of your diagram?...theres 2 wires coming from each output on the transmitter of which one output is linked internally between the power supplies negative and the 0's
 

eblc1388

Joined Nov 28, 2008
1,542
Hi, bowlingo. This is the current mirror I mentioned earlier.

If you are good with a soldering iron or can find someone who does, then the following circuit will provide a cheap and cheerful solution to your existing problem. All together it uses only 9 commonly available components(U1/U2 inside one IC), costs a few dollars and don't even need a separate power supply.

You can build a couple and have spare for future too.

What it does is to sense the input current and produce a current of similar value at its output, thus its name current mirror. Should you later wants to place a meter in series either remotely or locally for checking or calibration purposes, you are free to do so as accuracy won't be affected and there is no adjustment involved. A truly fit and forgot interface.

 

Attachments

praondevou

Joined Jul 9, 2011
2,942
Hello..

I dont get where the second connection between the transmitter and controller should go or is now connected as of your diagram?...theres 2 wires coming from each output on the transmitter of which one output is linked internally between the power supplies negative and the 0's
According to your very first diagram (first post) the Ampmeters are connected between + and the inputs S1 and S2. So the current to be measured is actually flowing from + through the ampmeter and then into the inputs S1 or S2. That means that the component that controls the output current is between "0" and "S1/S2).

If you have a closer look at my last diagram, I just replaced the Ampmeters with SMS controllers. So now the currents have to flow from + through the SMS controller and then into the inputs S1 / S2. Should actually work, unless the first posts diagrams are not correct. Of course the three devices need separate power supplies.

I don't know why they indicate "S1" together with "0" as an output and then their diagram says something else.... So, where do you connect your current measuring multimeter to, from "0" to S1/S2 or from "+" to S1/S2 ??
 

Attachments

Last edited:

Thread Starter

bowlingo

Joined Jun 29, 2011
162
thanks again for your help...

I am connecting the ampmeters between 0 and S1/S2 individualy so one meter I put between the 0 and S1 of the PH output and the other meter I put across 0 and S2 of the Cf output. This stays as it should when I use amp meter only but when I take the amp meters away and connect them to the individual sms controllers analogue input AGD (im asuming this is common) and AGD0 (there are AGD 0, 1, 2 and 3 on the sms controllers) the readings change and also go all over the place dependant on the output current of the other probe i.e PH or CF...as of earlier posts if the 2 x sms controllers are using a common point i.e a relay that is interconnected betweent he sms controllers this could be the problem.
 

Thread Starter

bowlingo

Joined Jun 29, 2011
162
Hi, bowlingo. This is the current mirror I mentioned earlier.

If you are good with a soldering iron or can find someone who does, then the following circuit will provide a cheap and cheerful solution to your existing problem. All together it uses only 9 commonly available components(U1/U2 inside one IC), costs a few dollars and don't even need a separate power supply.

You can build a couple and have spare for future too.

What it does is to sense the input current and produce a current of similar value at its output, thus its name current mirror. Should you later wants to place a meter in series either remotely or locally for checking or calibration purposes, you are free to do so as accuracy won't be affected and there is no adjustment involved. A truly fit and forgot interface.


Thanks for that :) My head is a bit fried right now so ill have to have a look at it tomorrow as my eyes arnt burning
 

Thread Starter

bowlingo

Joined Jun 29, 2011
162
According to your very first diagram (first post) the Ampmeters are connected between + and the inputs S1 and S2. So the current to be measured is actually flowing from + through the ampmeter and then into the inputs S1 or S2. That means that the component that controls the output current is between "0" and "S1/S2).

If you have a closer look at my last diagram, I just replaced the Ampmeters with SMS controllers. So now the currents have to flow from + through the SMS controller and then into the inputs S1 / S2. Should actually work, unless the first posts diagrams are not correct. Of course the three devices need separate power supplies.

I don't know why they indicate "S1" together with "0" as an output and then their diagram says something else.... So, where do you connect your current measuring multimeter to, from "0" to S1/S2 or from "+" to S1/S2 ??
looking at your new diagram with the current flow lines drawn on the circuit itself now makes sense to me..

This is exactly as its currently wired although the sms controllers are sharing a power supply. Over the weekend I tried the 2 sms controllers with there own power supplies and the transmitter has its own power supply this didnt work which is probably due to a common point that the sms controllers are sharing within the internal wiring of the system of which ive already built...due to the transmitter taking weeks to arrive from the US because they are probably built on order due to the rarety of the particular units I foolishly built the system on the basis it would work without problems. The shared point could be some of the many wires or one of the many wires sharing a negative somewhere. Last night I disconected all the negatives relating to the things its controlling but still had no luck. To do this properly I will have to strip the system down, draw it all out and start with 2 x sms controllers with there own power supply with nothing connected to there inputs and outputs then build it up from there.

I have ordered the 4-20mA loop isolators of which im hoping one on its own will work so it wont be so expensive (my problem is trying to understand if the loop isolators will work even if the controllers are sharing something) if not then ill use 2 as I am desperate to get the system working as its driven me nuts building it due to my knowledge in electronics being exceeded (although ive learnt a lot) if I cant resolve this problem then 4 months work and around £3000 will be down the pan as the CF and PH side is critical.

thanks
Alex
 

praondevou

Joined Jul 9, 2011
2,942
I am connecting the ampmeters between 0 and S1/S2 individualy so one meter I put between the 0 and S1 of the PH output and the other meter I put across 0 and S2 of the Cf output.
What you are saying does not correspond to what is in your first post diagram. There the ampmeter is between + and S1/S2.... :confused:

This is exactly as its currently wired although the sms controllers are sharing a power supply. Over the weekend I tried the 2 sms controllers with there own power supplies and the transmitter has its own power supply this didnt work which is probably due to a common point that the sms controllers are sharing within the internal wiring of the system of which ive already built
Yep, the two SMS controllers must not share AGD or GNDs because the AGD outputs go to separate outputs of the CDTX-300.
Are the AGD and GNDs of the SMS controller connected?

Don't worry, isolating the signals will do it.
 

Thread Starter

bowlingo

Joined Jun 29, 2011
162
What you are saying does not correspond to what is in your first post diagram. There the ampmeter is between + and S1/S2.... :confused:



Yep, the two SMS controllers must not share AGD or GNDs because the AGD outputs go to separate outputs of the CDTX-300.
Are the AGD and GNDs of the SMS controller connected?

Don't worry, isolating the signals will do it.

Sorry...they are connected between + and S1/S2 :rolleyes:

The sms controllers are wired in such a way that each units 8 inputs and 8 outputs are going into the system and are probably sharing the negatives or similar. I have it configered so that all the internal relays (controlled by both sms controllers) are connected to the power supplies negative point. The other day I tried disconnecting all these negatives and seperating them but still no luck. Its a question of stripping the system down and building it back up again I reckon as I kind of built it as I went rather than using planned circuit diagrams...it all started off as a simple concept but I got carried away :D
 

Thread Starter

bowlingo

Joined Jun 29, 2011
162
both types clip onto the standard 35.5 mm din rail.

the weimuller is thinner...and looks more like a weidmuller wire terminal block, whereas the RS looks like a module.
Terminal Blocks | www.weidmuller.com

I have just noticed the weimuller has an input range of max 15 volts..I am using a 24V power supply to power the transmitter.

please can you look at the 2 and tell me if theres going to be an issue with a 24V supply...

http://docs-europe.electrocomponents.com/webdocs/03b8/0900766b803b8ed4.pdf (£112.00)

http://docs-europe.electrocomponents.com/webdocs/0d57/0900766b80d575b0.pdf (£77.00)

the weidmuller is the MCZCCC

Thanks
 

praondevou

Joined Jul 9, 2011
2,942
For the MIN310 input it doesn't give the max voltage.

For the Weidmüller it says max input voltage/current is 15V/50mA. Does that mean there will be a voltage drop of 15V at 50mA? I don't know.

It also says at 20mA there will be a voltage drop of up to 3V. So there shouldn't be a problem.

More interesting: Does the CDTX output really sink 20mA if there is a load in series (the current isolator)?
Two options:
1. The CDTX output current is measured by the CDTX and regulated to the current it's supposed to give out, then everything will be fine.
2. The CDTX output current has a fixed/or adjustable ratio with regard to the input signal, then your output will be less than what you expect and the error will change according to the absolute output value because the voltage drop over the isolator will change.
 
Top