Samba High Speed Counter or Desecrate Input

[Dumbfitter]

Hello all, I have a question about the high speed counter on a Samba. I have a SM70-J-RA22. I have a project that I would like to use that or a Vision on. The Project has a Gas flow meter that has a Prox switch on it that is rated for 8VDC. Per the instructions they recommend running the prox switch through an amp , the amp they recommend has relay outputs. At full flow we should be about 2 pulses a sec.  So just to check to see how it would work I set a 1 sec pulse to output 0 and wired output 0 to HSC 0. I can probably just use a desecrate input and do a count and so forth but wanted to try this and see if it would work.What I noticed is the count is all over the place . Sometimes it adds 5 sometimes 10 and so on to the MI. If I look at the frequency its all over the place also.  So what I tried is just wiring a switch to HSC0 and pressed the switch by hand to see what it was doing and noticed when I let go of the switch the same thing would happen but sometimes worse. I'm assuming its not a clean break when I let go but would there be that much bounce on a relay? Since its not a real fast count I'm thinking I can use just the Input and do a count but would like to know why the counter would not be working.

Thanks for your help

Remember I'm just a Dumbfitter so please keep it simple.

     

[Ausman]

10 hours ago, Dumbfitter said:

At full flow we should be about 2 pulses a sec

For 2 Hz max you don't need to use the High speed, normal will be fine.  And if the "correct" amp is using relays this is very poor practice, they'll wear out in no time at all.   At full flow over 24 hours that's 172, 800 operations!  Should be solid state. 

cheers, Aus

[Flex727]

Even if only 2 pulses per second, what is the length of each pulse? Couldn't it still be missed unless a HSC is used?

[kratmel]

Can you provide more information about gas flow meter and amplifier.
User manual, or a link?

8VDC Prox switch is very strange ? !

[Dumbfitter]

If It works some info on the meter is attached. The WE77/EX-1 is a relay output. I was thinking of using a phoenix contact amp. The one I found is solid state. And before any one asks I would tie into the E200 Prox do to the flow reading would be to slow on the E1 reed switch. 

With all that said does anyone know why I had all the different readings on the High Speed counter?

Thanks Again

Ray  

EMS-DS4650-5.pdf

[Dumbfitter]

And if you read German to English translation. Every thing I found is showing 8v to the pulse generator and they list it as a prox switch.

Ray

 

100521_Elster_-_Quantometer_QA__QAe_-_Operating_instructions_-_BA_QA_QAe_DE_UK_FR_NL_IT_SP-1.pdf

[Dumbfitter]

Just to give background on the project.

We have 4 boilers that we are replacing the burners on. I'm planing on using the Vision to control the steam pressure of the boilers. 1 Vision per boiler . So just a basic modulating control. I'm also looking to program a warm up mode so we don't bend the boiler by running it in full fire when it's cold, we need to bring them up kind of slow.  I'm also looking to tie into the combustion controller and monitor status and alarms . Also looking go through a gateway so building automation system can monitor the state of the boilers also. The site had gas meters on each boiler to begin with and are being replaced with the QA's . As of right now the customer is happy with the mechanical display on them but would like add real time flow readings on the BAS. They are also looking at adding steam flow meters on each boiler, so they can see how much steam they are producing vs how much gas they are using. Sooo I'm trying to do all this with a Vision 700.  I just need to figure out how to get the gas flow reading as the rest I think I can work through. And no the meter does not have M-bus from what I saw . I will double check the next time I'm there. If it does I might just run that through a gateway, I just don't like having to many gateways . Already will need to go from Modbus to Bacnet.

Ray

 

[Ausman]

13 hours ago, Flex727 said:

Even if only 2 pulses per second, what is the length of each pulse? Couldn't it still be missed unless a HSC is used?

Yes, true enough might be possibly missed.....I was working on a same delay b/n on/off.  By the look of the inductive's specs it may indeed be a short ON time.

As for the different HSC readings, and assuming that all your hardware settings and wiring are correct, perhaps you are getting very fast bounces and as the scan read only relates to differences b/n each scan this may account for it.  For interests sake, if you do an immediate display lke you have already, alongside a cumulative one this might show this up.  A hard one to pin without being onsite.

And that inductive needing 8v is indeed a little weird given that it is meant to enable easy interfacing!  I noticed during my research that it looks like Honeywell have badge engineered the flow meters , and perhaps they may have a ready made supply module/interface that might be more user friendly and even better than a Pheonix.  Who knows?!!

PS...in my read of the literature they do have Modbus as an option.

cheers,

Aus

[Dumbfitter]

Ausman, where did you find Modbus as an option? Everything I saw was M-bus , which ,to my understanding is a different protocol. I can go with a gateway and go from Modbus to M-bus if needed.

Ray  

[Ausman]

I saw it somewhere in my document skimming...it may have been in the Honeywell stuff.  I'll have another squiz when I can and let you know.  Not at the moment.

cheers,

Aus

[sgull]

Hi

 

In your original post you said you tried to use the digital inputs and found that sometimes the count jumped by 5 or 10.

This will happen if you dont use a positive transition contact to increment the pulse counter.

if you use a nornal contact the counter will increment with each scan as for the duration of the pulse.

[Ausman]

OK, I've had a look at all I can find, and I think I must have misread M-Bus as Modbus in skimming things.  Apologies.

However, it also shows M-Bus as only fitted to the QAe versions, so my read is that you already have just the QA version so you might have an issue.  If you have the e types, you could use a gateway like the ones from Anybus (or similar) and this takes care of getting multiple reads from all your QAs via M-Bus, and then giving that info to the plc via Modbus.

However, further thinking (see how my brain has been ticking  over on this in the background without me knowing? .....thanks a lot!!) has me wondering how you are going to get accurate readings for your eventual linking into the BAS.  A variation up to 2 Hz is not going to give you much inherent accuracy if you are reading fairly often.  If you use the modbus info from the M-Bus gateway, I think it will be simply giving you a number it is currently up to at the time of interrogation.  Even 1 minute b/n calcs might be an issue, especially when going fairly slowly, and might lead to wildly differing quantity displays on each read which will likely not be the desired result! In some ways I think that you are still going to be best off doing an actual read of the count using the plc.  Have a look at this post where I showed how to get fairly good accuracy on something fairly low Hz....very similar to your situation but would need some variation on the concept.  Combine the plc read and this concept and it might be the best method.

cheers, Aus

 

 

[Joe Tauser]

Time to weigh in.....    I've been busy with a startup.

8 VDC is a common output level for a NAMUR (explosion-proof / instrinsically safe) sensor.

Relay contacts and other mechanical switches bounce a lot when they are closed, which is what you are seeing on your HSC input connection.  The bounces settle out within a millisecond or two of the relay closing.  I have seen this with my scope.

A regular digital input  has a 10 ms response time, so it never sees the bounces.  The response spec on an HSC input is 30 kHz, so it is capable of seeing 0.00003 second pulses.  Your previous post confirms that the HSC is confirming the bounce.

If you're only getting a couple of pulses a second I would stick with running them into a digital input with a solid state optocoupler terminal block.

I would not go the communications route unless necessary.  It really over-complicates things.  For your BACnet requirement, Unitronics does offer a gateway for this.

What was the part number of the Phoenix block you're looking at?

Joe T.

PS - "discrete"

[Ausman]

1 hour ago, Joe Tauser said:

8 VDC is a common output level for a NAMUR (explosion-proof / instrinsically safe) sensor.

Thanks Joe, having never worked with such devices I didn't know this.  I stay away from things that go bang.

cheers,

Aus

[Dumbfitter]

The one I bought is MACX MCR-EX-SL-2NAM-RO just to do a test as  the transistor version I would have to wait for. I did look at the meter and it is not the E version so no M-Bus. 

Joe as far as discrete goes look at my name. Lol 

Sorry I'm a real bad speller as everyone can see. Sometimes ( a Lot ) spell check has no idea what I am trying to say.

Thanks All for all your help.

 

Ray

[Ausman]

Just another comment on this.  I work with gas powered air furnaces a lot, but have never read anything from a flow meter as such, hence me not knowing the 8v stuff.  But all the furnaces I work with have modulating valves to vary the outputs.

It is a fairly easy thing to get reasonably accurate gas consumption by doing maths on where the modulation is set at.  If the pressures are correct and also more importantly to some degree, stable, (and they should be!) then this method is fairly accurate.  Not likely as accurate as the meter, but close enough for "pretty visuals" if that is what is wanted.  I use a combination of both the input and output the actuator is doing, which tends to allow for differences as the gear trains take up the slack.  (If the output pot is direct mounted to the shaft (rarely) then just use the output as it is the true position.)  To set the rate you run just that device and observe the mains meter over different settings.  Due to valve mechanical structures it is rarely linear, I use a series of steps picked up by compares on the modulation level.  Associated with this sort of situation is that I often work linearisations from other linearisations....it is a great way to get a sort of automatic compensation, adjustment, or fuzzy logic.

The only difference might be that edit...just reread your info that the Visions will control each boiler.... in your case the modulation is being done by the boilers themselves which might mean a bit of fiddling to get the modulation mA or more generally 0-10V amounts, whereas in mine the plc does all the level control which makes it easy.  If it is too hard to get the input, most systems I've seen don't actually use the output from the actuator, so just work with that.

cheers,

Aus