Why 4-20 ma is used in Analog Signals?

S

Thread Starter

sreeraj

Why 4-20 ma is used in Analog Signals? How come ISA selected this. Why not 5-21ma?
 
The early transistors that Honeywell's engineers used for the first transistorized circuits would not turn all the way off, so it wasn't possible to have a zero-to-whatever electrical current range with the leakage current.

"Fear not", said Marketing, we'll call it 'live zero', and sell it as a benefit. The concept of live zero was already established and recognized in pneumatics.

There were other contenders for a current loop 'standard': 10-50mA comes to mind.
 
B

Bruce Durdle

There is a long tradition dating from the early pneumatic instruments for a control range from 25% to 100%. If you want to know why, look at the pressure vs travel curve for a flapper-nozzle. The early electronic instruments used the same force-balance approach (many with a 10-50 mA range - easier to get a higher force with the higher current) which was eventually translated over into the pure electronic format.

4 mA is enough to supply the standing current needed for loop power and not enough to cause excessive power loss. Why not 5-21 mA? They had to decide on something. My pick would have been 5 -25 mA as it is slightly easier to convert a current into a % signal with the 20 mA span.

Bruce.
 
W
There are many reasons. It was a long time ago. One of the most important reasons was the need for a "live zero" for diagnostics. The European
standard was, and is, 0-20 mADC.

What I'd suggest is that you take it as a given.

Walt Boyes
Editor in Chief
Control magazine
www.controlglobal.com
blog:Sound OFF!! http://www.controlglobal.com/soundoff
_________________

Putman Media Inc.
555 W. Pierce Rd. Suite 301
Itasca, IL 60143
630-467-1301 x368
[email protected]
 
C

Curt Wuollet

Also, I believe it was used in teletype type current loop transmission and the currents were probably the current that guaranteed 100% of one type of relays would turn on and the current at which 100% of the same relays turned off while maintaining some minimum current both for the live zero and to bias the relays a bit. I read that on brittle yellowed pages somewhere before the information age.

Regards

cww
 
Does this selection have anything to do with selection of 250 Ohm loop resistance ie 4ma*250=1v
and 20*250=5v, which is the working range of most of the digital signals?
 
If you have a lot of power flucuations do not use this type of sensor. I promise you will be changinging out the sensor within 5 months to a year. Use a RTD type if you can.
 
M

Michael Griffin

In reply to Curt Wuollet: I believe we also had a discussion on this precise topic last year under the topic "NET: Current loop transmission rate". 20ma was indeed a common teletypewriter standard,
although there were also some (older ones I believe) which used other standards such as 62.5ma.

Teletypewriters were used in industry for paper tape punching and reading (for NC equipment), and for communications in the utility industry. In addition, teletypewriter technology was used for remote instrument telemetry. In other words, this sort of commercial equipment was adopted for industrial applications in much the same way that computer hardware was later (although teletypewriters were of much higher quality than the cheap PC equipment we see today).

Also, since teletypewriters were one of the larger volume electronic (or precision electromechanical) devices of the time, it is likely that components (e.g. power supplies) manufactured for them found a secondary use in other applications such as instrumentation.
 
C

Curt Wuollet

I got into the business about the time when "glass teletypes" were the rage but much of the test equipment and batch terminal stuff still had a TTY around with it's PTP and the smell of oiled paper tape. Well into the 80's many serial printers still offered a current loop option for long wire runs. The minicomputers and fledgling microcomputers were fully tty compatible as these were available surplus and often by simply scrounging. I possessed one of them for a while, but it went in a trade for some other long forgotten computer gear when I was playing with S100 bus machines and CPM. I should have hung onto it, it would be soothing to slowly but steadily cover a couple feet from the big yellow rolls and listen to the whine of the punch. As a side note, DOS still had reader and punch destinations built in long after the PC and magnetic storage made them obsolete. I still have a few rolls of paper tape and can faintly smell kerosene when I open the drawer. I can still fire up the CPM machine and marvel at how patient we were. Of course, in a race I did a while back, the CPM machine and WordStar were ready long before their modern counterparts, so I guess things don't change all that much. I should patch a driver for my laser printer just for the nostalgia. :^)

Regards
cww
 
M

Michael Griffin

In reply to Curt Wuollet: While we are talking about old hardware, I have some old computer hardware that is rather interesting from a historical perspective. I have an S-100 bus computer that was used to control some machinery with multiple servo axes.

It is a Compupro 8/16 which has an 8085 and 8088 dual processor, a RAM expansion board, an M-Drive-H RAM disk board, a math co-processor board, a memory mapped video board (it used a graphics monitor as well as a terminal), 3 single axis servo boards, and various other hardware. It came in a 4U rack mount case, and the 40 meg hard drive came in a separate 4U rack mount case. I got a WY-50 terminal to go with it from a separate source (the same as the original system used).

Unfortunately, the person who gave it to me had thrown out all the documentation, and all the 8 inch floppy disks including the boot floppy. It won't boot from the hard drive and I haven't been able to find another compatible boot floppy for it, so I've never been able to do anything with it (I admit though that I haven't put that much effort into it). The system was actually working at the time they scrapped it though.

What is interesting about it is that people were doing "PC base automation" before IBM came out with their first "IBM PC". The standard for business personal computer use in those days was the S-100 bus. Compupro (also known as Godbout) was the market leader, but there were others as well (Cromemco, Morrow, North Star, etc.). In those days, "real professionals" used the industry leaders - S-100 bus hardware and the CP/M operating system.

None of those companies survived past the end of the 1980s though. They went from market dominance to completely forgotten. The same thing has happened over and over again in the computing industry, and no doubt it will continue to happen in future. Most of the companies which are considered to be the industry leaders of today didn't even exist 10 years ago. The ones which which were the leaders of 10 years ago are the also-rans of today. It's a good reminder of the risks of basing your plans on the continued survival and dominance of any particular proprietary supplier rather than on genuine standards.

I've uploaded some pictures for anyone who is interested (the image hosting site requires javascript to be enabled to see much).

Both racks: http://tinypic.com/view.php?pic=66yp3k9

Mainframe top view: http://tinypic.com/view.php?pic=61vrtcn

Drives, top view: http://tinypic.com/view.php?pic=6gx2a12

S-100 bus racks were called "mainframes", although they weren't "mainframe computers". In both top views the front of the rack is to the right. The mainframe backplane can hold 21 boards, and this one is 2/3rds full. The three boards on the right are servo controllers. You can see the power supply at the top.

For the drive enclosure, the item at the bottom right is an 8 inch floppy drive. The item at the top right is a 40 meg hard drive. The power supply is at the back (left).
 
G

Gerald Beaudoin

Seems like there is a bit of nostalgia creeping into the list so I may as well put in my 2 cents worth too. (See Curt and Michael's comments)

I got into the business as a Field Engineer in the early 70s and cut my teeth on a Burroughs B3500. (Google that one for a look at the room full!) Model 33 Teletype was the order of the day for operator interface and they were just coming out with the new and exciting BIDS, binary input display system which had a CRT display! WOW... what progress.

Memory back then was measured in Kbytes and I recall doing a big upgrade from 48 to 96K. Memory back then was ferrite core technology which was actually pretty good and certainly a pretty cunning technique for its time. That was back when programmers had to program efficiently and every
instruction counted, or your program simply would not fit into memory.

The operating system allowed you to multi-program up to 10 jobs at a time... sort of like having a bunch of stuff sitting on your task bar today. Surprising to see just how many of the early concepts have survived to this day.

The hard drive was a 42 inch disk permanently mounted in a sealed enclosure. It had one head per track staggered across the surface so access time was equal to latency time, about 15ms. The were just starting to come out with removeable hard drives at the time.

Tape storage was the big thing back then with all its problems... stretched tape... pinch roller adjustments... vacuum columns... oh what fun.

Of course there was the usual assortment of card readers and card punches and line printers. To print a line of 132 characters, they used every technique from spinning drums to chains to slugs with embossed characters, all whizzing around at high speed, and being struck by carefully timed hammers to produce the characters on the page... all at thousand lines a minute.

It was at this time that I was first introduced to a Computer Aided Maintenance System. PMs (preventive maintenance) were all base on statistical analysis of all the field service reports recieved and new schedules were issued each month... hmmm... kinda sounds a bit like some of the stuff we do today.

We were very fortunate to get into the industry at that time as it has given us a unique window into how we have come to be where we are today.

Gerald Beaudoin
 
C
Hi Michael,

> In reply to Curt Wuollet: While we are talking about old hardware, I
> have some old computer hardware that is rather interesting from a
> historical perspective. I have an S-100 bus computer that was used to
> control some machinery with multiple servo axes.
>
> It is a Compupro 8/16 which has an 8085 and 8088 dual processor, a RAM
> expansion board, an M-Drive-H RAM disk board, a math co-processor
> board, a memory mapped video board (it used a graphics monitor as
> well as a terminal), 3 single axis servo boards, and various other
> hardware. It came in a 4U rack mount case, and the 40 meg hard drive
> came in a separate 4U rack mount case. I got a WY-50 terminal to go
> with it from a separate source (the same as the original system
> used). <

Yes, there was tremendous variety and innovation in those days before the IBM PC doomed the rest. It was a mixed blessing in that we would almost certainly be far more advanced if the IBM machine hadn't become a standard and froze things for decades. But the long standing standard did commoditize PCs.

> Unfortunately, the person who gave it to me had thrown out all the
> documentation, and all the 8 inch floppy disks including the boot
> floppy. It won't boot from the hard drive and I haven't been able to
> find another compatible boot floppy for it, so I've never been able
> to do anything with it (I admit though that I haven't put that much
> effort into it). The system was actually working at the time they
> scrapped it though. <

If you really get serious, I have some links to old hardware sites where you might find a boot image. Of course, you might have to patch it for your stuff. Computing in those days was really fulfilling because you knew a great deal about what was going on. My CPM machine is a latter day model because S100 wirewrap stuff will drive you crazy after many years of disuse. It's a Sanyo and uses 5.25 drives and such. The older stuff I have would require TLC to get running.

> What is interesting about it is that people were doing "PC base
> automation" before IBM came out with their first "IBM PC". The
> standard for business personal computer use in those days was the
> S-100 bus. Compupro (also known as Godbout) was the market leader,
> but there were others as well (Cromemco, Morrow, North Star, etc.).
> In those days, "real professionals" used the industry leaders - S-100
> bus hardware and the CP/M operating system. <

My first paid programming was on a Cromemco with Turbo Pascal. It was running test equipment over a IEEE488 bus. Later we moved it to a PC with a "huge" 60 mb hdd.

> None of those companies survived past the end of the 1980s though.
> They went from market dominance to completely forgotten. The same
> thing has happened over and over again in the computing industry, and
> no doubt it will continue to happen in future. Most of the companies
> which are considered to be the industry leaders of today didn't even
> exist 10 years ago. The ones which which were the leaders of 10 years
> ago are the also-rans of today. It's a good reminder of the risks of
> basing your plans on the continued survival and dominance of any
> particular proprietary supplier rather than on genuine standards. <

Yes, but in the computing world, some standards are like fish, keep them a little too long and they begin to smell. The PC architecture, for example, is getting to be a legacy problem.

> I've uploaded some pictures for anyone who is interested (the image
> hosting site requires javascript to be enabled to see much).
>
> Both racks: http://tinypic.com/view.php?pic=66yp3k9
>
> Mainframe top view: http://tinypic.com/view.php?pic=61vrtcn
>
> Drives, top view: http://tinypic.com/view.php?pic=6gx2a12
>
> S-100 bus racks were called "mainframes", although they
> weren't "mainframe computers". In both top views the front of the
> rack is to the right. The mainframe backplane can hold 21 boards, and
> this one is 2/3rds full. The three boards on the right are servo
> controllers. You can see the power supply at the top.
>
> For the drive enclosure, the item at the bottom right is an 8 inch
> floppy drive. The item at the top right is a 40 meg hard drive. The
> power supply is at the back (left). <

Yes, and you didn't feel like carrying any of these to a "net party" :^).

Regards

cww
 
M

Michael Griffin

In reply to Gerald Beaudoin: Ah! But I can provide a "connection" between your story and mine. The Burroughs B3500 was eventually replaced by the Burroughs B6700. A B6700 was used to develop the USCD P-System operating system. The Compupro computer than I have was replaced by a system running USCD P-System. (I know this is a bit of a stretch, but there we are, we're still all on topic). The following is a bit of history for the sake of those who are interested in these things.

The UCSD P-System operating system was one of the operating systems available for the original IBM PC (the others were PC-DOS and CP/M-86). It was already running on a wide variety of hardware at that time. Indeed it ran on on what was probably a greater variety of hardware than any other OS has, with the possible exception of the Unix (and Unix-like) OS family.

The P-System was much more sophisticated and user friendly than either PC-DOS or CP/M, but it was capable of running on what by today's standards would be considered very limited hardware. Many of the advanced features which UCSD P-System had appeared on other small computer operating systems only much later. It didn't have a GUI of course, as affordable hardware in those days was not up to the job of running one. None the less, it was menu driven, had a (rather simplistic) virtual memory system, and the programming interface was through a set of high level libraries rather than the interrupt (or trap) calls used in most other comparable operating systems.

The P-System was never more than a niche OS mainly because the various companies through which the marketing rights passed all wanted excessive prices for a user license. After the latest owners gave up on the conventional computer market they targeted the embedded market for applications, so it is possible that it may still be encountered in some older equipment. I don't know if anyone else has worked with this OS, but it was a real eye opener as to how dramatically different an OS could be.
 
as for as my knowledge, standard a to d converter of value 0 to 5 v requires 250 ohms for the corresponding zero and full digital o/p
 
generally in a transistor some amount of voltage is
required for turning it on. this voltage is the cut in
voltage. up to this voltage the exists a nonlinear in its
characteristics. beyond this cut in voltage the char. of
transistor is linear. this nonlinear region lies between 0-
4ma, beyond 4ma it linear. this is why 4-20 ma range is used.
 
B

Bruce Thompson

I think it is naive to say that the early engineers of electrical instruments had a vision and plan for the "choice" of 4-20 maDC as the electrical analog defacto standard. Rather, I am convinced that it was more of an evolution than a decision.

When I first got into the industry in the 60's there were only a hand full of instrument companies. Honeywell (from Brown Instrument Co), Taylor, Foxboro, Fischer & Porter, Moore and Bailey. Most of these companies used the highly skilled immigant labor pool to "manufacture" precision and often jeweled movements for motion balanced peunmatic instruments. Some of you may recall that we actually had two peunmatic analog signals; 3-15 psig and 6-30 psig, the later to impart more power at the valve actuator without the need to increase the size of the actuator.

The early electrical "motion" balanced versions of these used the vendor specific analog signal. 4-20 ma (F&P originated), 10-50 ma (Foxboro originated)to name but two. So why did we end up at 4-20 vs. any of the other (and there were many) signals. I think that it was for three primary reasons:

1) Live zero (as many have already mentioned)

2) Electronics component advances did not require the higher voltages (power) whether motion or force balanced

3) The emergence of the concept of Instrinsic Safety (difficult to acheive with an 80 volt PS for a 10-50 loop)

I also like Walt's approach to "accept" it as a given, albeit rather simpistic. Now I only wish we could have gotten the equivalent from a digital bus standard (aka ISA SP 50) and Wireless (aka ISA SP 100)

Bruce
 
That's true but irrelevant. In Europe 0-20 mADC was and is still used because there is more than just an IC in the circuit. But the real reason was that a "live zero" at 4 mA could be used as a diagnostic signal. If the current dropped below 4 mA, it signaled that the loop was dead and somebody needed to go fix it. That's why the original ISA50.1 called for 4 mA as zero.

Walt Boyes
Editor in Chief
Control and ControlGlobal.com
555 W. Pierce Rd Suite 301
Itasca, IL 60143

wboyes [at] putman.net
www.controlglobal.com
 
Hey Bruce! Good to hear from you.

You are correct about the evolution, and the migration to IS. Even the original SP50 analog standard was a three headed hydra as you mentioned.

As far as fieldbus is concerned, the various buses are now cooperating to find common ground. Unfortunately, ISA100's field layer (ISA100.11a) is a dead letter. First, it has been shown to not work. Second, NAMUR and the IEC have already anointed WirelessHART as interntional standards. And third, there are over 20 different companies now manufacturing WirelessHART devices that are all interoperable and interchangeable. We have a chance to create a single standard still. What should be done is for the ISA100 committee (of which I am a voting member) to simply adopt HART 7.1 as its field device standard. We can then concentrate on issues like energy harvesting, backbones for industrial environments, and stuff like that.

That's what the vast majority of end users want.

Walt

Walt Boyes
Editor in Chief
Control and ControlGlobal.com
555 W. Pierce Rd Suite 301
Itasca, IL 60143

wboyes [at] putman.net
www.controlglobal.com
 
> as for as my knowledge, standard a to d converter of value 0 to 5 v requires 250 ohms for the corresponding zero and full digital o/p <

... but why then are some transducers internally converted to 0.5 - 5.5 vDC; just a variation?
 
Top