Will this Amp meter work for model trains?

Like Dewey, I used analog meters and I think I only paid $6 for them. This meter looks like it will work. I am not sure about the usb but dc supplies are only a few dollars and you can always find dc power supplies from ho folks.

Looking over the specs given on the listing, yes, the meter will work, and yes a simple usb charger will work to power it.  

I would go for a very cheap charger here and not run anything else off of it simply because I have no idea how well this particular meter is made and wouldn't want to chance a failure sending track power through any other 5v devices.  That may not even be possible with the design, but for the cost of a 99 cent adapter, I wouldn't bother with the chance.  

As other noted, you may want to look into an analog meter depending on the look you are going for.  The spec sheet here says this meter has a 300ms refresh rate, which is not bad, but may not be enough to show quick spikes of current if you have a short, say with one roller on a switch or some such thing.  

For a simple, out of the box, digital display, this looks like a decent meter.  

Additional slightly related stuff:

If one were so inclined they could also use some pretty basic Arduino type modules to replicate the function of this digital meter.  For a single meter using the pre-made one is much easier and probably about the same cost, however if many such meters were wanted, the roll-your-own approach could add each additional track and display  at around $5 each.  you could also program such a meter, with a relay module added, to shut off power to a particular track if the current remained too high for some desired period of time, and maybe turn on a warning light.  

JGL

JGL's admonition about power supplies is well founded!  You should use a totally isolated logic power supply for this meter, as the measured voltage is partially common with the power supply.  You can't power any other devices that have any connection to your measured voltages with the same power supply.  Doing so will release the magic smoke from the meter in a hurry!

gunrunnerjohn posted:

... the measured voltage is partially common with the power supply.  You can't power any other devices that have any connection to your measured voltages with the same power supply.  Doing so will release the magic smoke from the meter in a hurry!

I'm not absolutely certain this is the case without having the meter to look at or a better description.  The "instructions" that are on the listing are a common set for the whole line of similar amp and volt meters.  I think I'm reading it as having a shared common on the volt meters, but I could be wrong.  

Having done a little more research, it seems likely this meter uses a hall-effect chip to measure the current, and if this is the case it is pretty well isolated electrically between the low power and high power sides.  

All that said, I still wouldn't run anything else off that particular power source.  Also, don't leave any magnets laying around near the meter.  Hall-effect sensors are magnetically sensitive and will read wrong if there is magnetic interference.  

JGL

Since I've seen one toasted by trying to combine the power supplies, I'd be somewhat concerned about running the volt and amp meter off a single supply.  However, it'll be interesting to see how it works. 

JohnGaltLine posted:

...

Having done a little more research, it seems likely this meter uses a hall-effect chip to measure the current, and if this is the case it is pretty well isolated electrically between the low power and high power sides.  

Tell me more about your research.    In the eBay listing the 2 page instruction PDF and the wording in the listing suggests (to me anyway) that you get a digital meter assembly and you attach an included shunt.   The current sensing and digital display functions may indeed be isolated but it's not obvious to me how it is done.

Stan, you're right.  After finding what appear to be the same meter listed elsewhere, it looks like this one uses a shunt, which should be included.  As a note, I've found the same meter listed for under $5.00.  

Also to add more to the confusion, about half of the places with this meter list it's effective range as 5-50 amps, suggesting it may not respond correctly on readings below 5 amps.  

As to where I got the idea that they might use a hall-effect sensor was mostly by getting sidetracked looking at sensor modules, most of which are based on the asc712 chip.  Many of the lcd displays are using these, or meters with ratings under 30 amps.  The sensor will actually handle more than 30 amps, but the modules sold are only offered scaled for 5, 20, and 30 amp reading ranges.  

In the end, after going this way and that, I think the only way to know for sure if the meter will work to monitor track current is to try one out.  

So will it work?  Maybe.  Can you power it with a cheap usb charger. yep, and you probably should, with nothing else powered off that charger. 

JGL

They are correct on the low current sensing, they don't measure at low currents.

Thank you everyone.  I appreciate all of the advice.  I am going to give it a try and will let everyone know how it works.  I am actually going with a red LED model that has very similar specs.  I am buying 4 Amp meters and 4 volt meters.  I am going to use separate power supplies for each meter, completely isolated.  

GRJ, I will try my best to avoid the magic smoke, but I am known for that...

George

Just be careful out there!

George S posted:

... I am buying 4 Amp meters and 4 volt meters.  I am going to use separate power supplies for each meter, completely isolated.  

I realize USB 5V adapters are quite inexpensive.  But if each requires its own AC wall outlet or spot on a power strip, it could get a bit untidy.  As shown in this thread, one alternative is to used DC-DC isolated power converters to isolate each meter from each other.

https://ogrforum.com/t...-digital-panel-meter

It does take a bit more assembly effort but you could then use one AC adapter (wall-wart) to drive all the DC-DC isolated converters thereby taking up only 1 AC wall outlet/socket.  Isolated power converters run about $2 each on eBay.  For example:

Stan,

That is very helpful!  Thank you!

George

George, I meant to ask why you chose 50 Amp (vs. something less) meters in an O-gauge model train application?

Stan, I really had trouble finding anything else at a reasonable price in digital.  I checked Mouser and the prices were extreme.  I may not be searching properly, but these seemed like the right price.  I did not see any 20 or 30 amp AC meters on eBay.  

Also, I checked the price of the DC to DC converters, and the price has gone up to $3.70 plus shipping each (over $6 on Mouser).  I had already purchased 10 5V DC power supplies on eBay for $18 plus $1 shipping.  I know this is not ideal, but I think I can manage the wire and power strip.

George

Hmm, I see what you mean on the 50A.  It looks the guy from the linked thread also used a 50A digital meter.  Not that my ponderings help you but it sure seems odd that these sellers just don't provide a different value shunt to make a 10A or 20A meter.  That is, it looks like the meters have settable jumpers on them which probably tells the electronics what scale/range to use.   Oh well.

If you already bought the USB converters then we're just chatting over coffee here ... but I noticed several of the eBay Asia suppliers had free incremental shipping for isolated DC-DC converters which suggested the unit cost could go lower when buying a few.  So, for example, here's one for $2 each when-you-buy 10.

Stan, I am not an expert here, but that is a 4.5 - 5.5 volt input model.  I can understand there is some ability to scale if you have an input power supply that is capable of 500ma and each converter only needs 100ma, but I would think the ones that use 10 -12 volt input could be powered with power supplies that could scale better.  I need to power 8 meters.  That is why I avoided the 4.5 - 5.5v input ones.

BTW, I just bought ordinary wall warts, not USB chargers. I thought about the USB ones, but the wall warts were cheap and will be easy to install.

George

I found dual output converters.  Am I right that the dual outputs would not be isolated from each other?  If they were, then those would be a good option.

Post the eBay auction number.  You can't post direct links here, but so far no objection to the auction number.

Be careful of "ordinary wall warts", the rated voltage could be only at load, many are unregulated.

Here is one.  There are several different options.

This is the power supply I bought:

They look like they're probably regulated, though I'd want to verify that to be sure.

George S posted:

Here is one.  There are several different options.

eBay item number:

161811041864

 

Looks like the outputs are isolated from each other. I've attached the datasheet for that part.

Thank you GRJ

George S posted:

Stan, I am not an expert here, but that is a 4.5 - 5.5 volt input model.  I can understand there is some ability to scale if you have an input power supply that is capable of 500ma and each converter only needs 100ma, but I would think the ones that use 10 -12 volt input could be powered with power supplies that could scale better.  I need to power 8 meters.  That is why I avoided the 4.5 - 5.5v input ones.

OK.  This was a helpful conversation.  I was not aware of the converters before, and given they are isolated, I can imagine a number of great applications that will protect my meters and other control circuits.  These converters would definitely save space and would potentially allow me to power my panel with one power-supply.  That is a nice benefit.

Best,

George

FWIW, the meters typically only require a pittance of current, I have a variety in my drawer, none take more than 20ma.  Even the combined LED volt/amp dual meter only took 20ma at 5V.  The LCD with EL backlights took from 8 to 12 volts at about 10ma.

GRJ, now that's useful information!  As you pointed out earlier, these DC-DC isolators typically have a MINIMUM load current to maintain its output voltage....for example 20 mA as shown in the table.  The application notes with the datasheet will say to add a load resistor to insure this minimum. 

So for anyone else following along, for example, if the meter only draws 10-20 mA (at 5V), a resistor could be added across the DC-DC converter output that draws, say, 15 mA to bump the load current to 25-35 mA.  Ohm's Law: 5V / 15 mA = 330 ohms.  The resistor would "waste" 5V x 15 mA = 75 mW of power (0.075 Watts) which is inconsequential in relative terms.  A 1/4 Watt resistor would be fine.

Note that some meter models wanted 8-12 VAC or DC, others wanted 5VDC.  It's important to match what you have to the power requirements.

So, I changed my mind and bought 4 of the dual output controllers.  They take 10 -  16v input.  I have an old laptop power supply that has 12v DC out.  It says 4amp, which is a lot of power.  I assume the controllers controllers will only draw what they need, correct? Or, do I need to find a smaller power supply?

Now I need to see if I can cancel the wall wart order...

Correct, the controller will only draw what they need.  Taking GRJ's numbers, it appears each meter requires no more than ~1/4 Watt (or so)...or, say, ~2 Watts for 8 meters.  So with DC-DC converter efficiency, you'd need ~3 Watts or so going in.  Well, I'd say your 12V, 4 Amp (48 Watt) supply out to do the trick!

Mind the comment about minimum required load on the converters.  Resistors are only a few pennies a piece. 

Great.  I have some resistors laying around.  I will see what I have in the 330 ohm size.  Seems like I have headroom for more current draw. I think the converter said it could handle 2watts, which would mean about 400ma. I tried to calculate the max size resistor I should use, and I got 12.5 ohms (5v/400ma).  That doesn't seem right.  I thought it would be a higher value than 330?  What am I doing wrong or misunderstanding?

Per the table, each converter can supply up to 200mA at 5V (1 Watt).  You have 2 converters in the package so that's 2 Watts.  The purpose of the resistor is to insure you are drawing at least 20mA.  The meter itself will draw something - as GRJ suggests it will be modest...maybe 10mA, 20mA, whatever.  The resistor is simply drawing so that the combined total of resistor+meter is above the 20mA minimum.  I simply gave an example where the resistor draws 15mA (330 ohm).

It could well be that the meter draws 20mA (or more) by itself...in which case no resistor is needed.  In any event, if you do need a resistor, it will draw no more than 20mA which means it will be no less than 250 ohms.

If you have a basic multimeter, it might/will come in handy when putting your panel together.  If not and you live near a Harbor Freight Tool store, you can get one for Free with coupon.

I have a multimeter, but I do want the free Harbor Freight one.  However, I can't find my coupon.

Thank you.  I realized my confusion.  A smaller resistor has less resistance, so it will draw more current from the transformer.  Got it and appreciate the advice to not try to exceed 20ma with the resistor.

George

Here is a typical Analog type ammeter available on Ebay.  This one has a range from 0 to 10 amp AC and does not require a separate shunt or separate power supply. Note that this one is cleary marked as AC amps (not DC amps).

While I haven't had anything to add, still been following along here all day, and as the project gets a bit more complex I'd like to re-propose the nuclear option from my first post, building your own micro-controller based meter unit.  Added to the complexity factor is that the proposed meters are not rated to give accurate results on loads under 5 amps.  While 10 amp supplies to each block are common, actually using more than 5 amps may not be, depending on what sort of engines and cars are being run.  Double heading dual motor postwar locomotives with a string of incandescent lighted cars?  You can count on pulling more than 5 amps, however with any modern engines, and non-lighted or LED cars, you may never see a 5 amp draw.  

Having priced out the parts and at least finding enough videos on youTube to prove the concept works, it is starting to look like a better option for accurate readings, and as the project to use the shunt meters gets more complex, using a bunch of auction site Chinese modules and an arduino is getting less complex by compare.  

If I understand correctly, George is planning on having 4 amp meters and 4 volt meters in this application.  I would assume for power from 4  transformers(or separate throttles on the same transformer) to 4 blocks or loops of track.  This can be accomplished with some easy to find parts from the auction site and some basic components.  Aside from being substantially less expensive, and only a little more complex to implement, you gain the benefit of having the high current voltage completely isolated from everything else, letting you use a single power source and not have to worry about the failure of one current sensor taking out anything else.  You also gain the ability to have the meter panel automatically cut power to the track in the event of rising above any preset current level, for any preset amount of time.  You could even make that adjustable with a little display and control panel if you wanted.  

The sensors I've found appear to have a .1 amp sensitivity, which may be less than the shunt meter offers, however it is plenty good enough for casually monitoring track current draw.  I did not see a sensitivity listed on the meter linked at the top of this thread, and every similar meter has the same documentation.  There is 75mv shown, but I do not know how to translate to how many decimal places it reads in amps. 

Anyway, the hardware involved would be as follows priced from china on the auction site.)

^{1 Amp USB charger:  $4.99}

^{Arduino Nano(clone board):  $1.84}

^{4x ACS712 based current sensor modules:  $1.36 each, $5.44 total}

^{4x 8-digit MAX7219 based 7-segment display:  $1.18 each, $4.72 total}

^{16x 1n4001 diodes: $0.01 each.  16 cents total.} 

^{4x 1uf 50v capacitor:  $0.10 each.  40 cents total.}

^{8x Undetermined resistors, about $0.05 each. Adding 6 bucks for a full assortment pack.}

^{4x fuse holder:  $1.01 each.  $4.04 total.}

^{4x fuses:  $0.10 each.  40 cents total.}

^{Various wire for hook up and high current:  apx. $5.00 ?}

^{Terminal strips for connections: $1.99 each, $3.98 total.}

^{(optional) :}

^{Arduino jumper wires:  about $1.00}

^{830 point solder-less breadboard:  $2.03}

Grand Total:  $40.00  oddly enough it came out to a nice even number, Couldn't do that if I tried.  

While LED's take very little power each, one may think you need a bigger power supply to light up 8 displays, after all you've got , say 7 segments per digit, 4 numbers per meter and 8 meters, so you could have up to 224 LED on at 20mA each, or 4.48 amps to light them all. But wait, you say, you're only using a 1 amp power supply!  Sure am.  The LED's use a multiplexer driver which is built into each set of 8 digits.  On each set of 8 digits only 1 led is actually turned on at a time, it just switches from one to the next faster than the eye can see.  there is never more than 4 LED's turned on at a time on the whole setup, drawing 80mA max.  that's nothing for our little power supply.  

I chose the NANO specifically for this project rather than the UNO or pro-mini arduino boards because the nano has 8 analog input pins where the larger board has only 6.  The code to read proper RMS amps is the most complex part of the project, but it has been demonstrated in many places on the web already.  the rest of the code is simply shifting data out to the displays and is straight forward.  

The ASC712 modules are Hall effect sensors.  they come in 5A, 20A and 30A versions, but there is actually no difference between them for our application except the max current we can read.  I'd use the 20 amp version to make the math simple in the programing as it is 100mV per 100mA.  All three versions will source a continuos 50 amp current with out damage, even though they can not measure the current, and will handle at least 100 amp spikes.  Being hall effect sensors they are electrically isolated between the sensor and the current being measured.  

The MAX7219 based 8 digit displays use a simple 3 wire interface to the arduino.  2 of those wires are shared across each display, and the last is separate for each, taking up 6 of the digital I/O pins of the NANO

The rest of the parts are just basic stuff.  Fuses as a CYA feature, incase something fails, diodes, caps and resistors to make a rectifier and voltage divider do the Arduino can measure the AC voltage, Wire and terminal strips for the connections and hook-up to the transformers and track, And some little wires and a bread board to plug everything together.  

I would probably choose a slightly more complex route using tiracs and make a fold back circuit, but with a $5.02 relay board and a 10 cent 74595 chip you could also add the function to turn off power if current goes higher then a set level.  Might need a couple of transistors in there, but not sure on that.  the relay boards are rated right around the max current the shift register can provide.  

Nuclear option over now, return to your regularly scheduled conversation.  I'm gonna build one and see how it works.

JGL

pro hobby posted:

Here is a typical Analog type ammeter available on Ebay.  This one has a range from 0 to 10 amp AC and does not require a separate shunt or separate power supply. Note that this one is cleary marked as AC amps (not DC amps).

Pretty sure he wants a digital meter, there's plenty of analog meters on eBay.

JohnGaltLine posted:

... I did not see a sensitivity listed on the meter linked at the top of this thread, and every similar meter has the same documentation.  There is 75mv shown, but I do not know how to translate to how many decimal places it reads in amps. 

..

I'm gonna build one and see how it works.

If you look closely at the photos in the eBay listing for the meter you can see the module uses the 7107 IC chip.  This chip is to low-cost meter circuits what the 555 IC chip is to timing - been around forever, used everywhere, inexpensive, etc.  So as the listing says, 3-1/2 digits or up to 1999 display counts with the leading "1" being the 1/2 digit.

Since you're going to build one, I suggest you study the 7107 datasheet:

http://www.intersil.com/conten...7/icl7106-07-07s.pdf

I understand the attractiveness of the Arduino with its multiple A/D converters.  But read the datasheet to work through how it solves classic measurement/instrumentation issues such as making differential/floating measurements, how it performs auto-zeroing, how it integrates (filters) measurements, etc.

If you want to bring something new to the party, in addition to your over-current detection feature, I'd be impressed if you implement true RMS and (since you have both voltage and current) true power displaying Watts going to the track.

JGL, watch out for Stan, he will stick you with "feature creep" if you're not careful.

GRJ, feature creep is my speciality!  

I like the idea of implementing a proper watt meter, and it would not be particularly difficult once you have real time volts and amps being measured anyway.  I've already been planning on adding a second set of LED displays for each meter to show a peak level over time, say the last 30 seconds, such as my multi-meter offers.  

The main limitation is insuring the processor in the arduino has enough time to sample the sine wave enough for an accurate reading.  as the code gets more complex for things like displays and floating point math, it really starts slowing things down.  I'll have to experiment to find if it becomes a problem, and if it is my solution will probably be to use a second arduino.  one will only handle reading the data from the current sensors and converting to an RMS value, then send that information to the other that will process the information and combine it with volt readings and send it to the displays.  

The 7107 is a neat chip.  looking over the specs it offers two scales from what I see, a 200mV range and 2v range, with 2000 counts, or steps, on each range.  That works out to 1mV per step on the 2V setting, and 0.1mV per step on the 200mV range.  Pretty good... but only if you only need to show 0-1999 on the display.  

With a 5-50 amp scale on this meter, the left most digit becomes entirely useless, without some added circuitry for auto-ranging, as it can only show a 1 or nothing (and even with that circuit would only be useful in the 10-20 amp range).  Left with 3 real digits, the meter must show, at best, the same 1/10th amp sensitivity that the Hall Effect sensors can detect, so, in the end the effective sensitivity appears to be the same, 100mA. 

With 100mA sensitivity you'r not looking to replace a multimeter for testing, but it is plenty good enough to monitor the current draw on each block of your layout, and is also enough to be useful in detecting over-current, and turning off the power.  with a 5 microsecond response time to changing current, the sensor will report changes much faster than the rest of the system could ever keep up with.  Also, with a 100mA sensitivity, I think the arduino's math will be good enough to report results that are usefully accurate.  it might come up with a figure that is off by one or two milliamps, but that will not matter when the smallest increment is 100 to be reported is 100.  I'm also tend to do my math with long integers rather than floating points when ever possible.  I don't think this is an option with squares and roots, however.  In any case, I think the error will be small enough not to matter in the application of track power monitoring.  

Guess I'll find out when parts are ordered and get here. Probably 3 weeks or so.  

One more question for you folks here:  When sampling for a RMS reading, how many samples per wave is enough for a useful reading?  

JGL