How much current is drawn by a consist?

[The_Ghan]

I think this would make a good sticky at the top of the DCC section: how much current does a consist draw?  Of course, it depends on what the consist is!  I've been doing some searching of the forum and I know that KenS has done some testing and in earlier threads has mentioned the following:

1. Kato E231 with an EM13 motor decoder, two FL11 cab light decoders and non-DCC LED lighting in all cars.  A six-car train peaked at 70 mA, and usually drew around 50 mA in motion;

2. LED lighting in a single car draws about 3mA.

Does this mean we should allow, say, 30mA (50mA) peak per motor car?  What about sound equipped locos, for example?

It would be good to have a table of measurements that we can use to gauge the amount of power our layout is drawing.

KenS, can you help with this?

Cheers,

The_Ghan

[KenS]

This sounds like a good idea.  I'm not sure I can accurately measure very low currents (low tens of mA), but a full consist is another matter.

Right now my layout is still apart for electrical work, which has been "ongoing" for months, but I haven't done much due to other projects. I probably need to get that back together to measure large consists.  But I'm on vacation next week and had been planning to work on that.

I have made a couple of past measurements, and I'll dig those up and post them.  I also have a couple of smaller things I can try measuring, and see how well that works for low-amperage devices.

[The_Ghan]

Thanks KenS,

A table of results would be tops.  I think measuring several consists from each manufacturer would probably be enough for us to get a pretty clear picture.  Take the shinkansens, for example.  Kato do single motor 16 car sets but Tomix often has a second motor car.  Thus, the Tomix one is likely to draw more current.  But how about locos?  EF63, for example, or D57?  How about differences between two of the same consist from the same manufacturer?  Or the same consist using to different multimeters?  This might help establish tolerances.  Then, we probably need to allow an "anti-bugger" factor to cover unanticipated and inexplicable things that happen on a home layout.  It might be a bit of work, but it might lead to some very simple rules of thumb that we can all apply.

On a side note: I bought a cheap multimeter from eBay and regret it.  I think it was a waste of twenty bucks.  It is conisistently off on voltage measurements.  It has a calibration feature but it won't adjust to anything like a true reading - unless all 5 of my controllers plus my Digitrax booster are all out of whack!

I'm now budgeting $100 for a new multimeter.  Trouble is ... I could also use that money for rolling stock!

Cheers

The_Ghan

[KenS]

A good multimeter is expensive, but worth it.   I use a Fluke 179, which replaced an earlier Fluke I "lost" (still wondering where that one went).  Both have served me well.

Anyway, I managed to do some testing using a Kato power pack (U.S. version) and a 360mm 2400mm (360 scale m) loop of Unitrack on my coffee table.  Here are some numbers:

Kato EH500 (3037-1)

Loco alone:
28 kph 60 mA
59 kph 80 mA
130 kph 110 mA

Pulling 9 WAMU:
28 kph 60 mA
59 kph 85 mA
130 kph 115 mA

The error range on this was less than +/- 5 mA, but not by a lot.  It's possible it drew a couple more mA with the WAMU at a scale 28 kph, but I couldn't tell.

(update: things aren't this clear-cut; see my next post for additional info)

Kato EF210 (3044)
21 kph 65 mA
56 kph 110 mA
130 kph 135 mA

Interesting that the smaller loco drew more power.  I didn't try it with the WAMUs since I'd put them away.  I'll try that later.

Kato Centram (white, 14-802-1)
32 kph 20 mA (the slowest it would go without stalling)
40 kph 25 mA

Modemo Tokyu 300 (NT85)
14 kph 25 mA
20 kph 30 mA
32 kph 35 mA
40 kph 45 mA

The Modemo ran much smoother at lower speeds.  I suspect there's more momentum (mass) in the motor, which accounts for both the smoother ride and the higher power draw.

More to come...but not tonight.

Update: fixed a typo.

[KenS]

I did some more testing, and decided the numbers with and without a load weren't as simple as I'd thought.  Part of the problem is that I can't exactly replicate the same speed in both cases, and even a small varation is introducing enough difference to swamp the apparent change due to load.  The other part is that the current draw is varying pretty wildly as the train runs.  While variation is usually under +/- 3 mA, at times, and for no apparent reason, it was > 5 mA.  I've been rounding to the nearest 5 mA, and that introduces some error since 83-87 and 85-89 both end up looking like 85. I'm also not clear just how consistent the meter is at these low currents. Some of the "variation" I'm seeing may just be in the meter itself.

All of this leaves me thinking that I'm not going to get meaningful numbers for "with" and "without" load, and can really only speak to the approximate power requirements (+/- 5 mA) for a given speed.

I made some more measurements with the EF210 trying to be more careful, but still rounding, and this is what I came up with:

First number is "with WAMU", second number is "without".

21 kph: (couldn't replicate) / 65 mA
24 kph: 70 mA / 65 mA
34 kph: 80 mA / 80 mA
56 kph: 110 mA / 105 mA
118 kph: 135 mA / 140 mA
130 kph: (couldn't replicate) / 135 mA

While this seemed to show the +5 mA cost of hauling the WAMUs, it wasn't always the case.  I think the extra power needed to haul the nine WAMUs is less than my error range, and may be closer to 2 or 3 than five.

I did a slightly different test using a Kato EF65 (3019-9) pushing the WAMUs, because it didn't have a Rapido coupler so it couldn't pull them.

(with/without)

28 kph: 60 mA / 60 mA
62 kph: 75 mA / 75 mA
86 kph: 100 mA / 80 mA
144 kph: 105 mA / 100 mA

Similar to the EF210 +5 variation (except for 86 kph; don't understand that one)

Next I tested a Greenmax Tokyo Metro 10000 series subway train (4128/4129) both with a full 10-car set and motor-only. This had head/tail LEDs on the cab cars, but no interior lighting. It also showed about 5 mA extra with 9 extra cars.

(full / motor only)

50 kph: 90 mA / 85 mA
72 kph: 100 mA / 100 mA
100 kph: 110 - 115 mA / 100 mA (for some reason this one was bouncing all over the place current-wise)
118 kph: 120 - 125 mA / 105 mA (this one two)

I think the Greenmax train may need lubrication.  It was rather noisy, and the variation with the full train makes me think the drag from the cars was variable.

I did another test with a Micro Ace E231 (A4013), but this time tested only the motor car and more precisely recorded power.

72 kph: 60 - 62 mA
93 kph: 72 - 75 mA
130 kph: 80 mA (+/- 1)
 
Interestingly, the MicroAce motor changes current much less with increased speed than the others.  I'll have to try some other MA trains and see if that's always true.

The bottom line is that these are all drawing 100 - 120 mA at scale maximum speed without added lighting. Which means that even with LED lighting, a 6 - 10 car train is probably in the ballpark of 150 - 175 mA.

[The_Ghan]

KenS,

Thanks for these results.  This is very informative.  I think some trends are being established but have one more question: How about consists that usually have two motor cars, such as the Tomix Shinkansens?  Results might be different in opposite directions too as motors push/pull each other.

I've given this some thought and a basic rule of thumb is the best way to handle this.  If we all know a single motor consist without lighting will draw upt to 150mA, for example, and a LED lit car will draw 3mA ... thus a 16 car Kato Shinkansen shouldn't draw more than 198, say 200mA, right?  I think a simple equation along these lines is the way to go.

Cheers

The_Ghan

[Martijn Meerts]

Age and wear of the motor, and the grade it's going up/down also play a rather big part in how much current is drawn by a train. Also, flywheels and size of flywheels will likely affect current draw, especially at slow speeds.

The ESU ECoS actually has a current meter as feature. It's really meant for the system to check how many amps are still available (it does this both for the main system and for each booster.) I've looked at it once in a while when running a train, and like KenS says, it fluctuates a lot, even on a simple oval. I think it depends on the rail connectors/feeders. If connectors are bad or the train is far away from a feeder, it'll likely get less track power. The decoder will have to compensate for that by adjusting the motor speed, which means current draw goes up.

[KenS]

I don't have a Tomix train other than a DE10 (which I can test to see how their motor compares).  I expect that two motors take very close to twice the power of one, given that all those empty cars are only adding about 5% (or less) to the total power demand.  Most of the energy is being lost in friction within the motor and drivetrain, not in rolling resistance of the cars. I can try running a 15-car Kato commuter with two motors and compare it to a 15-car Kato with one, but I'm not sure my 2.4m test track is long enough.

Now on a grade, things might be different.  I don't have space on my test loop to evaluate that.   There's certainly more power needed, as Martijn points out, but whether it simply spreads the excess across both motors, or if there is some efficiency in having two I have no idea.

Taking 200 mA as a "worst case" will work with the SMD-based LED lighting.  Older LED lighting using 1.5mm or similar LEDs with higher power draw could be a different matter.  Potentially up to 200 mA just for the lighting in a 10-car train. And older cars with bulb lighting are likely to draw a lot more.  I think I have a bulb-lit train I can test.

Martijn, I've seen the exact power drop-off effect you describe on a large Unitrack loop.  On a short one, it's less pronounced and I think the variation is more likely from dirt on the wheels (I cleaned the track, but didn't clean the wheels since the trains haven't been run much). It could also be due to the structure of the motor (DC motors have poles and as they rotate the intensity of the magnetic field varies) or to problems in the current pickup from the wheels (brass bouncing on brass making intermittant contact as the wheels go over track irregularities like joints).

What I've seen on a ~10m loop of Unitrack with DCC is that voltage on the far side with one feeder can drop by up to two volts (from 12 to 10).  If the train needed 150 mA @ 12V (1.8 mW), then it will likely need 180 mA at 10.  Of course my measurements here have been made at voltages around 5 - 7 volts for normal DC running.  How that will translate to DCC is hard to guess given extra loss in the decoder and the very different behavior of a motor driven by a PWM signal from one driven by straight DC, but I think comparing a pre-conversion train on DC to a converted train on DCC should provide some interesting data.

Looks like I have something to do this weekend in addition to watching fireworks. 

[KenS]

Just a quick note to compare two DE10's, Kato (7011-2) and Tomix (2222).  Not a perfect comparison because the Kato has a DN163K1D DCC decoder in it. My other Kato DE10 I haven't yet converted is hiding somewhere, so I couldn't test it.

The Tomix is noisy as all get-out, but runs fairly well.  The Kato was having some trouble with DC, with the speed sort of oscillating slightly (and the engine sound going up and down as a consequence). It runs much nicer on DCC, although I don't have track set up to check that.

Tomix DE10
  46 kph 95-102 mA
  72 kph 115-120 mA

Kato DE10 (w/ DCC decoder on DC power)
  45 kph 52-61 mA
  76 kph 64-70 mA

I haven't had the Tomix apart to see what it has for a motor.  The Kato has one of the their newer three-pole skew-wound motors, the same kind found on the "DCC Friendly" EMUs, but equipped with flywheels as well.

[The_Ghan]

KenS,

Thanks for your great work.  I'm going to buy a new multimeter over the coming week or two and test my own roster.  I think we need a few of us to build up more data on the subject so that we can more accurately record how much current is drawn.  I don't have ANYTHING that you tested on my roster, but I do have 2x several consists, which I'll test to compare things coming out of the same factory at the same time.

I'm also green with envy to learn that you own a complete Greenmax Tokyo Metro Series 10000.  At over $400 its a bit beyond my budget at the moment.

Cheers

The_Ghan

[KenS]

I'm also green with envy to learn that you own a complete Greenmax Tokyo Metro Series 10000.  At over $400 its a bit beyond my budget at the moment.

It's one of my favorite trains to look at.  It's not quite as smooth-running or quiet as a Kato though.  I still really want to get a second so my subway line can run two at a time, but the cost has deterred me so far.