Large Scale Central

Battery Voltage Choices

General Discussion
#21

The three battery packs I listed are all 6700 mah. Airwire’s 14.8 is 6800 mah, but I have been installing the HJE ones as of late as the Airwire ones have been harder to get a hold of. The 14.8, 18.5 and 22.2 are all 6700 mah. I would think that regardless of the battery, if I were to run the engines at a lower speed that I would get the same or very close to the same run time out of the batteries? In this scenario, the speed would be such that its lower than the max speed of the 14.8 v battery regardless of the battery capacity.

Referring back to my first post, if I could fit a 18.5 or 22v battery in the engine without difficulty and the batteries are all the same mah, then I would think that having a higher voltage battery gives me the chance to run at a faster speed if I want, even if most of the time I run it at slower speeds.

#22

Referring back to my first post, if I could fit a 18.5 or 22v battery in the engine without difficulty and the batteries are all the same mah, then I would think that having a higher voltage battery gives me the chance to run at a faster speed if I want, even if most of the time I run it at slower speeds.

Nicolas I thank you can answer this yourself with the info here: 1 don’t go over voltage to the airwire I know don’t what a full charge is on a 22v and 2 as long as you don’t get any radio interference. Then yes to your question.

have fun Richard

#23

For what its worth, I run airwire in a usa trains GP9 with an 18 volt LI battery. Yes, Top speed is kinda fast, lol. But with a 6,000 mah battery it has run over 5 hours pulling 11 usa train beer cars on flat and level track at a typical 35-40 scale mph… This included running a phoenix sound card and lighting at a moderate level. My locos are used monthly and there is no sign of the 3 year old batteries deteriorating. I use 14 volt li batteries in my steamers and prototypical speeds are just fine. Most people don’t run this long, but it will do it.

Bill

#24

Yep, the 2 axle motor blocks in USAT are plenty fast, the trains that do really run slower than prototype on lower voltages are the 3 axle Aristo blocks, like SD45 and E8, Dash 9.

If I remember correctly, the USAT 2 axle blocks are worm on motor and worm gear on axle, the 3 axle blocks have an intermediate gear between the 2, and the 3 axle blocks have varying wheel diameters.

Greg

#25

Assuming the same load, the increase in speed is roughly based on the square of the voltage difference.

18.5^2 / 14.8^2 = 56% faster speed

22.2^2 / 18.5^2 = 44% faster speed

#26

I tested my engines today. I have some preliminary results to share.

Comparing SD70’s only as I only have track powered SD70 for the moment.

  • 22 Volts - 5 seconds
  • 18 Volts - 6 seconds
  • 14.5 Volts - 6.6 seconds
  • Battery Power - 7.9 Seconds with 14.8 volt battery

For these tests, lights were turned on in all cases. Battery power was the best time I got. Engines are note fully charged. Will update with fully charged results. Voltages for track power were measured with volt meter. Measurements were made using two cameras recording video at different ends of a straight stretch of track. Engine is allowed to get up to full speed before passing the first camera and stays at full speed until it passes the second. Videos were synced, a position was picked on the track and time measurements were made from the moment the coupler passed the first marker until it passed the second marker.

I intend to test again with lights off, and with a fully charged SD70.

#27

Nice test Nicolas. What is the distance between cameras markers? If known, scale MPH can be calculated.

#28

Distance between measured points is 248 inches (very close to it). The scale mile per hour for those results

  • 22 volts - 81 miles per hour
  • 18 volts - 67 miles per hour
  • 14.8 volts - 62 miles per hour
  • Battery powered - 51 miles per hour

I definitely do not need the train to go 80 miles per hour. When I ran it at that speed it made me nervous going around the curves. I am sure it would stay on but it makes me nervous none the less. 60 miles per hour though would be nice.

Its interesting to note that track powered at the same voltage as battery was lower. Perhaps because the lights were not drawing as much power? Also possible that its due to a not fully charged battery. The wheels on my other engines are all much more exposed to the track then the new SD70 that is track power. I will probably test it again once I convert it to battery. I am tempted to get a 18.5 volt battery for the engine. Seems like it might get me close to 60 miles per hour if we take into account that the battery power doesn’t seem to provide the same power as track power for similar voltage.

I am also going to measure voltage from wheels of battery power at full throttle. I can’t measure the batteries as they are tucked away.

#29

There is most likely some internal voltage loss in the electronic speed control. Your input is the battery voltage, but if you measure the voltage at the motor outputs it’s probably several volts lower due to the electronics added in the power path that the track power input does not have. I agree that if you can fit the 18V pack that would be your best choice if you want capability of 60 scale MPH or so.

#30

so track power goes directly to the motor from the rails, no power lost.

Battery power needs something to control the full voltage of the battery, and that is done with transistors, and usually there are 2 of them active in the circuit, called an H bridge. The voltage lost is equal to the voltage drops of the transistors acting as “power control”. These are usually FETs, the voltage loss can be about 0.7 volts times 2 or good FETs will be less.

Interestingly your voltage drop seems significant based on the fairly large loss in speed, what controller are you using?

Greg

#31

Good point Greg. Besides the motor control bridge, the rest of the board must consume some of voltage, even if trivial. I may just meter my Railboss board and see how efficient it is. Pretty sure Nicolas said he is running Airwire.

#32

Yup. I am running Airwire.

#33

Yep, the board will use current to operate the microprocessor and other stuff, but that will not affect the voltage to the motor, it should be only the H bridge drop.

Perhaps Nicolas has added some diodes to protect the Airwire unit, that will usually lose 0.7 volts each.

That might explain it.

Greg

#34

I didn’t add anything to the board. I used Airwire’s drop in boards for the USA Trains engines specifically because they were easy to do. Installation is straightforward. You open the engine up, remove the board that is there, put the board from Airwire in its place, and connect up the motors/battery using the connectors. Very plug and play.

I am still waiting for one of the test engines to fully charge. It could be, given how long its taking, that the batteries were quite low. Does the voltage drop as the batteries drain? I know they do for AA, AAA, and 9V but I don’t know much about these. Will be testing again once the battery is charged.

#35

Looking at the illustrations in their manuals, there are 4 large devices that must be passing current, but they are not all identical, which would imply not a standard H bridge… so maybe they are not using FETs for all 4 and that would explain the higher than expected voltage drop.

Greg

#36

Nicolas Teeuwen said:

Distance between measured points is 248 inches (very close to it). The scale mile per hour for those results

  • 22 volts - 81 miles per hour
  • 18 volts - 67 miles per hour
  • 14.8 volts - 62 miles per hour
  • Battery powered - 51 miles per hour

I definitely do not need the train to go 80 miles per hour. When I ran it at that speed it made me nervous going around the curves. I am sure it would stay on but it makes me nervous none the less. 60 miles per hour though would be nice.

Its interesting to note that track powered at the same voltage as battery was lower. Perhaps because the lights were not drawing as much power? Also possible that its due to a not fully charged battery. The wheels on my other engines are all much more exposed to the track then the new SD70 that is track power. I will probably test it again once I convert it to battery. I am tempted to get a 18.5 volt battery for the engine. Seems like it might get me close to 60 miles per hour if we take into account that the battery power doesn’t seem to provide the same power as track power for similar voltage.

I am also going to measure voltage from wheels of battery power at full throttle. I can’t measure the batteries as they are tucked away.

The wheels should be isolated from the circuit. They should not have any voltage from the battery control circuit on them.

#37

I don’t see that the voltage drop is higher than expected and would seem to be ~1.4 volts.

Interesting…, I have a tilda (~) in front of 1.4 and it displays as a raised dash (_).

#38

You are correct David. I don’t know why I thought I could measure the wheels to determine voltage. The only way I can determine voltage is on the SD70 when I convert it before I put it back together.

I ran a separate set of tests last night with a fully charged SD70 with and without lights on. I am going to collect all my results together and update as soon as I do.

#39

I have gathered all the data, double checked it and updated my results. Hopefully the embedded graphs are visible.

On a full battery, the SD70 on battery power with the 14.8 volt batteries performed as well as the track powered SD70 at about 17 volts. It was hitting about 65 miles per hour and from the SD70 Max Speed, is 70. When it was not fully charged it performed slower then 14 volts. It took hours to charge so it may have been quite depleted.

My GP38’s go about 55 miles per hour. The GP38 Max Speed was 65 so mine can’t go max speed of a real engine. I have not tried any of these yet at fully charged. I plan to take a new set of measurements once with a fully charged engine.

You can see all my data here

#40

Todd, I would hope AirWire was using FETs for output transistors, in an H bridge you would expect .6 to .7 total drop (2 transistors), but the voltage drop appears to be along the lines of a silicon transistor of .7 (times 2 is the 1.4).

Again, maybe someone with an AirWire unit can read the part numbers from the output transistors.

Nicolas, I would expect your measurements to not change even with fully charged batteries, once you hit them with a load, they should be at a nominal voltage pretty much through the usable “charge life”.

Greg