My first experience with Power Functions

[22kane]

Last night I ran my first Power Functions train. I recently bought the latest Cargo train set and assembled the engine then hooked up some of my older 9v cars. My overall thoughts are mixed. For some reason though disappointed comes to mind. The controls are jerky and a couple times the engine got away from me and derailed. I would turn the control wheel once and nothing. I would turn it twice and nothing. One more time and the engine would shoot off like a rocket. I like to run trains at lower realistic speeds and this seems like a problem for PF systems. I do like the fact that there are no wires but that's the only pro I can come up with right now. Also, in order for me to power my Emerald Night and Maersk set it's going to cost an additional $100.

Now I understand what all the fuss was about. Was this really the best direction to take the Lego train line? I guess as time goes on I will master the controlling of PF and won't have these "rocket" issues. I will be very upset if I invest in PF for all of my trains and they change it all out to something else in the very near future.

Almost forgot, I do like the fact that the track is much cheaper than the 9v rails and the fact that I can still use my 9v rails with the current system.

So the real question is where can they go after PF? Aside from Lego trains I am an N scale modeler that uses DCC. Not much has changed in the latter except for the addition of microchips controlling each cab by a specific address. The installation of these DCC boards can be very difficult for adults as well as programming so I can see why it wouldn't work for Lego. In fact, one of the biggest topics in HO scale is battery powered locomotives and the possibility of them being readily available in the near future. Powered track even in DCC remains an issue because the connections must be even greater for the signal to remain constant. On my large N scale layout I had to solder every rail joint to ensure premium conductivity. The longer the track stretches the more feeder wires you need to keep the electricity from diminishing and that would not be an easy task for Lego trains. Now with PF, kids can set up a large layout and not have the problem of trains stuttering half way through their layout.

Would I abbandon PF entirely? Absolutely not. All in all battery power is a good idea for Lego but here are some things that need improvement without changing the entire PF system.

1. Stronger signal strength

2. Higher power batteries

Both of these will ensure smoother operating features. Would it drive the cost up? Maybe, but at the rate of technology it shouldn't be an issue for too long.

[Pingles]

I'd be interested ot hear form others if they have the same issues.

Mine is very responsive and my speeds range from a crawl to outrageous. Could IR interference be an issue?

I suppose it would be nice to have more steps. I could see the benefit of a true potentiometer type of response.

But I certainly have no delays or stacked responses.

[Toastie]

Now I understand what all the fuss was about. Was this really the best direction to take the Lego train line? I guess as time goes on I will master the controlling of PF and won't have these "rocket" issues. I will be very upset if I invest in PF for all of my trains and they change it all out to something else in the very near future.

Hi 22kane and welcome to the club!

Yeah it was and is an issue for me, not so much the PF stuff, in contrast, but I can't adjust to batteries. Whatever.

1. Stronger signal strength

2. Higher power batteries

Both of these will ensure smoother operating features. Would it drive the cost up? Maybe, but at the rate of technology it shouldn't be an issue for too long.

Completely agreed.

I guess this is your rocket problem: The IR signal. Maybe there is some additional noise in the room you were playing; I had some issues in this regard with energy saving lamps. They can severely limit the IR range. Upon turning them off it worked much better; classic heat balls don't do that.

Range extension by stronger signal strength: Well line of sight would remain an issue. I have built some custom IR/RF transceivers for my RCX powered loco's and some other RCX fun stuff. Even with the very low RF power I am using on these (the RCX does not have that much power on the active inputs, 5V/10mA max., and I am using a back channel for handshake operation), I have much longer range than with IR.

I am entertaining the idea of making a 1x4 footprint, 4 bricks high RF receiver only. This one should attach directly to the PF receiver and would be PF powered. On the transmit side one could use rather strong 38kHz range extender thingies or the like.

Would that work for you as well?

Regards,

Thorsten

[F0NIX]

Have you tried pointing the remote control up into the ceiling? It spreads much better in your room and you get better control of the train. The reason is the IR receiver on top of the train pointing up. And even if it has receptions to the side it is best from straight above the receiver.

[Sokratesz]

DCC is probably still too expensive for LEGO / too complicated to set-up for the average person that buys LEGO trains.

Personally I would like to see a return to the 12v era train style with a lot more railroad-related structures (stations, crossings, industrial buildings etc) and all the powered accessories that they had back then: remote controlled points, stops, crossings and extensive possibilities for custom lighting. They should also use more bricks in their trains, I hate large pieces more than anything. Remember the locs of 7745? They had a bloody nice number of bricks in them, not the entire huge window and wall pane sections that we have nowadays. (although recently this has changed for the better)

- Sok.

Edited August 20, 2012 by Sokratesz

[22kane]

Thanks for the tips, Toastie and Fonix. I tried a few different things with the placement of the controller and got a better response but like I said before it's still just not right. There is something about the old DC boxes that give you a better feel for train control. I'm not very well educated on wireless capabilities and was wondering if my wireless Internet may be playing a roll as well? Thanks again for the tips

[hoeij]

Thanks for the tips, Toastie and Fonix. I tried a few different things with the placement of the controller and got a better response but like I said before it's still just not right. There is something about the old DC boxes that give you a better feel for train control. I'm not very well educated on wireless capabilities and was wondering if my wireless Internet may be playing a roll as well? Thanks again for the tips

I think that there is a design flaw in the PF remote+receiver. If you turn the dial, instead of broadcasting "dial is in position #3" the PF remote broadcasts "one step up" "one step up", etc.

But if you turn it quickly, you might turn the dial 5 notches while it only broadcasted 2 notches. So the position of the dial now no longer corresponds to the speed setting.

With the usual DC controllers, the position of the dial corresponds with the speed, so it's easy to put it at speed setting #3. With PF, the position of the dial doesn't matter. All that matters is whether it is turning left or right. So if you want to be sure to get the train in speed setting #3, you have to look very carefully and count the steps you make.

As for accidental derailings, in one of my PF trains, I put only 4 batteries in the AAA battery box (and added some wiring). This way the top speed is only 6V, and it doesn't derail at that speed. A simpler thing to do is to use rechargeable batteries, their nominal voltage is lower and this might be just enough to prevent derailings (haven't tried that with PF, but with the previous passenger train, the one that uses AA instead of AAA batteries, this slows it down just enough to prevent derailings). Note that rechargeable batteries are also preferable for other reasons, they can produce more amps, so they are better for long trains. For long trains they'll also last much longer than alkaline batteries.

PS. the wireless internet does not interfere with PF, it's a completely different signal.

[AndyC]

I think that there is a design flaw in the PF remote+receiver. If you turn the dial, instead of broadcasting "dial is in position #3" the PF remote broadcasts "one step up" "one step up", etc.

But if you turn it quickly, you might turn the dial 5 notches while it only broadcasted 2 notches. So the position of the dial now no longer corresponds to the speed setting.

I think that's a very intentional design decision, it means the remote doesn't need to be constantly sending out signals in order to ensure the reciever has picked up the correct speed setting.

As to the longer term plans, I'm hoping TLG come up with a Bluetooth controller for Power Functions, although I imagine the difficulty there would be finding a way to reliably and simply pair with a controller.

[architrains]

The only problem I've had with PF is that with a long train stopped on a curve, sometimes on speed setting 1 the loco just sits there and hums. I look at it positively though and see it as being more realistic, as in the real world sometimes the throttle has to be opened all the way to get the train moving. But the train still slows down perceptibly in curves, which as far as I can tell has always been a problem with Lego trains (my 9v had the same problem), so broader curve geometry should fix both problems.

I've had the loco "rocket" every now and then, but I think what's happening is that the "clicks" on the tiny dial are too close together, which means that sometimes you can stop in between two clicks, and it goes ahead and clicks one (or two more, if your finger is still on it, adding weight), either up or down. I've considered trying to design something a little more appropriate for controlling trains, using those technic slots in the middle of the dials. By taking the radial movement out to a larger circle, the clicks would be farther apart. Maybe a miniature train-styled control panel similar in size to the 9v controller could be built? I'm not a good enough MOCer yet to attempt it, but it's an idea.

I haven't had many line-of-sight issues with PF, the only time was when the room was carpeted and people were bending over the train. I think the hardwood floor I use it on now helps the signal bounce. I had a track dead-end behind the couch in flex track so I could hear it approaching the end of the line, and the controller still worked.

As for accidental derailings, in one of my PF trains, I put only 4 batteries in the AAA battery box (and added some wiring). This way the top speed is only 6V, and it doesn't derail at that speed. A simpler thing to do is to use rechargeable batteries, their nominal voltage is lower and this might be just enough to prevent derailings (haven't tried that with PF, but with the previous passenger train, the one that uses AA instead of AAA batteries, this slows it down just enough to prevent derailings).

I have 6 rechargeable AAA in my 7939 loco, and it can still derail at top speed by itself or with a few cars.

[Tearloch33]

I never really experience 9V LEGO trains. I recently got back into LEGOs, and I have found PF quite enjoyable. I have the 7939 Cargo Train, and have yet to have any derailment issues, even at the top speed (I started with rechargable AAA, then moved to the LiPo battery).

The one thing I can suggest on the speed control is to just be sensitive to the dial position. I think the method PF uses to transmit speed change signals makes the most sense from a design aspect. One thing they could do is change the controller, so instead of a dial (with no real meaning to the actual dial position), they could using a up/down push button. This would eliminate some of the "skipping" of speed change signals (if you turn it too fast). Every time you push the button, it sends the speed up/down signal to the train.

[fred67]

I think that's a very intentional design decision, it means the remote doesn't need to be constantly sending out signals in order to ensure the reciever has picked up the correct speed setting.

As to the longer term plans, I'm hoping TLG come up with a Bluetooth controller for Power Functions, although I imagine the difficulty there would be finding a way to reliably and simply pair with a controller.

It wouldn't have to repeatedly send out signals... instead of saying "go plus 1" it would say "go speed x" and the train would continue at that speed (like it does now) until getting different input.

That does mean that you might turn it up or down and the train doesn't change speed because it didn't receive the signal, but then that's what happens now anyway. An extra button to resend the signal would be better.

Better still would be, as you suggest, bluetooth - or something else. Anything else, really.

[Toastie]

It wouldn't have to repeatedly send out signals... instead of saying "go plus 1" it would say "go speed x" and the train would continue at that speed (like it does now) until getting different input.

That is true, but the dial hardware on the remote may not allow this. This one just says: I was turned clockwise/counter clockwise and the electronics has to detect that. For a dedicated setting (speed X) it needs to know the absolute position of the dial, which is more elaborate to realize. And kids cannot go nuts on the dial because there are two end points ...

Well I am guessing, maybe wrong.

With respect to RF: I don't know but I could live with plain vanilla 1:1 IR to RF mapping with just a little checking on pulse length maybe to minimize RF noise issues - that would be a simple RF receiver and maybe an 8-bit super low cost PIC or something.

Best regards,

Thorsten

[skaako]

That is true, but the dial hardware on the remote may not allow this. This one just says: I was turned clockwise/counter clockwise and the electronics has to detect that. For a dedicated setting (speed X) it needs to know the absolute position of the dial, which is more elaborate to realize. And kids cannot go nuts on the dial because there are two end points ..

Hi Thorsten,

Yes you are correct about the PF Train Remote, it only sends to the receiver for the channel to step +1 or Step -1 with the dial, and to stop with the button. Although this type of transmitter operates in this way, the PF Receiver has the ability to receive commands in many different modes.

From reading the Lego PF Document it describes in detail all aspects of the RC Protocol.

The different modes are:

Extended mode

This mode is able to control: Brake, increment and decrement PWM in 7 steps on Output A and toggle Forward/Float on Output B. Toggle bit is verified on receiver. No timeout for lost IR.

Combo direct mode

This mode is able to control: Two outputs float/forward/backward/brake.

This is a combo command controlling the state of both output A and B at the same time.

Toggle bit is not verified on receiver.

This mode has timeout for lost IR.

Single output mode

This mode is able to control: One output at a time with PWM or clear/set/toggle control pins.

Toggle bit is verified on receiver if increment/decrement/toggle command is received.

This mode has no timeout for lost IR on all commands except “full forward” and “full backward”.

Combo PWM mode

This mode is able to control: Two outputs with PWM in 7 steps forward and backward.

This is a combo command controlling the state of both output A and B at the same time.

Toggle bit is not verified on receiver.

This mode has timeout for lost IR.

I assume this PF Remote must operate in Single Output Mode, as it can increase/decrease the PWM and also it has no timeout for IR. Also this mode is capable of directly sending the power value directly to the receiver ie. Set Channel A power to Step 3 Forward etc. Therefore it would be possible to make a IR transmitter that was capable of sending speed values to the receiver. The document contains all the details of the IR protocol too.

As for using RF transceivers.. now that's what i'm talking about :) 2 way communication with trains, excellent non-LOS range. I can't wait until i can get back with playing around with my trains again :)

Thanks

Mike

[Toastie]

Yes you are correct about the PF Train Remote, it only sends to the receiver for the channel to step +1 or Step -1 with the dial, and to stop with the button. Although this type of transmitter operates in this way, the PF Receiver has the ability to receive commands in many different modes.

From reading the Lego PF Document it describes in detail all aspects of the RC Protocol.

Hi Mike,

thank you very much for the info! I was just guessing from reading the PF document you referred to. Actually there is an updated firmware version for the PF controllers TLC called V1.10; Philo has that as well on his website. And now we have full access to all 16 channels ... apparently as of late 2009, but I don't see any current remote capable of handling this.

With a lot of help from Xander I was able to get access to all the modes you were mentioning with the HiTechnic NXT IR device. All 16 channels. Kinda cool. Expensive though of course (NXT + HiTechnic "sensor"); when it comes to cost/benefit decisions, all LEGO solutions are - well - unreasonable, I guess.

Transceiver wise: I think with a custom "receiver only brick" (RF to IR), put in front of a PF receiver, we'd get a lot of freedom. On the remote (TX) site, we'd just need the TX part of a "TV remote extension" RF device (about $20, since you are into cost efficient solutions ). The transceiver approach would require to design a new PF intelligent brick as well. May be a little tough.

Yeah, this thread is fun ...

Regards,

Thorsten

[skaako]

Hi Thorsten, your welcome about the info :) I just remembered i had a copy on my hard drive and when i posted the link from a quick search, i realised i have given you an older version. Then i saw the version i have on my hard drive is 1.20 dated 26/02/2010. I found the link to this file here after a more thorough google search. I havn't had the time to see what differences there are from the 1.10 version, but it looks quite similar.

That is really interesting about the extra channels available, i have heard that it was possible to expand the channels but i havn't looked into how this is achieved. I thought from looking at the protocol it would be possible to extend to 8 channels as there is the extra 'address bit' that could be used. I would be interested in reading about it if you have some links? Your very right about the all Lego solution being quite costly.

I havn't needed to build a RF-IR transceiver for power functions yet, but i don't think it would be too difficult. I think it would be more easier than the RF link i made for the RCX.

I had such a headache working around the parity that is used for communications, but it turned out that was why the transceivers were so cheap at $20 each. They didn't create a parity bit as they were suppose to, so i had to calculate and send the bit with the ic.

Edited July 16, 2011 by skaako

[JopieK]

It is not hard to make an IR booster or such a link. I have done that already using atmel uC's.

[nesquik]

The trouble with any situation such as the 9V vs P.F is that it's almost entirely out of the consumers hands when it comes to R&D. Sure, the AFOL community can contribute their ideas and experience to Lego Ambassadors, but only TLG have the final say, and most of the time it's a decision based on economic advantages.

http://news.lugnet.com/announce/?n=3637

It's easy to understand why TLG made the transition to P.F. For the primary child market, it's offers a cheaper way to expand upon existing sets because, as you rightly pointed out, the all plastic track is far cheaper. Not having any wires outside of the train or controller is an added bonus for the ease of set-up, and the reduced cost of the motor (as well as options) is welcome.

But outside the youth system, P.F really is quite flawed. The rechargeable battery box is ridiculously priced, and even if you get the regular battery box, means that you are required to add two large components within the train housing, thus altering the design of the train. In the 9V system, the motor is just in the bogie, so it doesn't alter the design. And whilst the controls are more 'convenient' being wireless, they are indeed more clunky to use.

I don't see TLG altering their current strategy in the near future, because it does 'work' for their target audience. Regardless of what they say, AFOL's are sadly not in the front of their minds.

[tbroyd]

Although not perfect, I do feel that the ability to control PF trains does improve with practice - mainly how the control is held and pointed. I also only ever adjust the control one notch at a time and check to see if the train has responded before adjusting again

I rediscovered Lego too late for the 9v trains which is perhaps what has forced me to persevere with the PF system

[Frank STENGEL]

But outside the youth system, P.F really is quite flawed. The rechargeable battery box is ridiculously priced, and even if you get the regular battery box, means that you are required to add two large components within the train housing, thus altering the design of the train. In the 9V system, the motor is just in the bogie, so it doesn't alter the design. And whilst the controls are more 'convenient' being wireless, they are indeed more clunky to use.

I agree and I disagree on these points.

The rechargeable battery is indeed expensive. However the price for a correct (that is *safe* and holding many cycles) Lipo and protective circuitry is not that small and integration not that easy. Furthermore if one wants to go another route one can do a bit of soldering and use smaller (less powerful) standard lipos or 9V cells that very often fit where a battery pack would not.

The other thingie that could be construed as being large is the IR receiver. In my experience it isn't that enormous and very often blends with exterior detail (the same way it does on the 7939 locomotive for instance).

There is another thing that I remember from the 9V era and that is the time spent on cleaning the track, making sure that every single connection is good and that the voltage drop is minimal. PF eliminates this at the price of an integration that is more difficult.

Both systems have their pros and their cons. Neither is really that superior to the other. One last thing us AFOLS tend to forget : Lego trains are designed to be *toys* and are supposed to be easy to use by a 10 year old kid without adult supervision. with PF no problem, with 9V, well...

Edited September 23, 2012 by Frank STENGEL