Musings on a 'real' PS3 supercap mod. Update - successful bench test.

I'm going to get a bunch of supercaps and boost regulators and see if I can convince a PS/2 or PS/3 locomotive to skip some dead spots.  Next would be Lionel Legacy, that may be easy or hard.

@sinclair  I have a few Legacy locos.  I've mostly tested them in command mode with the Legacy base.  It's been several years, but I'm pretty sure if you press DIR when running at speed, they stop gradually (at least a foot of continued operation) and then shift into "neutral" awaiting further commands.  I think there's a momentum setting that's programmable on the Legacy CAB-2.  Again it's been a while, but I always found their operation to be forgiving and kid-friendly.

One of my locos is the 11203 Berkshire.  Like the Vision Hudson, this has TWO flywheels and back-drivable gears.  In one specific case someone turned off the track power while it was running near full speed with a train.  It coasted at least two feet!

DCS also has programmable acceleration and deceleration settings.  From what I recall, they come from the factory set at 4, 2 and I usually change them to 2, 1.  On PS2 this worked pretty well but I've never tested a PS3 model in this regard.

sinclair posted:

Problem with pushing the direction button on the remote is that it's a function built into the electronics, so they will behave as they are programmed to do, which on MTH PS-3 running in DCC is to slow down and then speed up in the other direction.  Legacy will almost come to a complete stop with very minimal coasting.  So to run your test you truly need to cut track power.

Yep, the only way to verify is to cut track power.

Unfortunately, we are at the mercy of the control electronics in terms of the resulting behavior. As I mentioned above, when I cut power while running on analog AC, the PS3 board stopped the motor immediately. When running on DCC, it ran for almost 10 seconds.

On the ESU Loksound decoders, for example, this behavior is configurable up to 4 seconds, but with the stock PS3 (or other!) electronics, we have no such option.

gunrunnerjohn posted:

I'm going to get a bunch of supercaps and boost regulators and see if I can convince a PS/2 or PS/3 locomotive to skip some dead spots.  Next would be Lionel Legacy, that may be easy or hard.

I'm pretty sure at this point the PS3 will work. Legacy, well.... let us know. 

I'm thinking on how to charge the capacitors in the first place.  I can't charge them from track power without a totally isolated power source as the DCS or Legacy electronics doesn't have a common DC ground.  One possibility is to charge them from the same place I'm providing the power when track power is gone, but I have to deal with the issue of not charging when the caps are powering the electronics.

gunrunnerjohn posted:

I'm thinking on how to charge the capacitors in the first place.  I can't charge them from track power without a totally isolated power source as the DCS or Legacy electronics doesn't have a common DC ground.  One possibility is to charge them from the same place I'm providing the power when track power is gone, but I have to deal with the issue of not charging when the caps are powering the electronics.

Hey John, the caps will charge as long as the engine is powered, provided the output voltage of the boost regulator is set correctly, as I describe in the other thread. Since they are powered after the rectifier in the decoder it doesn't matter if the track power is AC, DCS, DCC, or whatever. The power/ground for the keepalive is internal to the decoder.

Am I misunderstanding your question?

Tested with another engine, on the track and with a train this time. You can see in the video when I cut the power by when the green light in upper left goes out. About 9 seconds runtime.

Here are a couple pics of the rectifier from the ES44:

Here's the datasheet from mouser:

https://www.mouser.com/datashe...(GBU)-V1-1626282.pdf

So if you can see the corner with the chamfer, you don't even need to remove the heat sink or test with a voltmeter - that is the positive terminal. The one opposite that is ground.

I wonder what happens if the smoke is on? Must die sooner?

I taught our grandson about smoke. Now he won't run without it. I just need enough time to get across my dead #7 Atlas 2 rail frogs. The solution would be to figure out a minimum speed to make it with the keep alive and smoke on. I didn't have problems when I only had #5 switches. So I think I just need a minimum time addition out of this. Right now the trains are rocking as each engine stalls for a second.

Could you do a quick test with the smoke on? Even 2 or 3 seconds would be great.

Well, you potentially have a Catch-22 situation with the charging of the caps.  The caps are charged whenever there is voltage on the output side of the rectifiers in the locomotive electronics.  However, there's always going to be voltage there until the caps are discharged.  I thinking on how to keep the charging circuit inactive when the caps are supplying the power, kinda silly to try to charge them from their charge.

gunrunnerjohn posted:

Well, you potentially have a Catch-22 situation with the charging of the caps.  The caps are charged whenever there is voltage on the output side of the rectifiers in the locomotive electronics.  However, there's always going to be voltage there until the caps are discharged.  I thinking on how to keep the charging circuit inactive when the caps are supplying the power, kinda silly to try to charge them from their charge.

I've done it! I've created a perpetual motion machine! 

Being a realist, I want to avoid trying to create such a machine as I know how the story ends!

gunrunnerjohn posted:

Well, you potentially have a Catch-22 situation with the charging of the caps.  The caps are charged whenever there is voltage on the output side of the rectifiers in the locomotive electronics.  However, there's always going to be voltage there until the caps are discharged.  I thinking on how to keep the charging circuit inactive when the caps are supplying the power, kinda silly to try to charge them from their charge.

On a more serious note... 

They do actually 'recirculate' a little charge when running off the caps. However, the charge current is limited to 50mA or less when the caps are charge to a usable level, so it's a minor effect considering the total energy available.

What is detrimental though, is if the boost regulator voltage is set too high - then the keepalive will be trying to power the decoder during normal operation, and the cap will never charge.

So... who are we mailing our engines to?  Or who will be making/selling a drop in module? 

Engineer-Joe posted:

I wonder what happens if the smoke is on? Must die sooner?

I taught our grandson about smoke. Now he won't run without it. I just need enough time to get across my dead #7 Atlas 2 rail frogs. The solution would be to figure out a minimum speed to make it with the keep alive and smoke on. I didn't have problems when I only had #5 switches. So I think I just need a minimum time addition out of this. Right now the trains are rocking as each engine stalls for a second.

Could you do a quick test with the smoke on? Even 2 or 3 seconds would be great.

Good question, I just tested this. Interestingly, with the smoke on, the engine stopped after about 3 seconds, despite plenty of charge remaining in the capacitor. So it's not a question of energy storage, although with smoke on it of course draws more.

We are probably hitting a limitation of the boost regulator in this case. It is rated for 2A input or output. Using my estimated power of 7.5W from the other thread, and adding a few watts for the smoke, we would be at 10+ W. Which, at 5V input, would be at or above the 2A limit of the regulator. It is thermal and overcurrent protected, so my guess is the thermal protection kicking in after a few seconds.

Since the boost regulator is powered from a low voltage and is outputting a considerably higher voltage, you're probably current limited on the input side.  FWIW, O-scale smoke typically uses 5-8 watts, not 2.5 watts.

Found a solution to tapping directly into track power for the charge, an isolated step down converter.  Add a bridge rectifier and cap, and all is well.  This eliminates the issue of the perpetual motion charging.

gunrunnerjohn posted:

Found a solution to tapping directly into track power for the charge, an isolated step down converter.  Add a bridge rectifier and cap, and all is well.  This eliminates the issue of the perpetual motion charging.

Wondered if that would be an answer. They can be quite small and fit in.

This one is 1" x 1" x 1/2".  Not as small as I'd like, but getting true isolation takes real estate.

gunrunnerjohn posted:

Found a solution to tapping directly into track power for the charge, an isolated step down converter.  Add a bridge rectifier and cap, and all is well.  This eliminates the issue of the perpetual motion charging.

 

Unless you're worried about the universe ending as a result, this is a solution looking for a problem. The lost energy due to charging while unpowered is around 3-5% of the energy stored in the cap. I mean, we're talking demonstrated 8 to 10 seconds runtime as it is, in a tidy two-wire package.

But if it's keeping you up at night...  

Exactly what rate are you charging the caps at?  The charge rate would determine the losses.

Well it's 5V through a 56 ohm resistor, so the charge current will of course vary depending on the capacitor voltage. When the keepalive first starts discharging, the 'recirculated' current will be essentially zero, and will increase as the cap discharges.

But for easy figuring, let's assume a worst case constant charge current into the cap at 2.5V (pretty typical for the 2.5F version at the end of a long boost). That means the power dissipated in the resistor (the main source of energy loss) is 110 mW. If we assume, again worst case, a 10 second run time, that's 1.1J energy lost. From the energies tabulated in my other post, the 2.5F version stores 26.3J. So that's ~4%, as a worst case bounding value. The actual loss will be less.

It would be elegant if we had a constant current charge regulator that could charge at a fixed 100mA or so. But, it would have to be small, run off a wide range of input voltages, and be configurable for the required charge current and cutoff voltage. Then the loss would be only the inefficiency in the regulator itself. But unless the implementation of that was as cheap and simple as the regulator, I wouldn't trade the complexity for a couple percent efficiency. Just use the 5F version if you have some monster that needs more runtime, lol.

It'll take a long time to recover the charge on a large supercap through a 56 ohm resistor.  When I did my YLB, I charge the 1.5F 5V supercap using a 10 ohm resistor.  This was, of course, directly from track power, so I wasn't worried about the available power. 

Engineer-Joe posted:

I wonder what happens if the smoke is on? Must die sooner?

I taught our grandson about smoke. Now he won't run without it. I just need enough time to get across my dead #7 Atlas 2 rail frogs. The solution would be to figure out a minimum speed to make it with the keep alive and smoke on. I didn't have problems when I only had #5 switches. So I think I just need a minimum time addition out of this. Right now the trains are rocking as each engine stalls for a second.

Could you do a quick test with the smoke on? Even 2 or 3 seconds would be great.

Hey Joe, I meant to follow up on this earlier and got sidetracked.

Check out this 8A boost regulator as an alternative to the one suggested in my other post. This would easily run an engine with smoke for as long as the capacitor had charge. It's a little bigger and more expensive, but if you gotta have smoke...  

I like the idea. I was thinking that if it was too hard, or got too expensive, it would be better to just power the frogs on my layout. It's mainly the #7.5 Atlas switches giving me issues. There's another area where a couple of switches are close together that seems to be acting up too. I think I have a dead rail for a few inches? I got tired of stuff not running right and quit for awhile.

 Ever since I ballasted my layout, things have changed. I need to work towards how smooth everything used to run. All of a sudden it seems like things went south. Running toy trains has me getting used to this type of thing. I don't jump at it in any direction. I have to methodically attack it to get it right. If not, I have to do it again.

 I found a few of my steamers all of a sudden won't run smoothly. I think running over glued track with the smoke on, has over taxed some power pick-up points and they need attention now. I found my 1 MTH Dreyfus wouldn't run with a train. I noted that it only got power from one rail, on one set of tender wheels. That point wasn't working right so the engine kept stalling. The other rail gets picked up by both the tender, and the main drivers of the engine. When the power pick up is spread out, each point seems to hold up better. When smoke is on, power points fail easier.

 Stuff like that has me pulling my hair out because the grandson just wants to run everything all at once. I stop to look at why and he's already on the fifth or sixth train. He finally pushed me too far one day and I stopped everything. We must have run around 25 different trains. Six engines went down and I had enough. He wanted more!

How's that for a simple reply??? 

Engineer-Joe posted:

Stuff like that has me pulling my hair out because the grandson just wants to run everything all at once. I stop to look at why and he's already on the fifth or sixth train. He finally pushed me too far one day and I stopped everything. We must have run around 25 different trains. Six engines went down and I had enough. He wanted more!

Sounds like a serious situation! 

Totally agree though, engines stopping when they shouldn't just really kills the illusion for me. Hence this thread. The keepalive is a great option to have available to help with that, although it is really my third line of defense, after trackwork and engine continuity.