ATLANTIC CENTRAL I solder all my rail joints, and glue my track with adheasive caulk, so the springyness of ATLAS flex has never been a problem. I do use a spike, or nail, and some weights, as needed during installation. When I was young, I hand layed track and built my own turnouts. But when ATLAS code 83 came out, I switched. I now reserve those skills for special trackwork only when needed. I have run on layouts with code 70 rail, I have hand layed code 100 thru code 70. Code 70 can be "fragile"........ For me, code 83 everywhere is an acceptable compromise. I use paint, scenery, ballast and ballast profile to create the difference between the mainline, a yard, and a siding. I take no issue with those who want to be more exacting, I was like that when I was a younger modeler. I'm a lot more "casual" now. And I ditched all the code 100 30 years ago......there is no real saving track that is hand layed. Sheldon
I solder all my rail joints, and glue my track with adheasive caulk, so the springyness of ATLAS flex has never been a problem. I do use a spike, or nail, and some weights, as needed during installation.
When I was young, I hand layed track and built my own turnouts. But when ATLAS code 83 came out, I switched. I now reserve those skills for special trackwork only when needed.
I have run on layouts with code 70 rail, I have hand layed code 100 thru code 70. Code 70 can be "fragile"........
For me, code 83 everywhere is an acceptable compromise. I use paint, scenery, ballast and ballast profile to create the difference between the mainline, a yard, and a siding.
I take no issue with those who want to be more exacting, I was like that when I was a younger modeler. I'm a lot more "casual" now.
And I ditched all the code 100 30 years ago......there is no real saving track that is hand layed.
Sheldon
Thanks for the tip on the code 70. The reason I'm considering it is because my closest LHS, stocks ME code 70 and had a good pile of flex and #6 turnouts. Its no more expensive than more popular code 83.
Honestly, I couldn't tell the difference between 70 and 83 even upon close inspection, but he had all the supplies right there at close to internet prices so it would be easy to just order more and pick it all up at once and have it. The difference might show up in photos if I ever decided to do the internet pics thing.
But if there are downsides to it, I wanna know.
- Douglas
Yeah, I have am using code 70 also in yards and branch lines as well and all the code 70 I am aware of is stiff track, so I adapt where necessary.
On curves with the springy flex track, I nail down all but about the last 8 to 10 inches. I tack it over so that last part is about straight, then solder on the next piece and it all bends like one long smooth piece of track. For turnouts, I don't solder it but I tack it down using ME spikes to hold it's shape, that seems to work very well. On the other end one rail sticks out and I trim it flush with the other rail as it would be fastened down in that orientation. Yes, if track stayed bent, it would be easier in one way but for bending nice smooth curves, it's a lot of massaging it. I have both stiff (code 70) and springy (code 83 Atlas) I've worked with. I guess there is no perfect world.
Rio Grande. The Action Road - Focus 1977-1983
riogrande5761 The springy nature of Atlas flex track is what makes it so easy to use. You can bend it to a nice smooth curve without having to massage it over and over and over. The springy nature is no problem if you use track nails or spikes because you can spring it nice and smoothly to fit the centerline and then nail or spike it down. If you object to the nail heads, then no problem, after the track is ballasted pull them out. And Bobs your uncle as the Brits say.
The springy nature of Atlas flex track is what makes it so easy to use. You can bend it to a nice smooth curve without having to massage it over and over and over. The springy nature is no problem if you use track nails or spikes because you can spring it nice and smoothly to fit the centerline and then nail or spike it down. If you object to the nail heads, then no problem, after the track is ballasted pull them out. And Bobs your uncle as the Brits say.
Yeah, I've used nails and tacks to bend track even when adhesive was the ultimate security, but it was still a hassle for me to keep the one end bent and secured while positioning the rest of the track straight.
The best way for me was to solder the flex to the diverging leg of the turnout, that makes the one end very secure and steady and it also allowed the sliding rail to slide down the proper end.
I was just thinking that track that stayed bent at one end would be easier to work with and position it for aligning spurs and sidings.
riogrande5761 Doughless My stock #6 insulfrog Peco code 83 turnouts make a yard ladder with dead frogs that are nearly exactly 8 inches apart. My six axle Kato SD38 fits nicely in the live space at about 7.8 inches truck end to truck end. The stock Pecos have little excess track beyond the frog, so the centerline of the yard tracks are about 1.75 inches apart. Adding more tangent track to widen the centers to 2 inches would of course widen the dead space. I won't be buying any track for a while. I have diesels of varying lenghts from SD40T-2's, to SD40's to SD45's, to GP40's, GP30's and many others. If I had a row of dead frogs, might just be my luck I might have the same problem as "he who is not to be named". As for Peco code 83, if money were no object I'd go cadillac. I might as well go Micro Engineering code 83 #6 which I'm hearing a lot of good things about and they cost 7 dollars less each, which adds up. Back to staging, I have a lot of code 100 I want to re-use, which is why I am strongly leaning toward Peco code 100 large, to replace the ancient Atlas code 100 #6. The choice there of course is insul frog vs. electo frog - 2 dollars difference and proponents for each.
Doughless My stock #6 insulfrog Peco code 83 turnouts make a yard ladder with dead frogs that are nearly exactly 8 inches apart. My six axle Kato SD38 fits nicely in the live space at about 7.8 inches truck end to truck end. The stock Pecos have little excess track beyond the frog, so the centerline of the yard tracks are about 1.75 inches apart. Adding more tangent track to widen the centers to 2 inches would of course widen the dead space.
I won't be buying any track for a while. I have diesels of varying lenghts from SD40T-2's, to SD40's to SD45's, to GP40's, GP30's and many others. If I had a row of dead frogs, might just be my luck I might have the same problem as "he who is not to be named".
As for Peco code 83, if money were no object I'd go cadillac. I might as well go Micro Engineering code 83 #6 which I'm hearing a lot of good things about and they cost 7 dollars less each, which adds up. Back to staging, I have a lot of code 100 I want to re-use, which is why I am strongly leaning toward Peco code 100 large, to replace the ancient Atlas code 100 #6. The choice there of course is insul frog vs. electo frog - 2 dollars difference and proponents for each.
I'm probably going code 70 for my branch line layout this time, and ME seems to be more available than Walthers code 70 or Peco 75.
DoughlessMy stock #6 insulfrog Peco code 83 turnouts make a yard ladder with dead frogs that are nearly exactly 8 inches apart. My six axle Kato SD38 fits nicely in the live space at about 7.8 inches truck end to truck end. The stock Pecos have little excess track beyond the frog, so the centerline of the yard tracks are about 1.75 inches apart. Adding more tangent track to widen the centers to 2 inches would of course widen the dead space.
I won't be buying any track for a while. I have diesels of varying lengths from SD40T-2's, to SD40's to SD45's, to GP40's, GP30's and many others. If I had a row of dead frogs, might just be my luck I might have the same problem as "he who is not to be named".
As for Peco code 83, if money were no object I'd go cadillac. I might as well go Micro Engineering code 83 #6 which I'm hearing a lot of good things about and they cost 7 dollars less each, which adds up.
Back to staging: I have a lot of code 100 I want to re-use, which is why I am strongly leaning toward Peco code 100 large, to replace the ancient Atlas code 100 #6. The choice there of course is insul frog vs. electo frog - 2 dollars difference and proponents for each.
rrinker I'm just the other way - just as I am plannign the biggest layout I've ever built, too. But it's not about buying cases of turnouts at once. I've always used Atlas. Never had a problem. They've worked very well in the last two layouts I built, even without poweing the frogs. I even added the wires for frog power, but never hooked them up because none of my equipment had any power issues. It's a bit harder to power the Atlas frogs - you need to use a brass screw in the hole provided int he frog casting because the pot metal is not readily solderable. Peco already has a wire soldered to the frog for power. The main reason I am going Peco this time - Peco 83 has a wider variety of turnouts, and I've tried my hand at handlaying turnouts and it's just not for me. So it would be Atlas + hand lay anything Atlas doesn;t amke, or pay more than the price of a Peco for someone else to hand lay what I need, or go with Peco and very likely not have a situtation that can't be handled by an off the shelf piece. Also the Peco to me looks better, there's more detailing cast in. ANd their flex track is a perfect medium between Atlas and ME. It bends easily, more like Atlas, much better than ME, but it doesn't spring back as easily as Atlas. That was my main stickign point, I did not decide to go Peco until I had a few pieces of their flex track to try out. --Randy
I'm just the other way - just as I am plannign the biggest layout I've ever built, too. But it's not about buying cases of turnouts at once.
I've always used Atlas. Never had a problem. They've worked very well in the last two layouts I built, even without poweing the frogs. I even added the wires for frog power, but never hooked them up because none of my equipment had any power issues.
It's a bit harder to power the Atlas frogs - you need to use a brass screw in the hole provided int he frog casting because the pot metal is not readily solderable. Peco already has a wire soldered to the frog for power.
The main reason I am going Peco this time - Peco 83 has a wider variety of turnouts, and I've tried my hand at handlaying turnouts and it's just not for me. So it would be Atlas + hand lay anything Atlas doesn;t amke, or pay more than the price of a Peco for someone else to hand lay what I need, or go with Peco and very likely not have a situtation that can't be handled by an off the shelf piece.
Also the Peco to me looks better, there's more detailing cast in. ANd their flex track is a perfect medium between Atlas and ME. It bends easily, more like Atlas, much better than ME, but it doesn't spring back as easily as Atlas. That was my main stickign point, I did not decide to go Peco until I had a few pieces of their flex track to try out.
--Randy
Randy, you raised a point with ME flex track. I'll probably go with ME code 70 for everything on my new layout, so I was wondering what your experience is with working with it.
My experience with Atlas flex is that it was always a bit of a hassle to bend the very end of it for spurs and sidings since it wants to snap back straight. Something that stays bent might be a little easier for that task. Of course, Atlas would be advantageous for making broad sweeping curves.
So I was wondering just how difficult working with ME flex track is?
Randy,
The choice of PECO flex track is the best reason yet presented to choice PECO turnouts, considering the differences in tie thickness, rail profile, etc.
Not sure what you mean by better selection?
Yes they have a #5, and yes they have one curved turnout. I never saw the #5 thing as a big deal. I built a yard years ago with Atlas #4 (4-1/2), it was ok but since then #6 is my minimum except industrial trackage.
I can make that curved turnout, 36" inside radius, 60" outside radius, from an Atlas #8. Cut some webs in the ties and give it a little flex, works great.
Has PECO made and code 83 slips yet? I use Walthers when I need those, they work well with ATLAS flex track.
But again, if you like the PECO flex, then go for it. I do agree about the detail, but after paint and ballast, not enough to make me spend the money or do the extra work to the turnouts.
riogrande5761 I'll re-post the info that went out with the bath water. If you have a series of turnouts in a row, such as in a yard ladder, and they happen to be spaced at a certain distance, unpowered or plastic frogs can essentially create dead spots if an engine is picking up power at those locations it may have a greater tendency to stall. If you install keep alive capacitors - dead spots are far less of a problem because engines can operate through dead spots with the power from the capacitor. In my case, I am planning later this year to rebuild a staging yard, which will have ladders, and because it will be in a semi-hidden lower level, appearance is not important. Thats why code 100 track is planned for that area - it is more durable although not as fine. Generally code 100 costs less and also I have a lot of track I can re-use, which also keeps cost down. If appearance is important, code 83, discussed above, may be something you would find worth it to use. Some of the most common code 83 turnouts include Atlas (lowest cost), Micro Engineering (mid level cost), Walthers made by Shinohara (higher end cost) and Peco (highest cost). As some have pointed out, the Atlas code 83 #6 are pretty good for the money - they don't have the stock rail points but if you weather them they blend in pretty well, as Rob Spangler has demonstrated in yards where he used Atlas code 83 turnouts.
I'll re-post the info that went out with the bath water. If you have a series of turnouts in a row, such as in a yard ladder, and they happen to be spaced at a certain distance, unpowered or plastic frogs can essentially create dead spots if an engine is picking up power at those locations it may have a greater tendency to stall.
If you install keep alive capacitors - dead spots are far less of a problem because engines can operate through dead spots with the power from the capacitor.
In my case, I am planning later this year to rebuild a staging yard, which will have ladders, and because it will be in a semi-hidden lower level, appearance is not important. Thats why code 100 track is planned for that area - it is more durable although not as fine. Generally code 100 costs less and also I have a lot of track I can re-use, which also keeps cost down.
If appearance is important, code 83, discussed above, may be something you would find worth it to use. Some of the most common code 83 turnouts include Atlas (lowest cost), Micro Engineering (mid level cost), Walthers made by Shinohara (higher end cost) and Peco (highest cost).
As some have pointed out, the Atlas code 83 #6 are pretty good for the money - they don't have the stock rail points but if you weather them they blend in pretty well, as Rob Spangler has demonstrated in yards where he used Atlas code 83 turnouts.
Jim, that advice was certainly logical, but I thought I'd provide some FYI with actual first hand info before you bought anything.
My stock #6 insulfrog Peco code 83 turnouts make a yard ladder with dead frogs that are nearly exactly 8 inches apart. My six axle Kato SD38 fits nicely in the live space at about 7.8 inches truck end to truck end. The stock Pecos have little excess track beyond the frog, so the centerline of the yard tracks are about 1.75 inches apart. Adding more tangent track to widen the centers to 2 inches would of course widen the dead space.
If an SD45 or F(P)45 is significantly longer than an SD38, the end of each truck might be on a frog, but then there would still be two wheels per truck in the live space.
Maybe an SD90MAC will have problems, if you plan on running something that long and modern on your layout. I would think steam locos with driver pickup wouldn't have an 8 inch gap between pickup points.
Following the logic, #4 turnouts would create a shorter dead space between frogs. Of course, the locos traversing the ladder would then tend to be shorter than an SD38.
Maybe compact #5 Pecos or other brands of #4s would cause problems for a GP loco. I dont have those to test fit.
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Deane Johnson Sorry Sheldon about omitting the rail size. It will be Code 83. I try to be thorough and my 82 year old brain definately capable of spacing things. Your comentary is excellent for my needs and I appreciate your taking time to write it. Deane
Sorry Sheldon about omitting the rail size. It will be Code 83. I try to be thorough and my 82 year old brain definately capable of spacing things.
Your comentary is excellent for my needs and I appreciate your taking time to write it.
Deane
Deane, you are most welcome, and Douglas added in some great points.
Regarding the ATLAS frogs, yes, sometimes they are a little high. I like to polish the blackening off of them anyway, so if they are a little high I just file them down.
In another recent discussion on here we went into great detail comparing PECO code 83 to ATLAS code 83, and the simple fact is that the PECO is smaller and has a slighly, only very slightly, tighter closure radius because the distance from the frog to the points is shorter.
Take care,
Doughless Deane Johnson Thank you Douglas. Another wealth of information that's very helpful to me at this point in my project. These posts really get it nailed down. I'm using Atlas flex track, so that's another vote for the Atlas. Your discription gave me a much better feel for the fit issues. I don't anticipate being concerned with photography, mostly overall appearence from a visual standpoint. I truly admire some of the awesome trackwork appearence that's being done, but I'll probably not focus that tightly on it. Thank you. If you're building a large layout, its probably going to resemble Sheldon's main line running layout rather than my smaller switching style of layout, so the Atlas is probably the way to go for you. The finger-flicking sprung nature of the Peco as well as the overall compactness makes them competitive for my situation.
Deane Johnson Thank you Douglas. Another wealth of information that's very helpful to me at this point in my project. These posts really get it nailed down. I'm using Atlas flex track, so that's another vote for the Atlas. Your discription gave me a much better feel for the fit issues. I don't anticipate being concerned with photography, mostly overall appearence from a visual standpoint. I truly admire some of the awesome trackwork appearence that's being done, but I'll probably not focus that tightly on it.
Thank you Douglas. Another wealth of information that's very helpful to me at this point in my project. These posts really get it nailed down.
I'm using Atlas flex track, so that's another vote for the Atlas. Your discription gave me a much better feel for the fit issues.
I don't anticipate being concerned with photography, mostly overall appearence from a visual standpoint. I truly admire some of the awesome trackwork appearence that's being done, but I'll probably not focus that tightly on it.
Thank you. If you're building a large layout, its probably going to resemble Sheldon's main line running layout rather than my smaller switching style of layout, so the Atlas is probably the way to go for you. The finger-flicking sprung nature of the Peco as well as the overall compactness makes them competitive for my situation.
There is no doubt that for a small ISL that the hand thrown over center spring is a good idea and PECOS are great for that application.
My industrial areas and yards have hand thrown turnouts, but in DC I want power routing and live frogs. So I mount sub miniature DPDT slide switches in the roadbed and make my own "spring rod".
The slide switches handle frog power and power routing, so without DCC, I need the electrical connections anyway.
When any turnout is thrown against a dead end track, that track is dead.
Several dozen slide switches being cheaper than 135 decoders........
The slide switches are easily operated with the same little screwdrivers I use for uncoupling.
In my yard, the power routing controls a bank of relays that provide an advanced power distribution that allows two crews to work the yard simply by route selection.
One crew from one end, one from the other end, two DC throttles, no block toggles, just select your route and go as long at it does not conflict with the other crews route.
Folks, drop the acrimony or I'll be forced to lock this thread. Your opinion is your opinion, not Unassailable Truth.
Also, we do not discuss that fellow or his website here on the MR Forums.
--Steven Otte, Model Railroader senior associate editorsotte@kalmbach.com
Deane. I built and operated a switching layout using both Atlas and Peco code 83 #6 turnouts (insulfrog). There are some differences between the two that may or may not matter to you. My personal opinion is that price probably outweighs any of the differences:
Pecos are more compact from point to frog. It will save about an inch of benchwork. Overall footprint of the stock Atlas turnout is much longer, but that is simply excess track beyond the frog that can be cut off. You'll be trimming other tracks with a Xuron rail nipper (hopefully), so trimmind an Atlas turnout to fit in tight space can be done. The important distance is from Point to Frog, and that can't be adjusted.
Pecos have sprung points, which means you can use a finger to throw the points and they will stay put. Atlas are unsprung, so they have the potential to...and often do...drift towards the middle as the vibration from the train moves the points. This can be solved by sliding a strip of styrene or cardstock under the throwbar to produce friction. If you use switch machines, the points will be held and the unsprung nature of the Atlas means its ready to go out of the box, where as with Peco you have to remove the "spring".
The frogs of Atlas switches are made from pot metal, and production realities sometimes cause the frog to be higher than the surrounding rails. They may need to be filed down to level, and that can cause scratching of the surrounding rails if you're not careful. I think the tolerances in production batches over the last few years are much better than some new-old-stock you might find. Pecos tend to have tighter engineering and more consistency, IMO.
The profile of Pecos tend to be different than Atlas, Walthers, or Micro Engineering. Whenever I mated a Peco turnout to Atlas flextrack, I had to crimp the Peco side of the rail joiner much more than the Atlas side. Easier to come loose if you don't do it correctly. Of course, Atlas products match well with each other. If I were to use Peco turnouts, I would want to avoid the profile difference pain-in-the-neck and would also want to use Peco flextrack, which has its own differences from Atlas flextrack in its bendability.
Peco makes a 6, and #5. Atlas makes a 6, a #4 (which is really a 4.5). Walthers Shinohara makes a true #4, (and a #5) and its mates with Atlas products better than the Peco. Myself, I use only #6 turnouts so matching a variety of products really isnt a concern for me. A steam or transition era layout might want to use 6s, 5s, 4.5s, and 4s, so I would exclude Peco products from the lot just because of the matchup issue (which is more of a pain than a real problem btw).
Atlas and Shinohara also make a greater variety of crossings and wyes than Peco, so there is more swapability and layout planning options...again...if the joiner crimping issue is a big deal for you.
Others report that Pecos have an engineering tightness that causes occasional shorting at the frog, which is resolved by a bit of nail polish. I have never ever experienced that, but I run Atlas and Proto diesels. Perhaps those who run steam locos with not a lot of play in the drivers have experienced more of this.
Again, a bunch of gives and takes that probably even out and become outweighed by price and availability, IMO.
Deane Johnson Just dropped in to this excellent discussion for the first time, and I have a question. I'm building the benchwoork now for a large layout with lots of switches (HO & DCC). With a higher number of switches, cost per unit becomes an issue. It appears to me that the Atlas Customline switches are signifianctly less than the Peco switches. I know the Atlas are larger and would require some redesigning of my track plan, but I can handle that. Anybody willing to post their thoughts on what I'd give up going from Peco to Atlas? And, what I would gain, if anything, other than reduced cost. Thanks for your thoughts. They are appreciated. Deane
Just dropped in to this excellent discussion for the first time, and I have a question.
I'm building the benchwoork now for a large layout with lots of switches (HO & DCC). With a higher number of switches, cost per unit becomes an issue. It appears to me that the Atlas Customline switches are signifianctly less than the Peco switches. I know the Atlas are larger and would require some redesigning of my track plan, but I can handle that.
Anybody willing to post their thoughts on what I'd give up going from Peco to Atlas? And, what I would gain, if anything, other than reduced cost.
Thanks for your thoughts. They are appreciated.
Deane, are you talking about code 83? or code 100?
Two completely different products with PECO.
PECO code 83 line is a straight frog, North American practice turnout, only very slightly smaller than the ATLAS Custom Line.
So, in code 83, the PECO has slightly better appearance, I will give it that. How much that matters after paint, weathering and ballast is up to the individual.
ATLAS Custom Line advantages in my opinion:
The metal frog is insolated but easily wired.
The throw bar is reversable and has no spring that needs removal for slow motion switch machines.
Out of the package they make up into 2" center crossovers and 2" center yard ladders without fillers or cutting.
They have feed through wiring which in my case is desired for my Advance Cab Control system.
The #8 version can be modified to make large radius curved turnouts very easily. The #6 can also be curved slightly.
As you noted, they cost much less.
Code 100:
The ATLAS product is nearly identical to their code 83 product - same features and geometry.
The PECO code 100 is completely different from their code 83 line. They are a continous curve, curved frog European style turnout offered in three "sizes", not frog numbers. The sizes relate to the radius, but it is not stated in inches, or mm.
They appear to be largely "in between" the North American sizes, with the medium being bigger than a #5 but not really a #6, and so on. This does give them a more compact footprint, which mat or maynot be a problem with some large equipment.
They have built in over center springs on the throw bar, and are offered insolated or electrofrog.
I don't think you give up anything with ATLAS, with the possible exception of some appearance issues, which paint and ballast can largely take care of.
I think this is at least partially decided by the age or quality of equipment the OP is using. Old equipment with few electrical pick-ups, or items of questionable quality may have issues. Fugate's example of wheels lining up perfectly over plastic frogs is possible, but given the size of the frog on the Peco turnouts, one would have at most two wheels not touching metal track. The potential for this to be an issue is one the OP needs to decide based on his equipment. For what it's worth, he admitted in his first post that he is "electrically challenged" so for simplicities sake, the insulfrog may be best. If he has concerns about his equipment, which may be part of the reason his previous layout had reliability issues, installing electrofrogs and either switches or frog juicers to power them should not be terribly difficult, but again, only the OP knows for sure how comfortable he is with that. I patiently await his decision and hope he shares and explains what brings him to his conclusion.
Mike
So on the electrical issue, I'm a live frog guy, so if I used PECO, they would be electrofrog.
As a DC operator with ATLAS custom line turnouts, all my frogs are metal and powered by contacts on the switch machines or by the miniature slide switches I use as ground throws.
Y'all stick with your Insulfrogs, I'm going with Electrofrogs.
richhotrainThe answer is:You want Insulrogs.
Jeff1952 So my question is...do I want Electrofrog or Insulfrog for my DCC locomotives?
So my question is...do I want Electrofrog or Insulfrog for my DCC locomotives?
Jeff1952 I'm still not sure I got an answer to my original question
I'm still not sure I got an answer to my original question
The answer is:You want Insulrogs.
Rich
Alton Junction
Thanks Bagal.
WOW! I had NO idea my "simple" question was going to devolve into all this. And yes, I did stay out of the discussion as it became more about geometry and less about electrical simplicity. I'm still not sure I got an answer to my original question, but thanks to all for your inputs....
richhotrain Doughless If you're running anything larger than a small loco, and at speeds greater than 5 mph, I can't see where stalling over a Peco Insulfrog is going to happen. Heck, you can even run locos at slower speeds than that over an Insulfrog as long as you have front and rear powered trucks. Rich
Doughless If you're running anything larger than a small loco, and at speeds greater than 5 mph, I can't see where stalling over a Peco Insulfrog is going to happen.
If you're running anything larger than a small loco, and at speeds greater than 5 mph, I can't see where stalling over a Peco Insulfrog is going to happen.
Heck, you can even run locos at slower speeds than that over an Insulfrog as long as you have front and rear powered trucks.
Considering that I build switching layouts mainly, I actually do run switcher locos at speeds less than 5 mph quite a bit. I'm probably one of the few members who actually have to worry about stalling over frogs as a matter of normal operations, and I've never had to seriously be concerned about it. Certain locos like the Bachmann S2 and S4 gave me fits. If I ever had a track mobile, or tenderless 0-4-0 or 0-6-0s, I'd probably power the frog on the few switches it might traverse (can you imagine a trackmobile running over an enitre layout?), but I have yet to have problems with normal locos running in situations they normally would run.
Atlas crossings with long frogs, short locos, and slow speeds are another matter.