First off, the drop across a diode is dependent on the current through the diode.  More current, more voltage drop.  Lightly loaded, the drop across a silicon diode is far less than when it's carrying more current.

You don't need more than 3A diodes for this task, unless you're really hauling some big loads with a large locomotive.

For the scheme to work, you need pairs of diodes back to back, as the polarity of the motor voltage changes with direction.  So, for each .5-.6 volt drop, you actually need two diodes.

What kind of locomotive is this that you don't have room for a few 3A diodes?

The 2 engines I am currently playing with are Lionel's Berk Jr. as in Polar Express. And a Lionel GP20 with Trainsounds. I can probably get the 3A diodes in both. I have managed a string of 5, well actually 2 strings of 5 each direction as in the thread on

J&C Studios O Gauge Archive.  

It made some difference but not as much as I'd like. I didn't know if different diodes have more or less drop than others. If not I'll just have to find a way to get more in there. Both these engines draw less than 1A on my layout. So I tried what I had on hand which was the 1N4003 1A diodes. This morning I strung 5 together as a test and connected it to a fixed 12VDC power supply and checked drop across each one, that's where I came up with the .36 to .37 drop. 12.24 VDC at input 10.41 VDC at end of string, with .36 to .37 drop at each diode.

I know the basic function of a diode, but looking at several electronic supply houses I see different types of diodes. This, and the specs, (beyond amp rating) are what lose me..........

Forgive the bolt type, not sure what happened here........

This is a perfect application for a higher voltage diode, as they'll also generally have a higher forward voltage drop.  I just checked three 1N4005 diodes with about 400ma going through them, all of them were between .8 and .85 volts dropped with that current.  Those are the standard ones I stock for general use.

A 1N5401 3A 100V diodes at 400ma gave me a .75 volt drop, the second exactly the same.

Two 6A4-T 6A 400V diodes at 400ma gave me a .73 volt drop, one was a tiny bit more, .74 volts.

I'm not sure why you're getting such low readings.

Thanks that's exactly what I need. I believe I can squeeze 3Amp ones in, not sure the 6 amp package will fit in that Berkshire. I was reading another thread on an electronics site, about using diodes a voltage drops. The ones this fellow was using probably similar to the ones I have, he gave a figure of aprox .4 drop per diode. I need some more diodes for several projects and I'll just get some of the numbers you are using.

When I want low diode drops, I normally buy Schottky diodes, they're a special type that has a lower forward voltage drop.  I'm impressed that the 1N4003 has such a low drop, I wouldn't have expected that.  I wish I had a couple to test, I'd like to see that in person.

If you were measuring the drop without a substantial current going through the diode, you will see only a small drop.  Add a 1 amp load and try it again.

The drop across a silicon diode varies exponentially with the current.  Usually a diode will drop .7-1.0 volts at full rated current.  For a 1N400X, that is about 1 amp.

Dale, all the different types I had readily available had at least .7 volts at 400ma, even the 6 amp ones.  I couldn't find anything that had a .3 volt drop except a Schottky diode.

The voltage drop is related to the current density in the junction.  In other words, a diode with 10 time the junction area and 10 times the current passing through the junction would have about the same drop as a diode with an area of 1.  For the 1, 3 and 6 amp diodes, some of the rating increase is junction area, but the heatsinking of the package and leads is also a major factor.  The 6 amp probably doesn't have 6 times the junction area of the 1 amp.

In addition to the exponential rise due to the semiconductor junction, there is also an ohmic drop through the bulk resistances of the various components.

I have some 160 amp diodes that I used in a power supply that I built in college (yes, they had electricity back then!), and their forward voltage drop at rated load is also slightly over a volt.

The 1N4002 is rated at about 1 volt max drop at 1 amp DC.  The drop will vary with temperature, going down in voltage as the component heats up.  As you indicated, higher voltage ratings can also increase the forward drop spec.

All good info Dale, and I don't see anything I disagree with.  Apparently he never loaded the diodes, I misread the second post as saying there was a load with that drop.

Yeah, that's why I asked. Well I did a test with the 1N4003's a 12VDC wall wart, and an 1157 bulb as a load. I got anywhere from .79 to .94 drop per diode. This is not what I wanted to find out. This means I'll need more by at least 2 pair to get where I want to be. It made some difference with 5 pair, but still starts to move before voltage is high enough to activate sound. The PE is close so I will probably have to live with it. I barely got the 5 pair in. Funny thing about the GP 20 with trainsounds, I have a similar, (same chassis, & boards) GP 38 and it takes 4 times, (or maybe more) to start to move the loco. Both close to new, and both have been lubed. The GP 20 appears to have the 2 can motors wired in series. Not sure on the GP 38. Any sugestions? or just more diodes. All draw under 1 amp at low to med speed with 8 or 9 cars, a mix of PW, Rail King, and RMT.

If one is wired in parallel, and the other in series, there will be a big difference in starting voltage.  Check your other unit.  This is confusing since the series connection should take more voltage to start, not less.

I was wrong on the electronics. They both have the same sound board but different reversing boards. Both appear to be wired in series. Hard to tell with the gaggle of wires, but it looks like the blue from one motor connected to the yellow of the other. But with the GP20 I believe I can get some more diodes in it. I don't guess it matters that much, but it bugs me when things don't work as they should. Thanks for the diode numbers though. And I needed more of higher amp rating anyway.

I believe that the blue and yellow wires are normally connected together since the motor assemblies are "pointed" in opposite directions on the chassis.  The question is whether one set of yellow and blue wires is connected together by themselves and are not connected to the circuit board or anything else.

Ok, I hadn't thought it out. But that would stand to reason. These are the first ones I've taken apart with a gaggle of wires. And not seeing a schematic it was very hard, (especially with my eyesight) to tell how they were wired. But I hadn't thought it thru. They would have to be wired like that otherwise they would just have a pulling contest with each other. There is something to be said for those simple PW locos. My grandson really likes the sound, Ok so do I..........

I suppose the other solution would be to install a reversing board (same as the GP38) in it. That one takes a lot more voltage to get moving, so the sound comes in before the loco starts to move. They are conventional and appear the same with the exception of the reversing boards.

I believe to connect them in series for those you'd connect the blue wires (or yellow wires) together alone, and then connect the other color wires to the two connections of the reversing board.  That will give you series operation.

I can't imagine the reversing board "taking a lot more voltage to get moving", you'll have to explain that to me.

Try John's series connection instructions.  That would certainly raise the starting voltage.

It's not the reversing board. It's getting the sound board started before the engine is moving. I am trying to slow down the motors (ie: more voltage to start the engine moving) there by allowing the sound to start before the engine actually starts moving. The PE is real close now. The GP 20 you can goose it to activate the sound board then drop the throttle (voltage) back down and the sound stays on. It's a minor irritation. If I had to live with it I would. But the "Tinkerer" in me wants to improve the performance.

I may try that. That would help the GP20, and I might be able to squeeze in a couple more diodes in the PE, it's really close now. Maybe just another 1/2 volt. I looked up the parts. Both the GP38 and the GP 20 use same motors, and sound boards, but the GP20 is original board and the GP38 is Rev 3. So other than the plugs that should be an easy fix (well 37.00 for board + plugs). I'll try again to find a wiring diagram had no luck the other day. Only found one for a Lionel with electronic rev. board, but that loco did not have sound.

I should clarify John. I do not know how much more actual voltage the GP 38 takes to get moving, it's about 4 notch's more on my controller than the GP 20, but it's more than enough to activate the sound board well before the engine starts to move. The GP 20 will move along at a rate of several feet per second but the voltage is not yet high enough to activate the sound.

And thank's to all for your replies.

I'm trying to imagine why there isn't room for a couple more sets of diodes, this is a dual-motored unit, right?  You can also use bridge rectifiers as two diode pairs if they'll save room.  Here's a 3A model and the dimensions.

3A Bridge Rectifier at Digikey

Yes the GP 20 is dual can motors. The PE is single motor, it's just tight inside. I can get a couple more pairs into the GP 20 but with as little effect 5 pairs have I was afraid I'd have to double that. That one has a reversing board that is three times the size of the Rev 3 board in the GP 38. I'll probably tinker with it some later. I think most of the problem is that the amp draw is so low, causing the voltage drop to be minimal. I couldn't figure out the wiring without taking it apart more, so I put a string of 5 pair in each motor. If I can figure out the wiring where I can put 1 string in and have both motors pull thru it, it should up the amp draw, also will give more room for additional diodes.

Well, regardless if they're in series or parallel, you should be able to run both motors through one set of diodes.  This will do two things, double the current draw, and half the number of diodes required for the job.

If you check, I'm pretty sure you'll find that at least one lead from the motors is common for both, break it there and insert the diode string there.  If they're in parallel, usually both motors are directly in parallel.  I don't know that specific engine, but I've seen lots of them with the motors connected directly to the PCB, and they're connected together on the PCB.

Break out the ohmmeter, or just remove the PCB from the mount and check the underside for the copper traces.

What's the exact model numbers of these two units?

Dale and John I must say you guys are a tremendous asset to the forum.  I'd love to see the two of you write a "How To Do It" booklet covering 3 rail electronics.  Thanks for being here for us.

'

Ok well I'm tinkering...... I got the wires untangled where I could see where they go. Had to unsolder a few and reroute and resolder. But then I was able to trace the wires. This was the reason I just put a set in series with each motor, (I couldn't tell where the wires went. Too many blue wires. Good to go on this count.

First I wired the motors in series, good results on start, sound on one to two notches before engine movement. However top speed a little slow.

Second wired a set of 5 pairs in red wire off rev board, (in series with both motors). Result fairly good, slightly touchy between sound and movement, but still better than ever, and close to performance of GP 38.

Third, (just because I already had them made up) tried 10 pair. Right about where the GP 38 is (ie sound to movement), but maybe a little slow. Still probably almost as fast as I would want to run them.

Probably will try 8 pair. Now I'm playing with another thought. Just a little faster than 10 pair, and maybe I can hit the magic number as a speed limiter. I mean while I'm at it, that way if my 4 year old grandson cranks it wide open it won't fly off the track. He doesn't like it when they go fast, I think he is afraid they will crash.........ME Too.

The Specific numbers of these locomotives,

GP 38 6-28878 Dale Earnhartd Jr. (rebodied to L&N)

GP 20 6-38504 BNSF

Both dual can motors, and Trainsounds.

Oh and thanks again to both John and Dale for some helpful sugestions. I learned a little more about diodes as voltage reducers.......... to both of you.

I think you're on the road to a solution, glad we could assist in the effort.

I discovered the difference between 5 pair and 10 pair, as related to the voltage it takes to start the engine moving is not noticeable. This I suspect this is due to low amp draw at that point. But It does change the top speed. With 8 pair the top speed is faster than I want to run it but not fast enough to derail at top end due to excessive speed. At some point I may put the current board in it (Rev 3). I have another project I can use that board in. I really liked the way it behaved wired in series, if the top speed was just a little faster. But what fun would it be if you just left everything alone...................

You have it, tinkering is half the fun.

And John and I are the mentors encouraging tinkering.