E-unit rectifier, July CTT

[lionelsoni]

I just read the article "Installing a rectifier--close up" in the July, 2015, CTT.  It has some problems:

The text says to solder the + and - rectifier terminals to the e-unit terminals and to disconnect the coil wires and reconnect them to the ~ rectifier terminals.  This is just backwards from the way it should be.  Not only will it not work this way, but there will be a heavy load placed on the transformer, that is, until the rectifier burns out.

The photograph shows the right connections, except for a typo in the wire labels:  "Negative wire to black coil" should be "Negative wire to back coil wire".

The rectifier that he used is way overqualified at 400 volts and 4 amperes.  There's no electrical reason not to use it; but a 50-volt 1.4-ampere part the size of an aspirin tablet will work just as well and can be mounted by its leads directly onto the e-unit, alleviating the space and fragility problems that he mentions.

[servoguy]

Bob, I am missing something.   "The text says to solder the + and - rectifier terminals to the e-unit terminals and to disconnect the coil wires and reconnect them to the ~ rectifier terminals."  This sounds correct to me.  The + & - terminals are full wave rectified DC which should go to the E unit coil.  The ~ terminals should go to the center rail and the outer rails.  The problem I do see is connecting the E unit off/on switch.  It grounds one end of the coil and it appears the coil would have to be disconnected from the rivet of the switch.  

[cwburfle]

I had to read Bob's post several times.
According to Bob, in effect, the article instructs the reader to connect the AC terminals of the bridge rectifier to the E-unit coil leads, and the DC terminals of the bridge rectifier to power and ground via the e-unit terminal plate.
This is backwards, as he wrote.

I did not see the original article.

[lionelsoni]

It's on page 66.  Here is the relevant text from the article, with my comments in brackets:

"First, I disconnected the coil wires attached to the lug on each side of the E-unit.  I soldered the positive (+) wire to the back lug and the negative wire (-) to the front lug [Note that the lugs are still connected to the rest of the locomotive wiring, that is, to the rails eventually; so he's got the rectifier's output terminals connected to the source.], but it really doesn't matter which one goes where.  [I agree that the + and - rectifier terminals can be interchanged in this application, but only if they are connected to the coil, not the rails.]

"Next came the second and more challenging part.  I had to disconnect the coil wires attached to the lugs on each side of [sic] E-unit.  [He had already disconnected these in the previous paragraph.]  Those wires were extremely short, and if I broke one off, well, good luck trying to find it!

"Once I got them out I made life easier by attaching some 22-gauge wires to them to extend their length.  I soldered those extensions to each one of the two AC leads on the inside of the rectifier (the ones marked with the two squiggly lines) [He's got the coil--the load--now connected to the rectifier's input terminals]."

[servoguy]

With that detailed explanation, it is clear he is confused and has things hooked up backwards.  Also, as you mention, the coil wires are somewhat fragile, and if they are broken off, the coil is scrap.  He said he extended the coil wires with some 22 ga. wire.  I hope he tied these wires down so they don't vibrate and break off the coil wires.

[lionelsoni]

I think he actually wired it right, as the picture shows.  But anyone who does it according to the text is in for a surprise!

[cwburfle]

 He said he extended the coil wires with some 22 ga. wire.  I hope he tied these wires down so they don't vibrate and break off the coil wires.

I have spliced extensions onto the end of short coil wires. I usually use some scrap magnet wire, or a few strands of wire cut from a length of stranded wire.
Just have to make certain the new section is the equivalent of a heavy enough gauge.