Railpower 1300 testing

[mmagliaro]

Hey Max, would yo please show/tell me how to hook the scope up to get those measurements?  I've got an old Tektronix Model 465 that I've never used 'cause I have no idea how to.  Any help would be greatly appreciated.

Doug

Let's assume your scope works... I'm looking at a photo of the 465 front panel...

Plug your probe into CH1.  Clip the ground clip of the probe to the negative from your throttle, and clip the main probe tip to the positive.

Set the channel 1 control (below the "VOLTS/DIV" to "DC")
On the right of the display, find the "TRIG MODE" and set it on "Auto".  We're going to have the scope trigger on the
signal itself, not some external trigger.

Next to that: Coupling: AC,    SOURCE: CH1

To the right of that, there's an "A TRIGGER" knob.  Remember where that is.  You'll need it.

Turn up your throttle a few volts.  Hopefully, you see a flat horizontal line on the scope.  This is the DC voltage from your throttle.
If you have pulses turned on, you should see some ripples or bumps on top of that line.

Use the CH 1 "VOLTS/DIV" line to set the vertical scale - this sets how many volts each horizontal line of the display represents.
So if your track voltage is turned up to 6, you might set the volts/div on 1 or 2

Now, if what you see if just a jumble, you need to adjust two things:  on the right, the TIME/DIV knob.  That controls how many seconds of time each vertical line of the display represents.  For pulses around 60 Hz,  that's 1/60 = about 16 msec,
so something like 5 - 10 msec/div is a good amount of time to let you see a few pulses across the display.
Then there's that "A TRIGGER LEVEL" control I warned you about.  If you still aren't getting a good picture of pulses
on top of your DC, adjust the trigger level up or down until you see the scope "lock on" to your pulse wave form.
You should get an almost static picture of the pulses.

Here are a few shots of my scope display to give you an idea of what you might see and what you should look for.

BE AWARE, the pulse waveform you see here might not be anything like what your pack puts out.  In fact, if you have no pulses, you might get nothing but flat horizontal lines (DC).  Not very exciting, but good to know if that's what your throttle is supposed to be doing.

Also BE AWARE... ha ha... I am a complete novice when it comes to using an oscilloscope.  I understand a few rudimentary basics, but that's about it.  So if there are settings I should be using to do this better, believe me, I appreciate any advice!

[C855B]

Max, you're understanding basic 'scope operation, but the problem you've taken on is likely not a repeated event of frequency X, which is what triggered 'scopes are best at. I suspect MRC has spent plenty of 'scope time proving their pulse shape is not to blame. I think we're looking for a "high-excursion transient", a single event (maybe with a "ringing" component) where the image will disappear as the phosphor trace fades away.

I wish we knew more about which component or components in the ESU decoders were failing. It easily could be a situation, as Pete mentioned, where a transient was exceeding suppression specs, with downstream consequences.

[tehachapifan]

Max, you're understanding basic 'scope operation, but the problem you've taken on is likely not a repeated event of frequency X, which is what triggered 'scopes are best at. I suspect MRC has spent plenty of 'scope time proving their pulse shape is not to blame. I think we're looking for a "high-excursion transient", a single event (maybe with a "ringing" component) where the image will disappear as the phosphor trace fades away.

I wish we knew more about which component or components in the ESU decoders were failing. It easily could be a situation, as Pete mentioned, where a transient was exceeding suppression specs, with downstream consequences.

Just smile and nod like you understand what he's talking about.

[BCR751]

Thanks much for that info, Max.  I don't have one of the MRC packs.  I'm just interested in checking out the wave form on my NCE system occasionally.  I also would like to use it for other electronics troubleshooting but, as I said, I have no idea how to use it.  It's one of those, "you need to use it often to know how to use it" things.  I have a Tektronix instruction manual for the scope but, it is very short on practical application problems.  I could read about it all day long and not get anything out of it.  Show me an example and I'm good to go, so to speak.

Thanks again, Max.

Doug

[MK]

Just smile and nod like you understand what he's talking about.

LOL!!!!!!!!!!!!!!!!!!!!!!!!! 

[mmagliaro]

Max, you're understanding basic 'scope operation, but the problem you've taken on is likely not a repeated event of frequency X, which is what triggered 'scopes are best at. I suspect MRC has spent plenty of 'scope time proving their pulse shape is not to blame. I think we're looking for a "high-excursion transient", a single event (maybe with a "ringing" component) where the image will disappear as the phosphor trace fades away.

I wish we knew more about which component or components in the ESU decoders were failing. It easily could be a situation, as Pete mentioned, where a transient was exceeding suppression specs, with downstream consequences.

Indeed... without a digital storage scope (which is not what I have), it will be almost impossible to see something like that.
I'm not convinced MRC has been looking for this.  They should surely have a storage scope and they could let it run for hours while somebody fools around with the throttle until the event happens to trigger the scope.  Their response to the original accusation (talking about safety and UL listings, which have nothing to do with the problem) does not sound to me like somebody who is trying to find out if the problem is real or not.

[draskouasshat]

My personal opinion is that MRC has a minor defamation case against @rapidotrains personally. I know they have found an issue, but in my opinion should have worded things differently.
Perhaps something to the tune of, we have found that our decoders are not compatible with the mrc 1300 powerpacks. We cannot warranty any claims if you are using one.
This might not sound so bad abd ruffle any feathers.
Drasko

[nkalanaga]

"Just smile and nod"

Translation of what he said - you're not looking for the normal waveform, you're looking for "transients", voltage spikes, or something else intermittent, abnormal, and probably rare.  Standard 'scopes are very good for showing normal, or even abnormal, waveforms, but not so good for transients.  They usually come and go too fast.

[OldEastRR]

Well, if Max loops in an ESU decoder to run off the same input the scope is reading, he'd get a very definite sign if a transient spike got past the scope. Like, a puff of smoke. Not entirely kidding here.

[mmagliaro]

Well, if Max loops in an ESU decoder to run off the same input the scope is reading, he'd get a very definite sign if a transient spike got past the scope. Like, a puff of smoke. Not entirely kidding here.

I doubt the result is an actual puff of smoke or even any heat.   The decoder probably just silently quits working.  Blowing out a small internal component inside a chip usually happens without any catastrophic physical clue.  Anecdotal incidents of tantalum caps exploding or housings melting are rare.  Most failures don't happen that way.

About all I can do with a conventional analog scope is look at the pulses that normally come out of the throttle, which I expect will not look like anything dramatic.  I can also see if I see anything when I swipe the speed pot around, flip the direction switch around, or power it on and off.  I might get lucky and see something there.    And of course, I'll look at the circuit layout to see if there's anything odd about the way it is designed.  I'm expecting to see your basic voltage regulator circuit with some simple supporting discrete components.  But who knows?

[peteski]

Thanks much for that info, Max.  I don't have one of the MRC packs.  I'm just interested in checking out the wave form on my NCE system occasionally. 

Old-fashion analog scope (like the 465) will not be very useful for looking at the DCC signal.  DCC waveform is constantly changing (since it is carrying the digital information by flipping polarity of the track power many times  second).  You will most likely a jumble of square-shaped pulses. 

[mmagliaro]

A quick update.  I have received the 1300 from a metalworkertom (thank you) and have started examining it.

First off, the internals are very simple.  Transformer, 4 diodes to rectify the AC, and a single TIP102 Darlington power transistor whose base current is controlled through the speed control pot and a single fixed resistor.

The output is always half sine wave pulses with lower-amplitude half since wave pulses in between.
That is, the output is big pulse, small pulse, big pulse, small pulse, etc.

I'll post photos of the wave form from the scope later.

I have a theory about what's going on (regarding the DCC decoders) but I'm keeping my mouth shut until I can test and be more certain. 

[strummer]

A quick update.  I have received the 1300 from a metalworkertom (thank you) and have started examining it.

I have a theory about what's going on (regarding the DCC decoders) but I'm keeping my mouth shut until I can test and be more certain.

...cue drum roll... 

Mark in Oregon

[Doug G.]

Ah, the old big pulse, small pulse trick and we all fell for it! Would you believe...



Doug

[DKS]

You might also want to consider running these tests on similar throttles that are not reported to be destructive. You need to get a baseline on what DCC will tolerate in order to possibly identify the potentially damaging behavior of the 1300. Otherwise you won't have any basis of comparison. You may have taken on quite a challenge...

Also, maybe this belongs in DCC/Electronics... (ducks and runs)