PS2 trouble shooting on SD9

Stan,  It is the educational value I am after.  I have blown the horn, bell when I was first testing.  Did not seem to change anything.  One thing I tried in conventional was to change direction and start motor movement while sounds were still normal and that did work.  Once volume dropped conventional motion control was lost.  Not sure why the first few seconds of start up everything seems normal.

This is from a newer diesel.

I agree about the battery drain.  I have been hesitant to run it with reduced volume on the tester.  It did catch me off guard about the battery drop, but a dead or shorted battery is a known issue to cause reduced audio volume according to MTH.

I am assuming there is a device isolating the battery circuit from the 5V regulator output circuit, besides the power trenches.  I believe the power trenches are rated at 7 to 11 amps, so maybe the battery is just being shorted to ground.  I have run it with no battery in DCS and the symptoms are the same.

Optimistically, I always hope it is a simple passive device or microcontroller that has failed so the board can be restored.  In this case while the battery seems to be shorted, the 5V regulator is normal.  Since the audio amp should have 5V supply available, something else is causing the reduced volume.  G

One thing you will note is each connectors has a different number of pins: 2,4,5,7,8, and 12 which should help to not getting them mixed up.  But I suppose with enough brute force determination one could effect an incorrect mating.

I'd say the 0.5V battery is now "it" in this game of tag.  Clearly an AA NiCd cannot sustain this rate of discharge for long and I can't imagine this stress adding to battery life! This is a troubleshooting pickle as I agree you don't want to take a coffee break with the circuit powered like this.

That's pretty clever you thought to cram in a few conventional direction changes during the several seconds before volume drops. You said battery voltage dropped from 2.4V to 0.5V instantly upon turning on the electronics.  It's curious that 0.5V is enough to be stepped up to a suitable voltage to keep the electronics running during the Direction button power interruption so at least this suggests the battery circuit is not totally toast.

If I were approaching this as a educational exercise, we know the 2.4V battery is boosted to supply audio and processor voltage using a step-up switching regulator which means a high-current FET switch charging an inductor by pulling it to circuit ground and then releasing it to step-up the voltage.  If the battery is being shorted as you suspect, I'd think this switching FET is somehow stuck "ON".  Since there are only 15-pins between the two boards and you have already identified some of them, that ought to simplify things somewhat but I think some luck will be involved.  If MTH advised you that it is likely a bad DSP chip, I'd want to use a scope (vs. a DMM) to probe the signals since it's likely the control signals would be pulsing as the TMS24... DSP chips do not have analog outputs.

Stan,  I have mapped some of the other 9 pins but I don't really know their functions because I don't know how the board is designed to work at that level.  It is not easy to reverse engineer this board.

What I do find perplexing is that the battery circuit short is not pulling down the 5V regulator circuit.  My assumption is that a component has this isolated.

So the question is why does a bad battery or Battery circuit drag down the audio amp output.  I had assumed it was because the 5V circuit was being loaded down, but that does not seem to be the case with this board.

So my only hope for a repair is that the DSP is sending a signal to a transistor or FET that controls the Battery Circuit power Trenches; and that that component is shorted on if I read what you are saying correctly?

I have inspected the board closely and there are no obvious damaged chips or SMD devices.  G

G,

I hear you on the battery-short pulling down 5V but I'd think that if a rechargeable battery is charged/good at 0.5V and discharging, it is supplying current to the "black-box" circuit that would help vs. drag down the supply.  OTOH, if the battery is defective/shorted at 0.5V and charging, then the "black-box" circuit would have the smarts to limit current and hence not put an undue load on the power supply.  I base this on the instructions that you charge the battery overnight (i.e., modest charging current) - and, that if you hook up a BCR which is effectively a "short" when first turned on, the voltage climbs at a modest rate (rather than instantly if unlimited charging current was available).

Yes, I too was going down the path that the volume drop was based on the +5V sag and hence I was curious if you could affect the volume by fiddling with the load on +5V such as changing the baseline volume.  Doug made an interesting observation that he used DCS to adjust the volume of his "good" engine and that a volume setting of 40 was equivalent to the "bad" engine volume setting of 100 after the volume drop.  While we don't know the exact relationship of volume setting to audio level in dB or whatever, it stands to reason that a drop from 100 to 40 in any scale has got to be more than what a +5V to +4V supply drop would cause in an audio amp (spec'd to operate down to 2.7V).  In other words, the audio level is being lowered through another mechanism other than +5V supply.

Since you have swapped PS boards and batteries to rule them out, it does seem like the Easter Egg hunt is on the processor board.  I assume you'd start by observing behavior on a known-good board-set to see what the signals are doing when power is applied, removed, charging, etc. Since you apparently have some kind of test fixture to give you better signal access, you might want to monitor battery current so you can confirm whether it is charging or discharging.  Given the DSP is involved, I think the biggest mystery will be to unravel how it handles the odd cases such as no battery connected, battery shorted, etc.  I'd think a BCR might also help when troubleshooting since its terminal voltage moves proportionally to current much more dynamically than a battery if that makes sense.

Again, given the apparent symptom of the PS board shorting the battery (rather than having a shorted battery), I'd say there's a good chance the power switch that charges a boost inductor (Wikipedia Boost Converter) by shorting it to ground might be "on" possibly from a faulty processor control signal.  Or at least that's one place to start.

Stan,  Funny you posted the wiki link as I spent the morning brushing up in Boost and Buck circuits.  I wrongly assumed the Power trench FETs were turned on or off, vice cycled.

I went back and took measurements again and looked at the notes I had created on the pin outs.

The 5V supply is rock steady at 5.03Vs.  The battery circuit seems to step down quickly to about 1.4V then continue to .5 to .7VDC depending on which battery is used.   So it is not instant drop to .5V.  I can still engage the motor at least once in conventional before the sounds go to half volume.  So it must be while the battery terminal voltage is still up around 1.5Vs.

I also measured the 3.3 V Regulator and it flucuates between 3.15 and 3.22VDC.  When I plug the battery in it seems to go up to 3.3VDC but still fluctuates some.

Removing or adding the battery while powered up doesn't change anything other then voltage drops to millivolts at the battery post when the battery is removed. 

I was looking at pin 4 of 6 voltage and found out it is near zero when the battery is out but up to .7V with battery installed.  I had previous static test done on the PS removed from the processor and determined .7V was a good reading.  So I guess it is sending a signal to the processor when a battery is install or not.

Pin 3 of 9 is the pin that activates the transistor circuit that  turns the Power Trench FETs on and off for the battery circuit.  That pin goes directly to one of the DSP legs.  I can't find any shorted component between the DSP and pin 3.

So it does seem like the DSP is at issue.  I thought about lifting it to see if it had an effect, I also thought of lifting the control leg of the small transistor that controls the Power FETs, but ultimately I don't think it helps me fix the processor. Since I can't replace the DSP.

I guess the only curiosity questions left, are why the reduced volume as you asked?  Has the DSP sensed a problem and reduced a major load on the 5V system by commanding a volume drop?  Why did the design guys feel the need to monitor the Battery charging circuit and have it listed in DCS under info.  This board does display a battery charging fault in DCS mode on the remote.

I don't have a special test fixture.  I have the MTH test fixture that replicates the engine.  That way I can half the problem quickly and visually see what works and what doesn't.

I do have some small test probes so I can hook a pin or leg to take measurements safely. 

While I don't think this one can be fixed, I now have some more insight to how it works and that the processor can't be ruled out if a battery circuit isn't working.

As a side note I did have a BCR installed in a previous Power Supply (totally different engine) and it would not charge rapidly.  The BCR worked with another PS board but not in the one I was testing.  The voltage would climb up about a few .01 of a volt a second.

All the time the BCR was charging the volume was extremely low, and there was no control of the engine.  Could not blow the whistle or bell.  Took several minutes for the voltage to get up above 2V or so.  At that time the volume came back.  With a battery installed the board seemed to work fine.  I never fully tested it, but I imagine there was some issue with the charging circuit.  G

Update Just finished swaping out the PS board and was able to determine that the PS board was bad. Volume  did not change that much. I also just finished PS2 swap from my SF PA1 to SF E6 that had loco sounds. Intersting note. Before the swap there were not any issues with the volume control . In conventional mode very loud. In DCS mode volume dropped Same comparison with SD24 SF 100 = 60 on the SD24 . On the SD9 the only thing I can think of is that the sound file has been corrupted somehow. The other idea is that the processers that control volume could be corrupt in DCS mode.

Doug, Post is a little confusing, are you talking about the PS swap on the SD-9 or the PS-2 swap for your E-6?

If there is improvement on the SD-9 but still volume control issues something else is not quite right.  You can try a Sound File reload.  Nothing to loose, because if the processor is bad, it won't be a repairable problem at the component level.   G

Let me know if you need a PS, I have one available when I needed the process to repair an engine.  I just need a turn in.  G