Weird TIU Voltage (AC to DC)

Hi everyone,

Thanks for your replies. 

Based on the information received, I will go ahead and replaced the 4FETs feeding VAR 2 Out.  If anybody knows the exact FET model that I have to order, please let me know.  Otherwise, I will find out when I remove the old ones. 

If replacing those FETs correct the problem, it will prove that these can fail in a way that they start acting as a rectifier somehow.  This would also solve the mistery of post https://ogrforum.ogaugerr.com/t...388#9595974411703388 reporting the same problem, but with no solution.

Daniel

If you "do" eBay and are in no rush you can find good deals on FETs...

If you have one bad channel, why are you replacing 4 FETs?

1.  You should redo the scope test with a load.

2.  You should use your volt/ohm meter on the "diode" setting and probe the FETs.  You can compare the readings on the good channels with the bad channel to determine if one or both FETs on the bad channel are shot.

Dale Manquen posted:

If you have one bad channel, why are you replacing 4 FETs?

1.  You should redo the scope test with a load.

2.  You should use your volt/ohm meter on the "diode" setting and probe the FETs.  You can compare the readings on the good channels with the bad channel to determine if one or both FETs on the bad channel are shot.

I do not have any schematics, so I go with the information I have... So, from the information I got, the total of 8 FETs are:

4 FETs for VAR1

4 FETs for VAR2

If someone has other information, you are absolutely welcome to share and contribute.

Thanks,

Daniel

stan2004 posted:

If you "do" eBay and are in no rush you can find good deals on FETs...

Thanks for the information !

I get from your message that all 8 FETs are identical.

There seems to be two versions.  Is there one preferable to the other one ?

Versions:

http://www.vishay.com/docs/91291/91291.pdf

http://www.vishay.com/docs/91292/91292.pdf

Thanks,

Daniel

Personally, I think you're jumping the gun.  I have observed this condition, or one similar, once or twice over the years, and tried a different meter. 

Apart from these crazy readings, how do the trains run?

Sorry, I have been servicing too many Lionel controllers that only use 2 FETs per channel.

RJR posted:

Personally, I think you're jumping the gun.  I have observed this condition, or one similar, once or twice over the years, and tried a different meter. 

 

Apart from these crazy readings, how do the trains run?

Here are the observations to help clairify things:

1) First, let`s forget about the meter.  The osciloscope reading can`t lie, and it clearly shows that the VAR 2 out is pushing a DC signal, while all other channels are pushing an AC signal, even if all INPUTs are on EXACTLY the same AC source (Z-500 or Z-4000).  If I put the meter reading picture in my post, is was simply to make a link with the other "Weird" post mentioned above.  I do know that I am NOT getting 55 volts AC.  The meter is just "confused" because the signal is somehow DC, as shown by the osciloscope.

2) This issue was found because of trains behaviours problems.  So, to your question "how the trains run ?", here are the observations:

a) For the one LIONEL loco I tried, it just does not move. It works fine on the other TIU channels.

b) For the MTH locosss (many tried), it does move normally, but the DCS signal is unreliable, or not working.  It works fine on the other TIU channels.

These two observations (problems) are the reasons why a meter and an osciloscope were put on the TIU.

Hope this clarify things,

Daniel

Well, the FETs are the same on the TIU I looked at - easy to read the part markings.  As for slight variations between parts, I'll simply observe that this is a 50 Amp FET running at 60 Hz - so I don't think the TIU application is particularly demanding.

First I still don't understand the 2 sets of power leads feeding Input 2.

By enlarging the 1st picture it appears that your jumper is incorrect, that is Fixed one Red is feeding fixed 2 Black and vice versa. That would certainley lead to a false output reading and a bad sine wave.

That sine wave doesn't look right anyways as a Z4000 produces a chopped sine wave and the TIU chops it again. That sine wave pic is to smooth.

Since your cheap meter isn't a True RMS (Fluke) meter then switch it to read a DC waveform as that's obviousley what you have.

I won't volunteer either.  As I said above, I have experienced this.  I can try to recreate it. I 

!

I think the OP has abandoned the meter reading issue.  As I mentioned above, this is/was probably due to the AC measurement "technology" used in his meter.  For example, I'll bet if he took a 9V battery and measured it in AC-mode it would read 18V (or so) in one polarity and 0V in the other.  Try it on YOUR meter...you may be surprised (or horrified)!

stan2004 posted:

I think the OP has abandoned the meter reading issue.  As I mentioned above, this is/was probably due to the AC measurement "technology" used in his meter.  For example, I'll bet if he took a 9V battery and measured it in AC-mode it would read 18V (or so) in one polarity and 0V in the other.  Try it on YOUR meter...you may be surprised (or horrified)!

 

Yes, thanks Stan2004. Please, let`s all forget the meter reading... it is unfortunately and obviously confusing the topic, and that is not the point here.

The point is that the TIU is damaged and trains do not run normaly anymnore on VAR 2 Out.  "Coincidentally", the Output on VAR 2 is abnormal, as shown on the ociloscope (DC) and on my "cheap" meter ;-)

I will order the FETs is a few minutes, and share the results next week.

stan2004 posted:

Well, the FETs are the same on the TIU I looked at - easy to read the part markings.  As for slight variations between parts, I'll simply observe that this is a 50 Amp FET running at 60 Hz - so I don't think the TIU application is particularly demanding.

Thanks stan2004, appreciated.  Ordering now.

Casey Jones2 posted:

First I still don't understand the 2 sets of power leads feeding Input 2.

By enlarging the 1st picture it appears that your jumper is incorrect, that is Fixed one Red is feeding fixed 2 Black and vice versa. That would certainley lead to a false output reading and a bad sine wave.

That sine wave doesn't look right anyways as a Z4000 produces a chopped sine wave and the TIU chops it again. That sine wave pic is to smooth.

Since your cheap meter isn't a True RMS (Fluke) meter then switch it to read a DC waveform as that's obviousley what you have.

1) Here is how the connections are done at the input.  There is no mistake on the input connections. (i.e. the black and red are NOT mixed)

2) For the "sine wave" shape comment:  I took a picture of what the ociloscope was showing, but it is true that a Z-500 was the source of power at the picture time (not a Z-4000). 

I do not have a picture with the Z-4000, but the general result was exactly the same: i.e.  It is AC-sine-shape on all channels, except VAR 2 Out (Strait DC).

Hello everyone,

As advertised last week, I have replaced the 4 FETs driving the VAR 2 Out, and the problem is now fixed.

The VAR 2 Out is back to normal with an AC signal, like all other outputs.  (see pictures below). 

The train tested is running good too, with good DCS signal on all TIU channels.

Conclusion:  The "DC" problem has been fixed by replacing the 4 FETs.  I can`t tell if 1,2,3 or all 4 FETs were damaged, as I simply changed all 4 FETs in one operation.

Thanks to GGG and Stan2004 for guiding me to the FETs.  That is really appreciated.

Daniel

Well done!  Next project: upward beam tilt on your scope CRT

stan2004 posted:

Well done!  Next project: upward beam tilt on your scope CRT

A little OCD there Stan.

For those not familiar with scopes, OCD = Oscilloscope Crt Divergence

Well, actually Stan, I was talking about Obsessive-compulsive disorder, but that's an interesting twist.

Daniel Auger posted:

Hello everyone,

As advertised last week, I have replaced the 4 FETs driving the VAR 2 Out, and the problem is now fixed.

The VAR 2 Out is back to normal with an AC signal, like all other outputs.  (see pictures below). 

The train tested is running good too, with good DCS signal on all TIU channels.

Conclusion:  The "DC" problem has been fixed by replacing the 4 FETs.  I can`t tell if 1,2,3 or all 4 FETs were damaged, as I simply changed all 4 FETs in one operation.

Thanks to GGG and Stan2004 for guiding me to the FETs.  That is really appreciated.

Daniel

Daniel, thank you very much for posting your FETs  replacement.  After testing 24 TIU Outputs there was one Output with 21 DC volts out with 17.5 AC volts Input.  When tested there were only 2 bad FETs (part # IRFZ44N)and I went ahead and replaced all 4 in that Variable 2.

I have no idea how the DC voltage might have been affecting the signal output.