Very Simple DCS Booster by stacking ACT244s

Great find ADRIAN. I have some solder paste and extra ACT244's. I'll bet this will really light up the LED's on our signal testers. Is it possible to just solder paste new ACT244's over weak or dead ACT244's ? It would make for a much quicker repair.

Dave Zucal posted:

Great find ADRIAN. I have some solder paste and extra ACT244's. I'll bet this will really light up the LED's on our signal testers. Is it possible to just solder paste new ACT244's over weak or dead ACT244's ? It would make for a much quicker repair.

Those chips have a low output impedance so you want to take the dead ones away or they'll degrade the drive strength of the added ones a lot. You can get a good hot air rework for $115 that makes short work of changing these chips. If you solder a lot of SMDs, it's a good investment...

will you be at the York DCS meeting

That's very interesting.  I think the way to make the retrofit would be to remove the existing chip and have the stack on a little pcb.  There are some nice SMT mounting pins that solder to the flat pads for a chip that are used for debugging. 

You could also add the diodes and TVS that I did on this board to have a total protection package.  That is a really simple solution!

Clem, he's on the West Coast, I'm not sure he's coming to York.

Clem, he's on the West Coast, I'm not sure he's coming to York.

Where/when?  No promises though.

Thanks john...sure would be nice if he could make it sometime. I went out to the Angels Gate Club once, nice bunch of people, Matt was my terrific host.  

Clem 

Adrian

Oct. 17 to 20 this weekend.   York eastern division TCA. The annual DCS meeting wednesday

clem k posted:

Adrian

Oct. 17 to 20 this weekend.   York eastern division TCA. The annual DCS meeting wednesday

Uh I was just away from the office more than a month for a space launch. I'd really like to, but that's too soon to work. Maybe next time.

I understand, there is always next time, watch Barry's post on the DCS forum, He is the one that puts it together. It was twice a year, spring and fall. All the brains of the outfit are there, you would fit right in, ask Matt he was there a couple of years ago. 

That's really cool, great idea! Looks like you have sparked GRJ's interest as well. I think this may be going somewhere soon. 

I have said this before, but it really is nice to have you folks around here with all these innovative ideas that you share with (and try to teach to) the rest of us less knowledgeable (me anyway). Very good learning experiences as well, Thanks!

More semi-useful info:

Stack <=4 good.

Stack >4 too hot.... time to start looking at heatsinking.

I think maybe going to that many make be bordering on overkill.  I think 2-3 would probably be plenty, and give them protection so they don't get whacked by transients.  At some point you may be taxing the internal power supply too much as well, I have no idea how much headroom there is with the power.  Adding a dozen or more ACT244's might be a bit much.

gunrunnerjohn posted:

I think maybe going to that many make be bordering on overkill.  I think 2-3 would probably be plenty, and give them protection so they don't get whacked by transients.  At some point you may be taxing the internal power supply too much as well, I have no idea how much headroom there is with the power.  Adding a dozen or more ACT244's might be a bit much.

You can always pipe in more supply power separately to the daughter-board if you wanted to go nuts (as long as you tie the isolated grounds together).  I could almost imagine a board the size of the TIU motherboard with a giant 5V supply and an array of maybe 64 ACT244s all in rows in and columns with a big heatsink on the back. The absolute max VCC is 7V on the datasheet so you can pick up 7/5 more drive right there raising the external supply a bit.  You could probably run an entire club layout off a single TIU channel with something like that. The only question is would the passives in the output network (the XMFR) hold up okay. That would be amps of signalling in the windings.

It's a funny thought though.

Remember, the engines have to talk back, and I know stacking those up on the PS/3 board ain't gonna' happen!   If the engines can't talk back, all the power in the world at the TIU isn't going to help.

I'm think that more than three or four at the most is probably overkill.  I'd be happy with more reliability than the existing design.

gunrunnerjohn posted:

Remember, the engines have to talk back, and I know stacking those up on the PS/3 board ain't gonna' happen!   If the engines can't talk back, all the power in the world at the TIU isn't going to help.

I'm think that more than three or four at the most is probably overkill.  I'd be happy with more reliability than the existing design.

 

A passenger car, but the passengers are ACT244s instead of people? (this is being silly)

Adrian! posted:

(this is being silly)

You got that part right!

With increasing the TIU signal output 2x - 3X, would this address the potential issue of the TMCC buffer stepping on the DCS signal?  The TMCC buffer provides ~3X and now the DCS ACT mod would also give you ~3X. 

rad400 posted:

With increasing the TIU signal output 2x - 3X, would this address the potential issue of the TMCC buffer stepping on the DCS signal?  The TMCC buffer provides ~3X and now the DCS ACT mod would would also give you ~3X. 

It's complicated. The DCS would be more "square" looking with sharper edges and the voltage swing would be larger. It's hard to say if it will improve how DCS performs with that 1-2V 450 KHz sine wave from the TMCC buffer superimposed on top of it. I plan to explore this at some point. The issue is it's easy to mod one channel with the ACT244 for a bench test, but at the club... yeah that's 20 channels in 5 TIUs, so a lot of soldering. I'll get there eventually.

Adrian, do you put GRJ's daughter board on top of the stack or is it not needed any more?

Susan Deats posted:

Adrian, do you put GRJ's daughter board on top of the stack or is it not needed any more?

Uh I have a stack of 3 ACT244s surface soldered on each-other. I also have the 1n4148 prototype boards and TVS wired into the output nodes from first post in the original thread.

Susan, my little protection board was intended to replace what Adrian did and be a bit easier to install and build.  You will still need to protect the stack of 244's from transients, that won't change.

Now that York is behind us, time to get back to stacking ACT244s

Has there been anymore info and or experimenting with stacking the ACT244?  I am ready to start modifying one of the clubs TIU's but wanted to get the latest info before I started.  

Adrain - When you stacked 3 ACTs (original plus 2 additional), did you bend the pins on the additional ACT SMDs down, so it would  touch the ACT below it, or did you use TTH version and cut the leads short?  Do you have pictures of the modification you can share?

GRJ - Have you looked into anything further of combining your TIU TVS/Diode board with the stacked ACT224s and using the SMT mounting pins you talked about above?

Thanks ,

Bob D

NJ-Hi Railers

Bob, you'd want to use the same part as is on the board and just bend the pins down.  The pin spacing on the thru-hole version is different, that wouldn't work.

I haven't done anything on combining all of the above yet.

gunrunnerjohn posted:

Bob, you'd want to use the same part as is on the board and just bend the pins down.  The pin spacing on the thru-hole version is different, that wouldn't work.

I haven't done anything on combining all of the above yet.

Got it

rad400 posted:

Now that York is behind us, time to get back to stacking ACT244s

Has there been anymore info and or experimenting with stacking the ACT244?  I am ready to start modifying one of the clubs TIU's but wanted to get the latest info before I started.  

Adrain - When you stacked 3 ACTs (original plus 2 additional), did you bend the pins on the additional ACT SMDs down, so it would  touch the ACT below it, or did you use TTH version and cut the leads short?  Do you have pictures of the modification you can share?

GRJ - Have you looked into anything further of combining your TIU TVS/Diode board with the stacked ACT224s and using the SMT mounting pins you talked about above?

Thanks ,

Bob D

NJ-Hi Railers

Exactly what GRJ said. Same part, pins bent. Sorry I didn't take a photo before I closed it up. I'm figuring out how to do the same thing on the PS3 board without goofing it up.

The PS/3 board may be trickier, I haven't looked at those to see.  I'm sure they used a finer pitch part, so I'm out of that game.

I straightened the pins on the add-on ACT but they don't touch the pins on the ACT below it.  ~.5mm space between the two set of pins.  This will be an interesting soldering process.  What is a safe temperature to set the soldering iron at?

Adrain-  Did you use a rework hot air gun to solder the ACT together?  If the pins don't touch, can a rework hot air work for this application?

Going back to the original post, Adrian stated, "there are still some situations where the excursion voltage still isn't enough for everything out there."  I don't recall ever having such an issue, although my layout is quite large and has many engines on powered tracks.  Can someone describe to me how this problem manifests itself?  Seems that in my case at least, we have a solution in search of a problem.

rad400 posted:

I straightened the pins on the add-on ACT but they don't touch the pins on the ACT below it.  ~.5mm space between the two set of pins.  This will be an interesting soldering process.  What is a safe temperature to set the soldering iron at?

Adrain-  Did you use a rework hot air gun to solder the ACT together?  If the pins don't touch, can a rework hot air work for this application?

Bob, set the iron for 260C (500F) or a bit below that.  That's the typical 10 second soldering temperature for many chips.