Skip to main content

Our club has been using the booster GRJ designed for years. It's really good (5 stars from me for sure). However one assumption made in the design is that your layout grounds makes sense (sadly, ours does not). What I mean by this is lots of "parasitic connections" exist between the track ground and building ground from people tapping off the layout for accessories and stuff over the years without knowing the difference. Despite my best efforts we just can't find every connection so the GRJ booster sees a lot of loading and never really saw it's full unloaded voltage swing or anything close.



So inspired by GRJ... to solve this, we are introducing the "Too Much Booster" booster to our layout. This gigantic class AB can drive 5 ohm to about 25V of swing at 455 KHz and can deliver about 150W real power (not complex). If we can't find the parasitic connections, maybe we can burn them out?



Thanks for the inspiration GRJ. I'll report back how this went if we don't burn the club down.

CLASS_AB

Attachments

Images (1)
  • CLASS_AB
Last edited by Rich Melvin
Original Post

Replies sorted oldest to newest

@Adrian! posted:

Well it's working. The layout is drawing about an amp. I mean an amp of 455 KHz signal... not power. It feels wrong but works well

Can you please elaborate.  I am working with a 30'x 200' multi level layout with probably numerous parasitic connections.  The GRJ/DM buffer is working fine, but always looking for ways of improving the TMCC signal.

Thanks,

Bob D

@rad400 posted:

Can you please elaborate.  I am working with a 30'x 200' multi level layout with probably numerous parasitic connections.  The GRJ/DM buffer is working fine, but always looking for ways of improving the TMCC signal.

Thanks,

Bob D

Remember Bob, if you boost the strength of the TMCC signal, you risk killing the DCS signal.  The whole reason the tuning pot was added to the TMCC buffer was to balance the boost to get reliable TMCC/Legacy signals but not stomp on the DCS signals.

Remember Bob, if you boost the strength of the TMCC signal, you risk killing the DCS signal.  The whole reason the tuning pot was added to the TMCC buffer was to balance the boost to get reliable TMCC/Legacy signals but not stomp on the DCS signals.

Understood, the current buffer is staying in place and not implementing anything that will affect the DCS signal.  Just want to see if there are any takeaways from Adrains experiment that can be used for diagnostic purposes.

@Bruce Brown posted:

If you add a longwire antenna to that 150 watt transmitter, you should be able to take over control of all Lionel command engines throughout the continental USA.

Hopefully, no one from the FCC--who is familiar with Federal Regulation Title 47, CFR Part 15--is an OGR forum fan.

I mean its funny ... but also not correct in terms of FCC things.



So no one gets the wrong idea:

1. We're not even close to a quarter-wave or resonant (600m / 1800 ft). If you look at an offset fed lambda / 18 structure (about where we are your dipole radiation efficiency is around 0.25%, like not even competing with a hertzian-dipole.

2. it's inside a building full of return wires so the E-field is going to be confined since the near field is shunted out by the near field ground.

3. It's only like 0.001 lambda over the ground so all we can do is induce an image current, not actually radiate (antennas over planes don't radiate). It's a capacitor not a radiator.

You're cool to develop whatever power you want, as long as it stays conducted and not in a radiating mode of propagation. The deflection amplifiers in an old 36" CRT are basically the same level of power and frequency as this thing. That's where I drew the inspiration for a class AB from.

Last edited by Adrian!

Understood, the current buffer is staying in place and not implementing anything that will affect the DCS signal.  Just want to see if there are any takeaways from Adrains experiment that can be used for diagnostic purposes.

Its  interesting actually. There were some unexpected benefits... before we hooked up this mega-driver... the common mode (potential from the layout ground to building ground) had a lot of components from 60 Hz stuff all the way up to several MHz (we have a lot of radars and stuff here in the harbor). So when you would look at the time domain on the scope you'd have a broadband fuzz of 2-3V, then a general wiggle to it that was the original booster of maybe another 0.5V superimposed, then the occasional ~12V pkpk DCS packets.

With the mega-booster in place the voltage is really really well defined and crisp now... a sharp sine wave on the scope. That makes sense since the Thevnin impedance looking back into the thing is basically zip so a lot of that fuzz can't couple in. So now I see a very clean 3V at 450 KHz that transitions into a very crisp and well defined DCS packet.

It feels like what we lose in mode conversion desensitization (Lionel going from the intended common mode to differential mode potential and coupling into the DCS packet) we more than make up for by having a cleaner and crisper waveform with less of the fuzz layer on top.... It's always hard to quantify broadband fuzz, but DCS actually felt more responsive with this thing in than out. I'm not thrilled by the fact we're circulating an amp of signal current, and the SPS modules that power the driver are making a noise that tells me they aren't thrilled either.

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×
×