https://cimg7.ibsrv.net/gimg/www.mia...f08bb6d096.png

Safety first (not mine)

https://scontent-ort2-2.xx.fbcdn.net...f6&oe=61B6A76F

https://cimg7.ibsrv.net/gimg/www.mia...53f9e0d50e.jpg

https://cimg2.ibsrv.net/gimg/www.mia...605e4a6575.jpg

18 empty Prestone concentrate jugs.

https://cimg8.ibsrv.net/gimg/www.mia...9a7a22ef01.png



55 gallon drum, fitted with an outlet at the bottom. Add 20 gallons of distilled water.

https://cimg8.ibsrv.net/gimg/www.mia...539255f48b.png
(Fun fact: this is how The Joker was created.)





And, after silver-soldering the flanges onto the copper tube, we're then immediately quenching them in a 5 gallon bucket filled with a mix of muriatic acid and water. So add the aroma of boiling acid to the aroma of churning antifreeze.

https://cimg8.ibsrv.net/gimg/www.mia...ba01f25ec4.png




Mind you, I am not @Erat. This kind of thing is supposed to be his domain. I normally just work with hilariously lethal high-voltage RF systems.




Because plumbing.

https://cimg0.ibsrv.net/gimg/www.mia...7fb1e0a3be.png

Which reminds me- Now that the system has been cleansed, purged, rinsed, purged, rinsed, purged, and final-filled, I need to close the manual bypass valves across those two thermo-mixing valves. Setting a calendar reminder for that.



The amplifier frames are still dry, but we assume those are clean from the factory. Planning to open the valves to fill those on Friday, which means yet more glycol pumping... I'm really getting my money's worth out of this puny little transfer pump I bought from Grainger.

https://cimg6.ibsrv.net/gimg/www.mia...e0bcbfc3cd.png

Gratuitous second-stage combiner porn:

https://cimg9.ibsrv.net/gimg/www.mia...7b55feeb3c.png

I've gotten pretty good at TIG brazing copper to brass during the course of this new transmitter build.

https://cimg9.ibsrv.net/gimg/www.mia...14f59c34cc.png

The secret turns out to be paying someone who is competent to do it for you.


Different part of the system from where we're silver-soldering similar joints. Here, I had the luxury of pre-calculating the dimensions, and ordering parts fabricated off-site. The ones we're doing on-site are the "Well, there are a hell of a lot of variables here..." connections.

And, just to be extra gratuitous, a bit of final-combiner porn:

https://cimg6.ibsrv.net/gimg/www.mia...c8fea2ca5e.png

This behemoth is where >all< of the power comes together on its way up to the master antenna. Each cylinder is a different TV station. Mine is the one on the far left.

The array is organized by frequency, and 503Mhz is right at the bottom end. It pays to be last in this scenario. (I assume that my predecessor paid quite a lot in actual dollars for this privilege.)

Why so short? 50% because I'm at the extreme low-end of the spectrum (so I have only one RF neighbor, not two), and 50% because I can afford to shove enough power into this array to say "Fuck a little bit of destructive interference. I have the most powerful transmitter in this chain by a large margin, and I can just keep turning the knob clockwise. My electricity is cheaper than your paid-by-the-minute design crew,"

The part which is hilarious, at least to me, is that every one of those seven filter cavities is, from a DC standpoint, a dead short circuit.

Like if you put an ohmmeter across one, you would die instantly. Because you were stupid and didn't shut off the transmitters first.

But if you did shut off the transmitters, and then put an ohmmeter across it, it'd read "0Ω", or as close to it as the probes on your meter are capable of registering.

Oh, and on the output side, it would read infinity. Because there is no center conductor. It's waveguide at that point, which basically means "electricity is literally traveling through space (air is irrelevant at this frequency), in a way that we can magically shape. But then, at the end, it rams into four more dead-shorts, which transform it back into conventional electricity again, until it gets to the antenna atop the pole on the 107th floor, at which point it turns back into magic again, at two different phase angles and two different polarities."

RF engineering breaks everything which first-year EE students learned in DC Circuits I.

On the standard Cu stuff, we are now using crimped connections, rather than solder / silver solder. Not sure of details as to pressure withstand.


https://cimg6.ibsrv.net/gimg/www.mia...952ca8f68f.jpg

Internally, the amplifier cabinets use those same crimp connections.

The pump cabinets use conventional soldered connections. As I used to work for the company that built this rig, I can attest to the fact that they have a super-fancy oven, which pre-heats the whole assembly in a jig to melting temp in a nitrogen atmosphere, and then squirts pre-molten solder into the joints while pulling a vacuum on the inside of the pipe. It's really cool to watch.


I can't say that I really trust the crimped connections, but they are what they are. The PA cabinets are way too complex for pre-joined assemblies to work. And, to be fair, 90% of them is threaded stainless. Only the end-connections are copper.

The building's rules gloss over anything inside a cabinet which has a UL sticker on it. So they don't care, for instance, that the internal copper is Type M, which is not code compliant in Chicago for pressurized water lines.

Outside of the cabinet, it's a whole different ballgame. Everything is specified to maximize cost. So Type L copper, and brazed connections. I'll admit that the code didn't require the fancy braided-stainless vibration isolators. I added those because I think they look cool as compared to short sections of rubber hose and worm clamps.


EDIT:

Found a photo I took a couple days ago of the backside of the amplifier cabinets with the doors removed:

https://cimg4.ibsrv.net/gimg/www.mia...ac727ede44.png


Ignore the black hoses at the bottom. Those are Home-Depot grade washing machine hoses, used to bypass across the amplifier cabinets so that we can flush / rinse the cooling system as a whole without needing to open the valves into the (presumably clean) PA racks.

All of this electricalicity-stuff makes my head spin. I am a lowly ME that cringes every time I'm forced to pick up a DVM to diagnose a problem on my car. I rue the day when the IC engine is rendered completely obsolete.

Seems like you could place an innocuous server rack down there and no one would notice. Plug it in and start mining... ;)

This will be a huge time suck for those of us "of a certain age":
https://www.radioshackcatalogs.com/f...k_catalog.html

https://cimg3.ibsrv.net/gimg/www.mia...d5a5e950c4.png

I definitely have sufficient power (600a 480v) and cooling (chilled water direct from the roof), I just don't have the time or the interest.



https://scontent-ort2-1.xx.fbcdn.net...90&oe=61B8121D

Very steampunk - now you can get away with coming to work dressed like this:
https://cimg1.ibsrv.net/gimg/www.mia...45942c374e.jpg

Radio Shack; for when you wanted Marantz, but couldn't even afford Sony...

Oh so true... I had a Radio Shack setup. Two small bookshelf speakers and a receiver unit that was dimensionally the same as the speakers. 5 Watts, maybe. In the early 70s, I wanted to upgrade. I wanted the Marantz, would have settled for the Yamaha, but ended up with a Wintec R1030. Interestingly, I still have it.

I also remember riding my bike from Wheaton, MD to the Lafayette in Rockville. Lafayette had higher end gear.