Friday, September 16, 2011

Exciting developments afoot, please stand by and forgive interruptions in the normal* programming stream.

From an early age I would crack the shell on stuff, just to see how it works (regardless of understanding anything in there or not). I continue, to this day, to continue this habit; not so much to see how it works, but to check condition. The bank of fuses pictured here was (actually, IS) housed within a reel to reel that appears pretty clean from the exterior. Not exactly the best substrate for contained current...

* HAH

Friday, September 09, 2011

Continued work on the Ramsa 8210. All remaining circuitry is supported on a removable frame seen here.

I pretty much only made it as far as the rectifier PCB of the power supply and the routing strip feeding the summing sections. It is here that the green wire bandit, who modified this mixer sometime in the annals of history, stepped up their game and made the prospect of undoing their work broach nightmare territory.

But before I tread those waters, let's have a look at this cute little stand-off which separates the rectifier/regulation/filtering PCB from the rest of circuitry (seen below). I'm really fond of external power supplies. Fortunately the AC portion is directly piggy-back the metering strips, so I doubt there's a lot of concern here, but still.. a bit close for comfort.

Besides, I've clearly got greater things to be concerned with.

It is at this point that I am pretty much prompted to do what I can to retrace what has been done so I can attempt to wrap my head around what exactly was trying to be achieved here. I foresee some hours of fruitless internet searches regarding published modifications, before throwing up my hands and deciphering. While the iron was hot, I went ahead and reversed this mod. I will be checking continuity between the severed legs before application of power, naturally.

It was at this point that I began questioning the logic and motivation behind all of the green wire mods (most of which involved cutting traces), see, all those green wires are just reflecting the printed circuit traces that originally tied everything together. So, I see this work as a pointless waste of time... which sort of guides my outlook a little on the rest of things.

Perhaps those reworked connections were due to an intermittent failure in functionality? This is the second punched pad I've come across. And if someone spent hours cutting traces and weaving a convoluted single color loom of wiring trying to correct THIS, I may have may work cut out for me here.

Monday, September 05, 2011

What follows is my tactic for repetitive rework, such as is apt to be found in mixing consoles, multitracks and really bad production runs. As far as recaps go, this approach makes rather high part count projects only a little more difficult than a simple tube amp recap. In not necessarily the most accurate order:

1st, I plot out the position of all parts to be replaced and mark the PCB so I can desolder everything on my bench (say, two or three channels) in a single pass. My preferred method of desoldering is to heat the joint with the iron and use a 7874B type pump to remove the bulk of material. Once most of the solder has been removed on everything at hand I will go through everything while the iron is still hot and twist each part looking for movement of both legs. If a leg is still soldered in, it will be heated while twisting to free it up before moving on to the next one. Once all parts are loose, I'll snip the bent ends if they don't pop out freely, and drop them out. This has resulted in minimization of pulled traces.

2nd, each identical channel has the same parts count, this unit per channel count has been marked on the bags containing the parts.

3rd, I begin with the lowest quantity per channel parts (typically "1") and load those all first, no higher part count parts are removed from the bags until it is their turn. Only the unique or super low part count parts are removed from the bags (and ALL of said quantity are removed prior to stuffing a single part). Pictured here is the 330uf, 100uf, 33uf and 4.7uf loaded into the "active" circuit, while an already stuffed part reference circuit (very important to this particular process, though after a few passes the example gets less and less attention) sits atop a bowl full of pull jobs. In the foreground are five 10uf that are to be stuffed as soon as the camera is set down, leaving all remaining seven slots for the 47 uf caps.

Observe polarity. This particular board is designed such that all "east-west" positioned caps had the ground lead to the left, and all "north-south" caps had the ground lead facing me, from the position that I was working on them. This made perpetual polarity checks less of an issue, but it's certainly not a feature one should count on.

Once the lot is stuffed it will be checked for shorts (a couple of the pads are very close to one another), cleaned, then reassembled.

160 some odd caps, minus a few known to have fled from the bench.

Channels are all done, next up is the summing, metering and PSU sections. Then a proper clean up & burn in.
After a short session of soldering I decided to tackle some of the mangled sheet metal on the Otari 8 track.

I still haven't conjured up a sequence of events that explains this condition while supporting the integrity of the surrounding machinery. Must've been a one in a million shot.

Here's another angle on the vent cover distortion, after extraction for hammering. Naturally, most of the hammering occurred while the sheet metal was not installed.

I popped the meter bridge face off to facilitate removal of the vent metal, only to discover that it too had taken on some stress.

Since a wood floor isn't a precision surface, here's a shot with the piece reversed proving the bend is in the metal. The repair for this was gently bending it back over my knee. Very little force was required.

Installation was not as easy as place & bolt, there was a mild interference fit at the butt joint between the vent shape and the transport shell. I'm not overly concerned about the skin of the machine though, so it was hammered into place.

While not minty fresh, this is a vast improvement over the former condition. Surface distortion, while very apparent to the touch, does not jump out and grab me by the eyes. However, while everything was disassembled a knob for input level struck out on the path of freedom, and remains at large. While I'm certainly not going to overlook a possible bout of personal absent mindedness in this, I'm hinging my investigations around my helper monkey who may have swiped it as a trophy.

Saturday, September 03, 2011

Dug out the bench today, in hopes of getting a little work done.

The contents of this pan were strewn over the bench. You could say my junk box situation is a bit unruly at this point in time.

First tier of the actual progress pile is working through the Ramsa 8210 and freeing up even more space on the bench while also paving the way to record.

Each of the channel strips has been marked with its position in the frame, which I find odd, seeing as how the channel strips are presumably all identical, and actual channel number is a function of physical placement in the console. I doubt this mark is stock, and I take it to be a sign that someone has been inside this thing in the past.

Doing some fairly drastic work, from the look of things...

Cut traces and jumpers, I haven't really done a huge amount of research and/or reverse engineering on this yet. I'm toying with the idea of reverting two of the ten channels back to stock for comparative analyses, since eight direct outputs will mate with the recorder.

Then again, perhaps I'll just recap and reassemble for brevity sake. I will ponder what it would take to return a channel to stock before making a decision. I have decided to leave the (groan) RCA connectors alone at the rear. Since the Otari is technically unbalanced, I'm not losing anything with the form of the plug. My plan is to loom up RCA direct out/return and XLR pin 3 hot unbalanced snakes to a latching hard wired 1/4" patch bay, which will allow a no fuss interface to the rest of world and keep the weirdo connectors intact.

I only got one channel recapped, after preparing the bench. Though another channel has been torn down, and is ready for more caps (tomorrow).

Side by side comparison of the new and old. I'm loading it with Nichicon KW series, purportedly audio grade caps and inexpensive. Each channel carries 7x 47uf @ 6.3v (upgraded to 25v); 2x 10uf @ 25v (upgraded to 100v); 3x 10uf @ 63v (upgraded to 100v); 1x 330uf @ 6.3v; 1x 100uf @ 6.3v (upgraded to 25v); 1x 33uf @ 25v (upgraded to 35v) and 1x 4.7 @ 25v (upgraded to 50v).

Rest assured, I will be bolting these channels down out of the sequence printed on their frames, just because I can.

Friday, September 02, 2011

MULTI-PIN

Here's the base of a Burroughs 6700 counter tube.


It's a ten digit Nixie driver tube, sadly, when I start to experiment I'll have to wire up a loom of female pin connectors since I don't have a proper socket for it.


Here's the functional diagram for this beast. Hails from the era of massive computers, I have machinations of deploying this in a repurposed manner, but am following my gut in terms of avoiding spouting too much nonsense before experimenting with it.

I'm hopeful I can utilize hardware from the same people that brought us the Bat Computer in the 1960s TV show. Here's a page with some nice top end shots of these: http://www.decadecounter.com/vta/articleview.php?item=313

Mine are similar to the two variants at the bottom of the page.