• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Parts and Pieces: A True ARC Mystery

dcgillespie

Fisher SA-100 Clone
Subscriber
I was originally thinking of using names like Humpty Dumpty or The Phoenix in the title of this thread. But as will be seen, they don't fully describe the full context of the project, although large elements of those names certainly apply. Humpty Dumpty implies attempting to put something back together again that was originally together in the first place -- which isn't exactly true in this case. The Phoenix implies building something new out of old ashes -- but the parts and pieces that make up this project are hardly ashes (in fact, many are new). And yet, the project very much involves putting something together that is quite complex (Humpty Dumpty), and new (The Phoenix), from well -- as I came to find out -- parts and pieces from........somewhere. And so the title of the thread as given came to be.

1983 -- By this time, ARC was already credited with single handedly saving the vacuum tube side of the high fidelity industry, when Bill Johnson and his fledgling little ARC company alone started producing very high quality, high performance vacuum tube amplifiers a decade earlier, filling the void created when McIntosh ceased production of their original tube amplifiers around 1970. But unlike Mac's well proven and elegant Unity Coupled design using a conventional if not simple power supply section, ARC went a different route, using a more conventional audio design, but with a highly complex power supply section.

With ARC's designs, a conventional OPT is employed, as is even the lowly balanced paraphase style phase inverter. Both Mac and ARC used cross-coupling to help enforce a balanced push-pull signal being presented to the output stage, although the ARC design is much more aggressive in this regard using active, dedicated, cross-coupled stages to accomplish it, rather than Mac's more passive approach by comparison. Mac and ARC both also used local push-pull feedback around the output stage as well. But here it was Mac who was much more aggressive with their Unity Coupled design, where as ARC's use of local push-pull feedback is far less impacting, using the loaded secondary output winding to supply the feedback. Both designs use direct coupling where possible with a bare minimum of coupling caps, with ARC using cap bundles for its coupling networks.

The one big difference in the audio section between Mac and ARC, is that Mac used a traditional single ended global feedback network, whereas ARC used a balanced, push-pull global feedback system. Theoretically, dB for dB, the ARC approach is more effective, which I hope to determine when all is said and done.

In the power supply section however, things couldn't be more different. The Mac approach is a picture of simplicity, using either a voltage doubler or full wave bridge rectifier, a choke, and conventional filtering and decoupling using electrolytic caps. ARC went down a very different and complex road, using massive electrolytic power supply capacitance at the output of the B+ rectifiers (1000's of uF), and then regulated the voltage to the output tube screen and control grids, and to all of the small signal stages as well. Significantly, the ARC approach does not use a single electrolytic cap to decouple any power supply point after the main filter caps at the rectifiers -- nor does it use exotic and large film caps in their place to accomplish that end, either. ARC uses the inherently low output impedance of its various regulators (a total of four discrete ones, plus a couple of 3T type regulators) to decouple the various small signal stages, so that no electrolytic caps are needed for that purpose. Back in the day, it was the height of the "no cap is the best cap" idea, and ARC ran with it to the max.

The ARC design then was no "me too" Mac imitator but rather, it's own uniquely different approach to producing high quality sound. The brand doesn't get a lot of chatter on AK for some rather obvious reasons (I believe), be it lower production numbers, higher that typical voltages, cost, and complexity to name a few. So with this project starting from basically less than square one, I thought it would make for an excellent opportunity to take a good look at, and document all that went into vintage ARC design and construction. With that then, on to the project itself.

It arrived on my doorstep via my good friend BuzzK who obtained it from a widow, whom along with her late husband, used to own a high end audio salon for many years in Arkansas. Besides both of them attending to the day to day operation of the store, she did the books, and he was the in-house repair for both Mac and ARC, whom they were authorized dealers for. Over the years then, she became quite knowledgeable of the various pieces of equipment they sold. When they retired, they moved to GA, bringing with them many of the pieces they retained when they closed up shop in Arkansas. Here in GA, they had sold off most of the pieces that were operationally ready to go, and were down to the items needing repair or restoration, which he was in the process of doing, when he unfortunately passed away. After a period of time, Buzz became aware that she was trying to sell off the remainder of the equipment (piece by piece), and that she was also very aware of the value of the equipment she had -- even in its compromised state. Buzz made a run to see her, which resulted (in part) in the purchase of my own very nice MC240 from her about 5 years ago, that ultimately needed only very minor attention to make it operational. Shortly after that, I then made my own run, which resulted in the purchase of my C11 preamp that needed a full restoration. Buzz had made one other run however that resulted (also in part) in the subject of this thread. Clearly being down to the crumbs of what was left, here's how it arrived here:

SAM_3071.JPG

SAM_3073.JPG

SAM_3072.JPG
Now how cool is that!! An ARC D115 kit!! -- and already partially built, too!! Except, this was no kit. As I came to find out, it was in fact a real audio mystery, with my initial research indicating it had been in this state of hibernation for nearly two decades. Right away, there were things that simply didn't add up, so to make any serious progress on it then, the mystery had to first be unraveled so at least I knew what I was dealing with. For vintage ARC devotees, there are already a few clues to be seen that should raise some eyebrows. Short of determining that however, because of its condition, I only paid about 10 cents on the 1983 dollar it sold for, so at the very least, its hardly a boat anchor but a dolly full of great parts. I am determined to see it through however, so what it really is for me, is the ultimate 2020 Covid project. It's still far from being finished, moving along with slow but methodical progress. But it's well under way, with enough progress having been made now that a thread can be started.

How a unit comes to be in the state it's in when it arrives here always fascinates me. The mystery as to how this piece came to be as shown when it arrived here is deep, but will start to unravel -- next time.

Dave
 
Last edited:
Register to hide this ad
Dave,

Based on your write up it seems like you know a thing or 2 about tube gear :biggrin: so I guess I'll follow along and see how things turn out.

Obviously I jest and look forward to your progress and as always appreciate the knowledge you impart.

I love these type of threads!

:lurk:
 
I think Mary Shelly can provide you with some inspiration for a title for this project. :banana:

I had no idea of the challenges I was passing along -- but BuzzK is typically "challenged" in one way or another, so I'm sure it was no 'surprise."

I'm particularly interested and impressed with your logical approach and keen attention to detail.

You are the perfect match for this project, as it's beyond the ability and bandwidth of most of us other mortals.

Looking forward to updates and progress reports. :lurk:
 
I'm glad something like this fell into the hands of someone who can actually do something with it. I'd be tempted, but I'd also be way in over my head.
 
Well, looking at the schematic for this beast I'm not sure I really understand how that differential feedback is working yet...but it looks interesting. Definitely gotta study that part more carefully.

This amp seems to be about as opposite as you can get from a Dynaco ST70. Even though both use EL34 power tubes.

And that power supply...that's one mammoth Tour de Force piece of engineering. There even appears to be a 6550-based regulator with 12AT7 error amplifier tucked in there. Appears that powers the screens. No small thinking allowed when designing this baby.

Okay...I'm getting tired just looking at the blueprints for this T-Rex and realizing what work lies ahead....but I can't think of a better person to tackle a project such as this. :)
 
link to stock manual / schematic just for anyone following along

https://audioresearch.com/ar-package/d115_manual_schem-pdf/

I see what you mean about the differential feedback. Grounded 4 ohm tap and feedback lines connected to 0 and 16. I do see that it ties the trafo secondary into the output tube cathodes, so its serving as both local and global feedback connections. This is a way more complicated amp than the typical "golden age" gear for sure. Heck, that power supply is more complicated than most amplifiers.
 
And so, the mystery begins. Understand that ARC aficionados may have been able to unravel parts of the mystery sooner than I, as while I have always been very familiar with the basic ARC circuit and theory of operation for their vintage models, I've never had any ARC gear actually cross my bench. Therefore, my working knowledge is quite limited in being able to immediately pick up on any physical irregularities. If anything then, this project has helped to fill in that void, if only by a little. To keep it all straight (if only for myself), I will try to present the mystery as it slowly unraveled to me.........

Chapter 1: The Chassis

BELOW: Hummm. This isn't like any D-115 I've ever seen (in pics that is) -- but that's clearly what the front panel indicates the unit to be, and the circuit board has the right tube count. OK. So somebody along the way wanted to install upgraded output connectors -- I get that. And, they only got one side of the upgraded output connectors done and the other side drilled out before work was halted for some reason. Now why they chose to make all four output connectors red is beyond me. The case can be made that either the Com or the 4Ω connector should be black for clear impedance and proper phase connections to be made. All the vacuum tube ARC designs use push-pull feedback from the output winding, so the 4Ω tap operates at ground level and therefore could support installing a black connector there. But since this fact is called out in print on the back panel, I plan to change the Com terminal to a black post so as to be traditionally color correct. But then, what's up with this balanced input provision and mount holes for (apparently) Cannon type input connectors? And no level controls? Off to the right (out of shot), and IEC power cord interface has also been installed -- something the D115 never came with -- but the power requirements printed next to it don't match up with D115 specs (700 watts indicated, but a D115 draws up to 1000 watts). And finally, as a premium bonus, the power transformer appears to be wound with dual primaries to support international service -- something that doesn't show up on any schematic for the D115 I've seen.

So, there's a combination of unique (but obviously) factory produced features (balanced input provision, international voltage capability), and after market work done (upgraded output and power connectors) to begin the opening chapter. But there's one other little detail sharp eyes might notice: Look at the city that (at least) the chassis was manufactured at: Minnetonka versus Minneapolis, as shown on standard D115 production models. Hummmm.........
SAM_3053.JPG

BELOW: And then there's this: No model number in the traditional location ARC provides it, and a stamped serial number that (through communication with ARC) they have no record of. And the darn chassis looks brand new to boot. This is starting to get interesting.........
SAM_3055.JPG

With no official record of the serial number, ARC's communication about the amplifier was, well, less than sparse. But besides no record of the amplifier, I was also able to at least glean from them that:

1. The D115 was never produced with a balanced input option, and
2. The numbers from the transformer set do in fact match to D115 pieces, and
3. International power transformers were available, and were used in both international and domestic units -- the latter maybe for units destined for industrial applications? Who knows.

4. And then there's this: When I knew this beast was headed my way, I went on line for a schematic. What I got was one indicating it used KT77s. Buzz was emphatic that it used 6550s. Humm. What appears to be a preliminary manual was provided with the unit (dated 6-20-84) -- made up of copies of typed pages bound with those black plastic binders as would be used with any presentation material, that was clearly labeled D-115MK II. And yep, it shows 6550s. So all that did, was muddy the waters even more. ARC confirmed that the original model came with KT77s, and that the MK II came with 6550s after that, but with no board or chassis markings as to the tube types required, who knows what this thing represents -- MK I or MK II? And the specifications page is titled as "D115MKII TENTATIVE SPECIFICATIONS" -- note the word "tentative". The tube complement of course changed from the original version, but so did other things: Sensitivity was now double the original unit. The amount of NFB applied was lowered, and there were other smaller changes -- but no mention of a balanced input option. And then even more muddiness: On the back of the manual, someone had written the serial number of the unit at the lower bottom left corner: "S/N 12460004 V-70" -- matching exactly that on the chassis...... except that it didn't. On the back of the manual, the first digit is a "1". That's what it looks like in the pic I provided. But close examination clearly shows that in reality it is a "7" -- but only close examination show that as the left side of the 7's top hat is only lightly stamped into the metal. Any casual look readily makes that digit look like a 1.

At this point, I'm wondering if this was a special factory request that got halted, or even a factory development or prototype model they decided not to continue forward with. I mean, with what (by all measures) appears to be a new chassis, new front panel (look back at the pic of it and you will see the protective masking tape still in place), and even new transformers (solder lugs never soldered to, and leads never cut with ends still tinned!), it's becoming abundantly clear that this thing never left the factory as a working unit, or was ever a working unit -- and we're still just on on the chassis at this point........

BELOW: A pic taken later on during the project, but shows the point that the transformer set connections are still as presented from the transformer vendor. The other output transformer looks identical:
SAM_3106.JPG

More research returned that in the 90's, ARC came to the revelation that balanced inputs were the wave of the future as the way to audio nervana. But apparently it was only ARC's revelation, as nobody else saw the light, and the corporation apparently damn near went bankrupt on that gamble. And then research happened to turn up a pic of the rear panel of a model V70 -- which I had never heard of and didn't even know existed. Now for me that was a revelation! There were the balanced input Cannon jacks, the 700 watts AC power requirement, and the Minnetonka manufacturing location! My chassis! No doubt about it, this was a model V70 type chassis (I say "type" since no model number was ever attached to it) -- which also then explained the V70 after the serial number written on the back on the manual provided. OK. Now some things are starting to come together: What I have is a front panel and an international transformer set for a D115, attached to a chassis designed for (at least) the V70 (which had been modified for upgraded output and power connectors), a tube cage that came from who knows what (only some of the attachment holes in the cage line up with those in the chassis), and a D115 MK II manual. As I said, parts and pieces, from.....somewhere.

Now this makes sense. Even though ARC quickly reverted back to more traditional input arrangements, it follows that they likely had a surplus of the ill fated "V" amplifier chassis as a result. The V series itself is a hybrid design using a SS input section with vacuum tube driver and output stages, and all of course employing balanced inputs. The V70 itself used push-pull-parallel 6550s in each channel, configured in triode mode, using a very similar physical layout to the D115. There may have been other models of the V type that used the same chassis since no model number was assigned to this particular chassis. By applying that later, it would allow for maximum use of the chassis with other models within the V line, so that probability is therefore likely quite high. But it still leaves the question as to how all of this came to be, and in the condition it was presented in. A possible answer to that came from ARC themselves in a final parting comment:

"We have no further information other than our service techs believe it is a prototype, engineering experiment, or something like that from a former employee. This was never an official ARC product. There is no documentation for it of any kind."

OK. I've squeezed out about all I can from the ARC source, but it seems reasonable enough: A former employee in later years was able to obtain a little bit of this and a little bit of that in the form of left over inventory replacement parts (front panel and transformer set), and production pieces (V chassis and cage), and was knowledgeable enough as to what would fit with what across model lines (sounds like a Parts Manager to me!!), so that he/she as able to pull all of this together. A 20 year term for retaining replacement parts is common -- which would have placed an origin for this project around 2003 (based on an introduction for the D115 in 1983), since any remaining replacement parts would have run their term at that time. A V model production chassis and a cage (from somewhere) would have (no doubt) still been available at that time as well. Against this possibility then, it is quite interesting to note another little detail in the form of a parts bag that was part of the "schtuff" in the box: A small press-seal plastic bag from Mouser for 300K, 0.5 watt, 1% resistors (very usable with this project), dated 07/22/03. Fits perfectly with the proposed time line for the origin of the project.

But how did all of this end up in a high end Arkansas audio salon? That's the last question to answer in this portion of the mystery, since I know the rest from there. But the likely answer to that is fascinating in itself, and came from a conversation with the salon owner widow herself. That, and Chapter 2: The Circuit Board -- and oh lord the story of it makes everything so far seem like small potatoes -- will be presented next time.

Dave
 
Last edited:
doesn't make sense that a non-functional display model would have a serial number though, unless whoever got their hands on the board decided to punch the chassis to match the number on the board?
 
One thing I forgot to add in the Chassis Chapter -- The bottom cover that came with this unit is clearly non-factory: A very nicely cut to fit piece of 1/16th inch solid aluminum -- much too thin for the weight of this beast (80 lbs -- with the chassis made out of 1/8th inch aluminum), so I'm going to annoy BuzzK for a proper one. Beside the weight factor, there is the heat factor: These things put out prodigious amounts of heat (the heaters alone disipate 115 watts of energy; the output tube and pass tube plates another 254 watts!) so that is a detail that ultimately must be properly addressed. I believe the factory pieces were steel with a very aggressive open venting design (almost entirely open), so clearly the piece provided was, I don't know -- either a casual after thought, or suggests someone along the way didn't realize the full reality of what this beast represents! At the very least, it means the factory was uh, out of stock!

Dave
 
This fits well the theory that someone at AR begun a prototype that was never finished, and that prototype
somehow got feets. Maybe it was shown/discussed with an importent dealer and forgotten.
 
Back
Top Bottom