Since it seems I may be doing some work on a D70 MKII in the near future, I just want to confirm that items 4-7 and 9-10 likely would apply to it as well since its basically the same core design? Figuring it probably doesn't need a KT120 for the screen reg since its running half the tubes.
About all I know of it's current operational state is that it has a hum, and the tubes are a mixed lot. The big power supply caps also look like they have been tanning thanks to their proximity to the screen regulator so I have a feeling its going to be getting a power supply rebuild as well. Just thinking if its going to be getting the usual 40 year old amp service stuff, may as well take the opportunity to try and make it more reliable going forward.
About all I know of it's current operational state is that it has a hum, and the tubes are a mixed lot. The big power supply caps also look like they have been tanning thanks to their proximity to the screen regulator so I have a feeling its going to be getting a power supply rebuild as well. Just thinking if its going to be getting the usual 40 year old amp service stuff, may as well take the opportunity to try and make it more reliable going forward.
EPILOGUE
I would be remiss if I didn't add these final comments -- important so you know exactly what I've been listen to. That is, while I have strove to make this every bit as much of a D115 MK II amplifier as produced by ARC (within the physical limitations of the V-70 chassis provided), I have none the less made some electrical modifications to the design which for ease of assessment, I am summarizing here. They have all been discussed in detail throughout the course of the thread, but this will serve as a one stop reference so they can all be assessed as a total modification package. None of the modifications were installed based on subjective or emotionally driven observations but rather, were installed to address specific issues of concern relative to the D115, and this particular project. A couple of the modifications were intended to improve performance -- one to achieve that which ARC intended in the first place, and the other to produce improvement beyond the original design under certain operating conditions. However, the vast majority of the modifications were intended solely to address the dependability and durability of the amplifier, and to make it more socially acceptable in the 21st century audio world. The total modifications then include:
1. The D115 100K Input Level Controls have been replaced with a fixed 100K resistor and wired for maximum sensitivity. This technically is not an electrical modification at all, since from that standpoint, it represents no electrical difference to that when the level control of the stock design is set for maximum sensitivity. This was necessary because there was no room to mount these controls in the V-70 chassis.
2. Matched, compensated external attenuators have been provided to adjust the normal 0.6 vac input sensitivity to 2.10 vac. Normally of course these would not be needed, since the level controls could be backed off to a point producing a 2.1 vac input sensitivity. But with no input level controls, the external attenuators are needed to achieve the reduced sensitivity required for a good signal/noise ratio with typical vintage preamps. The modification for improved performance then is that these attenuators have been compensated, so that there is no loss of HF response when inserting them at the input of the amplifier. If the levels controls could have been installed, they are not compensated in the circuit as designed, so that with the controls set to an intermediate setting (such as to produce a 2.1 volt input sensitivity level), the loss in HF response of the amplifier would be notable. The compensation prevents this loss from happening, while the attenuator itself produces the needed drop in input sensitivity required, which together then represents improved performance over that of the original design when adjusted for reduced sensitivity levels. An additional benefit of these attenuators is that they raise the input impedance seen by the preamplifier from the stock (measured) 65KΩ of the D115 circuit, to 225KΩ, which is much more agreeable with the capabilities of most vintage preamps. The attenuators are external to the amplifier proper for the same reason there are no level controls installed in the amplifier: There is simply no room in the V-70 chassis to include them.
3. The right channel (only) response tailoring caps have been increased from the specified 5 pF value to a total of 10 pF each. This modification restores the proper supersonic response and stability margin in the right channel as intended by ARC, so as to produce and match the ideal response displayed in the left channel that uses the specified value.
4. The coupling capacitors into the output stage have been replaced with quality film caps of the same value, but rated for 630 vdc versus the 400 volt pieces originally installed. At turn on, these cap are subjected to a 500 volt surge for several seconds. While film caps are generally self healing at a point where/when breakdown occurs within the cap, enough breakdown occurrences over time compromise the cap. Since these couple into the output stage, they have been replaced for the safety of the output tubes.
5. The output tube screen stopper resistors have been changed from 47Ω 2 watt wire wound resistors to 100Ω 0.25 watt carbon film resistors. Apart from being rated at much too high a wattage rating for the application, it is never a good idea to introduce any inductance (which wire wound resistors have) into the screen grid circuit, as it can promote output tube arcing. The resistors installed provide better protection against the formation of garden variety and damaging level parasitic oscillations, allows for the sustained, full rated power output of the amplifier to be achieved with no loss of power versus that of the original resistors, and also allows the resistors to act as appropriate fuses should an output tube have catastrophic failure.
6. The output tube cathode current sampling resistors have been changed from 1Ω 2 watt wire wound flameproof resistors, to 1Ω 0.5 watt carbon film resistors. As with the screen stopper resistors, these resistors were rated at much too high a wattage rating for the application. The resistors installed allow for the sustained, full rated power output of the amplifier to be achieved, while also acting as appropriate fuses should an output tube have catastrophic failure.
7. The output tube raw plate/screen supply film bypass caps have been replaced with quality film caps of the same value, but rated for 630/250 vdc versus the 400/200 volt pieces originally installed. In both of these power supplies, the original caps were being continuously operated over their rated voltage, and even more so during the turn on surge period. The new caps operate within their voltage ratings at all times.
8. The 6550 Screen Grid Regulator Pass Tube has been changed to a KT120 tube. No circuit changes were required to make this change. Under the conditions of full power output in both channels, this tube is required to dissipate nearly 70 watts from its plate and screen grid elements, and pass nearly 235 mA, yet as connected, the 6550 tube originally specified for this position is only capable of safely dissipating 48 watts and passing 190 mA. On the other hand, the KT120 is capable of dissipating 68 watts from these elements and passing 230 mA, making it a much more capable tube for the position, while only drawing 15% more current for its heater than the specified tube. Even under quiescent conditions, the pass tube is required to dissipate nearly 36 watts (passing nearly 100 mA), so it's hardly taking a walk in the park, but clearly working hard at all times. In fairness to ARC, the KT120 wasn't even a thought in 1983 let alone available, and under most normal use conditions, the 6550 will do the job anyway -- although its still having to work hard for its supper. And, the idea of using two 6550s would have really changed the physical landscape of the amplifier because the real estate is simply not available in it to add an additional pass tube as manufactured. Today however the KT120 is available, providing a real increase in pass tube capabilities with virtually zero down sides to its use, and safely allows for the amplifier's sustained full rated power output to be developed in both channels when operating at the same time. The simple change to this tube therefore represents a major improvement to the power supply's capability (as well as in dependability and durability), while its taller profile also still fits under the tube cage, so it's use represents a true win-win-win modification.
9. The two 3-Terminal Heater Regulators have had their Input and Output terminals partially bypassed together by external resistors. As designed, these regulators are each dissipating 2 watts, with each regulator mounted on a 2 watt heat sink, yet both are located very near to one bank of output tubes. As a result, the heat sink temps routinely hit 200˚F with even short 20 minute listening sessions. The regulators are therefore operating dangerously close to internal current limiting as designed. By partially bypassing their input to output terminals with an appropriate resistance, regulator dissipation has all but been cut in half, while the effective regulation performance of the regulators remains virtually unchanged. Noise on the output rail of the regulators even dropped slightly with the bypass resistance in place. As a result, heat sink temps now rarely exceed 150˚F -- with most of this being the result of the sinks absorbing heat from the nearby tubes -- even with several hours of extended operation.
10. A timer/relay circuit has been installed to delay the application of screen grid voltage to the output tubes by 45" from turn on, and instantly remove this voltage when power is removed. I must admit, this is a most puzzling aspect of this amplifier's operation, as produced and sold by ARC. Not puzzling as in what the problem is or how to resolve it, but that ARC would even allow this amplifier to be produced while having this problem, and particularly so for an amplifier of otherwise such high caliber, and not to mention, very great expense.
As designed, and principally because of the use of a single ended balanced paraphase phase inverter, there is a significantly major LF thump-pop-whallop-whatever-you-want-to-call-it from the speakers at shut down. When this happens with efficient speakers like mine connected(101db), I first seriously wondered if the bass driver's voice coil hadn't been pushed clear out of it's gap, the cat when flying, and the wife in the room above called down to ask if everything was OK. I was seriously startled. Without speakers connected, I have no doubt this event would have initiated an output tube arc. But whether using no speakers, efficient speakers or not, this event is seriously -- seriously hard on the output tubes and regulator pass tube, with all of them showing a major blue pulse when shutdown occurs. Now granted, it's a rather smallish, and very quiet listening room with big speakers -- so any event like this is going to be "amplified" so to speak. But I'm sorry, in calling it as I see it, this is a serious design flaw if only by the omission of addressing it. In my studies some 30 years ago regarding this very exact same type of issue (with amplifiers of my own design employing major power supply reserves, regulation, and large coupling time constants), you have a perfect storm that is extremely damaging to the output tubes: Huge reserve power supply capacitance that all but maintains full power supply voltage at the moment of shut down, cooling heaters due to the shut down (meaning a cooling cathodes), direct coupling and huge R/C coupling factors throughout the design for near unrestricted LF response, and a larger than full power pulse applied to the output tube grids, and you then have a recipe for output tube destruction. With the heater power removed, the electron cloud around the cathode is collapsing so that when a fully saturating pulse signal is then applied to the tubes with ample power supply voltage left to draw from, it means that the tubes then draw electrons directly from the cathode itself, which then produces the irreversible damage -- and it happens every time the amplifier is shut down. In my studies, output tubes in this type of environment were seriously compromised after just 250 hours of use -- this from quality brand tubes (GE 6550A) that should last at least ten times that long -- and still not be compromised to the level that the tubes in my study were. NO WONDER amplifiers of this type design were going through output tubes at a fantastic rate, not to mention the potential collateral damage caused with blown cathode resistors and burnt boards. I note too that at some point, ARC started mounting all the board components high off the board (they are in the D115), no doubt to prevent damage to it from component flame outs being a major reason why. Everyone always said it sounded so much better when they put in a fresh set of tubes -- and I bet it did! -- because the old ones were so thoroughly worn out so quickly. The answer isn't to change the tubes, but to fix the problem. The timer and delay relay are the perfect fix, so that at shut down, all you hear is a small click in the speakers, making the event such a non-event, that you don't even make notice of it. And, because of the timer, fast power cycling is a non-event as well. The amplifier merely shuts off, turns back on when the power returns (the power down interval can even be for just fractions of a second), and then 45 seconds later, its completely and uneventfully ready to go again. It is another win-win solution, with no down sides.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
So there you have all the modifications I did to my D115/V70 amplifier, and why. Some were necessary due to the nature of this project, most were small detail in nature -- and yet even small details can surely add up to have a significant impact on dependability and enjoyment -- while the last one was akin to fumbling the ball on the one yard line and the opposing team then running the ball all the way back for a touchdown to win the Superbowl. It was a major error. I note that some earlier ARC designs used a two-stage balanced paraphase inverter, that possibly helped to somewhat reduce the large turnoff transient, but either way, the solution presented completely resolves the issue, such that with all the modifications taken collectively, I am highly confident that the dependability and durability of this particular amplifier has been improved immeasurably through to a complete resolve -- and I know its grace and decorum in the listening room certainly has! Which sets up my final comment:
While Bill is an equal member of the three man innovation team I mentioned earlier -- and I hardly want the turn off transient issue to cloud his accomplishments and contributions in that area -- there is none the less one thing that does separate him from the other two. Back in the day, you bought McIntosh equipment, brought it home, hooked it up, and it just ran beautifully and uneventfully for years and more often, decades to come. You bought a Dynakit, built it, turned it on, and it too ran beautifully and uneventfully for years and years. Just look how many vintage examples of these brands are still around and available today -- even after all those that got (unfortunately) trashed, or sold out of the country. I can only wonder then, had the dependability issues with (some) of their early models been addressed early on, the rest of the build and design is such that there's no telling how much faster ARC might have grown at the time. Granted, they were building their business on a fastly declining concept at the time, but for the crowd they were playing to, that didn't really matter. Unfortunately though, the dependability concerns of the early power amplifier models set in, and the image -- retrospectively -- became what it was -- beautiful equipment to behold and wonderful to listen to, but temperamental and unpolished. And that is exactly how I would describe the D115 as designed and produced. But modified as detailed above, the temperamental and unpolished aspects completely disappear, making this an amplifier that operating as it does now, I surely would have saved for and seriously sought out back in the day. It even has me seriously considering looking for a suitable SP-3A-1 preamp now to go with it. But I digress. Next up, with all the emotionally (and real) distracting issues resolved, I'll present my final installment: my unclouded listening impressions.
Dave
