s-petersen
Scott
Does the circuit board look used?
Parts and Pieces: A True ARC Mystery
- Thread starter dcgillespie
- Start date
Origin Wrap Up -- I truly appreciate all the ideas and possibilities presented as to the origin of this piece. The idea of it being a display unit had not occurred to me, although I don't think this piece ever got built far enough to ever serve for that purpose. Ultimately of course, we'll never know for sure, but I know a heck of a lot more about it now than I did when it was first dropped off. As will be seen in the discussion on the circuit boards, they either provide some more answers, or only serve to deepen the mystery of the unit depending on how you look at it. Either way, when added to the preceding discussions, then the whole enchilada will have been presented. For now however, some final thoughts from the widow on the origin of this unit, as received over the course of a couple of emails from her (modified to protect her identity, and cherry picked for relevance):
"We traveled all over the country in our motor homes from 1973 till 2015. A lot of our travels were to visit factories of our suppliers. Among them was Audio Research. My husband and Bill Johnson became close friends after he visited our home in Mississippi and we became an ARC dealer. We visited him in California and several times in MN. Most of the talk was technical and what Bill should have done on this amp or preamp. They had a lot of friendly arguments. Between MN and attending every CES show in Vegas for many years he got to know all the sales and tech people from ARC pretty well.
I have no idea where the chassis or circuit board came from. I don’t remember it being in a box at any point which means it could have been put in the bay of our motor home on one of our trips. After he retired there was at least one box delivered daily with broken equipment, stuff to be refurbished and pieces of equipment that he just couldn’t resist the ‘bargain'.
The serial numbers are a real puzzle. If the board didn’t go with the chassis then how did the serial numbers get so close. Maybe this is something someone was trying to refurbish and after the chassis was silk screened they decided to sell it. My husband did not do any work on it to my knowledge and the only thing we ever had silk screened was McIntosh.
I sent you everything that was in the D115 file. I don’t know the origin or if it came with the chassis or if that schematic was something Bill sent my husband."
"I’m sorry I can’t shed any more light on the D115 kit. We moved it three time and it ended up with no box.
My husband was always improving on any electronics he could. This was probably one of his “I bet this would work. Some day I’ll do that”.
He was a wonderful dreamer but he couldn’t always get the doer going."
So there you have it. The salon owners became very good friends with the man Bill Johnson himself -- close enough that the most likely scenario is that this unit (such as it is in parts and pieces) passed directly from ARC to the husband, who apparently never got around to it -- or maybe started it with the upgraded output and IEC electrical connector work, but then stopped. Or, it was a project at ARC either as a company effort or an effort by one of the employees as a private project, that either way became abandoned -- with my feeling that it's the former scenario. Ultimately however, given the close friendship, it likely passed to the husband either directly via, or with the blessings of Bill Johnson, since the husband seemed to be pretty technically minded, which not all salon owners would necessarily be. Once the husband had it, it hibernated with him until it was transferred to me via BuzzK. Short of anything else or any new information that might come to light, that seems to be the most logical explanation - particularly when the peculiarities of the circuit boards are added to mix (which are plentiful) -- and which is coming up next.
Dave
"We traveled all over the country in our motor homes from 1973 till 2015. A lot of our travels were to visit factories of our suppliers. Among them was Audio Research. My husband and Bill Johnson became close friends after he visited our home in Mississippi and we became an ARC dealer. We visited him in California and several times in MN. Most of the talk was technical and what Bill should have done on this amp or preamp. They had a lot of friendly arguments. Between MN and attending every CES show in Vegas for many years he got to know all the sales and tech people from ARC pretty well.
I have no idea where the chassis or circuit board came from. I don’t remember it being in a box at any point which means it could have been put in the bay of our motor home on one of our trips. After he retired there was at least one box delivered daily with broken equipment, stuff to be refurbished and pieces of equipment that he just couldn’t resist the ‘bargain'.
The serial numbers are a real puzzle. If the board didn’t go with the chassis then how did the serial numbers get so close. Maybe this is something someone was trying to refurbish and after the chassis was silk screened they decided to sell it. My husband did not do any work on it to my knowledge and the only thing we ever had silk screened was McIntosh.
I sent you everything that was in the D115 file. I don’t know the origin or if it came with the chassis or if that schematic was something Bill sent my husband."
"I’m sorry I can’t shed any more light on the D115 kit. We moved it three time and it ended up with no box.
My husband was always improving on any electronics he could. This was probably one of his “I bet this would work. Some day I’ll do that”.
He was a wonderful dreamer but he couldn’t always get the doer going."
So there you have it. The salon owners became very good friends with the man Bill Johnson himself -- close enough that the most likely scenario is that this unit (such as it is in parts and pieces) passed directly from ARC to the husband, who apparently never got around to it -- or maybe started it with the upgraded output and IEC electrical connector work, but then stopped. Or, it was a project at ARC either as a company effort or an effort by one of the employees as a private project, that either way became abandoned -- with my feeling that it's the former scenario. Ultimately however, given the close friendship, it likely passed to the husband either directly via, or with the blessings of Bill Johnson, since the husband seemed to be pretty technically minded, which not all salon owners would necessarily be. Once the husband had it, it hibernated with him until it was transferred to me via BuzzK. Short of anything else or any new information that might come to light, that seems to be the most logical explanation - particularly when the peculiarities of the circuit boards are added to mix (which are plentiful) -- and which is coming up next.
Dave
Last edited:
Chapter 2: The Circuit Board
Geez, where do I begin? I guess my first effort was to identify what this thing was. By every indication, it is a D115 MK I board -- if there is such a thing. The original product was a D115, followed shortly thereafter by the D115 MK II, which was the main line production piece. The original D115 used 6DJ8s in all the small signal positions, and KT77 output tubes. The MK II replaced the KT77s with 6550s (most likely due to the availability of quality tubes at that time), and the driver and cross coupler small signal tubes with 6FQ7s -- an excellent move since the 6DJ8 in the driver position required all manner of protection components around it to deal with the over-voltage conditions (relative to the tube's ratings) that the tube could be subjected to during certain operating conditions. The cross coupler position is a high current position, and was cooking the 6DJ8 there dissipation wise relative to its rating, with both driver and CC positions requiring additional bias resistors to keep the current draw in line. I know the 6DJ8 is a more socially acceptable tube in high society settings relative to the work horse 6FQ7, but in this case, the 6DJ8 is no match for the FQ7's capabilities in these applications, so the move was a very good one. In the MK II, gone are all the protection components and extra bias resistors, making the circuit simpler, and the cross coupling action more effective, even though lower Gm tubes are now in place. The CC position is still a high current position in the circuit (each section passing 10.5 mA), causing each section of the 6FQ7 there to operate at 80% of its Design Maximum plate dissipation rating (based on both sections operating), so that puppy does get hot!
Quick sidebar -- it's interesting that the output tube sockets have all had their pin #1 terminals removed, and there are no facilities on the board to accept a pin #1 terminal were they installed in the sockets. This positively prevents the use of 6CA7/EL34 tubes, which was no doubt the goal of this move since their dissipation rating is lower than that of the KT77 -- a tube that is near universally accepted as a proper replacement for the 6CA7/EL34 -- and visa versa. This points to a significant difference then between the 77 and the other two types: In the KT77, the beam forming suppressor element is tied to the cathode pin (#8) as in 6550/KT88 tubes, where as in the 6CA7/EL34 it is not, but brought out to pin #1. But with no pin #1 present in these tube sockets, those tubes cannot be used then.........
The point of the previous discussion is to say that this board here was designed to support the original design using all 6DJ8 small signal tubes -- but -- it appears to have been factory modified to be a MK II board, as all the protection components for the driver stage are missing (no components were ever installed in those locations), and jumpers have been installed as appropriate (to bypass the original design's extra bias resistors) to configure the board for the MK II design. Since the jumper installations look easy enough to install (and appear to be installed when the board was stuffed), it's unclear whether there was a distinct MK I board and MK II board, or whether there was always just one board, with the MK II amplifiers having the appropriate adjustments made to their boards during stuffing. Either way, my board here is an original design board, but only partially stuffed with the MK II design components (maaany components missing) -- and is NEW. This board has never seen a millivolt of power applied to it, has never been operationally functional, and has never been installed in a working amplifier. There are far too may components missing for any circuit to operate at all, and where the components are missing, it is obvious to even the visually challenged that no components were ever installed in those locations since the board was manufactured...........in most locations, that is. Finally, this board shows zero discoloration from any exposure heat.
Uniquely (strangely?) however, all the components around the phase inverter tube were originally installed, but then at some point, removed! This design uses a single 6DJ8 as the phase inverter tube for both channels, but all of it's components were stolen in the night. The tube socket itself is present, but only lightly RE-soldered back in place on just one or two terminals. Now this is interesting. Remember the chassis is for a V70 design, with balanced inputs. If you remove the phase inverter tube from the D115 design, you could then apply balanced inputs directly to V2 and V3, the push-pull first AF Amplifier stages in each channel. Could that have been the original intention of the project back in 1990-something? To make a balanced input version of the D115? Of course it never got anywhere near far enough to try it out, so we'll never know for sure. BUT -- the grid return resistors for V2 and V3 (normally 475K) had 200K resistors installed, which then would have produce the 400K balanced input impedance specified for V type amplifiers and on the rear panel of the V70 chassis. Seems pretty reasonable to me! I mean, who would have intentionally wanted to modify the unit to a balanced input design, when history showed that nobody wanted it? So there's one more piece of the mystery to chew on. Maybe this board was in fact the factory effort to develop a "V" version of the D115, but when sales flat-lined, that's the real reason the project was scraped. New D115 front panel, new D115 transformers, new D115 board -- never completed and with the previously installed phase inverter components removed, a V type chassis with balanced input provisions to mount it all in, but sales of the concept go nowhere? This picture is really starting to come into focus! Whoever Dr. Frankenstein was however forgot that the heater of the phase inverter tube is wired in series with the heater of V3, so that if you completely remove V1, then V3 won't light and the left channel will be dead. Oops........ Wonder if that's why the phase inverter's tube socket was rather hastily reinstalled, soldered only on the heater pins and one other pin across from them to make for a reasonably secure re-installation...........
But there's so much more -- stuff that just doesn't make any sense. The board itself is fatally flawed in how some of the traces are run around the 3-T regulators (producing a direct short when the proper heat sinks are installed). This explains no doubt why wholly inadequate heat sinks were installed on the 3-T regulators (ones that don't secure to the board), and why the regulators themselves were all beat up with broken leads due to the board being tossed from here to there over the years. One transistor in the bias regulator circuit was actually installed backwards -- and clearly done so when the board was initially stuffed! A jumper was installed to place a dead short across the output of the bias rectifiers (maybe to prevent use of the board since it clearly failed to pass initial inspection?), and other components were of the wrong value by a digit or so (499K installed where a 4.99K should have been, 120Ω installed where 120K should have been, etc.). Also, this board has sockets installed for the IC error amplifiers -- which I'm not sure the production units used. Other components (unrelated to the phase inverter tube) that were originally installed had been removed as if to treat the board as a parts unit -- which it no doubt became in the years after the project stopped, and the unit found its way to the Arkansas audio salon. Heck it may even have been a parts board there as well.
In any event given all the previous information about the chassis and transformers, all of this strongly suggests that:
1. This was a D115 prototype board, produced before actual production boards were run. In pics I've seen of the D115, proper heat sinks were installed for the 3-T regulators, which is simply not possible on this board as designed, so with that, and incorrectly installed components, and incorrect value components installed, the board was likely shelved once the error were found and corrected for the production boards. (Edit: later information now suggests this is in fact a regular production D115 board, with a number of stuffing errors.)
2. During the 90s and the V period of ARC, things were (apparently) pretty lean if not bleak, so it follows that everything was scaled back significantly to keep the lights on. It also follows then that this board could easily have been pulled out of moth balls, and was in the process of being transformed into a V version of the D115 when the plug was pulled.
As a result, and with all the rest of the story that has preceded, the board leaves ARC an absolute mess, with layout errors, only about 60% stuffed, with some wrong components installed and some installed incorrectly, and stuck somewhere in between its intended design, and a V experimental board. No wonder it sat for years with no love. To that point, I have had to correct over 100 points of error on the board -- and I'm not talking about bad solder joints. Everyone of the points was of the type mentioned above, so that the only way to make the board properly operational, was to painstakingly go through it trace by trace, and component by component, so that all the errors were caught. But -- it came with a beautiful chassis, transformer set, and front panel!!
So where does the project stand today?
1. There are 3 boards to this amplifier: the main board, the driver regulator board, and the AC power relay board. The AC power relay board is MIA, so one will need to be made. The driver regulator board is clearly a similar -- and well used board (heat discoloration showing) -- but not the correct one for a D115. The voltages and terminal markings on the board do not match up with any D115 schematic, but as a regulator board, it is similar enough that it can be properly modified to do the job, as otherwise, it fits in place perfectly. That has been done and tested, so it is good to go. The main board has had all of its issues resolved, and has even had all of it's power supply and small signal circuits tested and proved out now, so that the output stages are the only elements of the board un-powered, and untested. But they have been thoroughly examined and verified to be without fault, although no tubes have been installed yet for operational verification. I did take the opportunity however to replace the insanely large (2W) screen and cathode resistors with appropriately sized components -- 0.5 watt cathode resistors, and 0.25 watt screen resistors (now 100Ω each). These resistor ratings will still allow the amplifier to produce full power output all day long if desired. Apparently, ARC amplifiers are somewhat known for these resistors blowing out -- and sometimes spectacularly so. Using the correctly sized components will minimize any damage done should this happen (sometimes burning a hole in the board!!). Increasing the screen stoppers from 47Ω to 100Ω will further minimize any tendencies towards arcing, and other potential issues relating to this tendency will also be examined as well once this beast is up and running.
2. In a real let down moment, it was determined that the board does fit in the chassis, but most of the mounting holes do not line up, as the circuit board for the V70 is slightly smaller -- it not needing the screen regulator circuits since it is a triode design. As a result, the chassis itself is slightly shorter (front to back by 0.75 inch), which will make things very tight, but with proper chassis pre-wiring planning, should be doable. The good news is that I have found that the general spacing between the board mounting holes is correct -- but just extended further out on each side on the D115 board. Once that was determined, I found that by using the original 5 mounting holes behind the transformers as a starting point to mount the new board, then the cage supplied could be made to work as well. But it requires re-drilling 11 holes collectively on the other three sides of the chassis to effect the board's proper mounting. That is in progress now.
So chassis work, chassis pre-wiring work, and a new AC power relay board are the order of the day before the main board can be permanently installed. It has been temporarily installed (using 5 original holes, and 5 new mount holes drilled) to check clearances, mark new mounting holes, and to power up the screen regulator circuits to ensure their proper operation, but has since been removed so the chassis work can continue. I'm also waiting on the material to make the other output terminal board to come in (slow boat from china) and, a proper bottom plate will need to be pursued as well. So things will slow a bit while the remaining logistics play out and come together. Until then, some pics to help illustrate the text discussion.
Below: Top side as received. Numerous parts never installed, various others removed. Circuit identifications on the board are mine.
Below: Bottom side as received. That is in fact 12 gauge stranded wire!!! Board notations are mine.
Below: Hey! Were'd yer phase inverter go???
Below: Various unrelated parts were removed as well. The project apparently finally got reduced to parts board status....
Below: Lots of parts were installed, but lots never were all over the board.
Below: As noted, the bias supply. But instead of a .22 uF 100 volt transient suppression cap installed, that black lead is a direct short across the output of the rectifiers!
Below: Bottom side of the board: Even with the short removed, the bias supply still has problems! That transistor on the left has already been replaced once before, but there's no way the circuit will work because as shown as received, it's installed backwards! The transistor on the right has obvious problems, and even with all of this, the bias regulator is still non-functional because of missing components that were never installed........
Below: Beat up 3-T regulators. Both have broken leads -- primarily because the wrong heat sinks are installed. The heat sinks attached are woefully inadequate for devices dissipating nearly 2 watts each even if they aren't in the D115 operating environment. You may note that the board is drilled to mount the proper heat sinks, but if installed, they will short directly across the IN and GND terminals of both devices! Woops......... Note in all the board shots the lack of any board discoloration.
Below: Note how much bigger the proper sinks are, and the mounting feet they have to secure them to the board, protecting the device mounted to them.
Below: A trial placement of the main electrolytic mount rings. Nearly two days worth of work was needed to carefully cut, trim, relieve, and orient the rings so that numerous parts on the board would not short to them, they would not interfere with the front chassis lip behind the transformers, and that the band tightening screws could still be accessed when installed. Incredibly tight clearances.......
More pics to follow.
Dave
Geez, where do I begin? I guess my first effort was to identify what this thing was. By every indication, it is a D115 MK I board -- if there is such a thing. The original product was a D115, followed shortly thereafter by the D115 MK II, which was the main line production piece. The original D115 used 6DJ8s in all the small signal positions, and KT77 output tubes. The MK II replaced the KT77s with 6550s (most likely due to the availability of quality tubes at that time), and the driver and cross coupler small signal tubes with 6FQ7s -- an excellent move since the 6DJ8 in the driver position required all manner of protection components around it to deal with the over-voltage conditions (relative to the tube's ratings) that the tube could be subjected to during certain operating conditions. The cross coupler position is a high current position, and was cooking the 6DJ8 there dissipation wise relative to its rating, with both driver and CC positions requiring additional bias resistors to keep the current draw in line. I know the 6DJ8 is a more socially acceptable tube in high society settings relative to the work horse 6FQ7, but in this case, the 6DJ8 is no match for the FQ7's capabilities in these applications, so the move was a very good one. In the MK II, gone are all the protection components and extra bias resistors, making the circuit simpler, and the cross coupling action more effective, even though lower Gm tubes are now in place. The CC position is still a high current position in the circuit (each section passing 10.5 mA), causing each section of the 6FQ7 there to operate at 80% of its Design Maximum plate dissipation rating (based on both sections operating), so that puppy does get hot!
Quick sidebar -- it's interesting that the output tube sockets have all had their pin #1 terminals removed, and there are no facilities on the board to accept a pin #1 terminal were they installed in the sockets. This positively prevents the use of 6CA7/EL34 tubes, which was no doubt the goal of this move since their dissipation rating is lower than that of the KT77 -- a tube that is near universally accepted as a proper replacement for the 6CA7/EL34 -- and visa versa. This points to a significant difference then between the 77 and the other two types: In the KT77, the beam forming suppressor element is tied to the cathode pin (#8) as in 6550/KT88 tubes, where as in the 6CA7/EL34 it is not, but brought out to pin #1. But with no pin #1 present in these tube sockets, those tubes cannot be used then.........
The point of the previous discussion is to say that this board here was designed to support the original design using all 6DJ8 small signal tubes -- but -- it appears to have been factory modified to be a MK II board, as all the protection components for the driver stage are missing (no components were ever installed in those locations), and jumpers have been installed as appropriate (to bypass the original design's extra bias resistors) to configure the board for the MK II design. Since the jumper installations look easy enough to install (and appear to be installed when the board was stuffed), it's unclear whether there was a distinct MK I board and MK II board, or whether there was always just one board, with the MK II amplifiers having the appropriate adjustments made to their boards during stuffing. Either way, my board here is an original design board, but only partially stuffed with the MK II design components (maaany components missing) -- and is NEW. This board has never seen a millivolt of power applied to it, has never been operationally functional, and has never been installed in a working amplifier. There are far too may components missing for any circuit to operate at all, and where the components are missing, it is obvious to even the visually challenged that no components were ever installed in those locations since the board was manufactured...........in most locations, that is. Finally, this board shows zero discoloration from any exposure heat.
Uniquely (strangely?) however, all the components around the phase inverter tube were originally installed, but then at some point, removed! This design uses a single 6DJ8 as the phase inverter tube for both channels, but all of it's components were stolen in the night. The tube socket itself is present, but only lightly RE-soldered back in place on just one or two terminals. Now this is interesting. Remember the chassis is for a V70 design, with balanced inputs. If you remove the phase inverter tube from the D115 design, you could then apply balanced inputs directly to V2 and V3, the push-pull first AF Amplifier stages in each channel. Could that have been the original intention of the project back in 1990-something? To make a balanced input version of the D115? Of course it never got anywhere near far enough to try it out, so we'll never know for sure. BUT -- the grid return resistors for V2 and V3 (normally 475K) had 200K resistors installed, which then would have produce the 400K balanced input impedance specified for V type amplifiers and on the rear panel of the V70 chassis. Seems pretty reasonable to me! I mean, who would have intentionally wanted to modify the unit to a balanced input design, when history showed that nobody wanted it? So there's one more piece of the mystery to chew on. Maybe this board was in fact the factory effort to develop a "V" version of the D115, but when sales flat-lined, that's the real reason the project was scraped. New D115 front panel, new D115 transformers, new D115 board -- never completed and with the previously installed phase inverter components removed, a V type chassis with balanced input provisions to mount it all in, but sales of the concept go nowhere? This picture is really starting to come into focus! Whoever Dr. Frankenstein was however forgot that the heater of the phase inverter tube is wired in series with the heater of V3, so that if you completely remove V1, then V3 won't light and the left channel will be dead. Oops........ Wonder if that's why the phase inverter's tube socket was rather hastily reinstalled, soldered only on the heater pins and one other pin across from them to make for a reasonably secure re-installation...........
But there's so much more -- stuff that just doesn't make any sense. The board itself is fatally flawed in how some of the traces are run around the 3-T regulators (producing a direct short when the proper heat sinks are installed). This explains no doubt why wholly inadequate heat sinks were installed on the 3-T regulators (ones that don't secure to the board), and why the regulators themselves were all beat up with broken leads due to the board being tossed from here to there over the years. One transistor in the bias regulator circuit was actually installed backwards -- and clearly done so when the board was initially stuffed! A jumper was installed to place a dead short across the output of the bias rectifiers (maybe to prevent use of the board since it clearly failed to pass initial inspection?), and other components were of the wrong value by a digit or so (499K installed where a 4.99K should have been, 120Ω installed where 120K should have been, etc.). Also, this board has sockets installed for the IC error amplifiers -- which I'm not sure the production units used. Other components (unrelated to the phase inverter tube) that were originally installed had been removed as if to treat the board as a parts unit -- which it no doubt became in the years after the project stopped, and the unit found its way to the Arkansas audio salon. Heck it may even have been a parts board there as well.
In any event given all the previous information about the chassis and transformers, all of this strongly suggests that:
1. This was a D115 prototype board, produced before actual production boards were run. In pics I've seen of the D115, proper heat sinks were installed for the 3-T regulators, which is simply not possible on this board as designed, so with that, and incorrectly installed components, and incorrect value components installed, the board was likely shelved once the error were found and corrected for the production boards. (Edit: later information now suggests this is in fact a regular production D115 board, with a number of stuffing errors.)
2. During the 90s and the V period of ARC, things were (apparently) pretty lean if not bleak, so it follows that everything was scaled back significantly to keep the lights on. It also follows then that this board could easily have been pulled out of moth balls, and was in the process of being transformed into a V version of the D115 when the plug was pulled.
As a result, and with all the rest of the story that has preceded, the board leaves ARC an absolute mess, with layout errors, only about 60% stuffed, with some wrong components installed and some installed incorrectly, and stuck somewhere in between its intended design, and a V experimental board. No wonder it sat for years with no love. To that point, I have had to correct over 100 points of error on the board -- and I'm not talking about bad solder joints. Everyone of the points was of the type mentioned above, so that the only way to make the board properly operational, was to painstakingly go through it trace by trace, and component by component, so that all the errors were caught. But -- it came with a beautiful chassis, transformer set, and front panel!!
So where does the project stand today?
1. There are 3 boards to this amplifier: the main board, the driver regulator board, and the AC power relay board. The AC power relay board is MIA, so one will need to be made. The driver regulator board is clearly a similar -- and well used board (heat discoloration showing) -- but not the correct one for a D115. The voltages and terminal markings on the board do not match up with any D115 schematic, but as a regulator board, it is similar enough that it can be properly modified to do the job, as otherwise, it fits in place perfectly. That has been done and tested, so it is good to go. The main board has had all of its issues resolved, and has even had all of it's power supply and small signal circuits tested and proved out now, so that the output stages are the only elements of the board un-powered, and untested. But they have been thoroughly examined and verified to be without fault, although no tubes have been installed yet for operational verification. I did take the opportunity however to replace the insanely large (2W) screen and cathode resistors with appropriately sized components -- 0.5 watt cathode resistors, and 0.25 watt screen resistors (now 100Ω each). These resistor ratings will still allow the amplifier to produce full power output all day long if desired. Apparently, ARC amplifiers are somewhat known for these resistors blowing out -- and sometimes spectacularly so. Using the correctly sized components will minimize any damage done should this happen (sometimes burning a hole in the board!!). Increasing the screen stoppers from 47Ω to 100Ω will further minimize any tendencies towards arcing, and other potential issues relating to this tendency will also be examined as well once this beast is up and running.
2. In a real let down moment, it was determined that the board does fit in the chassis, but most of the mounting holes do not line up, as the circuit board for the V70 is slightly smaller -- it not needing the screen regulator circuits since it is a triode design. As a result, the chassis itself is slightly shorter (front to back by 0.75 inch), which will make things very tight, but with proper chassis pre-wiring planning, should be doable. The good news is that I have found that the general spacing between the board mounting holes is correct -- but just extended further out on each side on the D115 board. Once that was determined, I found that by using the original 5 mounting holes behind the transformers as a starting point to mount the new board, then the cage supplied could be made to work as well. But it requires re-drilling 11 holes collectively on the other three sides of the chassis to effect the board's proper mounting. That is in progress now.
So chassis work, chassis pre-wiring work, and a new AC power relay board are the order of the day before the main board can be permanently installed. It has been temporarily installed (using 5 original holes, and 5 new mount holes drilled) to check clearances, mark new mounting holes, and to power up the screen regulator circuits to ensure their proper operation, but has since been removed so the chassis work can continue. I'm also waiting on the material to make the other output terminal board to come in (slow boat from china) and, a proper bottom plate will need to be pursued as well. So things will slow a bit while the remaining logistics play out and come together. Until then, some pics to help illustrate the text discussion.
Below: Top side as received. Numerous parts never installed, various others removed. Circuit identifications on the board are mine.
Below: Bottom side as received. That is in fact 12 gauge stranded wire!!! Board notations are mine.
Below: Hey! Were'd yer phase inverter go???
Below: Various unrelated parts were removed as well. The project apparently finally got reduced to parts board status....
Below: Lots of parts were installed, but lots never were all over the board.
Below: As noted, the bias supply. But instead of a .22 uF 100 volt transient suppression cap installed, that black lead is a direct short across the output of the rectifiers!
Below: Bottom side of the board: Even with the short removed, the bias supply still has problems! That transistor on the left has already been replaced once before, but there's no way the circuit will work because as shown as received, it's installed backwards! The transistor on the right has obvious problems, and even with all of this, the bias regulator is still non-functional because of missing components that were never installed........
Below: Beat up 3-T regulators. Both have broken leads -- primarily because the wrong heat sinks are installed. The heat sinks attached are woefully inadequate for devices dissipating nearly 2 watts each even if they aren't in the D115 operating environment. You may note that the board is drilled to mount the proper heat sinks, but if installed, they will short directly across the IN and GND terminals of both devices! Woops......... Note in all the board shots the lack of any board discoloration.
Below: Note how much bigger the proper sinks are, and the mounting feet they have to secure them to the board, protecting the device mounted to them.
Below: A trial placement of the main electrolytic mount rings. Nearly two days worth of work was needed to carefully cut, trim, relieve, and orient the rings so that numerous parts on the board would not short to them, they would not interfere with the front chassis lip behind the transformers, and that the band tightening screws could still be accessed when installed. Incredibly tight clearances.......
More pics to follow.
Dave
Last edited:
More pics:
Below: Carefully study will show that both rings have been serious relieved to accommodate this resistor. It is the bleed resistor connected across the 3200 uF main power supply caps. Good lord, we don't need a short here!!!!
Below: A couple of layers of hi temp tape insulate the IN terminal trace on the board from the bottom of the heatsink, which operates at ground level.
Below: The used driver regulator board, from something other than a D115. Someone at least took pity on the poor pass transistor at attached dual heat sinks to it. But again, the device dissipates 2 watts, and both sinks together are still smaller than the proper sink, which as before, has provisions to be mounted to the board -- and as before, will also short the board out. Am I missing something here? Remember, I'm a vacuum tube guy.........
Below: Proper size cathode and screen resistors have now been installed in each channel, and home made test point meter lead connector loops have also been added to the board to make the bias setting process easier. Still need to police the area to remove the excess solder flux from the board in this shot.
Below: Coming along. The screen supply filter cap (1700 uF) and one of the 4 main B+ filter caps (800 uF each) have been checked and installed. Both caps have less than 1 mA of leakage! The red caps are standard ARC items to protect against shock from the otherwise exposed terminals were the caps not in place. The mount rings have all been finished up and mounted in place now. The other 3 main B+ caps don't need to be installed at this point, as the amplifier will be perfectly operational without them -- they are simply tied in parallel with the existing cap installed. And hey -- who says (only) 800 uF won't allow the amplifier to still produce full power output? Other amps of just as much power operate with far less capacitance! Those three caps may likely be the very last pieces installed, and can easily be done with the board fully mounted in place.
Below: Give me clearance, Clarance!
Below: There it is! (Phase Inverter circuits reinstalled).
Below: By mounting the board by the 5 front mounting holes, it properly centers the board so the cage also still works. The 5 holes at the back of the chassis have been drilled so that only the 3 on each side (and the most difficult because they are so close to the existing holes) remain to be drilled. The bottom most tube is the phase inverter tube, and is in fact touching the rear of the chassis cutout. But, it would in a real D115 as well were the chassis not relieved at that point to fully accommodate the tube. I still have that yet to do.........
Below: Yeah, drilling these are gonna be a little tougher. The other side is the same way........
Below: No doubt about it. The output terminals will absolutely need to be pre-wired before the board is installed!
Dave
Below: Carefully study will show that both rings have been serious relieved to accommodate this resistor. It is the bleed resistor connected across the 3200 uF main power supply caps. Good lord, we don't need a short here!!!!
Below: A couple of layers of hi temp tape insulate the IN terminal trace on the board from the bottom of the heatsink, which operates at ground level.
Below: The used driver regulator board, from something other than a D115. Someone at least took pity on the poor pass transistor at attached dual heat sinks to it. But again, the device dissipates 2 watts, and both sinks together are still smaller than the proper sink, which as before, has provisions to be mounted to the board -- and as before, will also short the board out. Am I missing something here? Remember, I'm a vacuum tube guy.........
Below: Proper size cathode and screen resistors have now been installed in each channel, and home made test point meter lead connector loops have also been added to the board to make the bias setting process easier. Still need to police the area to remove the excess solder flux from the board in this shot.
Below: Coming along. The screen supply filter cap (1700 uF) and one of the 4 main B+ filter caps (800 uF each) have been checked and installed. Both caps have less than 1 mA of leakage! The red caps are standard ARC items to protect against shock from the otherwise exposed terminals were the caps not in place. The mount rings have all been finished up and mounted in place now. The other 3 main B+ caps don't need to be installed at this point, as the amplifier will be perfectly operational without them -- they are simply tied in parallel with the existing cap installed. And hey -- who says (only) 800 uF won't allow the amplifier to still produce full power output? Other amps of just as much power operate with far less capacitance! Those three caps may likely be the very last pieces installed, and can easily be done with the board fully mounted in place.
Below: Give me clearance, Clarance!
Below: There it is! (Phase Inverter circuits reinstalled).
Below: By mounting the board by the 5 front mounting holes, it properly centers the board so the cage also still works. The 5 holes at the back of the chassis have been drilled so that only the 3 on each side (and the most difficult because they are so close to the existing holes) remain to be drilled. The bottom most tube is the phase inverter tube, and is in fact touching the rear of the chassis cutout. But, it would in a real D115 as well were the chassis not relieved at that point to fully accommodate the tube. I still have that yet to do.........
Below: Yeah, drilling these are gonna be a little tougher. The other side is the same way........
Below: No doubt about it. The output terminals will absolutely need to be pre-wired before the board is installed!
Dave
Last edited:
More misc. pics:
Below: There is zero room for error in mounting the board. Clearance on the front side between the mount rings and the front lip behind the transformers is nil. If they touch, then a ground loop is formed......
Below: On the rear side, the phase inverter tube literally touches the chassis cut out. The lines mark where a scallop must be cut into the chassis to provide adequate clearance -- just as ARC had to do.........
Below: A great example of a factory board stuffing error: The 499K resistor shown is supposed to be 4.99K resistor, and is the reason that a first test of the screen grid regulator circuit failed by producing an output of only 275 vdc. Replacing the resistor with a 4.7K resistor made the regulator output spot on at 312 vdc.
Below: All wired up for the final screen grid regulator test. All is good with the incorrect resistor replaced with a proper one.
Below: Full length transformer leads with tinned ends, and output transformer lugs that have never been soldered to. Unbelievable........
Below: The board is now fully tested, repaired, and operationally ready to go once the chassis work is completed so as to be ready for it. The small signal DC Balance controls have even been properly set for the 6DJ8 tubes installed, the AC balance controls centered, and the output tube bias controls set for maximum negative voltage to start the output tube biasing process from a cool point of operation, once that exercise begins.
I will update the thread as more progress is made, but this is were it all stands at this point. Hopefully, Bill is smiling down on the project........
Dave
Below: There is zero room for error in mounting the board. Clearance on the front side between the mount rings and the front lip behind the transformers is nil. If they touch, then a ground loop is formed......
Below: On the rear side, the phase inverter tube literally touches the chassis cut out. The lines mark where a scallop must be cut into the chassis to provide adequate clearance -- just as ARC had to do.........
Below: A great example of a factory board stuffing error: The 499K resistor shown is supposed to be 4.99K resistor, and is the reason that a first test of the screen grid regulator circuit failed by producing an output of only 275 vdc. Replacing the resistor with a 4.7K resistor made the regulator output spot on at 312 vdc.
Below: All wired up for the final screen grid regulator test. All is good with the incorrect resistor replaced with a proper one.
Below: Full length transformer leads with tinned ends, and output transformer lugs that have never been soldered to. Unbelievable........
Below: The board is now fully tested, repaired, and operationally ready to go once the chassis work is completed so as to be ready for it. The small signal DC Balance controls have even been properly set for the 6DJ8 tubes installed, the AC balance controls centered, and the output tube bias controls set for maximum negative voltage to start the output tube biasing process from a cool point of operation, once that exercise begins.
I will update the thread as more progress is made, but this is were it all stands at this point. Hopefully, Bill is smiling down on the project........
Dave
Last edited:
looks like the idea was to mount those regulators 180 from how they are. Maybe its just the angle of pics but it looks like there are holes in the ground trace spaced in a manner that might accomodate heat sink legs. The one with the doubled heat sinks almost looks like a rectangle pattern of holes. Not sure if thats the same trace or not, but if it is they seem placed such that it would work for a support. No guarantee that the heat sink was intended to be tied to ground though. Possible yours were meant to be hot and an insulator between the reg and the heat sink took care of making it not let the smoke out?
Lavane
Lunatic Member
The way the traces were made, if the heatsink legs were soldered in it would have shorted out.
tube-a-lou
Addicted Member
Another great Dave thread, looks like a great project and kit to work on. ARC are known for there
great sound.
great sound.
just looks to me that the holes are in a common trace, so I don't understand. Possible I'm just not seeing what you mean though. I tend to be of the belief that circuit boards are evil and have no business intersecting with vacuum tubes anyway.
Lavane
Lunatic Member
I was going by what Dave wrote aways down in post #23.
The IN and GRND termianls would be shorted if the heatsink was soldered to the board. He put a strip of tape between the heatsink and board. Looking at the picture, it looks like if the heatsink was installed upside down it would clear the board better.
Thanks for all the comments/suggestions on the 3-T heat sinks -- that's one I look at to assess every possible option. what it boiled down to is this:
1. The sinks are meant to be soldered in place, to allow the traces to aid in the radiation of heat.
2. The devices themselves of course cannot be turned around as then their leads don't match up with the board traces properly.
3. The board is only drilled for both heat sinks to be installed in the direction they are installed in: The front most regulator (nearest the transformers) could possibly be turned around if the board were drilled for it, but it wouldn't do any good as the base of the sink would still be going across the IN trace. With the sink grounded by virtue of the device mount tab, then the potential for a short is still there. You could insulate the device from the sink, but then cooling is compromised, and the sink would still likely bang into the mount ring of the big electro cap behind it -- a big problem if the sink is (again) not insulated from the IN trace.
4. The other regulator cannot be turned around, as then it would bang into the 7.5Ω dropping resistor behind it.
5. Every relevant pic I can find has the regulators mounted as shown. Either the board was modified to prevent the potential shorts, or the sinks had some sort of insulation applied to their bottom side.
I just didn't see any other way to mount them that offered any better solution. Turning the sinks upside down then exposes them to more tube heat from the nearby 6550s, and also (again) prevents the devices from being well secured to the board (other than by their leads) -- particularly if unmounted, upside down heat sinks are attached to them.
Thanks again --
Dave
Edit -- For clarity, it's not the soldering of the heat sinks to the board that's the problem. Whether soldered or not, the mounting of the devices to the sink grounds them anyway. The problem is that on the top side of the board, the bottom edge of the sink would literally sit on top of the IN trace were no insulation provided. With the sink grounded (one way or another), then the fireworks would begin........
1. The sinks are meant to be soldered in place, to allow the traces to aid in the radiation of heat.
2. The devices themselves of course cannot be turned around as then their leads don't match up with the board traces properly.
3. The board is only drilled for both heat sinks to be installed in the direction they are installed in: The front most regulator (nearest the transformers) could possibly be turned around if the board were drilled for it, but it wouldn't do any good as the base of the sink would still be going across the IN trace. With the sink grounded by virtue of the device mount tab, then the potential for a short is still there. You could insulate the device from the sink, but then cooling is compromised, and the sink would still likely bang into the mount ring of the big electro cap behind it -- a big problem if the sink is (again) not insulated from the IN trace.
4. The other regulator cannot be turned around, as then it would bang into the 7.5Ω dropping resistor behind it.
5. Every relevant pic I can find has the regulators mounted as shown. Either the board was modified to prevent the potential shorts, or the sinks had some sort of insulation applied to their bottom side.
I just didn't see any other way to mount them that offered any better solution. Turning the sinks upside down then exposes them to more tube heat from the nearby 6550s, and also (again) prevents the devices from being well secured to the board (other than by their leads) -- particularly if unmounted, upside down heat sinks are attached to them.
Thanks again --
Dave
Edit -- For clarity, it's not the soldering of the heat sinks to the board that's the problem. Whether soldered or not, the mounting of the devices to the sink grounds them anyway. The problem is that on the top side of the board, the bottom edge of the sink would literally sit on top of the IN trace were no insulation provided. With the sink grounded (one way or another), then the fireworks would begin........
Last edited:
Man. It almost sounds like someone was just learning how to run a computer-based board/circuit layout program, and hadn't figured out how to turn on the safeguards, to prevent conflicting component placement... and then the file produced wound up being accidentally sent to production...
Regards,
Gordon.
Regards,
Gordon.
More pics added to my post #25.
Dave
Dave
Gordon -- Remember this would have been about 1982 when this board was being developed (the board itself is dated 1983). We're talking almost 40 years ago -- did such programs exist to prevent component interaction errors for CAD designed boards then? That is something I am completely ignorant about.........
Dave
Dave
Dave -- if they will fit, you might consider some nylon capacitor clamps and sleep a bit easier:
https://au.rs-online.com/web/c/passive-components/capacitors/capacitor-mounting-clips/
https://au.rs-online.com/web/c/passive-components/capacitors/capacitor-mounting-clips/
Thanks Buzz -- Now that the originals have been modified for good clearance, I'll stick with them since they are what the original product used -- but surely nice to know the nylon clamps exist!
Dave
Dave