Dynaco SCA-35 Restore
- Thread starter Dave451
- Start date
Well, here we are well past the mid point of the restoration of this SCA-35. Things are coming along pretty well. All three boards are installed, as are the transformers and much of the front and rear panel wiring is now complete. Please forgive my using blue wire instead of green for the new hook ups--it's what I had on hand (kind of like it, too--matches the PCB's?). What remains is to re-wire the pre-amp PC-11 board, including all its' connections to the rotary switch. I have removed all the original wires (carefully!) from the switch and cleaned it up physically to remove the surface tarnish and rust. Once everything is assembled and checked, I'll clean the switch contacts on it and all the other switches and controls.
As advised, I'm trying to pay close attention to the lead dress specified in the original assembly instructions. If I don't twist another pair of wires again soon, that will be OK. The transformer leads were tough because many were just barely long enough in the original install. As Dave G recommended, I desoldered them instead of cutting to preserve as much wire as possible and just made it without any extensions (the blue heat shrink on one OPT is where I accidentally cut it during the painting process
). The old, cloth-covered wire is fragile and I tried my best to handle it carefully and not overwork it during the reinstall.
I put in a new line cord. It's polarized, so I made sure the hot (black) lead went to the switch and the neutral (white) wire the other way. As you may notice, I installed a CL-80 current in rush protector downstream of the 600W unswitched AC outlet. I do this for all the amps and receivers I rehab. The fuse holder is new as well.
As advised, I'm trying to pay close attention to the lead dress specified in the original assembly instructions. If I don't twist another pair of wires again soon, that will be OK. The transformer leads were tough because many were just barely long enough in the original install. As Dave G recommended, I desoldered them instead of cutting to preserve as much wire as possible and just made it without any extensions (the blue heat shrink on one OPT is where I accidentally cut it during the painting process
). The old, cloth-covered wire is fragile and I tried my best to handle it carefully and not overwork it during the reinstall. I put in a new line cord. It's polarized, so I made sure the hot (black) lead went to the switch and the neutral (white) wire the other way. As you may notice, I installed a CL-80 current in rush protector downstream of the 600W unswitched AC outlet. I do this for all the amps and receivers I rehab. The fuse holder is new as well.
Attachments
Do these have fuses in them? If not, it wouldn't be a bad idea to add one. I like them between the hot lead of the cord and the power switch. In my PAS2, I used an internal fuse holder piggy-backed on an existing mounting screw. I have to pull the bottom off to change it, but it has one and it has no visible mods to the chassis.
I completed the wiring in three sessions as I had time, reading back and forth between the Dave Gillespie instructions and the original assembly manual. The prior assembler trimmed the leads close on the power transformer HV secondaries and the secondaries to the rear output transformer but got them re-soldered without extensions. The color codes from the old cloth-covered wires were hard to decipher, but I was pretty sure I got them right (I had labeled each lead before removing the transformers for clean up and painting). Last part was to wire the selector switch after installing all the wire leads to the PC-11 pre-amp board. Took care to route the low-level input wires (Tape Head and Phono) to the switch exactly as described in the manual to avoid hum. The instructions in the original assembly manual were very clear.
Well, I checked the boards and wiring again and found two errors underneath, which I corrected and decided it was time to re-tube and power up. I kept two of the 6BQ5's that came with the unit that tested best and added two I had on hand that were pretty close to the other two for mutual conductance on my B&K 707. One pair tested 92 and 93, the other 94 and 98 (these are all good, solid readings in the 120 max scale of the 707--it provides a "%" type reading, not an actual reading in micro mhos). Although the original two were labeled RCA, they had dots indicating GE under the tube designation and the ones I had on hand were also GE. I installed two NOS JAN Philips/EGC 6U8A I got from Jim McShane that were well-matched section by section. I used the original 12AX7's that came with the unit. They both tested very well and matched closely (RCA, "Great Britain", looked like Mullard, but I didn't examine them very closely--seams on top of the bottles).
I put in a new slow-blow fuse and hooked it up to the dim bulb tester and Variac and the bench speakers and crossed my fingers (my son-in-law was on hand to observe and provide moral support!). No smoke and everything worked fine as I brought the voltage up on the Variac. Sound came out when I hooked the iPhone to the "spare inputs." The amp was dead quiet at full volume into my bench speakers on the line level inputs, with a very slight hum and very low tube rush at full volume on Phono and Tape Head (low-level) inputs. Without the bottom plate and cover, I figured this was OK for now.
B+ voltages checked out well (VDC with specs in parentheses) : PC11 HV = 223 (235), PC10 6U8A HV = 331 (335), PC10 6BQ5 HV = 389 (390). These were with the line voltage at 121.8 before the CL-80 and 119.4 VAC after. Checked all the tube voltages and no anomalies. 6BQ5 plates were 384VDC, screens 387 VDC. Filament voltages good at 6.2 VAC. Interesting the screen voltages are a smidge higher than the plates--any cause for concern?
I set the bias using the test points on the Power Supply/EFB board. The two sets of tubes weren't awful on match, but not what you would call a matched quad. I went for 270 mV on average between LC and RC (as Dave G instructs), but got the average with the left channel at 250 mV and the right channel at 290 mV. This is no doubt due to the age of the tubes, particularly the left channel. This would give 25 mA cathode current for the left channel and 29 mA for the right. The difference is born out in the power test below.
I adjusted the new 'hum balance' pots for minimum hum in the Phono mode and everything responded well with a very distinct (and low level) minimum.
I tested output power an the onset of clipping (1 kHz, single channel driven, 8 ohm non-inductive load) and got 15.1 WRMS for the left channel and 15.8 WRMS for the right with old output tubes, compared to 17.5 WRMS specification. This is a little light, but I suspect is due to the age and state of the output tubes.
I put new feet on the bottom and put the bottom plate on the amp after playing the amp for a couple of hours on the bench into my very insensitive bench speakers. Volume was good and sound was quite good.
So, I took the newly-restored SCA-35 up to the listening room and hooked it up to my Klipsch KG4's. Threw in a couple of jazz CD's (Miles Davis, Miles Ahead, and Chet Baker with Paul Desmond), and the little amp drives the speakers well and the sound is excellent as everything plays in
. I'm going to be playing it against my Scott 299B and comparing, but I am very well pleased with what I'm hearing at this point. This is a great sounding little amp!
Last step is to clean up and paint the cover. I am running it in the stack above the 299B and under my Scott 350B tuner without the cover right now, but it's out of the way and I know not to put my fingers in it, since there is high voltage running around the circuit boards on top of the chassis.
As you can see from the initial photos, it's come a long way from the "before state" and will make a very good sounding and reliable integrated amp for years to come, I believe.
Dave
Well, I checked the boards and wiring again and found two errors underneath, which I corrected and decided it was time to re-tube and power up. I kept two of the 6BQ5's that came with the unit that tested best and added two I had on hand that were pretty close to the other two for mutual conductance on my B&K 707. One pair tested 92 and 93, the other 94 and 98 (these are all good, solid readings in the 120 max scale of the 707--it provides a "%" type reading, not an actual reading in micro mhos). Although the original two were labeled RCA, they had dots indicating GE under the tube designation and the ones I had on hand were also GE. I installed two NOS JAN Philips/EGC 6U8A I got from Jim McShane that were well-matched section by section. I used the original 12AX7's that came with the unit. They both tested very well and matched closely (RCA, "Great Britain", looked like Mullard, but I didn't examine them very closely--seams on top of the bottles).
I put in a new slow-blow fuse and hooked it up to the dim bulb tester and Variac and the bench speakers and crossed my fingers (my son-in-law was on hand to observe and provide moral support!). No smoke and everything worked fine as I brought the voltage up on the Variac. Sound came out when I hooked the iPhone to the "spare inputs." The amp was dead quiet at full volume into my bench speakers on the line level inputs, with a very slight hum and very low tube rush at full volume on Phono and Tape Head (low-level) inputs. Without the bottom plate and cover, I figured this was OK for now.
B+ voltages checked out well (VDC with specs in parentheses) : PC11 HV = 223 (235), PC10 6U8A HV = 331 (335), PC10 6BQ5 HV = 389 (390). These were with the line voltage at 121.8 before the CL-80 and 119.4 VAC after. Checked all the tube voltages and no anomalies. 6BQ5 plates were 384VDC, screens 387 VDC. Filament voltages good at 6.2 VAC. Interesting the screen voltages are a smidge higher than the plates--any cause for concern?
I set the bias using the test points on the Power Supply/EFB board. The two sets of tubes weren't awful on match, but not what you would call a matched quad. I went for 270 mV on average between LC and RC (as Dave G instructs), but got the average with the left channel at 250 mV and the right channel at 290 mV. This is no doubt due to the age of the tubes, particularly the left channel. This would give 25 mA cathode current for the left channel and 29 mA for the right. The difference is born out in the power test below.
I adjusted the new 'hum balance' pots for minimum hum in the Phono mode and everything responded well with a very distinct (and low level) minimum.
I tested output power an the onset of clipping (1 kHz, single channel driven, 8 ohm non-inductive load) and got 15.1 WRMS for the left channel and 15.8 WRMS for the right with old output tubes, compared to 17.5 WRMS specification. This is a little light, but I suspect is due to the age and state of the output tubes.
I put new feet on the bottom and put the bottom plate on the amp after playing the amp for a couple of hours on the bench into my very insensitive bench speakers. Volume was good and sound was quite good.
So, I took the newly-restored SCA-35 up to the listening room and hooked it up to my Klipsch KG4's. Threw in a couple of jazz CD's (Miles Davis, Miles Ahead, and Chet Baker with Paul Desmond), and the little amp drives the speakers well and the sound is excellent as everything plays in
. I'm going to be playing it against my Scott 299B and comparing, but I am very well pleased with what I'm hearing at this point. This is a great sounding little amp! Last step is to clean up and paint the cover. I am running it in the stack above the 299B and under my Scott 350B tuner without the cover right now, but it's out of the way and I know not to put my fingers in it, since there is high voltage running around the circuit boards on top of the chassis.
As you can see from the initial photos, it's come a long way from the "before state" and will make a very good sounding and reliable integrated amp for years to come, I believe.
Dave
Attachments
That is a super job Dave! That sucker was dead and nearly buried before you started. So neat to see something so far gone brought back from the edge.
With a good set of tubes, you should have no problem achieving more than 17.5 watts of power from each channel individually, and right at 17.5 watts with both channels operating.
Congrats and enjoy!
Dave
With a good set of tubes, you should have no problem achieving more than 17.5 watts of power from each channel individually, and right at 17.5 watts with both channels operating.
Congrats and enjoy!
Dave
Thanks very much, guys--appreciate the kind words! This actually was a lot of fun--it was by far the worst initial condition restore that I've ever taken on and it is satisfying to bring something special back into playing shape. It does take a lot of effort, though--I spent a lot of time with my Dremel and a wire brush removing tarnish and rust from AC outlets, RCA jacks, and other widgets, let alone rust-busting on the chassis.
Speaking of special, I was pondering the screen vs. plate voltages and went back to the schematic and was reminded that this is an ultra-linear power amp circuit. It's the only UL circuit unit I've owned or worked on to my memory and is cathode biased to boot. It really is a special amp, which is one reason I didn't want to do any signal path or other mods beyond the very well engineered upgrades in Dave G's designs. As Dave G noted in his SCA-35 paper:
In classic Dynaco fashion, the Z-565 transformers...used in the SCA-35 are excellent performers. The best examples are capable of delivering the
full 20-20 kHz bandwidth at full rated power with very low distortion, although on the low end, the figure is often closer to 22 Hz. They also provide an optimized Ultra-Linear output stage configuration that gives the SCA-35 a clear advantage in damping and regulation of power delivered to a speaker load, versus that of an equivalent pentode output stage. The output transformers always determine the ultimate capability of a given design, and in the SCA-35, the transformers set a very high benchmark indeed.
That from "A New Look At an Old Friend--Improving the performance of Dynaco’s SCA-35 & ST-35 amplifiers," by Dave Gillespie. I recommend it for anyone interested in the SCA-35 as essential first reading. As I am learning, it is often "all about the iron."
I was reading up on ultra linear design and found this in Wikipedia:
'Ultra-linear' is a special case of 'distributed loading'; a circuit technique patented by Alan Blumlein in 1937 (Patent No. 496,883), although the name 'distributed loading' is probably due to Mullard.[1] In 1938 he applied for the US patent 2218902. The particular advantages of ultra-linear operation, and the name itself, were published by David Hafler and Herbert Keroes in the early 1950s through articles in the magazine "Audio Engineering" from the USA.[2] The special case of 'ultra linear' operation is sometimes confused with the more general principle of distributed loading.
If I am not mistaken, the SCA-35 is a David Hafler design and use of this design helps explain why you get the kind of performance from this relatively simple amp that has made it so popular for so long.
So, thanks to you, Dave G, for your outstanding work on the SCA-35 that inspired and enabled me and to Flyquail56, who provided the excellent PDF's of the wiring diagrams that were enormously helpful in the re-assembly work. Thanks too for all the input in the posts above from AK'ers. I really enjoy being a part of this community!
Dave
Speaking of special, I was pondering the screen vs. plate voltages and went back to the schematic and was reminded that this is an ultra-linear power amp circuit. It's the only UL circuit unit I've owned or worked on to my memory and is cathode biased to boot. It really is a special amp, which is one reason I didn't want to do any signal path or other mods beyond the very well engineered upgrades in Dave G's designs. As Dave G noted in his SCA-35 paper:
In classic Dynaco fashion, the Z-565 transformers...used in the SCA-35 are excellent performers. The best examples are capable of delivering the
full 20-20 kHz bandwidth at full rated power with very low distortion, although on the low end, the figure is often closer to 22 Hz. They also provide an optimized Ultra-Linear output stage configuration that gives the SCA-35 a clear advantage in damping and regulation of power delivered to a speaker load, versus that of an equivalent pentode output stage. The output transformers always determine the ultimate capability of a given design, and in the SCA-35, the transformers set a very high benchmark indeed.
That from "A New Look At an Old Friend--Improving the performance of Dynaco’s SCA-35 & ST-35 amplifiers," by Dave Gillespie. I recommend it for anyone interested in the SCA-35 as essential first reading. As I am learning, it is often "all about the iron."
I was reading up on ultra linear design and found this in Wikipedia:
'Ultra-linear' is a special case of 'distributed loading'; a circuit technique patented by Alan Blumlein in 1937 (Patent No. 496,883), although the name 'distributed loading' is probably due to Mullard.[1] In 1938 he applied for the US patent 2218902. The particular advantages of ultra-linear operation, and the name itself, were published by David Hafler and Herbert Keroes in the early 1950s through articles in the magazine "Audio Engineering" from the USA.[2] The special case of 'ultra linear' operation is sometimes confused with the more general principle of distributed loading.
If I am not mistaken, the SCA-35 is a David Hafler design and use of this design helps explain why you get the kind of performance from this relatively simple amp that has made it so popular for so long.
So, thanks to you, Dave G, for your outstanding work on the SCA-35 that inspired and enabled me and to Flyquail56, who provided the excellent PDF's of the wiring diagrams that were enormously helpful in the re-assembly work. Thanks too for all the input in the posts above from AK'ers. I really enjoy being a part of this community!
Dave
BinaryMike
Pelagic EE
You might find this article from 1952 interesting: http://www.keith-snook.info/wireles...ss-World-1952/Amplifiers-and-Superlatives.pdf
rshep
Working my way to 1000 posts
Nice restoration. If you haven't purchased new outputs yet I've been happy with the Sovtek EL84m's I'm running in my SCA35. Sound good and I hear they are very durable. My restore is partial, ECB power supply. I have the boards and parts for the rest. You're thread has inspired me to get the project done.
Rshep, glad your project is moving Ahead! Agree completely on the Sovtek EL84M''s. I used them in both Scott 299B''s I restored and have been very happy with them.
BinaryMike, I looked over the article and it is indeed interesting. While not exactly a feud, there is some academic disagreement over the claims of Hafler et al for ultra linear operating between the American and British schools of amplifier topology. Looks like the British are saying Hafler and Keroes are merely exploiting a sub-category of distributed load operation, and not well, but how to argue with someone as distinguished as D.T.N. Williamson? See this for a defense of the Hafler case, Williamson's critique notwithstanding: http://www.r-type.org/articles/art-139.htm
Thanks for Forwarding!
Dave
BinaryMike, I looked over the article and it is indeed interesting. While not exactly a feud, there is some academic disagreement over the claims of Hafler et al for ultra linear operating between the American and British schools of amplifier topology. Looks like the British are saying Hafler and Keroes are merely exploiting a sub-category of distributed load operation, and not well, but how to argue with someone as distinguished as D.T.N. Williamson? See this for a defense of the Hafler case, Williamson's critique notwithstanding: http://www.r-type.org/articles/art-139.htm
Thanks for Forwarding!
Dave
BinaryMike
Pelagic EE
I've seen this material several times and realized it was a very great shame that triode operation with grid power drive wasn't included in the investigation. This mode is far easier to implement now than it was at the time, and yields pentode output power with triode distortion. There's no longer any need to accept the compromises of UL tapped transformers, at least in the context of creating new tube amp designs.
Assuming you want to add a little sand to the design that is...... I'm not particularly opposed to that notion, either. But unless you do that, its back to the driver transformer.........
Dave
I'm not particularly opposed to that notion, either. But unless you do that, its back to the driver transformer.........Dave
BinaryMike
Pelagic EE
A few power tube datasheets, like the one for RCA's 6L6/6L6G, give triode curves with the grid driven well into positive territory. Grid current begins flowing as the grid goes positive, but linearity doesn't suffer as long as the driver keeps up with that requirement, and the power tube goes on to conduct about as much current as it could in pentode mode at low anode voltages. Driver circuits using cathode followers or driver transformers were used with some success in the past, but modern MOSFET source followers work spectacularly well. I've also used Sziklai pairs of ordinary BJT transistors with excellent results. A substantial negative voltage supply rail is usually required. Pete Millett sells a little class A2 driver PCB that's flexible enough for many retrofit projects as well as new construction. Maybe that's enough to stoke those little grey cells...
Hey, folks. I'm going to replace the aging output tubes in the SCA-35. Looking for thoughts on new production EL84/6BQ5's. I'm thinking the Tung Sol EL84 Re-Issues, but would certainly appreciate thoughts on experience and recommendations with new production tubes. Thanks!
Dave
Dave
