Dynaco-STEREO120-pwr-mod维修电路原理图.pdf

Audio Electronics 2/991 T he fate of my old Dynakit Stereo 120 was in grave doubt (one chan- nel noisy, the other oscillating) when Kenneth Millers TAA 3/92 article arrived. It described a simple upgrade to a modern design, using a JFET input and MOSFET output devices while still meet- ing all of Dynacos original specifications. Fate would spare the old ST-120. Years passed, however, before that issue of TAA surfaced again, from under a pile of other projects waiting on my desk. Still, it presented an opportunity to try a MOSFET design without having to start from scratch. Why not? Isnt that part of what this hobby is about? The other part is listening. The mod was well worth the effortsmooth, tight, and honest, with no noticeable “sound” of its own. It was an excellent, modern, high-end amplifier hiding in the body of a 30-year-old classic (Photo 1)! Discontinued Parts However, device obsolescence had taken its toll, creating new challenges for a once-proven design. Meeting those chal- lenges was an experience I shared with Ken, who was helpful beyond reason as I forged ahead. This began when I discovered the N- channel output MOSFETs were no longer available. I found what looked like a nice substitute pair, but, unfortunately, they proved unsuitable, requiring a consider- able bias increase to overcome their crossover notch distortion, which low- ered their output-power capability. I wrote to Erno Borbely, whose ampli- fiers used the same outputs, and to Ken, to see what suitable devices theyd found currently available. Both responded quickly. Erno suggested several pairs, with some helpful information about them. Ken still had some of the original type but, he, too, would soon face the problem of obsolescence. I decided on 2SK1058/2SJ162 output pairs, suggested by Erno, which are inter- nally the same as the original types, but in a plastic TO-3P case. Because of the plas- tic, they have a slightly lower power dissi- pation than the metal-cased originals. This wouldnt be a problem. Ken had al- ready told me that only one output pair was needed for an 8 load, but that two pairs were necessary to drive a 4 load. I opted for the two pairs for load flexibility. I quickly discovered that most of the other devices in Kens circuit were also no longer available. Moreover, from Mo- torolas excellent website, I learned that it had discontinued its entire MPSU line. Ken began researching this problem and came up with new devices and some cir- cuit changes to support them (see side- bar, “Replacing the Obsolete”). Chassis Mods I forged ahead with great enthusiasm, planning changes more extensive than Kens new amplifier cards. I cleaned off the old chassis, replacing the original chassis parts with new ones. Since I like maximum control with grounding arrangements, I added a small switch to lift the new, three-wire AC cord ground, if needed. On the other end of the chassis, I added a five-way ground terminal to allow chassis exterior grounding where and as needed. Such flexibility in ground- ing schemes is particularly helpful to tune a system for minimum hum when it includes devices utilizing internal grounding to the neutral side of the line, and other such commercial schemes. I also replaced the original RCA-type input jacks with some high-quality ones from Old Colony, leftovers from an earlier project. The new output terminals would be K.E. Langs Ultimate Speaker Connec- tors (nice, solid-brass 30A six-way bind- ing posts, with noncaptive thumb nuts, which accept flat-ring terminal lugs). The power transformer was good, but C12 and one of the output C7s had lugs broken off, with the old leads sol- dered to the rivets at the lug base. Who built this? That certainly wouldnt do in a new amplifier! Fortunately, we have here in central Florida a builders heaven in Skycraft Rodney E. Cavin MOSFET MOD FOR THEDYNA ST-120 This article describes an upgrade for the Dyna ST-120 MOSFET modification (TAA 3/92), providing currently available transistors, the original authors new parts values, and some useful construction techniques for this highly satisfying amplifier. PHOTO 1: The old classic, but almost everything under the hood is new. About The Author Rodney E. Cavin became an audiophile at age 8, when, after buying a windup phonograph, he tried to make it louder and sound better. A graduate of DeFor- rests Training and Navy electronic schools, he began in radio and TV engineering, worked in electronics de- velopment at Cape Canaveral, and then went into TV and film production. He currently does audio and video work for business conventions. He also writes film scripts, occasionally directs, and continues build- ing amplifiers and speaker systems to make them sound better. RadioFans.CN 收音机爱 好者资料库 Parts, a surplus electronics (and other parts) store. A little shopping, and I had replace- ment capacitors, but they were slightly taller than the English- made originals. Their height would put the screw terminals within spark-gap distance of the top cover. Since I planned to use new -high rubber feet under the chassis, I found the needed room below. I nibbled some- what crude holes in the chassis to allow the capacitors to ex- tend below it, providing ample clearance between the termi- nals and the cabinet top. The chassis didnt clean up as well as Id hoped, so I had a plating company strip it, then nickel and chrome plate it. I painted the top cage flat black to com- plete the popular color scheme. I man- aged to preserve the Dyna logo, a re- minder of the amplifiers heritage. How could this new design, sporting a chrome chassis, fail to live up to expectations? Next, I planned my parts needs. Most of the capacitors came from Welborne, along with W and larger resistors. The output MOSFETs came from MCM Elec- tronics, and the rest from Newark Elec- tronics. Two of the values in W resis- tors were not available in the resistor se- ries from Welborne, so I doubled the value and paralleled two of them, using the W, 1% resistors from Digi-Key. I filled a few of my capacitor needs from Skycraft, quality names I recognized from previously purchased Old Colony kits. Making the PC Boards I planned to make my own PC boards using the copier technique with TEC-200 film from Welborne. While its possible to use the circuit view provided in Kens article, I would need to reverse it to use the film process, which would make the final film a second copier step, or anoth- er generation, away. (Now that Ive made a few boards with the film, I see that the extra step probably would not cause much image degrading.) I also wished to rearrange the pads to accommodate some of my new parts (for example, a good tubular 6.8F instead of a minielectrolytic) and to make larger pads, because they hold up better for the parts changes that are sometimes neces- sary in home-builts. I also duplicated a couple of pad pairs so I could use differ- ent-sized parts, depending on availability. I started with Welbornes Assist pro- gram. Laying out boards is easy with this DOS computer program, but be aware that an HP Laser printer is neces- sary for final printing. I have an inkjet, which the program doesnt scale prop- erly. Since the program allows you to draw schematics that you can convert to board layouts, it is an excellent one. I hope Welborne upgrades it to allow more selection in printers. I finally turned to TurboCAD, an easy-to-use CAD program for Windows that I already possessed. I laid out the board from the component-side view, which gave me the proper final view for using film to transfer to the PC- board copper. This program also al- lowed me to add mirror images of letter- ing, which would read correctly after I flipped the film to transfer the resist pattern to the copper. To provide for two boards on a stan- dard 8 11 film sheet, I used the CAD program to copy a second board image below the first. I printed it on my inkjet, then using a copier, transferred the page to TEC-200 film, which I used to heat-transfer the resist image to the PC board. Making your own PC boards is really an easy process. If I can do it, so can you. Mono Transformer At my surplus-parts supermarket, I had discovered a power transformer very similar to Dynas original, so I decided to make a mono version. It would be my Dolby center-channel amplifier and also serve occasionally as a speaker-test am- plifier (Photo 2). Being a fan of heatsink overkill, I found some high-dissipation sink blocks, which I centered on the original Dyna sheet metal. Theyre really unnecessary, but they give the output stage a killer look, and they barely feel warm when the am- plifier is pushed. The TO-3P case design of the new out- put MOSFETs requires different mount- ing holes than Kens, or Dynas, original TO-3-style devices. In case youre not fa- miliar with them, the TO-3P devices look like oversized TO-220 devices: one mounting hole, and three holes (or a slot) for the leads if you plan to poke the leads through the heatsink (Photo 3). I drilled mounting holes for two MOSFETs on each end of the brackets, skirting around the original TP-3 mount- ing holes. When drilling the new mounting holes, be sure to remove any burrs around the holes which might cut into the mica insulating washers that youll mount between the MOSFETs and the heatsink. I find a gentle hand twist from a sharp, oversized drill bit does a nice job. I mounted the MOSFETs on the same side of the heatsink where the original devices were located, using a plastic shouldered washer between the mount- ing screw and the device, and a TO-3P mica insulator between the device and the bracket. Placing a generous coating of silicon heatsink compound between the MOSFET, insulator washer, and bracket is a necessity. You may prefer to mount the MOS- FETs on the inside of the brackets, which would require drilling only a mounting hole, thus eliminating holes for the leads. I did this in my mono ver- sion. As a custom builder, its your choice. I bent the leads of the MOSFETs at right angles to protrude through the heatsink. Short pieces of Teflontub- ing on each lead prevents shorting to the heatsink. Wiring Advice A few words about wiring: I have become a firm believer in the Walt Jung approach to high-quality wiring (see “POOGE-2,” 2Audio Electronics 2/99 PHOTO 2: The covered mono version built on a commercial chassis. PHOTO 3: The TO-3P MOSFETs with the killer blocks on the original heatsink. RadioFans.CN 收音机爱 好者资料库 Audio Electronics 2/993 TAA 4/81, p. 9ff), namely, using a number of #24 wire-wrap wires paralleled in a Litz-type approach (Photo 4). It costs a lit- tle more, but is more commonly available than when Walts article first appeared. For PC boards, I usually twist the ends together, enlarging the pad holes as required, and solder in the usual man- ner. However, its more practical to make small loops of all the individual leads for attachment to the MOSFET leads (Photo 5). If you use ordinary single-conductor solid or stranded wire, be sure to obtain solid-copper wire. Much of the common hookup wire today is plated steel. In view of the audio quality you are at- tempting to approach in these projects, dont handicap them with poor wire! In any case, you need some method of clamping the MOSFET wiring to the heatsinks to relieve strain on the thin leads of the plastic devices, which could easily break off (Photo 6). Ken used a different approach. He mounted tie- point strips at the MOSFETs to junction the device leads to circuit wiring. This is an even better strain relief. I tested the power supply with the two 300 7W resistors from the old boardsin series, as Ken suggested. Note that you must use a minimal load on this power supply to bring it into its regulatory range. With no load, the out- put will be a few volts high. Although providing 72V, my old sup- ply was poor on regulation, so I re- placed the transistors. Again, the origi- nal transistors are nearly impossible to find. The common TIP31C and TIP32C transistors became replacements for Q8 and Q7, respectively. Several choices are possible for Q9, including the readi- ly available 2N3055 and the MJ802. You can also use one of the former NPN out- put transistors for Q9, as I did. My power supply was still soft. After a little troubleshooting, I discovered the problem to lie in some of the original diodes, so I replaced them all. I used 1N5404s (3A 400 PIV) for the 3A original diodes, D47, and 1N4004s for the 1A diodes, D8, 9, 11, and 12, all eas- ily obtained. New Zeners With all else being new, I replaced the zener, D10, on general principles. As 58V zeners are no longer common, I installed a 1N4758A (56V, 1W) zener. It provided good regulation and with Kens modification of R21 to 1.8k82V. In my mono version, I used a 1N4759A (62V, 1W) and lowered R21 to 1.3k. This arrangement gave somewhat better regulation than the zener/resistor choice in the stereo version, and yielded the same output voltage. I also needed to in- crease R26 to 47k. Both changes were re- quired by the slightly higher secondary voltage of my new power transformer. Next I completed the amplifier PC cards, following Kens check-out proce- dure (Photo 7). These simple checks confirm that the cards are in working order. Time spent here can save an in- vestment in output MOFSETs, the most costly part of this modification. Since Im usually behind in my build- ing, I like to scan subsequent magazine issues for letters and error corrections. Doing this, I found a note (TAA 2/93) about the diodes in Kens article schematic. This correction is wrong! The diode polarities are correct as origi- nally published. A quick look at the cir- cuit, and reading Kens thorough de- scription, makes it obvious. In accordance with a fairly common audio modification, I also added bypass capacitors, 5.0F and 0.47F polypropy- lenes, across C11 and C12 in the power supply, and across each channel output capacitor, C7. Ive been doing this automatically since reading of the tech- nique in TAA years ago. Again, its a per- sonal choice. As another personal choice, I added 0.01F, 400V disc capacitors across PHOTO 4: Inside the modified Dyna ST-120, showing the Litz-type wiring; not as pretty, but sounds good. PHOTO 5: How the multiple wires of the Litz-type wiring are at- tached to the MOSFET leads. PHOTO 6: With the PC board raised, the method of clamping leads to the heatsink is visi- ble. The killer heatsink fins are on the back side. 4Audio Electronics 2/99 Replacing the Obsolete By Kenneth P. Miller Id like to thank Rod Cavin, as well as all the other AE readers whove built my mod, for their favorable “reviews” of my redesigned ST-120 amplifier. Such testi- mony confirms that my project is a cost- effective conversion, fun to build, and easy to listen to. All the details of modifying the amp were published in the 3/92 issue of TAA. If you try to modify your ST-120 from the original article, youll find, as Rod has mentioned, that the output t