Currently I am repairing a SAE Mark XXXI power amp that is capable of 50 watts RMS per channel (can go up to 70 watts RMS before clipping, however this is unobserved by me with the use of a sine wave generator, an oscilloscope and a load). I hope to at a later date to find the benchtest graphs of this amp, if they exist among my voluminous files. A few months ago, I overdriven the amp playing Pink Floyd "Wish you were here" CD and the 2.5 A slo-blo fuse blew. I replaced the fuse and powered it up and the fuse blew promptly. I opened up the chassis and looked around, didn't see any problems except for two electrolytic capacitors with melted plastic coverings splitting off their metal cases. I unsoldered these from the PCB and checked both capacitor ratings and esr with a PEAK Atlas esr meter. Both capacitors were originally rated 64 microfarads at 64 V. The top capacitor on the PCB was found to be 68.20 microfarad with an esr of 0.20 ohms. The bottom capacitor was found to be 82.14 microfarad and an esr of 0.14 ohms. The top half of the pcb is judged to be the A channel with the bottom half judged to be B channel, both being on the component side. The component side is towards the front of the amp.
Let's describe the age of the amp and how the amp was obtained by me. I went to college from 1972-1977, at one point during a summer vacation, I built a Dynaco tubed power amp and solid state preamp from kits. And during that period people from the Mcintosh company came by a local stereo shop to benchtest any amp the public brought in to that store. Their way of convincing the public that their amps were state of the art, I guess. So I had the Dynaco amp benched tested against the McIntosh amp and to everyone's surprise, the Dynaco was found to be virtually the same level as the McIntosh standard. The owner of the stereo store made an offer that I could not refuse, namely the SAE MARK XXXI power amp. Apparently I bought the kit for around a hundred bucks and the owner told me the SAE was retailing for 200 bucks, and it was brand new and boxed (more on that later). At the time, I was a poor student, so gaining an amp worth twice the value of the Dynaco was a steal. Of course many years later I rued the loss of the Dynaco simply because it was a tubed amp (plus the fact that Dynaco no longer sell these kits and if you wanted a Dynaco, you can probably find these on EBAY but selling for many times what I paid for in kit form) and I now realized that tubed anything of auditory nature are to be prized and never parted with. So given the years I attended college, I'd say the amp was probably made in those years. Further research of the internet discovered a brochure for SAE components (wardsweb.org) saying the Mark XXXI sold for 200 bucks around the years 76-81 and a copy of the Mark XXXIb schematic (wardsweb.org) shows the amp was probably in its second iteration as of 9/30/77 (the schematic doesn't match what I can observe of the amp I am working on now, to wit, the amp I have doesn't have a biasing transistor, a time-delayed relay circuit, a thermal fuse, a choke after output transistors (Darlington'ed on the PCB) that the Mark XXXIb has. Other differences of note,on the Mark XXXI the output transistors are all NPN and ganged together with low ohm wirewound resistors feeding into the output terminals whereas the Mark XXXIb have same number of output transistors but half are PNP's and the other half are NPN's with half the number of low ohm wirewound resistor feeding into a choke circuit and passing through a time-delayed relay circuit on its way to the output terminals, and further the power supply is sized about 4 farads higher than the Mark XXXI.
Upon closer look of the components used and seeing telltale rosin deposits, I noticed things that are not consistent with my first belief that I had gotten a brand new amplifier in the box from the owner of the stereo store. To wit, most of the diodes on the A channel side between the pre driver(?) transistor (M SJ5262) and the driver(?) transistor (M SJ5261) which are obvious due to the black cases with white bands being replaced with glass cases with black bands and resin deposits left behind on the trace side of the PCB. The B channel diodes are original and clean. Additionally the pre driver (?) and driver (?) transistors were also obviously replaced (rosin deposits) on the A channel only. In the area close to the middle of the PCB close to the M SJ5262 transistors, there are 4 discrete diodes in a bridged formation (original black cases with white bands and rosin deposits on the traces). A teardrop capacitor is not consistent with the original rectangular capacitors and is found on the A channel side between the M SJ5262 and M SJ5261 transistors (again rosin deposits). Two capacitors were found on the trace side of the PCB and the lack of rosin deposits leads me to believe they were modifications made by the SAE manufacturing process. As best I can tell, the ceramic disc capacitor runs from the collector side of the 2N3738 to the base side of the 2N4036, being on the B channel side. The other capacitor (teardrop type) being on the A channel side runs from the base side of 2N4036 to the base of the M SJ5262. Further signs of repairs are found on the A side output transistors all marked 466 above the SAE701 above the 7330 (marked in black ink on the cases). Looking at the back of the amplifier with the heatsink covers removed. the A channel ouput transistors are exposed and from the bottom left transistor (A1) moving to the top left transistor (A2) to the top right transistor (A3) to the bottom right transistor (A4). A1, A2, A3 are marked by someone's hand with black slashes. On the A4, it appears to be written as 1-28-75. I surmised that around that time, it has been repaired and probably predates my receipt of this amp. On the B channel output transistors from the bottom right transistor (B1) to the top right transistor (B2) to the top left transistor (B3) to the bottom left transistor (B4) are found to be unmarked and probably original to the amp. There are differences among these though and are intriguing little mysteries. B1,B2, and B4 are marked 466/SAE701/7242 (green inked) but B3 is marked 466/SAE701/7251 (black inked).
PEAK Atlas DC Pro results are as follows: A1 NPN silicon BJT hFE=6 at Ic=5.00mA; Vbe=0.575V at Ib=5.00mA; Ic leak=0.000mA, A2 NPN silicon BJT hFE=5 at Ic=5.00mA; Vbe=0.567V at Ib=5.00mA; Ic leak=0.000mA, A3 NPN silicon BJT hFE=6 at Ic=4.99mA; Vbe=0.572V at Ib=5.00mA; Ic leak=0.000mA, A4 NPN silicon BJT hFE=5 at Ic=5.01mA; Vbe=0.567V at Ib=5.00mA, Ic leak=0.000 mA, B1 NPN silicon BJT hFE=11 at Ic=5.00mA; Vbe=0.571V at Ib=5.00mA; Vce sat=0.09V at Ic=5.00mA and Ib=1.00mA; Ic leak=0.000mA, B2 NPN silicon BJT hFE=13 at Ic=5.00mA; Vbe=0.572V at Ib=5.00mA; Vce sat=0.080V at Ic=5.00mA and Ib=1.00mA; Ic leak=0.000mA, B3 NPN silicon BJT hFE=13 at Ic=5.01mA; Vbe=0.582V at Ib=5.00mA; Vce sat=0.05V at Ic=5.00mA and Ib=1.00mA; Ic leak=0.000mA, B4 NPN silicon BJT hFE=12 at Ic=5.12Ma; Vbe=0.563V at Ib=5.00mA; Vce sat=0.079V at Ic=5.00mA and Ib=1.00mA; Ic leak=0.000mA. At first glance it appears the transistors are ok due to no leak results but I am bothered by the doubling of hFE's from the A side to the B side and that the B side shows Vce sat results, which I need to do some further research on. I would appreciate comments on which is the better replacement transistor (2N3773 versus MJ21194 or other) for these already in place.
I plan to unsolder from the PCB the M SJ5261's, M SJ5262's, and 2N3738's for testing by the PEAK Atlas DC Pro at a later date. I would also appreciate an actual schematic (free source or not).
I am planning to recap all of the electrolytic caps (Holland brand, blue plastic see throughs with dark blue spirals, not made anymore except for NOS on EBAY, probably taken over by Phillips) with Mundorfs for the following old values, new values: 1) 10 mF 63 V; 10 mF 100 VDC, 2) 64 mF 64 V; 68 mF 100 VDC, 3) 220 mF 16 V; 220 mF 63 VDC, 4) 330 mF 63 V; 330 mF 63 VDC, 5) 470 mF 40 V; 470 mF 63 VDC, 6) 6000 mF 50 VDC GE brand; 6800 mF 100 V Nichicon brand (esr meter shown positive rail can was at 5760 mF esr of .03 ohms, negative rail can was at 5964 mF esr of .03 ohms). The Sensitron bridge (SLA-200-28; ~200 V 3 A rated) will be replaced by a Vishay 400V 8 A bridge. In the process of rebuilding the power supply, I will check voltages of the transformer output to verify whether it is 45 volt rails or whatever (schematic for SAE XXXIb says 45 volt rails). And the last thing will be to check all resistors and replace if necessary. Any comments would be appreciated. RoveronMars