Dynaco SCA 35 amplifier modifications

schematics of the stock SCA35 are at the bottom of this page

Idea for getting DC for preamp heaters in the SCA35 amplifier

Thinking of a way to get DC for the 12AX7s in the SCA35 preamp, to reduce heater hum. By passing cathode current from the output tubes. Yes, the old "preamp tube heaters as cathode resistors" trick. The four 6BQ5s look to be drawing 147ma of current thru the 95 ohm cathode resistor, to get about 13VDC. That current is close to that used by 12AX7s in 12V mode, but with 2 12AX7s I'd need 25.2VDC. But I could make that happen if I place a positive bias (in reference to ground, not the cathode!) on the grid 1's. And do that by grabbing the voltage between the 2 12AX7 heaters in series (see diagram). That will make the top end of the 2nd tube's heater be at 25.2VDC, or 12.6V above the grid bias. This is a little like the trick used to make a 5V regulator chip produce say 9V output (the tube cathode = regulator output, tube grid = regulator's "ground" pin, which is biased by a resistor voltage divider; which here are the preamp tubes' heaters). The actual heater voltage is a tiny bit low, but that actually helps the linearity of the 12AX7s.

The triode section of the 12AV6 is the same as half a 12AX7

Yeah, that higher cathode voltage effectively reduces the B+ supply voltage, in terms of the amount of audio output power I can get, but as today's powerlines are higher than back in the day, that shouldn't be much of an issue. In addition to giving me DC on the preamp tubes' heaters, this trick should also reduce the load on the power transformer a little, a little less than 4 watts.

Used some surface mount aluminum electrolytic caps underneath to bypass the heater midpoint that now feeds the grid resistors on the output tubes. cut the trace to ground, and attach the cap there. Did this to not introduce crud into the audio.

The preamp has almost no hum (you need the cage in place to shield those ribbon cables), and just some hiss. Unpacked my turntable and played a few records, sounds fine.


Mod to bring up the highs in the phono preamp

I have a RIAA test record "FEWR-A" from BSR Electronics. "For use with models FEW II, III and IV" but it shouldn't matter what turntable you play it on, to feed my SCA35's phono preamp. Oh, and regular music LPs sounded good after I was done with the tests and mod. Before these results and my modification, I noticed that the phono preamp seemed weak on the highs, so after I figured out the phono preamp schematic (making sense of those switches was a bit of a PITA) I simulated it, and it did come out a little low on the high end. Played with various component values, but the only one that did what I needed was to change the 330pF cap to a smaller value, 240pF. Which looks to get me closer to a proper RIAA curve, as seen in my results measuring the waveforms off the test record. (The tone controls were centered, as they might affect the "tape output" jack some).
Yummy!
And using an o'scope to measure the "tape output" jack's audio signal (after doing the mod), I tabulated this table (normalized at 900Hz) with excel. The only mod that made a difference in the curve was the 240pF cap. With the original 330pF cap, the high end was down about 3dB. Which my ears noticed. The other two caps look to be coupling caps only, and as I was replacing the 0.1 ceramic disc and the Dynaco light blue cap 0.022uF with WIMA MKP10's (Polypropylene, which IIRC should be a good choice for audio work) in 0.068uF and 0.033, which I had lots of, and I was out of 0.1uF's and no 0.022uF's on hand. Simulation showed that these didn't matter, and my tests seem to verify that.

SCA35 mod: convert that "tape head" input to line level

What I did was change the function of the "Tape Head" input from a low level phono like input to a line level input. Biggest pain was figuring out a way around the oddball switching arrangement of the selector switch. Eventually figured out a way, which was to use the phono-tapehead equalization portion of the switch to do a preselect of phono or the new line level "tape head" (I stuck with calling it by its old name, as changing the silkscreening on the front and rear panels would be a bit difficult and make it look neat). The equalization RC (the 1. 5 meg resistor and 3. 3nF cap) was moved to the circuit board and hardwired into the now RIAA only phono preamp. To avoid crosstalk from unselected inputs, I kept the "short 'em to ground" scheme. Also note I hard wired the preamp tube input to point 17, instead of making it go thru the switch like it used to. But I kept the switch connection on point 17, so the short to ground when unselected would still happen. And the wire in the phono and tape head input braid that used to connect to the "tape head" RCA jack I tied to ground (as additional shielding for the phono input line). I ran new shielded cable from the old RCA tape head jack to the above new line level switching.

Inserting an inrush current limiting thermistor on powerline side of power transformer

Here I did a few simple things. Main one was to add, on the primary side of the power transformer, a thermistor of the sort found in old PC power supplies and VGA monitors. The one that feeds their internal power supply circuits (primary side rectifier and filter caps), not the degaussing coil. These are usually about 15 ohms cold, and when they warm up the resistance drops to around 3 to 5 ohms. They look like disc ceramic caps. The one I used is colored green, they are commonly black. This limits inrush curent when you turn the amplifier on. Inrush from cold tube heaters, and from initial filter cap charging.
Another thing I did, was to replace the old silicon rectifier diodes with damper diodes that are rated 1500 PIV and average current of max 5A and max peak 20A. I had these on hand: I doubt that the currents will come close to the specified max values. The coil has no real purpose other than to act as a jumper wire between the diodes and the 100 ohm resistor I used to drop the B+ a few tens of volts (the 6BQ5s I was using got unhappy with the B+ a shaded too high (higher powerline voltage today), but now I have 7189s in this amp now). Left this resistor in as I won't miss a watt or two of peak audio output power, and not work the tubes so hard.
A third thing I also did was to figure out what to do with the chassis hole created when the unswitched back panel outlet broke apart. No suitable outlets avaliable, so I did a rather lame modification: Use a switch to select the mode the other back panel outlet will be in, switched by the front panel power switch, or be always on. Yes, a stupid feature, but I wanted to fill the empty hole in a way that looks like it has a purpose.
Diagrams of the stock SCA35: