K&K Audio / Lundahl Transformers
Welcome! Need support, you got it. Or share you ideas and experiences.
I talked to Kevin over the phone and he cleared this post, as I changed a lot of stuff (I just can't leave well enough alone). As Kevin put it, " sure go ahead, after all it's DIY", so here goes.
My brother had an old ST-70 that he had put a Triode Electronics board on, and it went bad. As a Christmas present I told him I'd re-do the ST-70. I was thinking about doing a driver board myself, but when I saw Kevin's kit, it was so close to what I had up my sleeve, I thought why re-invent the wheel, so we bought the premium kit without the capacitors.
The kit is very professionally put together, and the instructions are excellent. Even newbies shouldn't have any trouble with them. I've seen some kits before, but this is especially well done. All the parts are packaged very nicely (most important every part was there), the instructions come in a binder, and I can find no typos or mistakes. It even comes with a photo CD, VERY NICE!
Let me say right off the bat I never thought an ST-70 could sound this good. I realize that there isn't much ST-70 about it anymore, but I'm amazed of what those old output transformers are capable of! To say my brother and are are pleased is an understatement.
Here's where I parted with the kit...I like carbon film resistors, so I put in a lot of KOA Speer carbon film resistors in the signal path (pretty much throughout). I had a bunch of 70's era Soviet 6p3s-e tubes (sorta like 6L6s), and matched up a perfect quad, at least as close as my Heathkit T-1 tester could determine. I also had a bunch of early 70's 6n23p-evs, and matched up a pair with matched triode sections. The other thing I did was to use .113uf Russian SSG silver and micas for coupling caps.
To get the coupling caps on the board, I had to cast little silicone bases on the top of the caps with a couple of silver wire leads sticking out. The bases allowed me to drill some holes through the board for the leads to go through, and mount the caps lead side down, bringing the leads up through the backside of the board, and soldered to the pads. This allows not only for the space for the big Russian caps, but for very short leads from the caps, as well as a certain amount of cap damping from the silicone bases. Don't know that the damping and short leads really help, but sure can't hurt.
My brother had a miss-matched set of EL-34s that were pretty trashed, so I didn't even try them. Unfortunately this means I couldn't do a direct comparison with the Russian tubes. The CCSs make the tube change easy, just drop them in, and adjust. I adjusted the tubes to run at 46ma. At the voltages the power supply puts out, they run right at 20 watts dissipation. They bias (42V) a little lower than EL 34s, so they only get about 8.2 watts rather than 10, but I can live with that. They sure sound nice!
Surprisingly the 6n23p-evs required to deviate from the kit instructions a little. With the CCSs set at 15ma, there was about 190V on the plate, and the bias was at about 5V. This put the operating point pretty much in a non-linear part of the plate curves. It turns out that the earlier 70s EV versions have substantially less RP than the later production 6N23s. Looking at the schematic, I realized that if I turned up the CCS that the increased current draw dropped the B+ through the plate resistor and dropping resistor. I upped the CCS from 15 to 18ma, the B+ came down to about 125V, with a bias of 3V. This put the operating point in a more linear part of the curves (about the same as the operating points would be with the stock tubes), and the amp sounded a lot better. Of course with the increased current draw, the load and dropping resistors are working pretty close to their limits (more about this later).
When I first fired it up it sounded pretty thin and congested, but after a few hours it really started to open up. The only speakers we had at the time were some old 87db efficient NSTs. I was amazed that the little amp, in triode mode, drove them pretty well. I have since finished my Straight 8 clones and WOW does it sound good.
I've included some shots of the amp, my bread boarded 6n1p-ev parafeed transformer output preamp (lots of cheap iron, 11 inductors), and the tweaked and modded Straight 8 clones (I guess they're twystd 8s now).
One last thing, I started thinking about the driver stage load and dropping resistors, and decided to up the load, and lower the dropping resistor to get pretty much the same operating points on the 6n23p-evs. The thought here is that with the increased load, I should get a little more mu out of the driver stage, and a little less distortion. At the same time I could up the wattage of the resistors for a better safety factor. I sprung for four 5 watt 27K Kiwame load resistors, and two 5 watt 5.6K Kiwame dropping resistors. I could get the 5 watt dropping resistors on the board, but had to hang the four 5 watt load resistors on the bottom side of the board upside down. The resistors aren't broken in yet, only about 5 hours on them. I can already tell the higher load added a fair amount of detail, especially noticeable in things like cymbals, and other high end detail. I think I'm going to keep this mod, as the added detail hasn't led to any listener fatigue, and is quite impressive.
We put the amp on the test bench, and found out that it puts out 8.2 watts before clipping, and at 1 watt it is 1db down at 20hz, and the 3db down point is 27Khz. BTW, the amp does invert phase, but that works out because so does the preamp, besides it's no trouble to change phase with speaker cables.
My brother and I are very happy campers, can't believe an "ST-70" can sound that good, and highly recommend the driver board to anyone. Nice job Kevin.
Edits: 02/20/09This post is made possible by the generous support of people like you and our sponsors:
Topic - K&K ST-70 driver board. (long) - twystd 22:31:58 02/20/09 (2)