• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Eico ST-40, Redo/Modo

Kegger

R.I.P. 1/12/1966 - 6/1/2017
Here are the mods/changes I made to this ST-40.
(Again another take on the ASL wave 20 build I did a while back)

First it's not fair to compare the original to the "upgraded" version the original was,
well original/all stock, caps, resistors, everything, BUT just a quick take, it was the
old school "very" vintage tube sound, thick an muddy, loose bass, not the very clear
sharp top end, no sparkle, no air. AGAIN it really is not a fair comparison as a proper
rebuild would have changed things a good amount I am sure. :) No I’m not saying all
vintage pieces sound like described above, some do and I was just using that for my
reference to try an put out there what I was hearing an how I would character some
of the old school tube setups that were produced.

(ON MOST TUBES)
As I’ve stated in the past unless I’m doing a regulated screen supply I’m not into the
pentode mode of operation, so here with no UL taps triode it was with a 130ohm res
connecting the plate to screen. In my experience when running triode output stages
feedback really is not needed so I did away with it an removed all the center channel
stuff along with it. The original grid resistors were off and I dropped them to 240K.
(pentode with no regulators does still work on certain tubes quite well though)


(next section help I did a little spice simulation for good values before building)
Others have mentioned this as well, the drive/gain in the output section is more then
is needed so I chose what I thought would give me less but enough drive and little to
change on the circuit, so I went with a 12AT7 same pinout as previous but 60mu both
sections where the original (Most know I’m sure) uses basically a 12AX7 100mu input (12DW7)
and 12AU7 20mu splitter in one bottle. All this took was changing the 2 plate resistors
on the inputs from 330k to 100k and the cathode resistor from 1.1k to 560ohm, I left
splitter values the same. In this config there is a few other tubes that can roll in here
as well with no changes needed like 5965, 6829, E80CC, E180CC ect. Grid resistors in
this section had drifted as well but were replaced with 470K.
(the splitter resistors could be 56k as well if wanted, better drive)

This amp is for a friend who has rather highish sensitivity speakers so power is not the
concern here and so I had no problems redoing this supply for more filtering but at that
cost of available voltage to the outputs. There are 2 RC stages of filtering before those
outputs see the voltage, consisting of 75ohm 12w WW resistor to 47uf cap then 33ohm
12w WW resistor to a 47uf and 180uf in parallel for roughly 230uf, nice good clean clear
power supply to start an amp off with, the next stage resistor was calculated in PSUDII
so the rest of the amp would get original voltages (or at least very close to it), screens
were no longer on the next section an today’s higher wall voltage actually made the 1.8k
resistor increase to 2.2k.

All 8 of the black .1uf coupling caps were changed to the green Russian .22uf K42 caps.

Here’s were I had to make a choice that I really went back an forth on, the 7591’s in this
amp were not very strong and I did not have any in my stash to put in here, but I do have
many 6L6 types around, and the voltage would be low enough that old school and the ST
shaped ones could be used, so yep I rewired the outputs for 6L6, 250 ohm shared cathode
resistor bypassed with 220uf running roughly 104ma per pair and the original grid resistors
had drifted, since I was changing them I opted for 240K, should keep any suspect vintage
tubes in line and since the coupling cap was upped were fine there to. (Running 6L6GC)

Only resistors that really seemed to drift in the amp were the grid resistors on the output
stage and driver stage plus the plates of the drivers.

All the controls cleaned with Deoxit, and I gotta say they are nice and quiet, also did the
loudness mod. I tapped straight into the grid of the 12AT7 driver from the Tape 1 Tape 2
jacks making the unit useable as strictly an amp or still using as an integrated going into
any of the other inputs.

I have left the 4ohm taps as ground and for the time being don’t see a need to change it,
amp performs well and sounds very good. If anyone has thoughts on this (or anything I'd
done here) I would be glad to hear about it.


---------------------------------------

No this is not a rebuild to keep the unit as original design, it was a mod for my preference
and what I felt would give a better performance for where this unit will be used, many like
the stock units performance and sound, I have no problem with that and if the extra power
is needed or original sound wanted then this would not be what someone would go after.

All in all I am very happy with how this unit turned out, it is dead quiet, not a hint of noise
on high sensitivity speakers, the mids are clean and clear, the high’s go out nicely with the
air I was after, overall a more “updated” and HIFI sound than most vintage amps tend to do
and the bass is not loose in the least, has nice detail and plenty there! I wish it was mine. :)

One kinda nice thing here as well when used as an integrated some of the prior “old school”
sound is sprinkled into the mix and could be quite pleasing at the same time.

Final tube compliment now is (1) 5AR4, (2) 12DW7 (now have 2 spares), (2) 12AT7, (4) 6L6.
(plus the phono section, (2) 12AX7)

New power supply and output section schematic is attached, I got the original voltages from
the build manual when I redid this unit, and these new voltages are marked on this schematic.
(I used the original schematic then made my changes, hopefully it is readable)
Original schem attached as well for reference.

Anyway hopefully someone finds this useful or at the least entertaining. :)
Keg
 

Attachments

  • top_989.jpg
    top_989.jpg
    81.4 KB · Views: 686
  • btm_cps_916.jpg
    btm_cps_916.jpg
    145.8 KB · Views: 635
  • ST_40_JP.JPG
    ST_40_JP.JPG
    72.8 KB · Views: 712
  • ST_40_m2_jp.JPG
    ST_40_m2_jp.JPG
    68.2 KB · Views: 686
Register to hide this ad
Hey, I have an old beat up ST40 in "the pile"
Looks like a worthwhile mod!
(and I'm a sucker for 6L6 sound so...)
THANKS for sharing Kegger!
WopOnTour

PS> Just curious,(regarding the removal of the feedback loop) did you "square-wave" the end result?
 
(thanks for the questions and the interest)
I swept from 30hz to 30K, looked decent an that's what I was concerned with.

When running Triode outputs I feel loop feedback isn't needed here, when running
in UL some of the time I don't feel the need, when running Pentode yah I am going
with some loop feedback more then likely if not other means of feedback as well. ;)

I build with low mu/low plate resistance triode drivers, so the gain stages certainly
are pretty well linear, if a quality output is used in UL or Triode not much is needed
to be done as far as feedbacks concerned unless your looking to stretch bandwidth
and or drop output impedance for those difficult loads that I try an avoid. :)

Basicaly my philosophy is less is more. Meaning the less you manipulate the signal I
feel you get better "sounding" performance, but some circuits just don't permit that.
It's just much easier to build a stable circuit without using feedback, and when I feel
the circuit doesn't "need" it then I just don't use it. (didn't feel the need here)

No knock on anyone who does want to use it, or feels that loop is there goal plus the
added benefits is what there after, just stating what my design goal was on here is all. ;)

The outputs in quite a few of these Eico's are very nice an perform rather well to boot. :thmbsp:
 
You talking an EICO ST-70 or a Dynaco?

Either one would work, only thing is when you remove the feedback the gain goes up,
the amp may be to sensitive meaning you can barely touch the volume pot for LOUD.

If you give me more info we may be able to get rid of some gain fairly elagantly then
removing the feedback would give you a "normal" driving condition.

It is also possible removing the feedback and using a passive pre would work for you.
 
Alright after some thinking and research, I say wire the outputs in triode and remove
the feedback from the tranny an that little cap from the bottom output tubes an see
how the gain works out and what you think of the overall sound of the amp.

It won't hurt anything if the gain is to high, you'll just see that your pre's volume will
be higher then normal with the knob being turned up less then normal.

Now unlike the Eico the Dynaco does NOT have excess gain from the factory, it takes
over a volt to drive the Dynaco to full power, an once the outputs are wired triode you
will be needing even more drive, SO removing the feedback just may be about right. :)
(but gain isn't the only things here, drive is, the original has limited "drive" capability)

If there is to much gain or the amp is more sensitive then you'd like I do have a rather
not drastic but somewhat involved mod to do that you may have to tweak. And that's
wiring the pentode portion of the driver in triode, changing the plate resistors so you'd
end up with close to 85v on it's plate is how you would dial it in.

I've attached this original Dynaco circuit and this modded version showing what I'd do
for wiring the driver in triode, also a quick show of the outputs being wired in triode an
the overall feedback removal all marked as well. (you'd need to check voltages though)
(should be able to see just what is needed for removing feedback and wiring outputs)

What I have done to the front end just may be exact values to get the 85v but won't
know until someone actually does it, I have to many projects right now to try myself.
If the voltage is low drop the plate resistors, if it's high raise there values to line it up.
It really would not be a very involved thing for someone to dial it in nicely, not at all.
 

Attachments

  • 70_orig.GIF
    70_orig.GIF
    22.5 KB · Views: 183
  • 70_orig_triode_driver_triode_output.GIF
    70_orig_triode_driver_triode_output.GIF
    23.1 KB · Views: 200
Alright after some thinking and research, I say wire the outputs in triode and remove
the feedback from the tranny an that little cap from the bottom output tubes an see
how the gain works out and what you think of the overall sound of the amp.

Thanks for the diagrams and the thought put into this. I have one more question. For testing purposes, could I just remove the feedback connection at the 16 ohm tap?

Thanks again,

Robert
 
Modding my Eico St40; reviving this thread with a question

So, I've got an Eico ST40 that I want to mod, and I'd like to do what you've done to the driver stage and power supply but without rewiring for 6L6's (keeping the 7591's). I don't plan on using the preamp at all, so I'll wire inputs directly to the grids of the drivers.

Kegger, if I do the 12AT7 driver mod and your power supply mods, will the voltages still be correct for 7591's wired in triode with no feedback?

Thanks in advance!
 
WRT 6L6s in the Eico ST40-

Since you have a 12AT7 and 6L6s, this might be a candidate for a change-over to the HK A250 amp stage topology? It uses 12AT7s into 6L6s... but uses some 'bootstrap' gain, to increase the amount of gain, and appropriately more feedback. I'd expect it (like all Stu Hegeman designs) to be very linear and accurate, with great dynamics...

It's like the HK250 Jay has, but with slightly more gain. IIRC, Tom Bavis has the service file for the A250... if you ask nicely, he might send it to you. :D

Regards,
Gordon.
 
Keg -- Very nice rework! The added bonus of your approach is the limited peak current draw now placed on the power supply, so in effect its regulation is automatically improved in doing the conversion to triode operation.

Dave
 
It's 2 years since I did this, if I studied it long enough there's probably something I'd
do a little different now, but as far as voltage an the 7591 goes those would be fine.

-----

Quite a "few" things going on over here I'm trying to get through lately, but any more
questions and I'll be glad to help where I can.

I've got some more of these amps coming my way, at least 1 will be converted to 6V6
and I'm sure another 6L6 as well, I'll see then if I'm thinking of changing anything... :)
 
Yes please do post back, only thing you'll probably have to tweak/adjust is the cathode
biasing resistor/resistors for the output tubes, somewhere between 45 and 50ma a tube
is about where I'd shoot if you get those same voltages in the supply that I had there. :)
 
Well, I finally got around to finishing the mod-- ended up doing it with 6L6's, as Kegger did. I gotta say, it sounds absolutely great. Sweet, sweet midrange, tight bass, plenty of power to drive Quad ESL 57's, and no noise or hum whatsoever. This amp is a terrific match to the Quads.

I completely bypassed the preamp and am running it as a straight power amp (through a 100K ohm stepped attenuator as a volume pot).

Thanks, Kegger, for posting this!
 
Hey nice to hear it worked out well, :thmbsp:

One tweak, if building the circuit from scratch, I'd change
the splitter loads to 56k on each verse those original 82k's
 
Back
Top Bottom