OK, if I'll feel (or should I say 'hear') that there is not enough bass I'll know where

to look
Here is the schematic which I hope incorporates all comments. Please let me know if anything else should
be changed before I start buying process (hopefully coming weekend).

I'm trying to collect in one place the info scattered across multiple pages. Here is the final schematic for PS
and the parts list.

Code:
T1 - GXSE15-5K
[URL="https://www.edcorusa.com/gxse15-8-5k"]EDCOR - GXSE15-8-5K[/URL]

V1 - 12AX7
[URL="https://www.ebay.com/itm/Sino-12AX7B-Preamp-Vacuum-Tube/152759915406"]Sino 12AX7B
Preamp Vacuum Tube | eBay[/URL]

V2 - 6P15P-EV
[URL="https://www.ebay.com/itm/6P15P-EV-4Pcs-Russian-Military-Long-LIFE-Vacuum-Pentode-tube-
OTK-same-year-EL83/292276425741"]6P15P-EV 4Pcs Russian Military Long LIFE Vacuum Pentode
tube OTK same year EL83 | eBay[/URL]

C1 - 0.082uF 400V
[URL="https://www.mouser.com/ProductDetail/EPCOS-TDK/B32642B6823J?
qs=sGAEpiMZZMv1cc3ydrPrFzRLALvOkyF1xqwV3JmB0SfGtnllqjsKTw%3d%3d"]B32642B6823J
EPCOS / TDK | Mouser[/URL]

C2 - 0.1uF 400V
[URL]https://www.mouser.com/ProductDetail/EPCOS-TDK/B32671L4104J000?
qs=sGAEpiMZZMtI26rxh1qzjhCu5rOhviVHVa96FMB1%252bC8%3d[/URL]

C3 - 680uF 16V
[URL]https://www.mouser.com/ProductDetail/Nichicon/LKG1E681MESYCK?
qs=sGAEpiMZZMtZ1n0r9vR22d%252blFeYd%252bH4EB5OPpYC2i4g%3d[/URL]

C4 - 180pF 1kV
[URL]https://www.mouser.com/ProductDetail/Vishay-Cera-Mite/561R10TCCT18?
qs=sGAEpiMZZMt1mVBmZSXTPFXppOBlV6Pk7PmGGS9GNQY%3d[/URL]

R1 - 100 kOhm 0.5W

[URL]https://www.mouser.com/ProductDetail/Xicon/293-100K-RC?qs=sGAEpiMZZMukHu
%252bjC5l7YQs4PzqcsJEK1gKCm%252btXr3E%3d[/URL]
R2, R12 - 100 Ohm 0.5W
[URL]https://www.mouser.com/ProductDetail/Vishay-Semiconductors/TZX3V6A-TR?
qs=sGAEpiMZZMui8icXhU1dEupM5jfX%252bQnA[/URL]

R3 - 820 Ohm 0.5W 1%

[URL]https://www.mouser.com/ProductDetail/Welwyn-Components-TT-Electronics/MFR4-820RFI?
qs=sGAEpiMZZMukHu%252bjC5l7YRDwiouGrKBmwSsVuz94HuQ%3d[/URL]

R4 -150 kOhm 2W
[URL]https://www.mouser.com/ProductDetail/Vishay-Dale/CPF2150K00FKE14?
qs=sGAEpiMZZMu61qfTUdNhGz34irntT%2fJFIKWZR%2fBAcb0%3d[/URL]

R5 - 100k 5W wirewound
[URL]https://www.mouser.com/ProductDetail/Xicon/280-PRM5-100K-RC?
qs=sGAEpiMZZMtbXrIkmrvidHiKlRtT248Uj8xqXsZy2zA%3d[/URL]

R6 - 150 Kohm carbon film 1W

[URL]https://www.mouser.com/ProductDetail/Xicon/294-150K-RC?
qs=sGAEpiMZZMu61qfTUdNhG3GWZmU6PwDPWeAE%2fkNyxUU%3d[/URL]

R7 - 220 Kohm carbon film 1W

[URL]https://www.mouser.com/ProductDetail/Xicon/294-220K-RC?qs=
%2fha2pyFaduiRHRkVJgnDOB3vlddufLrLIToe7QaFR9Q%3d[/URL]

R8, R10 - 10 kOhm 0.5W

[URL]https://www.mouser.com/ProductDetail/Vishay-Dale/CMF5510K000BEEA?
qs=sGAEpiMZZMukHu%252bjC5l7YWGhjK80auj0pDzGU6krlXY%3d[/URL]

R9 - 470 kOhm 0.5W

[URL]https://www.mouser.com/ProductDetail/Xicon/293-470K-RC?qs=sGAEpiMZZMukHu
%252bjC5l7YTF6je7BkLxK9noGLdhjO%252bM%3d[/URL]

R11 - 150 Ohm 0.5W

[URL]https://www.mouser.com/ProductDetail/Bourns/3362P-1-151?
qs=sGAEpiMZZMs26uXuX7MLMQDudKPvZ1vl[/URL]

Recommended speakers 3" 8 Ohm 91 dB:

[URL]https://www.parts-express.com/faitalpro-3fe25-3-professional-full-range-woofer-8-ohm--294-
1104?AID=1457483&PID=7598510[/URL]
E
There are errors in the signal BoM.

R1, R9, and R12 should be metal film, not carbon film.

R2 and R10 should be carbon composition. Use this and this.

A 0.1% tolerance part for R8 is costly overkill. Use this. The price is nice and KOA/Speer's reputation is solid.

A 1/2 W. rated part is cutting things too fine, for R11. Use this. "Better to be looking at it, than looking for
it."

To put it mildly, coupling caps. are a subject of considerable "debate". With some exceptions, my
preferences are discrete polypropylene film/metal foil and Soviet surplus K40 paper in oil (PIO) parts.
Polyester exhibits some electrical quirks and (for the most part) metalized film is "meh". This part gets my
nod for C2. We now come to subject of C1. Polystyrene is easily as good electrically as polypropylene, but
it's heat sensitive. C1 is not nearly as thermally "exposed" as C2. So, this is what I suggest for C1. Don't
loose sleep over the WVDC rating. Plate to grid shorts/arcs in small signal triodes are RARE occurrences.
Code:
T1 - Triad N-68X
[URL]https://www.mouser.com/ProductDetail/Triad-Magnetics/N-68X?qs=
%2fha2pyFadujgpvvi3%252bprbw33P0KLwfnamaApL%252b8cFew%3d[/URL]

T2 - Triad VPS24-1000
[URL]https://www.mouser.com/ProductDetail/Triad-Magnetics/VPS24-1000?
qs=sGAEpiMZZMvwUzoUXIIvydH8%2fj5UF54yTv9IKwPkmq4%3d[/URL]

T3 - Triad C-24X
[URL]https://www.mouser.com/ProductDetail/Triad-Magnetics/C-24X?qs=xrpI9KL
%2FXvsp8oLEd8wC%2FQ%3D%3D[/URL]

V1 - OA2
[url=https://www.ebay.com/itm/1PCS-SYLVANIA-OA2-0A2-OA2WA-6626-CV4020-Vacuum-tube-
NOS/152620246140?epid=1072830523]1PCS - SYLVANIA OA2 / 0A2 ( OA2WA / 6626 / CV4020)
Vacuum tube NOS | eBay[/url]

C1 - 0.01uF 2000V
[URL]https://www.mouser.com/ProductDetail/Cornell-Dubilier-CDE/DPPM20S1K-F?
qs=sGAEpiMZZMv1cc3ydrPrFwLJAmLhcfj35IIgJQNHda0%3d[/URL]

C2 - 20uF 400V
[URL]https://www.mouser.com/ProductDetail/United-Chemi-Con/EGXF401ELL200MJ30S?
qs=sGAEpiMZZMukHu%252bjC5l7YYOm8CqN7Zj%2foW0GsUvVneI%3d[/URL]

C3, C4 - 180uF 250V

[URL]https://www.mouser.com/ProductDetail/Nichicon/UCY2E181MHD6?
qs=sGAEpiMZZMtZ1n0r9vR22awqI6nzE%2fsTPmyizGVpPNg%3d[/URL]

C5 - 82uF 400V
[URL]https://www.mouser.com/ProductDetail/Panasonic/EEU-EE2G820?
qs=sGAEpiMZZMtZ1n0r9vR22dN7zJS18SGV4uFjDxRrTxw%3d[/URL]

C6 - 0.068uF 400V
[URL]https://www.mouser.com/ProductDetail/WIMA/MKP1G026804C00KF00?
qs=sGAEpiMZZMv1cc3ydrPrF2CPLAojp4yWTme4nl3ONOM%3d[/URL]

D1, D2 - UF4007GP-TP
[URL]https://www.mouser.com/ProductDetail/Micro-Commercial-Components-MCC/UF4007GP-TP?
qs=sGAEpiMZZMtbRapU8LlZDzWSnAblitvPZLDaoGuk%2f10%3d[/URL]

L1 - CL-140 Inrush Current Limiter

[URL]https://www.mouser.com/ProductDetail/Amphenol-Advanced-Sensors/CL-140?
qs=sGAEpiMZZMuqZeNK75brD9Lrm4Ls3H6A27r%252bQtFpb38%3d[/URL]

R1, R2, R3, R4 - ?

You need to calculate R2, not just dial it in. The 0A2 Needs at least 5 mA, and 30mA maximum. If you set
the 0A2 for 17.5mA (middle of its operating range, and add the predicted screen currents of the output
tube, you get the current that passes through R2.

B+ Volts - B++ Volts of 150V = the voltage drop in R2 Voltage Drop/(17.5mA + 2 screen currents) = R2
Ohms Be sure to use the current in Amp form, i.e. 0.0175 A.

But of course, you have to pre-calculate the B+ voltage. Either use a software program, or calculate it
manually (don't know if you know how to do that).

Until you get the +150V set, you can not just measure the B+ in the amp, because if the screen is anything
other than 150V, the plate current will vary accordingly.

You need to know the voltage of the B+ secondary, DCR of the B+ secondary, the DCR of L1 (what is L1? A
transient limiter?), and the DCR of T3. Then know your desired plate currents of the output tubes, screen
currents of the output tubes, and the plate currents of the 12AX7. Now, you can calculate the B+ voltage
under load.

If the B+ secondary is 110Vrms, and a pure sine wave, under load there will be somewhat less than 311VDC
at the input to T3 (T3 is a filter inductor, normally listed as L3).

Designing "by the seat of the pants" to use one power supply, and connect to another amplifier circuit,
requires either some calculations, or requires some really lucky guesses.

By the way, if you have the B+ unloaded, it will raise to maximum, the filter caps have to stand that voltage.
And if you then dial in R2, you could pass a lot more than 30mA through the 0A2 (and ruin it).

Be careful with C6. I know it is there to reduce the noise of the 0A2. The problem is you may have a
relaxation oscillator (R2, C6, 0A2), especially before the output tubes warm up and draw screen current. In
that case, you may need to remove C6.

Caution: You do not have any bleeder resistors on the B+. The power supply will drag down to about 150V,
and stay there when you turn it off (especially if you do not have the amp tubes in the amp and filaments
warmed up when you turn the amp off). The end voltage is unknown if only the amp tubes and 0A2 are the
only drain when the amp is turned off. Wire a bleeder resistor across the B+. Safety First! You Effectively
have about 200uF of capacitance on the power supply. Two 25k Ohm 5W resistors in series will drag the
voltage from about 300V down to 100V in 10 seconds. After you turn the amp off, by the time you get the
screws out, and the bottom cover off (20 seconds) the amp will be safe.

I've tried all my best merging Mr. Eli Duttman's changes to Mr.Gillespie's design. Please correct me if I
made errors and please address the question marks. What is the wattage of the resistors? Thank you!
I've tried all my best merging Mr. Eli Duttman's changes to Mr.Gillespie's design. Please correct me if I
made errors and please address the question marks. What is the wattage of the resistors? Thank you!

This is a full pentode mode amp. The ultra-linear (UL) tap on the O/P trafo's primary is unused. Thoroughly
insulate the end of that wire and tuck it out of the way.

Whether or not the R4/C3 Zobel network is necessary has to be determined experimentally. That issue and
the value of phase compensation cap. C2 are intimately associated with the properties of the specific O/P
trafo being employed. We do not know those important properties. Do you own a signal generator and an
oscilloscope? If you don't, brute force methodology can be employed.

R8 was, and should remain, 470 Kohms. An additional 10 Kohm carbon composition grid stopper, like R2, is
connected to g1 of the O/P tube.

In general, 1/2 W. rated resistors are fine. However, it is necessary to look where substantial current is
flowing to see if a more robust part is in order. R7 is carrying 'X7 plate current. Assume a 1 mA. plate
current and the part dissipates 0.47 W. To ensure an adequate margin of safety, use a 2 W. rated part here.
Current flows from B+++ to ground, via R6 and R5. A quick "crunch" suggests R6 dissipates approx. 0.43
W. Again, a 2 W. rated part is in order.

For an increased range of adjustment, make variable resistor R9 150 Ω. To simplify Ohm's Law calculations
of O/P tube cathode current, make R10 100 Ω. To prevent trouble with deep bass phase shifts, make
C5 680 μF. As it was blocked at the amp's I/P, infrasonic noise is not a matter of concern.

The 6Π15Π (6p15p) brings the suppressor grid (g3) out to the base. Show g3 in the schematic and its pin 1
connection to pin 3, the cathode. 6BQ5s, with their internally connected connected suppressor grids, work
in sockets wired for the 6Π15Π.

I would make C1 a 400V part. Also, more like 1uF unless you want to limit LF. (-3db at 19Hz instead of
1.59Hz). It will protect the preceding stage in case of an arc to grid. C2 can be as low as the speaker output
voltage (25V is probably enough).
Why is R6 there? R7 looks high. Usually around 100k.
I doubt this will have enough power to drive the screen grid of the 6P15P.

Repeat, the value of C2 is currently unknown. A signal generator and an o'scope are
needed to determine the correct value. The R4/C3 Zobel network may not be needed
and, if needed, the parts' values will change.

The suppressor grid (g3) "sits" between the screen grid (g2) and the plate. In a 6Π15Π,
g3 comes out to the base at pin 1 and pin 6. In a 6BQ5/EL84, g3 is internally connected
to the cathode. Connecting either pin 1 or pin 6 of the socket to pin 3 allows for proper
operation of both O/P tube types.
What about volume control?

The circuitry driving the amp has to comply with IHF "standards" and be capable of
driving a 10 Kohm load. Add a 10 Kohm audio (log.) taper pot. across the terminals of
the I/P RCA female. The wiper of the pot. feeds the 0.82 μF. high pass capacitor. Please
notice that the "textbook" 1:10 driving/driven impedance ratio rule is obeyed.

Also, more like 1uF unless you want to limit LF.

We very much want to block infrasonic noise. Keep that part 0.82 μF. A global NFB
loop encompasses the O/P trafo. If the amp tries to follow infrasonic noise, the error
correction signal will cause the "iron" to saturate.

I'll make C1 400V, C2 25V and add volume pot. I'm not sure if I need to change tube
drawing to reposition g3.
kodabmx, the presented schematic was based on the design by Mr. Gillespie (see the
post #38)
R6 and value for R7 came from that design.
I'll update the schematic later today.

Alas I have only multimeter in my disposal

FUGLY describes the power pentode symbols. Go back to showing only the control and
screen grids. During construction, don't forget the pin 1 to pin 3 jumper that allows both O/P tube types
to function correctly.

Alas I have only multimeter in my disposal

Then we use brute force methodology.

The open loop gain of the voltage amplifier is made to increase with increasing frequency by "peaking".
That means part of the anode load is inductive, along with being resistive. The schematic shows a 470
Kohm resistance. Make that 470 Kohms up by assembling a series wired composite of a 100 Kohm
wirewound part, a 150 Kohm carbon film part, and a 220 Kohm carbon film part. While the inductance of
wirewound resistors is frequently a liability, it's an asset, this time around.

The selected Edcor O/P trafo spec's state up to 18 KHz. Above that freq., its response rolls off. "Peaking"
the voltage amplifier increases gain, with increasing frequency. The O/P trafo's response decreases, with
increasing frequency. We are adding with 1 hand and subtracting with the other. We rely on the NFB error
correction signal to maintain linearity. However, we will "run out of steam", eventually. That is dealt with by
eliminating NFB beyond 80 KHz. Don't try to lift excessive weight. Remove C2, C3, and R4 from the
schematic. Add a 180 pF. C0G/NP0 ceramic cap. across the O/P trafo secondary.

My "fat finger" error. 0.082 μF. is the correct value. The "pole" at the I/P is 3 dB. down
at 19.4 Hz. Remember, the lowest freq. a "standard" double bass produces is 41 Hz.

Yes, but there's a pipe organ that makes 8Hz (Can't remember what it's called, but I
have some music from that organ on vinyl) and a lot of modern music has lower bass.
Even the low note on a 5 string bass is 30.5Hz. Just my 0,02$ My "standard" -3db
frequency is 1.59Hz and I don't have saturation issues with my OPTs.
EDIT: Maybe that's why my tube amp has better LF bass than my 300W Yorkville...

400V is a safety thing that I do but strictly speaking on a grounded cathode gain stage
you can usually omit the input cap altogether. We are talking about a 20 cent part
though. X2 MKP safety capacitors make excellent audio coupling caps.
Last edited: 2019-01-16 3:26 pm
The Atlantic City, NJ, Convention Hall organ has a 64 foot stop. Its "sound" is felt, not
heard.
OK, if I'll feel (or should I say 'hear') that there is not enough bass I'll know where to
look

Here is the schematic which I hope incorporates all comments. Please let me know if
anything else should be changed before I start buying process (hopefully coming
weekend).
Do you means the schematic on post #106? It does not contains the resistors I am referring on my
previous post. To avoid a humongous hum, the heater filament of the tubes must have a reference to the
ground. A simple way to do so is to connect one tap of the filament transformer to ground. This is the way
most tube radios were built, but a better way is to have both ends of the filament at the same potential
relative to ground. To reach this goal, you connect two resistors in series between the filament transformer
secundary winding, and then connect the point between the two resistors to ground. The usual value for
6.3 volt windings is 100 ohm. You have selected a 12.6 volt filament transformer, a 150 ohm resistor may
be a better choice.

Remember, R5 is now a series wired composite of a wirewound part and 2 carbon film parts. Connect the
wirewound part to the 'X7 anode.

Connect pin 9 on the 12AX7 socket to the junction of the O/P tube heaters. Initially, that "bus" will be
connected directly to ground. If, during the debugging process, hum levels are unacceptably large, that
"bus" will be grounded for AC, via a cap., and elevated to approx. +60 VDC, by adding a resistive voltage
divider between B+ and ground.

Permanently "jumper" pin 1 of each O/P tube socket to the corresponding pin 3. Then, it doesn't matter
which O/P tube type is in use, at any particular time. The O/P tube cathode current gets set, at installation
time. Recheck the cathode current at approx. 6 month intervals. Tubes are consumables that wear out, over
time.