The "$250" Amp? - 6LU8 Compactron SE, UL

[djnagle]

Here is a mock-up of the base for my amp. each piece is 2" x 2" wood. I did this just to make sure there will be room in the steel or aluminium 2 x 2 square tubing I will use for the real deal. Looks like every thing will fit well. The Rec tube I WANT to use is the AZ1 but it only take 4v not 5v so I don't know what to do about that. After the base is all done and before I wire it, I am thinking that it would look cool gold plated. The back leg is the PS and each front leg is the signal side. So from the front it will be RCA input, 6LU8, can cap (if aplicable) output transformer. From the back it will be troidal power transformer, choke, and AZ1 dead center in the middle. I need to finish up a couple projects first (Turntable) but I will start collecting parts now. Cheers.

 

[djnagle]

Here is the power supply the Squidward and Kegger came up with and I drew. Let me know if it needs correction. Cheers.

 

[Kegger]

I made some adjustments to the drawing Dennis,

The 7uf cap shouldn't have a polarity as I suggest polypropylene cap here.
Also a 400v rated cap will work here but it can see say 375v at times so if
you can get a 450 to 630v rated cap like some of the Solens I'd suggest it.

This is all just for the 7uf cap, also that same cap "tunes" the voltage here,
make it larger and your output voltage will go up, smaller and it goes down.
Anywhere between 290 and 310 volts would be quite acceptable output V.

 

[djnagle]

Cool keg. I just got off the phone with a metal company trying to find aluminium tubing to slice up for the gimbles for the DIY TT Newman and I am building. They also had the exact size square tubing in need for the base of this amp. I told you this morning that I was going to put this on hold for a bit, but finding the base materials I need kind of got me jazzed about it again. The more I look at the schematics, the more I realize the simplicity of this design.....good job. Cheers.

 

[ikoflexer]

Hello everybody. I would be interested in building this amp, but there are some parts which are confusing to me; I'd very much appreciate if someone could help out a bit. In particular, it is not clear to me what the operating point of the pentode is. The schematic seems to indicat e a bias point of -37V, but that seems out of the chart for any useful plate current. Could anyone indicate what the plate current is, if possible, for both valves?
Thanks in advance for any suggestions.

 

[ikoflexer]

I have finished a prototype of a variation of this amp. It really looks like 37V on the pentode cathode is a typo. Mine shows about 17V and a nice approximately 50mA plate current. I can add some more details about it if there is any interest. Initial impression of the sound is that the prototype sounds a bit dark, not as good as my valve buffered gainclone.

 

[Kegger]

Sorry been out of town, the 37v is correct, for just a bit more then 50ma, more like 52.5ma.

What plate voltage and cathode resistor are you running? With this setup and 17v your more
like half that current and 25ma, probably would be a bit dark sounding, with the correct setup
this amp is anything but dark, very much does it have top end sparkle an quite open/revealing. :scratch2:

 

[ikoflexer]

I had to "adapt" the values from your schematic because my supply voltage is somewhat different. Here's what I got:
Ebb = B+ = 253V
Ep1 = triode plate voltage = 144V
Ek1 = triode cathode voltage = 1.84V
Rk1 = triode cathode resistor = 510R

Ep2 = pentode plate voltage = 251
Ek2 = pentode cathode voltage = 16V
Rk2 = pentode cathode resistor = 255R
Rc2 = pentode screen resistor = 37.3k
Ec2 = pentode screen voltage = 125V
Irk2 = current through pentode cathode resistor = 63mA
Irc2 = current through screen resistor (pin 8) = 3mA

It just occurred to me that I used a larger screen resistor, trying to get the voltage Ec2 as close as I could to 120V, so that I could relate to the published characteristics in the data sheet. I see that your screen resistor is much smaller than mine, which will put a higher voltage on the screen, which in turn shifts up all the curves, and that's probably why a 37V on the cathode ends up with only about 50mA of current. Does this make sense?

I also use a 37k loading resistor on the triode. Since the resistor and voltage values I have are quite different than yours, I don't want to bother you with more questions unless I can stick to your schematic. Would it be possible, when you got some time, to publish some of the key voltages and currents on your schematic? I'm sure it would help many of us who are trying to replicate your work.

One thing is that I have some not so good output transformers (re-purposed power transformers) and that might affect the sound. I figured they're good enough for now, for prototyping. The amp certainly is quite loud, and has good base, I must say. By the way, I'm running 21LU8s, since I had a more appropriate transformer for their heaters, and I figure it might be easier to regulate 21V at .9A later on, when the main operation gets settled, than to regulate and smooth 6V at 3A.

 

[Kegger]

Hey no problem post away, when I get time I will answer all questions, as you say they
could help others as well.

I will run your values through spice and confirm what you are seeing, maybe even some
suggestions as well.

Are you running output tranformers that have UL taps? Or are you just running pentode?
If your running pentode then yah your's will be quite different from the original, and yes
your much higher G2 resistor will drop it's voltage and the current in the tube be off.

With the one in the schem the plate and screen are at the same voltage, but we are also
running in UL not pentode, an switch back and forth from triode.

Also all the key voltages should already be on the schematic, the currents can then be the
calculation of voltage devided by resistence, the driver 2v devided by 620ohm is 3.2ma, an
the output tube is 37v devided by 700ohm (both in paralell) is then 53ma.

 

[Kegger]

A quick look of the operating perameters, says yah your running pentode, and not
a bad setup where they are operating it looks like either, the front end tweaks for
the power supply you have look fine to me as well, what I would do is though the
270k plate to plate feedback resistor I would change down to 100K, pentode mode
can use a bit more feedback to sound better, also you supply the driver with some
more voltage, now more like 155v on the plate and 2v on the cathode of the triode.

That may make the tube a bit less dark sounding running in pentode, you may even
want to try triode strapping it using the 1.2k g2 resistor then upping your cathode
resistor to 510ohm, so if you'd use paralell 510ohm cathode resistors you switch in
the pair for 255ohm or single for the 510, then put yourself in a switch for pentode
have the 37k on it and triode have the 1.2k on it. Just a few thoughts there.

When running the pentode "triode" strapped I would switch the feedback resistor back
to 270k so the driver can drive the output to full power easier. (triode needs more drive)

Otherwise the last thing yu know that changes the sound is those output tranny's.

 

[ikoflexer]

Great, your feedback is very much appreciated.

Yes, indeed, until I get proper OTs I'm running it in pentode mode. Dropped the feedback resistor to 100k as per your suggestion, and changed the operating point of the driver also, to 155V plate voltage and 2V cathode voltage. Now the sound is brighter; I'll have to do a bit more listening, but for now it seems it lacks a bit of dynamic.

By the way, do you model the pentode as pentode in spice, or as triode (with fixed screen voltage)? I've been using ltspice and used curvecaptor to come up with a model for the triode, and then for pentode I traced the curves given 120V screen voltage, however, curvecaptor could not come up with very accurate lines for the pentode curves. Surprisingly, my spice simulation is very close to what I've been measuring on the real circuit.

 

[Kegger]

I'm a little out of it, I'll post more when I feel a bit better.

Try messing with the feeback resistor, in pentode I'd go as high as 200K, with the
low being the 100K your running now.

Basically I model pentodes as pentodes, when using them as triodes I just wire the
screen to the plate with a 10ohm resistor, (spice doesn't like a very big resistor there)
I just take pentodes that run close to what I'm working with an tweak them, in this
case a KT88 tweaked got me the same voltage and current the 6LU8 did on bench.

Another kinda fun thing to mess with is using an LED for a cathode resistor on that
triode driver, most any 1v - 2v LED should do, I did up a paralell version of this amp
with a pair of "sockets" (1 for each channel) to swap in different LED's.

 

[Glassweights]

$1 6lu8

Looks like http://www.vacuumtubes.net/index.html has the 6LU8
for $1 again.

I was buying some 6SF5 when I remembered I should get some of the compactrons next time I ordered there.

So I ordered 10 of them and 4 of the sockets . . . bet I wish I ordered more of the sockets later!

I just love a bargain! I will post what brand they send me.

Dan

 

[ikoflexer]

Hey I hope you feel better, no worries, I'm playing around with different things.

My simulations show pretty crazy distortion when using cathode caps.

I ran a signal from my HP signal generator and it seems that the effect of the output trafos is mainly a small drop in frequency response in the high and very low frequencies.

At the moment I'm listening to it on my MDR-V6 headphones. The amp is very detailed, a bit edgy, with huge dynamics, and very tight. The bass is very well defined, almost edgy. I don't think I've listened to an amp like this before, quite different. I'd say it has a sort of rough power to it, yet the sounds are clear and crisp. "Sweet", I'm afraid, does not apply to my version of this amp

 

[Kegger]

Knifefighter,

Thanks for the headsup on the tubes.

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

ikoflexer,

Good to see more "tweakers" out there.

I'd suspect the Edcor outputs to be a bit "smoother" then power tranny outputs, BUT
your right on when you say this not a "sweet" sounding amp, it is rather HIFI in that it
isn't an overly tubey, midrangy presentation, but rather detailed and full sounding amp.

 

[ikoflexer]

For the tweakers out there interested in spice simulation with this tube, I've optimized using Matlab the pentode equation parameters for the plate current and got a pretty accurate model. Can't vouch for the screen current. I also include below the triode model, for completion.

My simulation with ltspice (free download) is very close to real life.

Feel free to modify or otherwise use in any way you wish.

* 6LU8 model optimized to fit curves in GE data sheet
* 6LU8P P G1 C G2 PENTODE HALF OF 6LU8 OR 21LU8
.SUBCKT 6LU8P 1 2 3 4 ; P G1 C G2
+ PARAMS: MU=8.5070 EX=1.4137 KG1=551.8098 KG2=4200 KP=25.9952 KVB=21.5557
+ CCG=16P CPG1=0.5P CCP=9P RGI=1K
RE1 7 0 1MEG ; DUMMY SO NODE 7 HAS 2 CONNECTIONS
E1 7 0 VALUE= ; E1 BREAKS UP LONG EQUATION FOR G1.
+{V(4,3)/KP*LOG(1+EXP((1/MU+V(2,3)/V(4,3))*KP))}
G1 1 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG1*ATAN(V(1,3)/KVB)}
G2 4 3 VALUE={(EXP(EX*(LOG((V(4,3)/MU)+V(2,3)))))/KG2}
RCP 1 3 1G ; FOR CONVERGENCE
C1 2 3 {CCG} ; CATHODE-GRID 1
C2 1 2 {CPG1} ; GRID 1-PLATE
C3 1 3 {CCP} ; CATHODE-PLATE
R1 2 5 {RGI} ; FOR GRID CURRENT
D3 5 3 DX ; FOR GRID CURRENT
.MODEL DX D(IS=1N RS=1 CJO=10PF TT=1N)
.ENDS

* 6LU8T 1 2 3 P G C TRIODE HALF OF 6LU8 OR 21LU8
.SUBCKT 6LU8T P G C
CGP G P 6p
CGK G C 7p
CPK P C 2p
Gp P C VALUE={(0.03216728967m)*limit(V(P,C)*log(1.0+(-0.135902839)+exp((6.28862323)+
+(6.28862323)*((55.34449305)+(-909.0977301m)*V(G,C))
+ *V(G,C)/sqrt((84.92047116)**2+(V(P,C)-(11.20775662))**2)))/(6.28862323),0.0,1.0e16)**(1.222270641)}
.ENDS

 

[Glassweights]

6LU8 Tubes arrived

OK,

I received the 6LU8 tubes from vacuumtubes.net

qty=9 Magnavox made in Japan
qty=1 USA 6LU8 is all it says.

The sockets are almost red colored with nice holes for soldering, kinda elongated single hole per pin

Makes me wish I had ordered a couple more to get one more USA tube, but, I am a maggie fan anyway.

Dan

 

[ikoflexer]

Dan, good price on those 6lu8s. Mine are Japan too, and for this simple design and "el cheapo" amp I don't think I'll make an effort to get anything fancy. Are you going to do a prototype or plan to go right into a serious build?

I'm still tweaking it, but it already is very listenable.

 

[Glassweights]

Wondering

ikoflexer:

Well,

I am pretty interested in this amp. I can probably just build one if I got the outputs. Serious build, and then tweak maybe?

However I am sorta wondering if the BOM idea/group build of a standard design has petered out?

If not I might just wait a bit and see what comes up. (I am good either way)

I bought a Scott 222c and I am really anxiously awaiting for it to come in. Looks like it needs caps at least, so my maggie 175 amp may give up a few of the parts I have stocked to build it just so I can get something tube into my main listening room!

A small amp like this has So many possibilities! I could not believe the cheesy speakers my daughter brought over so we could listen to her ipod thanksgiving r2d2 music. They are a full 1 inch square!

So I see a serious need to introduce my family to real tone and quit hogging it all in my guitar amps . . .

 

[Glassweights]

7uf

Oh,

I was having a hard time finding a 7uF cap for the front of this thing at mouser

Anyone got a part number or place you can get it? I am still thinking of waiting, but, I am curious now.

I have some 4.7, 8, and 10uF electrolytics in stock, but I read I want POL!

hah

Dan