diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   Happy New Year - A New Amp to Argue About... (http://www.diyaudio.com/forums/tubes-valves/180182-happy-new-year-new-amp-argue-about.html)

wrenchone 1st January 2011 07:28 AM

Happy New Year - A New Amp to Argue About...
 
1 Attachment(s)
Attached is the first schematic for another bratty push-pull amp using a pair of 6LR8 triode-pentode vertical deflection tubes per side. The 6LU8 would work just as well - same tube, different pinout. The schematic as shown has all the bell and whistles, and I'll decide how many of them I want to keep as I go on.

Options - balance control, regulated screen/not, also whether I want to use the electronic cathode bias circuit or a mess of LEDs instead (maybe Stu, just maybe). I bought the Electronic Goldmine jumbo pack of 2000 pieces of 3mm green LED (1 cent each!) so I'll be prepared if I decide to do that. I figure 5 parallel strands of 7 LEDs in series would make the nut and treat the LEDs pretty conservatively.

20to20 1st January 2011 01:26 PM

What is the load impedance for the pentode section? All I can find is the plate resistance. 14K. Perhaps the data sheet is using the two terms interchangably.

SY 1st January 2011 02:03 PM

Why not return the zener string to the junction of the OP cathodes instead of ground? Other thing is to take advantage of the pentode operation of the OP and stick some 50k or so grid stoppers to improve overload performance (not that I ever run MY amps too loud...:rolleyes:).

smoking-amp 1st January 2011 04:58 PM

Is the current mirror above the triodes actually producing matched outputs? I see a path thru two junctions to two 475 OHm resistors in parallel for the right side triode plate.

I did once see a schematic for a duplex current mirror by J.L. Hood which had high Z loading on both sides and complementary outputs (like a CT inductor load). I can't remember how it worked though, and in fact I couldn't figure out how it worked when I saw it. But I'm sure it must have. Sure would like to see it again.

kenpeter 1st January 2011 06:07 PM

I'm not sure why the cathode CCS with the mirror?
If you simply hold both plates at constant current
and put a resistor under the cathodes, the same....

Seriously, would a pair of Mu followers on top, and
a single cathode resistor have been any different?
Or even two entirely separate cathode resistors?

The current U-Turn at the cathodes is meaningless
if both paths held equal elsewhere.

You specifically asked for argument...

wrenchone 1st January 2011 08:08 PM

First - the whole purpose of the amp is to do it with just the 6LR8s and some semi helpers, so no extra tubes, no mu followers, etc. There's no room on the top of my chosen case for any extras, anyway. The topology of differential amp/splitter into pentode outputs will remain the same.

Smoking Amp - Google "Wilson Current Mirror".

I neglected to put the XFMR impedance in. I'll be using a pair of transformers stolen from old Wurlitzer organ amps. Each transformer originally worked off a pair of 7868s. They measure out as ~6200 ohms : 8 ohms, with respectable primary inductance. A bonus is that they fit inside the case.

Current U-turn at the differential cathodes? That's a cascoded JFET current source. I'll probably be replacing it with a single Supertex TO-92 depletion mode mosfet - high impedance, fewer parts. Set thetotal current in the two tubes with a rod of iron, make it equal....

smoking-amp 1st January 2011 10:29 PM

I don't doubt that the Wilson current mirror works to copy the current. Its that the input impedance to it, for the right triode, is very low, freezing the voltage swing there. You could put a resistor in series to it I guess, but then what is the point of the mirror? (the CCS tail, as Ken mentioned, is already making for complementary currents)

I assume the intention is to get the active equivalent of a center tapped plate load inductor. What you may want is the J.L. Hood complementary mirror, to get complementary Mu gains on each side. Or there is a simpler setup with two CCS plate loads (1/2 the bottom CCS) with cross-coupled resistors to control the current reference levels as complements.

wrenchone 1st January 2011 10:43 PM

1 Attachment(s)
I kinda had you anticipated. Here is an alternate scheme for the input stage. Another alternative would be to use a completely boring (but workable) current source for the tail with resistive loading at the plates - say, 68-75k.

20to20 1st January 2011 10:43 PM

Quote:

Originally Posted by wrenchone (Post 2417873)
First - the whole purpose of the amp is to do it with just the 6LR8s and some semi helpers, so no extra tubes, no mu followers, etc. There's no room on the top of my chosen case for any extras, anyway. The topology of differential amp/splitter into pentode outputs will remain the same.

Smoking Amp - Google "Wilson Current Mirror".

I neglected to put the XFMR impedance in. I'll be using a pair of transformers stolen from old Wurlitzer organ amps. Each transformer originally worked off a pair of 7868s. They measure out as ~6200 ohms : 8 ohms, with respectable primary inductance. A bonus is that they fit inside the case.

Current U-turn at the differential cathodes? That's a cascoded JFET current source. I'll probably be replacing it with a single Supertex TO-92 depletion mode mosfet - high impedance, fewer parts. Set thetotal current in the two tubes with a rod of iron, make it equal....

How did you determine the load impedance of the tubes would be well matched for your transformers?

smoking-amp 1st January 2011 10:58 PM

Ahh, this new "current mirror(s)" version has the look and feel of the J.L. Hood mirror I recall seeing in Wireless World some time ago. I assume Q4 sets the total current, and Q3 sets the base voltage level for the Q1 and Q2 mirrors. Still mystifies me that the Q1,Q2 collector voltages can convey any control over the collector currents. Maybe the Early effects of Q1,Q2 unbalance the teetery base voltages?

Hmm.. Or is this just two equal CCS loads for the plates? ( I think the J.L. Hood setup was claimed as complementary currents, sorta like a CT inductor)
If so, then what if you make R3 and R5 in the latest as the same resistor. Then the two currents have to be summing to the reference current.

Is there a name for this mirror(s) circuit if it is actually producing complementary currents?

If this complementary mirror scheme works well, then Sayonara to CT plate load inductors from now on. At least for low power stages anyway.

----------- Sorry about the streaming edits. Think I got it right finally.


All times are GMT. The time now is 04:04 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2