toob noob with GE 5670 and I'd like to do something with it. Help?

[danielwritesbac]

Well, here's this GE 5670 and the only thing I actually know about it is that its definitely awful in a circuit not designed for it.

I have need of a buffer, a preamp with a gain of about 6, or a phase splitter.

Currently, the project that I'm working on does happen to need a phase splitter. Its triple parallel LM1875 with hybrid variable current drive and capacitive multiplier power. Auto bridge is no good, DRV134 is not great, transformers are spendy, but here's this cute little tube. . .

However, I'd like to use the tube for what its best at and in a circuit designed for it, not a circuit designed for a near equivalent. Help?

[Eli Duttman]

Start with the data sheet. That's a GOOD looking twin triode. Its amplification factor (μ) is too large for your preamp application. However, its high transconductance of 5.5 mA./V. indicates service as either a cathode follower (buffer) or "concertina" phase splitter will be highly satisfactory.

GE's data sheet uses the absolute maximum rating system. Tubes operated under such conditions have short service lives. Back off 15% and get years of service out of the "bottle".

[danielwritesbac]

So, is there some sort of cathode follower or concertina circuit that's optimal for GE 5670?

Well, a start would be getting it powered. Can it operate on 53vdc or should it have more?

[mastodon]

Well, it will draw current (e.g. at a grid bias of 0V), but it really depends on what you're trying to drive if this will be a good operating point. What sort of amplifier are we talking about here? If you're designing a phase splitter, what does the power stage look like? 2x EL84/6BQ5 or something? Or two triodes? Or a bunch of triodes running OTL?

Designing an audio circuit is essentially an optimization task. That means you need at least two things:
1. A goal or target - what are you trying to achieve
2. Awareness of other relevant variables; this includes the investments you're prepared to make, any stuff you have lying around, the time and skills you have, etc.
Can you be a bit more specific on the above?

[Eli Duttman]

Originally Posted by danielwritesbac

So, is there some sort of cathode follower or concertina circuit that's optimal for GE 5670?

Well, a start would be getting it powered. Can it operate on 53vdc or should it have more?

A 53 VDC rail is way too "short". Look at page 5 of the data sheet, with its plate characteristics graph. A plate current of 9 mA. or so and a rail voltage around 160 V. looks good to me. Please notice that those numbers comport WELL with the data sheet's "typical" 8.2 mA. and 150 V. Allied Electronics own $16.35 6K1VF power trafo will take care of energizing the tube quite well.

[danielwritesbac]

Originally Posted by mastodon

Well, it will draw current (e.g. at a grid bias of 0V), but it really depends on what you're trying to drive if this will be a good operating point. What sort of amplifier are we talking about here? If you're designing a phase splitter, what does the power stage look like? 2x EL84/6BQ5 or something? Or two triodes? Or a bunch of triodes running OTL?

Designing an audio circuit is essentially an optimization task. That means you need at least two things:
1. A goal or target - what are you trying to achieve
2. Awareness of other relevant variables; this includes the investments you're prepared to make, any stuff you have lying around, the time and skills you have, etc.
Can you be a bit more specific on the above?

Yes, I'm not capable of designing a circuit for this tube, so I was hoping to find a circuit already optimized for GE 5670.

My amplifier input load is 100k.

A wide random selection of computer sound cards will have pretty treble with insufficient bass when driving ordinary transistors with 100k input load, or, at 10k there's dull treble, dull dynamics and a lot of bass. At every point in-between there yet more varieties of insufficient.
But, the problems are Less likely to occur when the "wide random selection of computer sound cards" drives a fet or tube.
When driving a tube, there's the opportunity for a "fresh" sound which is very helpful to the dull MP4 type codecs, such as AAC, Itunes, and HD radio. This "fresh" (not quite brash) sound exhibited by the harmonics of some tube buffers, saves the day for many computer sources and isn't particularly harmful to other sources. At about that point, you come up with a reason to use an amplifier on a computer source.

This description seems to call for a triode; however, maybe not the GE 5670. Due to the absence of designs specifically for that model tube, creating support circuitry for it seems like a wasteful shot in the dark. Searches for uses for it come up somewhere between indecipherable and nil, so by going on popularity alone, I'd have to believe that my little tube needs to stay in its box.

Thanks for your time.

[danielwritesbac]

Here's a developed circuit for it:
Untitled

[aardvarkash10]

hey daniel - your op is as odd as stating "i have a MJE350 what's a great circuit for it?"

Having said that, put it up front of your LM amp and it may be a visual (although not necessarily sonic) addition.

[danielwritesbac]

Originally Posted by aardvarkash10

hey daniel - your op is as odd as stating "i have a MJE350 what's a great circuit for it?"

Having said that, put it up front of your LM amp and it may be a visual (although not necessarily sonic) addition.

Well, I would like to know what tube buffer (and known-good applied circuit) would indeed make the sonic inflection as described above in post #6, and working correctly at dc voltage between 57 and 70?

A great circuit for MJE350? Are you reading my mind?
This is funny because I was searching through my parts box to find transistors with a High Voltage rating to make capacitive multiplier for this project and found MJE15032, MJE15033 pairs. . . as if BD139, BD140 that could run on much higher voltage. I can't really read/interpret the particulars on the MJE15033, MJE15032 datasheet(s), but it looks and smells workable.

While searching, I did manage to find a production circuit that actually uses GE 5760 tube.
Interesting "nested" arrangement has 6db preference for the sound of the 5760 instead of the sound of the output tubes and output transformer. Probably, this mild amount is for leveling out the frequency response? Can this be done to prefer the sonic signature of a tube instead of the sonic signature of a sand amp, by the same proportion as the global negative feedback aka nesting?

[mastodon]

Yes, the negative feedback is there to reduce distortion and to flatten the frequency response. However, I take it you will only be using the input stage and phase splitter portion of this circuit. I assume that the LM1875 accepts line level inputs. This would mean that using a concertina phase splitter (i.e. one 5670 half per channel) would be ok, since you don't need over-unity gain. You could decide to buffer the input with the other half of the tube (then you'll need 2 5670's for stereo) as is done here, but you won't need much gain. Less than in the schematic you posted above.

I don't think there's a ready-made and optimized schematic for your particular application. The topology of the 16LS kit (which I assume that is) you refer to will certainly work, but it might (will) need optimization for your particular purpose, since driving a PP (double) pair of 6005's is different from driving an LM1875.