DC Coupled Cathode Follower ANOMALY - Help.

[Glennjeff]

I have a very standard DC coupled cathode follower that is not behaving as expected.

According to The Valve Wizard here http://www.valvewizard.co.uk/dccf.html  (halfway down page) this 12AX7 should be biasing close to 0V but it is -3 Volt biased and acting pretty much as a perfect voltage follower. Schematic from The Valve Wizards site attached.

I have fiddled round with the first triodes bias to get between 170 and 190 V on the grid of the second triode to no avail. B+ is correct @ 300 V.

Anyone ???

[HotBluePlates]

I have a very standard DC coupled cathode follower that is not behaving as expected.

According to The Valve Wizard here http://www.valvewizard.co.uk/dccf.html  (halfway down page) this 12AX7 should be biasing close to 0V but it is -3 Volt biased ...

Well, Merlin said, "We know that for normal operation the cathode will be a couple of volts higher than the grid (which is connected to the anode of the previous stage and is therefore at 195V), but for simplicity we will assume it is at the same voltage as the grid..."

Sounds like everything is working just as it should. Does it not do what you need it to do?

[tubeswell]

And in addition to what HBP said, the resistor/diode pathway between the HT and the ground return will 'steal' some current from the plate resistor for the driver stage, so the triode won't behave as it would if that current pathway was less direct. On its own, that driver triode would sit at ~190 on the plate with HT = 300, Ra = 100k, ra = 62.5k, Rk = 1k5, but with the resistor diode chain of 100k>100R>10k>D1>47k, you will inevitably drag the plate voltage of the driver down more. When the CF biasing kicks in, this 'current stealing' pathway will be through the grid of the CF.

[Glennjeff]

HBP,

I should have re read the text eh. 

The stage is not giving the one sided compression it should, but instead being an accurate voltage follower. Even when the prior stage clips, it follows accurately.

I've been working on a LTPI with a constant current FET in the tail and compressive cathode followers for a distortion specific amp, and it's been a lot of trial and error due to my lack of experience. Only thing to do now is get these CF's to compress. I might just have brain drain and be missing something simple.

I was getting ready to elevate heaters and put more volts on the grid so thanks for the rescue.

[Glennjeff]

tubeswell,

With the cathode 3 volts higher than the grid the protection diode is reverse biased and not drawing any current.
I knew that diode was meant to be reverse biased for normal operation.

I managed to overlook this obvious point when incorrectly believing the text said cathode and grid would be at zero volts.

Definitely got brain drain.

All the voltages are as they should be apparently, so will try to analyse currents tomorrow.

Thankyou. 

[PRR]

It may be "linear" for -small- signal levels. IMHO, it must be "badly" asymmetric for signals more than a few volts. (2nd stage can't pass 4mA with only 110V plate-cathode.)

[Glennjeff]

PRR,

I didn't explain it too well. It is only linear for very small signals but my imaginary problem was that the distorted signal on the anode of the driver was not being compressed at the cathode of the follower. Of course that circuit work s through grid current in the follower dragging down the voltage on the drivers anode. Once I measured from the grid of the driver to the cathode of the follower the compression effect was clearly evident.

I was trying to think of too many things at once and got in an autistic spectrum disorder sinkhole.

Just got the whole LTPI with FET tail sink and Compressive CF's working and it has a promising sound to it.

Thanks a lot for bumping the stuck record folks.

[Glennjeff]

Might as well put it here, comments, thoughts and critique much appreciated.

Not all values provided as I need to experiment with different triodes and bias setups to get the preferred result.

I'm trying to get all stages of the amp to go into grid clipping as gradually as possible and all at the same level.