White cathode follower - DC bias?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

alk509

Well-known member
Joined
Sep 3, 2004
Messages
1,207
Location
MA, US
Hey guys,

OK, so I'm trying to finish off this mic pre with a White CF, but something isn't clear to me here... The WCF will be getting signals with peaks of up to about 10V (18dBm or so), so I'm thinking I should bias the grid of the upper toob to around -11 or -12 Volts, right? But then I start looking at other WCF's online and in books and things, and they ALL have these wimpy grid biases in the -1 to -3 Volt range.

I want to use a 6N1P tube for my WCF, so anything lower than -4V is pretty much out of the question... So my question is: How the hell do you settle on the grid bias of a WCF? Do you just set whatever grid bias you want and let the intrinsic feedback in the WCF deal with too-high input signal levels?

Thanks!

Peace,
Al.
 
[quote author="alk509"]I want to use a 6N1P tube for my WCF, so anything lower than -4V is pretty much out of the question...[/quote]

Just to clarify, when I say "lower than -4V" I mean lower. As in -5, -10, or -16V.

Peace,
Al.
 
> signals with peaks of up to about 10V (18dBm or so), so I'm thinking I should bias the grid of the upper toob to around -11 or -12 Volts, right?

No. You are thinking of grounded-cathode design. In a cathode follower, the cathode tends to follow the input.

So at idle, you may have 100V on the grid, 102V on the cathode (2V G-K). Swing the grid up to 120V, and the cathode will rise to 121V (1V G-K).

Actually, for a no-load cathode follower, the grid-cathode swing is about Vout/Mu. A good tube for this duty will have a Mu of about 20, so a 2V G-K bias is fine up to 40V peak swing. A full WCF even more.

The real question is: with tube shorted grid-cathode (maximum conductance), can you pass enough current to feed both the DC and AC loads at that plate voltage?
 
In most modern WCF uses, we are "wasteful", throwing 300VDC in to get say 10Vpk out. In that case the design can be approximated as "no signal voltage swing" and you can get a workable design with a glance at the plate curves.

Image45.gif


For starters: assume you have about 300V supply, about 150V per tube, and zero signal swing. At maximum positive current swing, the G-K voltage should go to zero (never positive; grid current is problematic and not justifiable until you get VERY high plate voltages).

Find 150V Vp and 0V G-K. Current is about 24mA. Idle current should be half or less than that: 12mA. Still at 150V, the G-K voltage for 12mA is about -1.8V. In a full WCF, this tube can give 24mA peak at gross clipping.

BTW, you do not set the G-K voltage on the top tube. Set the cathode resistor on the bottom tube to establish the desired idle current. 1.8V/12mA= about 150Ω. Then bias the top grid to about half the total supply voltage, and the top cathode will find itself at the right place.

Now, 24mA peak in 600Ω is about 14.4V peak. So where we assumed 150V, the tube will really only have 150V=14.4V= 135V across it. We also need to subtract the resistor voltages, a few volts. So the peak current is more like 20mA, 12V peak, at gross clipping. Idle current should be 10mA, idle G-K voltage now about -2V, cathode resistor should be around 200Ω.

While I've only estimated the gross-clip point, a well-balanced WCF is very-very linear right up to the point that it gives up. 10V peak in 600Ω is reasonable. Getting a WCF to balance properly can be hard. But to over-simplify: the top plate resistor should be about the size of the bottom cathode resistor (about 200Ω here). A full derivation for low-Z loads suggests a slightly smaller value like 160Ω, but larger values work very well in higher-Z loads (but then it can stop working as a WCF and just be a double-cost Cathode Follower).

Always check the max ratings! Idled at 150V and 10mA, plate dissipation is 150V*10mA= 1.5 Watts. Specs on Russian tubes are hard to be sure of, but the one I found says 2.2W per plate. Nominally OK. Plate voltage max is 250-300V, OK. Cathode current max is 25mA, OK. Heater-cathode voltage max is 120V, so the top tube can not be heated with a grounded heater supply. In fact this is one real problem with the WCF: cathode voltage can spike very high at start-up, puncturing heater insulation.
 
Thanks PRR! I knew you wouldn't disappoint. :grin:

Lemme read that post and do some math and get back to you.....

Peace,
Al.
 

Latest posts

Back
Top