White Cathode Follower sound

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

MaxDM

Well-known member
Joined
Oct 23, 2018
Messages
206
I'm considering building a mic pre without an output transformer, inspired by the Mastering Lab ML-1 preamp.

I was wondering if any of you had already experimented with such a circuit and if in the end it pleased you, or conversely you found it strident or annoying?
 
Manley uses WCF's in almost everything. To my view Manley was certainly influenced by TML's approaches. See this thread: Needed Manley All Tube Mono Mic Preamp Info

On the early Manley stuff it was WCF with 6414 with no output transformer. This is what's in my Elop. Later they added an output transformer, but provided the option of using the unbalanced outs without transformer, or balanced out with transformer. Now on the Force preamp they are utilizing a MOSFET WCF. I am interested to hear this sometime.

Directly coupled WCF will be "faster" and "more open" sounding than an output transformer. It will have more headroom. It has a sparkly hi fi clear sound that I personally like on lots of stuff. Whether you like that is up to you. The next gear in line will also matter. I find that I can "slow things down" with transformer based gear inline (think Pultec) if I like, or somtimes I'll use Crane Song's Phoenix plug in. But if something is recorded too sludgy to begin with, it's nearly impossible to "clean it up" later. YMMV.
 
Sparkly, fast and hi-fi is fine with me.

You don't notice any disagreeable distortion, or exaggerated transients?

One thing I am cautious about, when circuits have a lot of correction, is how it's going to sound when reproducing the transient of the pick hitting the string on an acoustic guitar, and other instruments with sharp transients.
 
A word I've heard used about the ML-1 is "effortless". It can do +30dbU at 0.009% THD, with a proper set of tubes. +24dBm. It has such stupid headroom that it's usually not the problem. If you play guitar, it's like playing a very high gain, fast amp like a Dumble the first time. It makes everything you do right sound better, and everything you do wrong worse. If the player is clickly and not smooth, and the mic might be a bit stiff, and the strings are brand new, a fast preamp of any kind might be something to avoid. But as I said, all the Manley stuff uses this topology, for good reason. We've all heard it a thousand times. I strongly doubt you'll be disappointed with WCF output, in and of itself. Oh wait, I forgot LA-2A is WCF I think.....

Regarding recording acoustic instrument players, Gary Paczosa says he uses the ML-1 on everything. Listen to his work with AKus and you'll get a big dose of that sound. They are known to be great on acoustic instruments, which I've found to be the case too. Stellar, in fact. George Massenburg said in a interview "on vocal I either use one of my preamps or a Mastering Lab".

I know a guy that just used two ML-1's for his acoustic record. His name is Eric Wiggs. He's on all the places to get music. It's a great example of how when they are right, they are pretty unbeatable.
 
Last edited:
A well designed 'White Follower' should not give any untoward artifacts.

One thing to mention: the White follower in the LA-2A is NOT one to copy, that top resistor is far too big a value for the gm of the 12BH7 as used.

There's lots of discussion online on how to properly design one with the correct amount of signal voltage being sent from the top to the bottom half.
 
wow, sounds like something I would be interested in using!

From what I've read, it would seem like the Mastering Lab's pre uses a simple two-triode, cascaded gain amp, DC-coupled into the white follower, of which I already have the anode resistor value, being 809 ohms.

I suppose the top and bottom half need to null, gain-wise?

Perhaps, an anode choke on the second stage might improve the design further?
 
Perhaps, an anode choke on the second stage might improve the design further?
Balancing a WCF requires controlling "the correct amount of signal voltage being sent from the top to the bottom half.", as Winston wrote. This voltage is primarily governed by the upper anode resistor and the load.
Two consequences:
A) The upper tube load must be of the same nature as the load. Since we are considering the load as resistive, the upper tube anode load must be resistive.
B) The upper tube anode resistor must be optimized for the load, which implies making sure the output stage sees a rather constant load.
 
The optimum value of the top resistor is 1/Gm and this is irrespective of the load on the complete stage. What we're looking for is a balance of the top and bottom triodes such that each one delivers the same output impedance to the external load.
The composite output impedance being Zo = 1/2Gm
 
No worries Abbey. You're probably not fiddling with White followers that often so, easy to forget.

There are folks who do fiddle with White followers, for a living, any they didn't even know this in the first place to have temporarily forgotten it now :cool:
 
No worries Abbey. You're probably not fiddling with White followers that often so, easy to forget.

There are folks who do fiddle with White followers, for a living, any they didn't even know this in the first place to have temporarily forgotten it now :cool:
and they also erroneously assume that this (small signal) output impedance is identical to driving capability.

Cheers

Ian
 
Haha, also true.

Best uses for a White would be driving modern type higher input impedances.
Or driving a low ratio output transformer such as 3.5:1
 
Haha, also true.

Best uses for a White would be driving modern type higher input impedances.
Or driving a low ratio output transformer such as 3.5:1
Like the SRPP it is a true push pull output stage so with about 6mA of quiescent and a 2:1 output transformer is should be capable of over +20dBu into a 600 ohm load. Just don't try this with a 12AX7!!

Cheers

Ian
 
Sounds good to me Ian. I picked 3.5:1 as a ratio merely because it popped into my head from personal experience. A 2:1transformer with appropriate standing current , H.T supply, and valve (yes, no 12AX7's please!!) would be nice.

One benefit with an optimized and perfectly balanced follower over a standard, single cathode follower when driving a transformer is that the output impedance is halved: Zout = 1/2Gm

Transformers always benefit from being driven from a lower impedance source.
 
Last edited:
P.S. I haven't tried this, but a high Gm pair of small-signal pentodes would most likely make a really nice White follower.
However, to get the same signal swing ability as a triode-ed White, you'd need to have the ability to raise the H.T. supply to account for where the two screen grid voltages sit.

I suppose, if you were dealing in a world where your average signal is 1 - 2V rms, then a typical 300V H.T. line would still be more than enough.

Anyway, just for clarification purposes if anyone ever tried this: the top pentode's screen grid resistor would need to be bypassed with its cap returned to your output signal. With a 300V H.T. supply this would typically be at about the halfway point of 150V DC. Don't return that cap to ground.
We want the top screen grid to ride along and be referencing its respective cathode.

The bottom pentode screen grid's bypass cap would go to ground of course.
 
Last edited:
P.S. I haven't tried this, but a high Gm pair of small-signal pentodes would most likely make a really nice White follower.

I wonder if the 6EJ7/EF184 might work well here? I've admired them from afar, but never used one. I've considered trying it in the front end of a guitar amp head, but since their Gm is 15mA/V at 10mA Ia and they were intended for IF amplifier duty out to 40 MHz, they'll (A) probably oscillate just sitting in the box unless strict preventive measures are taken, and (B) I fear they could be rather microphonic, especially considering their insane gain.

Configured as a White CF, they'd indeed be able to blow fire through 10km of cable. Has anyone tried them? They're only about $6 USD each for NOS, and the Russian 6J15P equivalent is quite inexpensive as well. The only thing better than "great" is "great and cheap."
 
I wonder if the 6EJ7/EF184 might work well here?
....
Has anyone tried them? They're only about $6 USD each for NOS, and the Russian 6J15P equivalent is quite inexpensive as well. The only thing better than "great" is "great and cheap."

I haven't messed around with those in particular, but they're exactly the type of high Gm pentode I'd think would be great configured as White followers.


My experience is just playing around with these types as a standard follower, not White, and the valves I used were E180F's and E83F's - purely because I had some lying around as spares for old Siemens V72A and V76 mic amps.
I probably slipped into 'heretical' territory because there was a solid-state constant current sink below them in my experiments.
A White follower of course will give you a 'hollowstate' constant current so you still get to hold your head up in polite company then.

Your thinking on being "great and cheap" is also where my thinking is.

I don't know what there is lying around in the USA, but in the UK you can get a nice NOS Mullard E180F or Philips E83F in the £5 range and they are plentiful because no one really wants or particularly knows about them.

Anyway, I don't see why an EF184 wouldn't also be great, and $5 is cheap as chips :)
 
Last edited:
the anode choke would be placed on the second stage, not the WCF

I am trying to quote on this forum, but must be doing it the wrong way
 
Last edited:
One benefit with an optimized and perfectly balanced follower over a standard, single cathode follower when driving a transformer is that the output impedance is halved: Zout = 1/2Gm

Transformers always benefit from being driven from a lower impedance source.
Which is a good reason to choose a high gm tube for such a stage. My preference is the 6922 (E88CC).

Cheers

Ian
 
I have always liked this circuit. I was just wondering in the past few days why a similar solution hasn't been used in the Motown direct Amp we are discussing in another thread. The patent was published in 1944, in the following 20 years it should have had a good circulation.

Cheers
JM
 

Attachments

  • GB564250A_Original_document_20210524134200.pdf
    681.3 KB · Views: 76
Back
Top