Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: separateness on August 21, 2023, 08:45:27 am
-
Can any one tell me if the cathode follower in the attached picture is grid-leak biased? If it is not, then how is it biased? If it is gird leak biased, could this fact account for the output of this triode having quite a bit of white noise/hiss? Further, could it be improved with a different biasing scheme?
If this question is too general for this forum, I apologize.
-
That circuit looks wrong. Look here for the proper way to bias an AC coupled CF...
http://www.valvewizard.co.uk/accf.html
-
I agree that it looks wrong. In fact I am quite confused about it, but it is indeed the circuit which exists within a commercially produced piece of gear that I own and use regularly.
I had considered changing it but I am smart enough to doubt that I know better than a circuit designer.
Edit: attached full schematic for context
-
but it is indeed the circuit which exists within a commercially produced piece of gear that I own and use regularly.
Can you check that your circuit actually matches the schematic you’ve provided? As the schematic is not an original and appears to be the only one available online it would be nice to confirm if it’s drawn correctly.
-
It is accurate. I know the input capacitor has some variation, some are 150p others 250p. I do not recall what I put in mine when I went through it but the circuit is accurate. I did change a few things, like the heater circuit, but that CF is indeed as pictured.
-
Interesting unit. I had never heard of it before. I don't see any reason to muck with the cathode follower. Good demo on youtube.
-
I love mine. Having a motor running inches from a pretty high gain pentode stage is somewhat unfortunate but it is mostly very well behaved and the delay sound is to die for for my money. It's just when I turn the echo or reverb pots up past about 4 it has some hiss/noise and I wanted to try and minimize it for recording.
I was trying to brainstorm ways I could minimize it and since the common denominator seems to be the CF I thought maybe I could fiddle with that but I can't even seem to figure out how it works or is biased. I don't think I've ever seen a cathode follower such as this, though the catalogue of circuits I have seen is not vast. For such a low-parts count circuit, I have found it somewhat tricky to parse.
-
The heater elevation does not make any sense, too.
-
It is "wrong", but the signal level off the playback head is VERY small, so it works OK.
-
The 2.2M resistor directly in the signal could be noisy? Perhaps you could experiment with a low value snubber cap to ground?
-
Well, the hiss is only present when the Reverb and Eco knobs are turned up, The 2.2M is in the signal path at all times and would not be effected at all by the Reverb and Echo.
If the CF is 'wrong' is it worth my time to make it 'right'?
-
The noise is "tape noise". Yes, it is not tape and not magnetic but it is surely non-uniform, random.
-
So is this to say that it is an unavoidable artifact of the crude recording medium used by the device, and that nothing short of inserting a dbx circuit between the oil can and the cathode follower is likely to make any meaningful difference?
The post AlNewman's suggestion of a snubber (around the 10kΩ load resistor, I presume) had occured to me but I have no idea how to calculate the gm and thus Zo of this circuit, as my feeble attempts at determining the bias point gave me ridiculous answers.
-
Have you considered types of components , meaning, carbon comp , carbon film, metal film resistors,
And different poly caps, ceramic, silver mica, etc
?????
Maybe some hiss killers hiding in the different types.
✝️🤪🎹🎸
-
ehhh, i was thinking right on the back of the 2.2M resistor, maybe a 10-100 Pf cap to ground.
But you're recording, it's probably easier removing those frequencies in post production. Why mess with a nice thing?
-
When I first got this thing I replaced pretty much every component in it. Metal film resistors, silver mica small caps etc. I believe PRR is wholly correct that this is tape hiss, or electrostatic disc hiss as it were.
If I put a cap in the clean signal path I risk effecting negatively the bandwidth of the the clean signal and still the signal which goes through the echo pot would remain unaffected, I predict. If I make an intervention I would prefer its effects to be isolated to the wet signal path. You are perhaps correct that some post-recording eq might be my best course. The thing plays great. I rarely play without it on.
Back to the original question, perhaps purely academic: can any one tell me if this CF is actually grid-leak biased, or how in the heck the biasing is happening here? The best I can tell the grid to ground voltage is right around 0V and the cathode is sitting around 15-20V which according to the datasheets I have seen puts the quiescent bias point darn near shutoff.
-
Yes, that CF is grid leak biased. If you want to try other biasing methods, go ahead, whatever floats your boat. You can always put it back to stock.
-
Would you predict that another biasing method would be less liable to noise? I have read that grid leak is somewhat associated with picking up stray noise (because of the massive input resistor, I presume), but if it is only passing on what it is getting from the can, then probably filtering might be my only course of action.
-
Would you predict that another biasing method would be less liable to noise?
Try it and see.
-
Is "the sound" in the tubes? Or is that just the only technology of that time?
And isn't that a LOT of functions to pile into one bottle?
Get a couple FET opamps and do the stages in a slightly less troublesome technology.
-
I think part of the sound is the tube but it is mostly the ridiculous stuff going on in the can. I think I do rather like the way the pentode responds to playing and setting the screen voltage with a pot on the back is pretty neat for tailoring a sound. Another thing about it is if you turn the reverb up a lot and dig in the echoes get quite loud and hairy. Perhaps this is from overcoming the tiny bias voltage of the CF, I'm not sure. I kind of like that but also the onset of it is pretty sudden, going from peaceful reverberation to banshee-in-a-canyon with only a slight difference in touch. No, I do not believe I want to forego the tube completely. I think I might just try a shunt cap right around the 10kΩ load resistor (though I still have no idea what the Zo is here), and maybe, if that does not work, try designing a self-biased cathode follower stage, something I have never done. One worry I have with that is after building one on paper earlier, I realized that a well designed CF would idle at a much higher current than what is in there (which idles somewhere between 1 and 2 mA), which might increase PS noise.
I have it apart on my bench, only I got sidetracked trying to fix my (Philips 3215) oscilloscope. I had to take out the 'Focus' pot to bend the wiper blades and man they do not make oscilloscopes easy to service.
-
The type of oil in the can plays a huge part in the sound. Did you listen to the demo video I posted? I thought it sounded very good.
Catalinbread makes an Adineko pedal that I thought may even sound better.
-
I did and it does sound quite good. I use only UCON LB-65 oil as that is what is supposed to go in it. I've read of people using brake fluid or all kinds of stuff but I wouldn't dare use anything but what Tel Rey intended.
-
... designing a self-biased cathode follower stage, ...
See example (a 7199 triode is similar enough to a 12AU7 triode that you could just adopt these circuit values).
-
Well, I hate to be a disappointment (something I ought to be used to by now) but having messed around with it quite a bit yesterday I think I am just going to suck it up and accept the little bit of hiss. I bought this thing knowing it was a rowdy old piece of gear and that is what I am going to live with. Yesterday I hung a 33nF cap across the 10k load resistor and was done with it.
I think putting in a legit cathode follower would tax the ps too much and ripple would become an issue. We have a halfwave rectified ps with a current draw of about 1.2mA per stage stock, making 2.4mA total. This is peanuts—those filter caps probably barely feel it. A properly designed CF with this tube would put me at 10mA quiescent current draw minimum. I think this would create PS ripple which was intolerable.
A fellow once said 'nothing is often a good thing to do' and I think I will heed that advice here. I appreciate all the responses here gentlemen but special props to user AlNewman for this sage wisdom:
But you're recording, it's probably easier removing those frequencies in post production. Why mess with a nice thing?
I am quite pleased to finally understand why this cathode follower looks so weird and how it works. For that I am very grateful.