| voivodata |
HI ALL
Here’s an interesting experiment:
The preamp section of my modified Fender Tremolux Head has a clean and an overdrive channel (lost the tremolo). The OD channel has four gain stages and is based on commercial amp design such as Groove Tubes, English, Bogner, etc… The main difference is that the third gain stage cathode is shorted to ground (initially I did this by accident but I liked the sound and decided to leave it like that). Compared to cathode resistor and cap values of say 1k5 and 1.5 uF and all else being equal, when that cathode is grounded there is rather significant (30 – 40 %) increase in overdrive (distortion) and in the same time a noticeable decrease in noise . That’s a quite a bargain cause usually more preamp clipping is achieved by more gain resulting in more noise. I don’t have knowledge to calculate and understand what exactly is happening there, but I suspect that the third stage does not provide more gain but only clips more (sort of like the diodes in solid state distortion designs). This mod leads to some decrease in bass response, which is only audible at lower overdrive settings (like for blues) and at lower volumes, but this doesn’t bother me because the bass was anyway too much - it caused the amp to lose clarity in higher distortion settings for rock, metal, etc…Other than that the mod does not have drawbacks such as more harshness or loss of character. I should note that grounding the cathode works as described only if done at the third gain stage. At any other stage the result is less gain, less clipping, less noise, as could be expected.
My questions are:
1. Can anybody provide a simple (not too scientific) explanation why the above experiment leads to more distortion and less noise at the same time?
2. Does grounding the cathode of ½ of an ECC83 shorten the tube’s life?
Thanks & regards!
V. |
|
|
| SY |
You are certainly getting more gain and at the same time decreasing overload (i.e., more distortion). I don't know if there's negative feedback wrapped around that stage, but if so, the greater open loop gain (with a fixed closed loop gain) will increase the feedback and decrease the noise.
As for tube life, it depends on how much current is being run through it. If you're running that 12AX7 balls-out, it will die sooner. That may not be a big deal if you like the sound; those tubes are cheap and easy to get. If you increased the current from (say) 0.5mA to 1mA, it's unlikely you'll affect the lifetime very much. |
|
|
| voivodata |
Thanks SY for you reply!
There's no negative feedback around the preamp so the decrease in noise still bugs me especially since there's an increase in gain as you note and since this happens at the third gain stage only.
Maybe it's related to the parts placement or my messy point to point, star ground wiring attempt causing miracles. I'm currently building a preamp (only) and can't wait to see if the same "trick" works there too.
Don't know how to figure the increase in current but will make sure to replace the NOS RFT with something else just in case. Thks
Regards! V. |
|
|
| SY |
| Could be that you got rid of a noisy (defective) cathode resistor. |
|
|
| voivodata |
I thought I should report back.
I tried the "trick" with a preamp only, which is an older unit I had landed to a friend (haven't finished yet my new "ultimate" :) tube distortion) and the results are the same - more gain and less noise and works at the third gain stage only. So I guess it was not defective resistor related. Could be the design because when building the preamps I was pretty "imaginative" and trusting my ears (rather than following a particular design) in choosing values for grid stoppers, filtering caps, etc... between stages. Will try to draw and simulate the schematic with some free CAD one of these days.
Well... that's it... and cheers to you all, good tube people!
V. |
|
|
| aletheian |
| Sounds like you inadvertantly converted the stage to grid leak bias and shifted the operating point to a bit more of an assymetric swing. Fender actually used to use it in very early amps.. typically you'll use a 2-5Meg grid leak resistor. It is not too reliable though, and drifts around a lot, so for the sake of consistency from amp to amp, and other factors, they ditched the idea. you can read about it in the RDH4 or just about any old tube textbook. |
|
|
| fdegrove |
Hi,
For a typical ECC83 circuit you find gridleak bias resistors more in the order of 10M.
This method was used in older, lower end Counterpoint preamps, the more expensive ones used a seperate negative bias rail to set Vg-.
This method offered several advantages, no bypass caps with their slow transient response for one. Another advantage is that you can fiddle with the sensitivity of the first stage to accomodate oddball MM cartridges for instance.
If you just short the cathode to ground bias will be practically inexistant with a typical 100K gridleak resistor, it will initially impress with a quite funky distortion pattern rendering kickdrums and lead guitars, for example, a quite livelike sound. Coinciding distortion patterns?
I never really looked into it further but indeed, noise goes down (no bypass cap, no cathode resistor, no NFB locally?) and gain seems maximized but overload suffers no end rendering the whole idea pretty useless except perhaps for a few special cases....
Ciao, ;) |
|
|
| Merlinb |
Firstly, no it will not shorten the lifespan of the valve. It probably has a large (100k) anode resistor, and that means only a couple of milliamps can EVER flow through the valve, so it should last for years (it's a different story for power valves though- don't remove THEIR bias!)
Distortion increases because you've basically wired the valve like a diode- a half wave rectifier. It can amplify negative input signals easily, but when the input swings positive, most of it is shunted to ground through the cathode. It's not as harsh as an SS diode of course, but the same principle.
It works ok at lower levels, but at higher levels you get pretty hard asymmetric clipping, which will usually result in a fuzzy mush, as you seem to have noticed.
The reduction is noise is odd. What sort of noise did you mean- hiss, hum? If it's less hum then it may be because you've placed the cathode firmly at zero volts, so there can be no interference from the heater (AC). |
|
|
| aletheian |
| quote: | Originally posted by Merlinb
Firstly, no it will not shorten the lifespan of the valve. It probably has a large (100k) anode resistor, and that means only a couple of milliamps can EVER flow through the valve, so it should last for years (it's a different story for power valves though- don't remove THEIR bias!)
Distortion increases because you've basically wired the valve like a diode- a half wave rectifier. It can amplify negative input signals easily, but when the input swings positive, most of it is shunted to ground through the cathode. It's not as harsh as an SS diode of course, but the same principle.
It works ok at lower levels, but at higher levels you get pretty hard asymmetric clipping, which will usually result in a fuzzy mush, as you seem to have noticed.
The reduction is noise is odd. What sort of noise did you mean- hiss, hum? If it's less hum then it may be because you've placed the cathode firmly at zero volts, so there can be no interference from the heater (AC). |
Yeah, with the cathode grounded, you are decreasing the available headroom for the negative signal swing. With the 100k plate resistor and what i *assume* to be the standard Fender 250v B+ supply, you are probably drawing just under 2mA... which won't hurt the tube, but if you crank it, you are probably driving the tube (and I am just speculation here) into A2 diode grid clamping or blocking distortion... which is ugly sounding. |
|
|
| voivodata |
Hi
It's nice to hear a differing opinion as to effect on tube's life.
That stage's anode resistor is 220k and B+ is like 375v. I see I didn't initially clarify that this head is not an original Fender but a knock off or a clone - lookes like a factory job thought, not a DIY.
| quote: | | It works ok at lower levels, but at higher levels you get pretty hard asymmetric clipping, which will usually result in a fuzzy mush, as you seem to have noticed. |
Merlinb, by lower and higher levels did you mean levels of the power amp (master gain) or levels at that particular stage (preamp gain)? Asking because with the cathode grounded the amp sounds better at higher preamp gain levels - not fuzzy and mushy bur rather tighter and more focused - think System of a Down on steroids ... well, almost maybe :). I'm not sure if the same applies for master amp levels though. Generally, I'm having more of a problem getting the lower gain blues type of sounds. I think the next time I open up the amp, I'll implement a switch for grounding the cathode so I can compare the sounds at different levels with different speakers, etc...
| quote: | | What sort of noise did you mean- hiss, hum? If it's less hum then it may be because you've placed the cathode firmly at zero volts, so there can be no interference from the heater (AC) |
I'm not quite sure but I think it's neither hiss nor hum. It's more like a quiet ..brrrrrrrr.... To me it sounds like the guitar pickup's noise is being lowered by the cathode grounding at the third stage (or maybe just kept at the same level as at the previous stage) while at the same time distortion increases.
Btw, I had an epic battle with noise in this amp and thanks to help from this forum was finally able to reduce it to very low (at least for me) levels considering the amounts of shredding distortion coming out of this thing. Here's a link to my previous post: http://www.diyaudio.com/forums/show...&threadid=80031
There's a picture of the amp's insides attached there but I think it's taken before my cathode grounding experiments, before I replaced most resistors and caps with better quality ones, and before I added a 1/2 of ECC83 in parallel with the input stage.
Thanks and regards.
V. |
|
|
|
