Paralleled V1 question

[kagliostro]

The simple way is to use a switch to connect the second triode grid to the input

but assuming to go a step further on, is possible to "control" the insertion of the second triode, I mean something like using a pot ?

if the tube wasn't V1 I'm sure this can be done, but as the signal present at the grid of V1 is the lowest signal in the amp

is possible to use a pot in the position of the switch ?

Do you see any other way to achieve the progressive insertion of the second triode ?

Thanks

K

[shooter]

found this on the net, tried it, seemed way "overdriven" for my app, but it is sorta what you're askin....I think

[tubenit]

Check out the "FAT" control on the original Tweed Overdrive Special.


With respect,  Tubenit

[sean_thornton]

All those ideas are kinda cool. In my current build i have V1 paralleled with a relay switching in different cathode and plate resistor values along with the bypass capacitor. Originally i designed it so that V1b could be switched in and out via the plate and cathode, leaving the grid connected. I worked fine but i prefer the tone of a paralleled setup so i dumped the switching idea.

[kagliostro]

Thanks Shooter - Tubenit - Sean_Shornton


Jeff the idea starts exactly from your FAT control on the TOS 

only, here, I'm not sure about the use of a pot because here we talk about V1 and the input signal is very low

the only pot in front of V1 grid I see around are on HiFi circuits where the signal is supposed to be higher than the signal from a guitar

or isn't so ???



Franco

[shooter]

Not sure what a guitar norm is, I use .5vac from a tone generator, volume seems close when a guitar is plugged in.  I think hi-hi's uses a standard  like line in, line out, 1vac?

sean_thornton did you do your component switching on the fly, a foot switch?

[sean_thornton]

Yeah Shooter, double footswitch, one for reverb, one for "fat". I will post a schematic once I have the bugger finished, then it's on to the next build, LOL!

I was actually going to ask kagliostro the same thing about his idea, are you going to use a footswitch and a relay, or just a switch on the front panel?

[kagliostro]

The idea was to use a switch on the front panel


of course, if a pot arrangement is possible, there may be a switch too


K

[kagliostro]

I'm in doubt

I've think to the circuit and a second option come out

this second version provides for the use of a separated plate resistor for each triode and a second switch to connect or disconnect the second plate output

do you think this is necessary to avoid changes in tone of V1 when it is alone, I mean to have V1 that sounds as if in the circuit there isn't V2 when V2 is not paralleled (I hope I've explained this thing in a comprehensible way)

Thanks

K

[DummyLoad]

switching at the grid, the relay wiring will add *some* noise; the plate switching will POP and possibly exceed the relay DC volt rating but probably won't matter since it's "floating". bottom schematic is what i'd build. fewer switching elements.   


have you considered switching the cathode of V1b off ground?


--pete

[kagliostro]

Ciao Pete

Do you think that this arrangement will be of benefit trying to avoid pops ? (look to the attached schematic)

---

One other thing, the original schematic uses a V1a Plate resistor of 100k and 1.5k cathode resistor

the schematic that we are discussing is the Alembic F2b  to wich is added a CF on the output and having an odd triode this is used as to


have a paralleled input



Which do you think will be the best value for the Plate resistor using a single plate resistor for both V1a & V1b ?

Grazie

Franco

[sean_thornton]

If you have the plate and cathode switched out there is no need to mess with the grid. My own experiments showed me that leaving both grids connected in a parallelled 12ax7 had no detrimental effect when switching the cathode and plate of one triode in and out. It also taught me that i prefer the tone of both triodes switched in so I dont even bother switching them anymore, I have come up with other ways to manipulate gain and tone. I also suggest using relays, this will also help to eliminate "pops" when switching.

[kagliostro]

OK Pete & Sean_Thornton

Do you think this will be a better arrangement ?

Franco

[DummyLoad]

should work. however, it is likely sean is on track: you'll eventually stop using the switch.

[sean_thornton]

Well, it looks like you are trying to imbalance the 2 triodes. Not sure why. The Alembic preamp is a gorgeous sounding one, much loved by Gerry Garcia, and is hard to improve upon. What is it that you are actually trying to achieve? Have you checked out Matchless Amps? They have a parallelled V1, look at their arrangement.

[DummyLoad]

paralleling triodes will bump gain up roughly 15%.

[kagliostro]

What is it that you are actually trying to achieve?

All starts from the fact that the F2B is a good preamp but is missing of the necessary impedance adapter at the output

if the buffer is add via a single triode (6av6 - 6at7) the counterpart is the availability of spare parts "on the road"

if the buffer is add via a double triode (12ax7) the counterpart is an unused triode

as to have easily spare part available is obviously welcome but what to do with the "odd" tube ?

To add a gain stage will upset the original F2B and looking to other options one is to use this triode for a CF

between V1a and TS, one other to use it paralleled to V1a, paralleling triodes, as Pete say, will add a 15/20% of gain

and increase the noise rejection, the increased noise rejection is the principal reason for which I chose to go parallel

the switch is think as to permit to use the preamp more like the original if desired

K

[sean_thornton]

Ahhhhh! NOW i understand! Okay, so basically you wish to keep the Alembic as close to original as possible but have the option of switching the spare triode in and out.
Okay, kep the V1b grid connected, just switch the cathode and plate in and out. For simplicity i would suggest that the cathodes share a common resistor and the plates do the same. this can easily be achieved either with a switch or via a relay. You CAN incorporate separate resistors for cathodes and plates but I found that it isn't needed.
I hope this helps, let me know if you need any more info.

[HotBluePlates]

If you still haven't tested to see whether you will like the parallel-triode option, I would wire up the volume control as you had in your very first post.


If you decide you want to incorporate switching that is guaranteed not to pop, you will want to look at solid-state switching, using series and shunt JFETs as the "switch elements" (that is, in place of switch/relay contacts) and solid-state control circuitry to activate the JFETs. I'm not a solid-state expert, but LooseChange (and probably others) has used this in at least a couple amps, and Kevin O'Connor covers it in some detail in his books.

[kagliostro]

@ Sean_Thornton

kep the V1b grid connected, just switch the cathode and plate in and out

Switching the plate is enough safe ? No pop or other contraindications ?

@ HotBluePlates

If you still haven't tested to see whether you will like the parallel-triode option, I would wire up the volume control as you had in your very first post.

The pot option is on standby because I'm afraid that the pot will load to much the input that is very low level signal, do you think is possible to give it a try ?

SS switching intimidates me, I've no familiarity with SS circuits

Franco

[PRR]

> add via a double triode (12ax7) the counterpart is an unused triode

EL84 is an excellent single triode with LOTS of possible output and readily available on-the-road.

[kagliostro]

EL84 is an excellent single triode with LOTS of possible output and readily available on-the-road.

Ciao PRR


do you mean an EL84 connected as triode for a CF output ?


Something like this ??





Franco

[HotBluePlates]

@ HotBluePlates

If you still haven't tested to see whether you will like the parallel-triode option, I would wire up the volume control as you had in your very first post.

The pot option is on standby because I'm afraid that the pot will load to much the input that is very low level signal, do you think is possible to give it a try ?

Use a 1MΩ pot, wired exactly as you first drew your schematic. The pot will appear to the input signal exactly the same as the typical 1MΩ grid reference resistor, and will be no extra load at all.

[Platefire]

Yep, interested to hear your opinion/report of the results once installed. I've got a couple of amps with unused V1b gain stages I could possible apply this to. Platefire

[DummyLoad]

@ HotBluePlates

If you still haven't tested to see whether you will like the parallel-triode option, I would wire up the volume control as you had in your very first post.

The pot option is on standby because I'm afraid that the pot will load to much the input that is very low level signal, do you think is possible to give it a try ?

Use a 1MΩ pot, wired exactly as you first drew your schematic. The pot will appear to the input signal exactly the same as the typical 1MΩ grid reference resistor, and will be no extra load at all.

how so? the 1M fixed R is in || with pot + grid leak. so less than 500K load. leave out the 1M fixed and it's identical to original ckt, no?


--pete

[kagliostro]

Thanks for the advice HotBluePlates


The 1M resistor connected into the switch of the input jack in reality isn't connected there in the original F2B circuit


it is connected between the 68k resistor and the grid of V1a


so I think HotBluePlate is right, to use a pot instead of a resistor in that position will change anything on the input circuit


now is to be considered if a switch that connects the grid of V2b to ground is to be preferred to the switch on the cathode


or if is better to have both


K

[PRR]

Switching the cathode will pop. {EDIT} *Violently*. 90V transient at the plate which has roughly 30V peak signal.

Switching grid to ground, but leaving the plate in the audio path, will load-down the one active tube and give much lower gain than the "original".

[kagliostro]

So the option will be (as from Sean_Thornton) to switch plate and cathode ? See attached schematic.

---

Without the switch of the plate don't you think that the 100k resistor on cathode of V1b will be enoug as to maintain the cathode cap charged in order to prevent pop ?

Of course if there is a swith at the plate the 100k resistor can't do the job for which it was intended

Franco

[PRR]

IMHO....

The only happy way to do this is:

Drive both grids
Connect each cathode to its own R-C network
Connect each plate to its own resistor
Take output from ONE plate for "original"
Short the plates together for "double".

If the two triodes are Matched and there is no load, there is no pop and no change of gain.

If the two triodes are Matched and there is a load, there is slight pop and slight change of gain.

If the two triodes are not Matched and there is a load, there is slightly more pop and slight change of gain.

The simulation below shows this. For "switch" I used time-switches. One closes at 4mSec and the other opens at 6mSec, marked on the waveform by arrows.

To approximate the effect of mis-match I edited one fake triode for mu of 90 instead of 100.

I know I have seen one of Doug's 12AX7s show 213V and 214V on plates, fantastically good matching. But reading voltage charts of other amps I have seen much worse matching in a single 12AX7.

So the scheme could be very critical about the 12AX7 you use in it. If the two sides are not well matched, you might not dare to switch unless you mute the amp (volume down, or standby). Maybe fine for headbanger party but not when playing the waltz while bride dances with father.

[PRR]

Your V2B stage is all wrong.

No plate resistor needed or wanted. 

Grid resistor is needed.

A typical grid resistor won't need 0.1uFd to drive it.

The 2.2K cathode resistor will clip for signals above a Volt or two.

10uFd is not needed for 10K output pot and likely loads. A non-electrolytic will give less trouble.

If you must use 12AX7 for an output device, it is hard to beat the Fender 5F6a's tone-driver amplifier. Steal both the voltage amp and the cathode follower. Note the unusual value of the volt-amp cathode resistor (820 Ohms); this is needed to get the plate down low so the cathode follower has some voltage to work in. You may bypass this if you wish, to get closer to the Alembic/Blackface 2nd stage gain.

The 12AX7 in this circuit (any practical circuit) only flows about 1.8mA, and can not put all of that to a load. A high guess would be 1.5mA. If you load with a 10K pot and a 10K mixer, 5K total load, that is 5K*1.5mA= 7 Volts peak or 5 Volts RMS. This "should" be ample; however distortion may be significant.

Whereas a EL84 can loaf at 10mA and easily put 15V into 5K, 5V into 1K, drive any "line input" you run into.

[kagliostro]

One of the first option to which I've think trying to plan this thing, is in the attached schematic

I was thinking to disconnect the input of V1b from input and short it to ground, at the same time the plate of V2b will be disconnected from V1a plate

one option to which I was thinking at that time was to short to ground (via a decoupling cap) the plate of V1b when V1b plate wasn't connected to V1a plate

---

About the V2b circuit, I'm a plagiarizer and that circuit is the send circuit of the Engl Fireball (as draw by Mr. Merlin)



such FXLoop was added with success to a 5e3 circuit (see schematic on this post http://el34world.com/Forum/index.php?topic=17360.0)

and I was thinking it will be OK also for the F2B output

Do you think that a circuit like this will be more suitable ? (see attached schematic)

---

I'm interested on the use of an el84 as output CF can you give some values for components ?

---

One other idea will be to use a paralleled double triode as CF (so no mod to the circuit as tu use an odd triode), what do you think about ?

Franco

[HotBluePlates]

how so? the 1M fixed R is in || with pot + grid leak. so less than 500K load. leave out the 1M fixed and it's identical to original ckt, no?

Sorry, I saw what I wanted to see... 

I ignored the fixed resistor from input jack to ground, because it's not needed; the 1MΩ pot would already be an adequate grid-reference. So what I really meant was delete the fixed resistor, and use only the pot.

At that point, the first triode sees only a non-changing 1MΩ to ground, as does the input jack. The second triode gets a varying percentage of the input jack voltage, according the the setting of the pot. And the resistance from wiper-to-ground is irrelevant to the 2nd triode operation, apart from varying the signal input level.

[kagliostro]

Any thoughts about this version ?

Thanks

Franco

[DummyLoad]

you don't need the plate switch relay - switch from cathode to cathode

--pete

[PRR]

I say use the plate switch, not the cathode switch.

Switching the idle plate to a large cap kills the audio there, but that audio does no harm. OTOH when you switch *to* the large cap thing get out of whack for some seconds.

Build it. Wire-in (well-insulated) switches. Try it.

[tubeswell]

Just to add more confusion


Switchable cascade:parallel V1 by tubeswell, on Flickr

[kagliostro]

OK I think that this thing must be wired and tried

But at the moment I'm in Toscana for some days ( just the time for a small taste of wine    )

So a little standby is on at the moment

Thanks my friends

K

[DummyLoad]

i have the standel spin on the breadboard, (see thread "good clean fun") that had an unused 1st stage triode. i did not wire in a pot - too many parts to move to insert one.

the perceived increase in volume is somewhat noticeable. IMO, forget the whistles and bells, don't switch anything. just build the F2B as is, unless you can stand the popping an clicking. 

if you must that triode, then switching the triode in/out as a paralleled gain stage may be of greater gain benefit, but that changes the preamp topology significantly that requires a mixing circuit that is no longer the alembic design. your decision.

see schematic attached for results.

--pete