12AU7 parallel driver for SPring reverb

[seabstudio]

Hi everyone ;)
I m new here
Hope I can ask question on this forum related to my project

I built a 12AU7 parallele driver and follow the Merlin book consideration
But I have a doubt about witch circuit I need to use in front of this driver
Merlin talk about 100V p-p
How I can archieve this level with a 12AX7 ?

My reverb will have a transformer input ( like a reamp ) to allow Line balanced level in front of Preamp stage
At the output after Recovery for Reverb , I put a direct box ( 100K load ) and put this on my mixing desk
The goal is to have a nice reverb for reggae and Dub Production
Thanks in advance !
Seb

[sluckey]

You only need 12V_pp if you use the lower input.

[seabstudio]

You only need 12V_pp if you use the lower input.

I m not sure to understand, so i don t need to connect circuit in front of 470p capacitor
It will float ?

[terminalgs]

Merlin gave you C1 and that 1M to show you how you might connect it to any given circuit and attenuate a signal appropriately to feed the driver.  to adapt his driver (assuming you have the tank and B+ and everything right), you want  12VAC p-p and 0VDC on the grids.  Since you need 0VDC you will need *something* to isolate it from previous circuitry VDC.  hence the coupling cap.


the 1M+170K form a voltage divider and attenuate 100V to 12V.  Merlin gives you the low down on that divider when he's talking about a pentode driver:


"so the potential divider in the schematic has been chosen to suit a 100Vp-p guitar signal from the previous stage."


First step: figure out where in the circuit you will tap your audio signal and what voltage you have at that point.   If you have 36VAC p-p, you need 3:1 divider.  so maybe a 330K instead of the the 1M.  If the place you are "tapping" the signal from has VDC voltage, then select a coupling cap based on the low frequency roll-off that C1 and a 500Kohm (330K+170K) form.  Elsewhere in the chapter Merlin writes:


"at very low frequencies which we don't need- anything below about 200Hz is too muddy to be of use."


He isn't talking about the coupling cap specifically at that point, but it certainly applies. If you figure the low freq. roll off the 470pf & 1170K compute,  you get about 290Hz.  for the 330K+170K you might pick .002uf or .001uf.


On the reverb recovery end, you'll need to do some similar signal VAC p-p figuring.  Where you want to return the wet reverb signal to the circuit will determine how much gain the  recovery stage must provide and (if you use mix resistors) what ratio mix resistors will need to be to blend the wet with the dry.

[seabstudio]

Merlin gave you C1 and that 1M to show you how you might connect it to any given circuit and attenuate a signal appropriately to feed the driver.  to adapt his driver (assuming you have the tank and B+ and everything right), you want  12VAC p-p and 0VDC on the grids.  Since you need 0VDC you will need *something* to isolate it from previous circuitry VDC.  hence the coupling cap.


the 1M+170K form a voltage divider and attenuate 100V to 12V.  Merlin gives you the low down on that divider when he's talking about a pentode driver:


"so the potential divider in the schematic has been chosen to suit a 100Vp-p guitar signal from the previous stage."


First step: figure out where in the circuit you will tap your audio signal and what voltage you have at that point.   If you have 36VAC p-p, you need 3:1 divider.  so maybe a 330K instead of the the 1M.  If the place you are "tapping" the signal from has VDC voltage, then select a coupling cap based on the low frequency roll-off that C1 and a 500Kohm (330K+170K) form.  Elsewhere in the chapter Merlin writes:


"at very low frequencies which we don't need- anything below about 200Hz is too muddy to be of use."


He isn't talking about the coupling cap specifically at that point, but it certainly applies. If you figure the low freq. roll off the 470pf & 1170K compute,  you get about 290Hz.  for the 330K+170K you might pick .002uf or .001uf.


On the reverb recovery end, you'll need to do some similar signal VAC p-p figuring.  Where you want to return the wet reverb signal to the circuit will determine how much gain the  recovery stage must provide and (if you use mix resistors) what ratio mix resistors will need to be to blend the wet with the dry.

Many thanks for this great explanation , it help me a lot !
Ok now I understand more, we need 12Vpp at the grid , so need to adapt the voltage divider before
I admit i m a begineer ( but carefull with Voltage etc )

For preamp in front of unit  I use a simple 1/2 12AX7 stage
To calculate de VAC I just need to put a multimeter in AC mode in the tapping point ( and send a 1khz signal at 1V )
I put the stage I use before the driver
Or maybe I can wired a 1W pot ( 1 Mohms )as a variable resistor and turn it to hear the sweet point before clipping ?

[sluckey]

Or maybe I can wired a 1W pot ( 1 Mohms )as a variable resistor and turn it to hear the sweet point before clipping ?

This would be called a "DWELL" pot. I like to permanently mount the pot on the front panel for better control of the reverb.

[seabstudio]

Yes I know it for DWELL it s nice  ,but  i mean  the first resistor of voltage divider to fix the futur value of resistor  ( potentiometer just to test and find a rapid value and use ear to find the sweet stop value )

[seabstudio]

Merlin gave you C1 and that 1M to show you how you might connect it to any given circuit and attenuate a signal appropriately to feed the driver.  to adapt his driver (assuming you have the tank and B+ and everything right), you want  12VAC p-p and 0VDC on the grids.  Since you need 0VDC you will need *something* to isolate it from previous circuitry VDC.  hence the coupling cap.


the 1M+170K form a voltage divider and attenuate 100V to 12V.  Merlin gives you the low down on that divider when he's talking about a pentode driver:


"If you have 36VAC p-p, you need 3:1 divider.  so maybe a 330K instead of the the 1M.  If the place you are "tapping" the signal from has VDC voltage, then select a coupling cap based on the low frequency roll-off that C1 and a 500Kohm (330K+170K) form.

I try 330K instead of 1M , it s really better :) but i have little saturation again, so try to optimise it a little
But to have 36VAC p-p at the plate output ( with a capacitors to stop DC ) with a half of 12AX7 ( classic values arround ) what is the voltage output typicaly ? or it need to be cascaded ?
Thanks again :-)

[Willabe]

Take a look at this on 1 tube reverb. (I know you have a parallel driver.)

Jeff talks about the  difference in tone between using a 12AU7 and a 12AX7.

http://el34world.com/Forum/index.php?topic=24216.msg260710#msg260710

[seabstudio]

Take a look at this on 1 tube reverb. (I know you have a parallel driver.)

Jeff talks about the  difference in tone between using a 12AU7 and a 12AX7.

http://el34world.com/Forum/index.php?topic=24216.msg260710#msg260710

Thanks I read it , but I can t find the information I need

How to know the plate output ( in Vp-p ) of a 1/2 12AX7 or cascaded 12AX7 ?

[PRR]

> How to know the plate output

Depends what you put into it, and what you load it with.

You say Line from a Transformer, but "line" can be 0.2V or 2V. Transformer can be 1:1 or 1:10.

If the "line" is a strong source and you have 1:10 line transformer then that's all you need. If lines in your studio are lower level and you have 1:1 transformers, put a "Drive" pot then half a 12AX7 in front of Merlin's "12V" input. Adjust "Drive" for happy result.

[seabstudio]

> How to know the plate output

Depends what you put into it, and what you load it with.

You say Line from a Transformer, but "line" can be 0.2V or 2V. Transformer can be 1:1 or 1:10.

If the "line" is a strong source and you have 1:10 line transformer then that's all you need. If lines in your studio are lower level and you have 1:1 transformers, put a "Drive" pot then half a 12AX7 in front of Merlin's "12V" input. Adjust "Drive" for happy result.

Thanks for great info !
I have just a 1/1 transformer ( 600:600) Edcor , my mixing desk send +4dbu , so I think it s 1.25V RMS
I have to lower my aux send ( in middle ) to match just under saturation

When You say half 12AX7 ,Do I have to change the traditional coupling capacitor ( 0.022uF ) for the 500p directly to 12V input  ?( so I Remove the voltage divider too ? )

Thank

[seabstudio]

I just test now
The reverb can' t have a big punch if I compare to the recovery channel volume
When I touch the spring I have a full clean volume so the output is really good
But When I send signal to the driver tru the half of 12AX7 ( Aux send > transformer 1:1> to preamp input ) i have an heavy "peak transcient  " saturation, like it clip the grid of 12AU7 very fast

on plate output of preamp ( 12AX7 ) I remove the traditional 0.022uF  &  replace with  a 500pF capacitor follow with a 100K resistor before the grid ( 12V input in Merlin Shematics ) maybe i m wrong hear

[tubeswell]

What are your idle voltages on the 12AU7?


If you are getting grid current limiting (or cut-off), try adjusting the 12AU7 to be more centre-biased.


The recovery stage will sound 'strong' of you twitch the spring to tickle the pan's output transducer, but that is applying much more force to the spring than it would get in normal operation.


If you otherwise have a weak wet signal in the signal path, then you either need another gain stage after the recovery stage to boost the wet signal prior to blending into the main signal path, or you need to attenuate the dry signal more before you blend the wet signal back in.

[seabstudio]

What are your idle voltages on the 12AU7?


If you are getting grid current limiting (or cut-off), try adjusting the 12AU7 to be more centre-biased.


The recovery stage will sound 'strong' of you twitch the spring to tickle the pan's output transducer, but that is applying much more force to the spring than it would get in normal operation.


If you otherwise have a weak wet signal in the signal path, then you either need another gain stage after the recovery stage to boost the wet signal prior to blending into the main signal path, or you need to attenuate the dry signal more before you blend the wet signal back in.

Maybe is that .
On 12AU7 i have 270V before the  Ra 22K resistor, Merlin say it need 300V , but i don t think 30V will change the headroom too much ?

On other spring reverb , I have a good balance beetween signal send and recovery "thunder " hit . I mean is almost the same volume, that why it . surprise me on the design i try to build .Because Merlin say that 12AU7 have it s a perfect current source for high impedance tank

I use only WET signal , So I send Aux send tru Mixing desk to a Transformer >  first stage 1/2 12AX7 > 12V input in 12AU7 parallel DRIVER  > Recovery channel ( 1/2 12AX7 too ) > Output ( other 1/2 12AX7 ) then I plug the output to a good direct box to m mixing desk via mic preamp

[seabstudio]

Hi Everyone

I finish my build of the spring ,but i have still some questions
the positive things is now the reverb is dead quiet ,no hum , no buzz and recovery part give an excellent sound

yesterday I try to lower the resistor ( come from V1 half 12AX7 tube plate ) so I was used 10nF capacitor + 1K resistor before the 12AU7 parallel reverb drivers
But after some minute , when I push my aux from mixing desk , a strange noise appear ...and disapear few second , i think i reach the limit of the 12AU7 grid ...( shot effect ? )

My question is about this ...I really want to "smack " the reverb tank :)
If i put a cathode folower after the V1 half 12AX7 , the impedance will be very low ( maybe less than 1K ) if I refer to the Merlin book
actually the 12AX7 have a very high impedance to drive a 12AU7 no ?

My idea is to add simply a IRF820 at the end of V1 and send directly the signal to the reverb driver ( with a coupling caps of course ) , good idea ? or another tube preamp in front ?

Thanks in advance
sorry for my english , i m french :)

[tubeswell]

Got a schematic?

[seabstudio]

Got a schematic?

Yes its mostly the same as this one :
https://www.mzentertainment.com/pics/fisher_k_10/dr_zee_fisher_k10_spacexpander_spring_reverb_schematics.gif

Exept the drivers, i choose the 12AU7 to drive high impedance tank
 http://www.valvewizard.co.uk/reverbdriver.html

I change the psu to have 300V for the parallel 12AU7 anode , rest of tube work little harder ( the 12AX7 V1 in Fisher K10 shem  ) but not to much

Thanks

[seabstudio]

Hi

So I put a moset IRF820 after the V1 tube ( like a classic Cathode Follower for an EQ )
I think it help to lower impedance at the grid of Parallel 12AU7 reverb driver
I check the signal to be sure to have a full 12V input
So at input of jack i have 0.4V AC , it give me 35V at the end of cathode follower , so i put a voltage divider with a capcitors in front of 12AU7 , i have now beetween 12 - 15Vac at the grid
The 12AU7 is very "hot " compared to the 12AX7 in preamp and recovery section
I have a strange random noise sometime ( 1 or 2 second ) when I push maximum signal in front of 12AU7 ...:( maybe a bad solder or cold joint :( :(  but reverb sound really nice

[tubeswell]

hematics.gif

Exept the drivers, i choose the 12AU7 to drive high impedance tank
 http://www.valvewizard.co.uk/reverbdriver.html

So are you using a transformerless 12AU7 driver? (Did you remember to use a 22k plate resistor, for the load?)

[kagliostro]

Don't know if this can be of some interest, I've find this in my archive





and this (this may be not)



Franco

[seabstudio]

hematics.gif

Exept the drivers, i choose the 12AU7 to drive high impedance tank
 http://www.valvewizard.co.uk/reverbdriver.html

So are you using a transformerless 12AU7 driver? (Did you remember to use a 22k plate resistor, for the load?)

Yes , i follow Merlin recommandation , Ra and Ri both 22K

[seabstudio]

Don't know if this can be of some interest, I've find this in my archive





and this (this may be not)



Franco

Yes thanks !
When I tale a look on Magnatone , I see that the grid leak resistor is 1M , can it help to give more power at the output of driver ?
Because now I have the 12Vac and want more power ...I think Merlin calculate the "perfect " sweep spot  , but i want to drive the tank harder ?
it s possible ?

[kagliostro]

Reading Merlin

Obviously the drawback of this method is that we must use both triodes to obtain sufficient power since no OT is used

he seems to consider that to drive via a transformerless solution a reverb with a 12au7 is not a must

he say also that an improvement can be reach using an ecc99 tube as driver
(if you can get russian tube https://tubes-store.com/product_info.php?products_id=319)

--

I've some examples of transformerless reverb drivers and the most part use pentodes

--

Have you also experimented with the recovery section ? May be it will help

--

BTW, which are input and output impedances of your tank ?

Franco

[seabstudio]

Reading Merlin

Obviously the drawback of this method is that we must use both triodes to obtain sufficient power since no OT is used

he seems to consider that to drive via a transformerless solution a reverb with a 12au7 is not a must

he say also that an improvement can be reach using an ecc99 tube as driver
(if you can get russian tube https://tubes-store.com/product_info.php?products_id=319)

--

I've some examples of transformerless reverb drivers and the more use pentodes

--

Have you also experimented with the recovery section ? May be it will help

--

BTW, which are input and output impedances of your tank ?

Franco

Yes , thanks again
Initially i want to use ECC99 but my PT is an Hammond 290 WEX ...only 1.05 amp for heater, I have 3 tubes , so 900mA used , a ECC99 can t work
But i m not sure about the translation : " might be used on its own. " , I keep  the parallel design or use just a half of ECC99 ?

I use a 4FB tank ( Fisher Gibbs and TAD )
Recovery section of Fisher K10 sound really nice and ( i think ) the design is smart ( with the NFB ) ...if I touch the spring with finger , it sound really clean and punchy

[kagliostro]

About ecc99 I guess he mean both triodes in parallel

an unexplored way may be to use as driver an EL95 pentode (low heater consumption tube - 200mA), note, EL95 NOT EL90

http://www.r-type.org/exhib/aaa0406.htm

http://www.r-type.org/pdfs/el95.pdf

Franco


p.s: I mean an unexplored way related to the type of pentode, not that it cannot be done with a pentode, with a pentode it has already been done

After a search I've find this, it isn't a transformerless driver, it uses a transformer, but we can see that the EL95 tube was used as reverb driver





--

pentode transformerless reverb drivers example

https://el34world.com/charts/Schematics/files/Fbt/Fbt_500_r2.pdf

https://el34world.com/charts/Schematics/files/Fbt/Fbt_500_r.pdf

[seabstudio]

Yes ! really nice idea , need to test on the next spring reverb built
For now , change completly the driver will be long , so i wait the ECC99 and post the difference
But EL95 seem to be a nice option for a compact build
I imagine a preamp with fet and moset at input , then EL95 driver and another fet + mofet for the recovery . That way , just 200 ma needed for total heater , it can reduce the size and price for the power transformer
Maybe a 220>12 12<220 transformer like i see in lamington reverb will be a cool idea too
thanks for share again

[kagliostro]

....  i wait the ECC99

OK, but remember ECC99 require 800mA @ 6.3v and 400mA @ 12.6v for heaters

Franco

[seabstudio]

Yes , I change the transformer inside
Now I put a toroid 240V primary > 250V secondary 0.16mA
and 3A for the heater ! So it have space

[kagliostro]

3A is a good choice 

This give you the chance for other tubes to be added if you want

Franco


EDIT: May be this is interesting to be read with attention

http://dougcircuits.com/reverbtanks.html

[seabstudio]

Greetings

I receive the ECC99 tube
So after swap the power transformer ( Hammond 290WEX ) I put a toroid transformer with 3A
The ECC99 is really good , it s like it add a "step " to the output and drive the tank better
but i want more :) if possible
It s not for a guitar use but studio use
I put a reamp in front of this reverb , so 0.4Vac output in front of first Tube  ( because i don t know how to put a real balanced line input )


can you advice some tweaks arround this Tube ? It s a driver for reverb , so i think audiophile perfect bias isn t the solution
If I want more power, I need to change the cathode value ? actually it s 820 ohms , i have in my disposition a 680ohms 1W resistor
I just check some voltage with the ECC99
Anode voltage ( after 22K is 165V DC ) and before it it s 335V DC

Thanks

[seabstudio]

3A is a good choice 

This give you the chance for other tubes to be added if you want

Franco


EDIT: May be this is interesting to be read with attention

http://dougcircuits.com/reverbtanks.html

I read it now, thanks Franco...really technique for me , but i will try to understand :-)

[seabstudio]

Any ideas ?