Hello.
I built a nice 5W guitar amp, and I'm checking how the stages are working with a pure sine input and oscilloscope.
It has a two stage 12ax7 based preamp.
The next oscilloscope capture shows input (yellow), first pream output (blue) and second preamp output (magenta). All of them AC coupled.
View attachment n5x_input_v1a_v1b_outputs.BMP
Look at the "clipped" bottom of the magenta signal.
The grid is not able to approach cathode voltage, and stays 0.5V appart. The next capture shows it (yellow: grid, green: cathode).
View attachment n5x_v1b_grid_cathode.BMP
I think this is causing the asymmetrical output.
The spice simulation does not behave this way. The grid-cathode voltage can reach -20 mV, where I suppose grid current starts to flow.
The simplified schematic and simulation:
This is my second tube project, and the first time I'm really trying to understand all steps in detail, so I'm not very experienced yet. But I would say the 12ax7 should reach Vgk = 0. Could there be something wrong in my circuit?
Thanks for your help and support!
I built a nice 5W guitar amp, and I'm checking how the stages are working with a pure sine input and oscilloscope.
It has a two stage 12ax7 based preamp.
The next oscilloscope capture shows input (yellow), first pream output (blue) and second preamp output (magenta). All of them AC coupled.
View attachment n5x_input_v1a_v1b_outputs.BMP
Look at the "clipped" bottom of the magenta signal.
The grid is not able to approach cathode voltage, and stays 0.5V appart. The next capture shows it (yellow: grid, green: cathode).
View attachment n5x_v1b_grid_cathode.BMP
I think this is causing the asymmetrical output.
The spice simulation does not behave this way. The grid-cathode voltage can reach -20 mV, where I suppose grid current starts to flow.
The simplified schematic and simulation:
This is my second tube project, and the first time I'm really trying to understand all steps in detail, so I'm not very experienced yet. But I would say the 12ax7 should reach Vgk = 0. Could there be something wrong in my circuit?
Thanks for your help and support!
Last edited:
Hi
In a 12AX7, grid current starts to flow at Vgk=-0.5 V.
As you have a voltage divider in the input, it reduces the sine voltage applied to the grid as current flows.
Simulation of grid current in tube models often doesn't work when grid voltage is negative. So simulation likely not correct.
Can you post the listing of your 12AX7 model ?
Jacques
In a 12AX7, grid current starts to flow at Vgk=-0.5 V.
As you have a voltage divider in the input, it reduces the sine voltage applied to the grid as current flows.
Simulation of grid current in tube models often doesn't work when grid voltage is negative. So simulation likely not correct.
Can you post the listing of your 12AX7 model ?
Jacques
Ok, I see. I didn't know about that.
The model I use is this:
**********************************************************************
* GENERIC: 12AX7 / ECC83
* MODEL: NH12AX7
* NOTES: No heater model
**********************************************************************
.SUBCKT NH12AX7 A G K
XV1 A G K TRIODENH
+PARAMS: LIP= 1.5 LIF= 0.000016 RAF= 0.076498 RAS= 1 CDO=-0.53056
+ RAP= 0.18 ERP= 1.5
+ MU0= 87.302 MUR=-0.013621 EMC= 0.00000111
+ GCO=-0.2 GCF= 0.00001
+ CGA=3.90E-12 CGK=2.40E-12 CAK=7.00E-13
.ENDS
.SUBCKT TRIODENH A G K
+PARAMS: LIP=1 LIF=3.7E-3 RAF=18E-3 RAS=1 CDO=0 RAP=4E-3
+ ERP=1.5
+ MU0=17.3 MUR=19E-3 EMC=9.6E-6 GCO=0 GCF=213E-6
+ CGA=3.9p CGK=2.4p CAK=0.7p
************************************************************************
*
* Anode/grid model
*
* Models reduction in mu at large negative grid voltages
* Models change in Ra with negative grid voltages
* Models limit in Ia with high +Vg and low Va
*
************************************************************************
Elim LI 0 VALUE {PWR(LIMIT(V(A,K),0,1E6),{LIP})*{LIF}}
Egg GG 0 VALUE {V(G,K)-{CDO}}
Erpf RP 0 VALUE {1-PWR(LIMIT(-V(GG)*{RAF},0,0.999),{RAS})+LIMIT(V(GG),0,1E6)*{RAP}}
Egr GR 0 VALUE {LIMIT(V(GG),0,1E6)+LIMIT((V(GG))*(1+V(GG)*{MUR}),0,-1E6)}
Eem EM 0 VALUE {LIMIT(V(A,K)+V(GR)*{MU0},0,1E6)}
Eep EP 0 VALUE {PWR(V(EM),ERP)*{EMC}*V(RP)}
Eel EL 0 VALUE {LIMIT(V(EP),0,V(LI))}
Eld LD 0 VALUE {LIMIT(V(EP)-V(LI),0,1E6)}
Ga A K VALUE {V(EL)}
************************************************************************
*
* Grid current model
*
* Models grid current, along with rise in grid current at low Va
*
************************************************************************
Egf GF 0 VALUE {PWR(LIMIT(V(G,K)-{GCO},0,1E6),1.5)*{GCF}}
Gg G K VALUE {(V(GF)+V(LD))}
*
* Capacitances and anti-float resistors
*
CM1 G K {CGK}
CM2 A G {CGA}
CM3 A K {CAK}
RF1 A 0 1000MEG
RF2 G 0 1000MEG
RF3 K 0 1000MEG
.ENDS
The model I use is this:
**********************************************************************
* GENERIC: 12AX7 / ECC83
* MODEL: NH12AX7
* NOTES: No heater model
**********************************************************************
.SUBCKT NH12AX7 A G K
XV1 A G K TRIODENH
+PARAMS: LIP= 1.5 LIF= 0.000016 RAF= 0.076498 RAS= 1 CDO=-0.53056
+ RAP= 0.18 ERP= 1.5
+ MU0= 87.302 MUR=-0.013621 EMC= 0.00000111
+ GCO=-0.2 GCF= 0.00001
+ CGA=3.90E-12 CGK=2.40E-12 CAK=7.00E-13
.ENDS
.SUBCKT TRIODENH A G K
+PARAMS: LIP=1 LIF=3.7E-3 RAF=18E-3 RAS=1 CDO=0 RAP=4E-3
+ ERP=1.5
+ MU0=17.3 MUR=19E-3 EMC=9.6E-6 GCO=0 GCF=213E-6
+ CGA=3.9p CGK=2.4p CAK=0.7p
************************************************************************
*
* Anode/grid model
*
* Models reduction in mu at large negative grid voltages
* Models change in Ra with negative grid voltages
* Models limit in Ia with high +Vg and low Va
*
************************************************************************
Elim LI 0 VALUE {PWR(LIMIT(V(A,K),0,1E6),{LIP})*{LIF}}
Egg GG 0 VALUE {V(G,K)-{CDO}}
Erpf RP 0 VALUE {1-PWR(LIMIT(-V(GG)*{RAF},0,0.999),{RAS})+LIMIT(V(GG),0,1E6)*{RAP}}
Egr GR 0 VALUE {LIMIT(V(GG),0,1E6)+LIMIT((V(GG))*(1+V(GG)*{MUR}),0,-1E6)}
Eem EM 0 VALUE {LIMIT(V(A,K)+V(GR)*{MU0},0,1E6)}
Eep EP 0 VALUE {PWR(V(EM),ERP)*{EMC}*V(RP)}
Eel EL 0 VALUE {LIMIT(V(EP),0,V(LI))}
Eld LD 0 VALUE {LIMIT(V(EP)-V(LI),0,1E6)}
Ga A K VALUE {V(EL)}
************************************************************************
*
* Grid current model
*
* Models grid current, along with rise in grid current at low Va
*
************************************************************************
Egf GF 0 VALUE {PWR(LIMIT(V(G,K)-{GCO},0,1E6),1.5)*{GCF}}
Gg G K VALUE {(V(GF)+V(LD))}
*
* Capacitances and anti-float resistors
*
CM1 G K {CGK}
CM2 A G {CGA}
CM3 A K {CAK}
RF1 A 0 1000MEG
RF2 G 0 1000MEG
RF3 K 0 1000MEG
.ENDS
Thanks Elerion
In the model you use, grid current starts when Vgk > -0.2 V.
I suggest you try others values more realistics in my opinion :
In the model, you have 2 parameters describing grid current :
GCF change from 0.00001 to 0.000003
GCO change from -0.2 to -0.5
Furthermore, if you want to avoid this effect in you real circuit, remove the voltage divider between first and second stage.
Jacques
In the model you use, grid current starts when Vgk > -0.2 V.
I suggest you try others values more realistics in my opinion :
In the model, you have 2 parameters describing grid current :
GCF change from 0.00001 to 0.000003
GCO change from -0.2 to -0.5
Furthermore, if you want to avoid this effect in you real circuit, remove the voltage divider between first and second stage.
Jacques
V(v1b_a) swings 157V to 72V, 82V peak-peak.
The 6BQ5 is probably running 8V at cathode, bypassed, so can't take a lot over 8V peak (16V p-p).
The "asymmetrical output" at 2nd plate happens only LONG LONG after the 6BQ5 has gone totally square-wave. It is in 5X overdrive before the 2nd stage gets bent.
So not a problem for Clean.
And generally "asymmetrical" is not a big problem in guitar amps. It adds "flavor" to the dry sound of a naked steel string.
I agree the Vgk is interesting, but that IS a high impedance at Grid, and 12AX7's tight-strung grid flows some grid current before it gets positive. And different in different brands.... Neumann and Irving's Guitar Amplifier Overdrive has some data. However I do not think your 2nd stage is the least bit stressed when the 6BQ5 is banging itself rail-to-rail.
The 6BQ5 is probably running 8V at cathode, bypassed, so can't take a lot over 8V peak (16V p-p).
The "asymmetrical output" at 2nd plate happens only LONG LONG after the 6BQ5 has gone totally square-wave. It is in 5X overdrive before the 2nd stage gets bent.
So not a problem for Clean.
And generally "asymmetrical" is not a big problem in guitar amps. It adds "flavor" to the dry sound of a naked steel string.
I agree the Vgk is interesting, but that IS a high impedance at Grid, and 12AX7's tight-strung grid flows some grid current before it gets positive. And different in different brands.... Neumann and Irving's Guitar Amplifier Overdrive has some data. However I do not think your 2nd stage is the least bit stressed when the 6BQ5 is banging itself rail-to-rail.
I tried. But the same happends. Just the cathode signal shape changes a little bit, but the 0.5V gap is still there. This is not bypassing the cathode resistor.
With voltage divider:
View attachment nocap_nby.BMP
Without voltage divider:
View attachment nocap.BMP
When using a bypass capcitor, the grid-cathode voltage does reach 0V, and when removing the the voltage divider, the voltage even gets a little bit positive.
I don´t simulate but have been building tube amps since 1969.
The waveform you show is exactly how a triode behaves when driven hard, and, as mentioned by PRR, an important part of the "flavour" which makes tube amps preferred by most Guitar players.
If you are annoyed by it, just go Solid State
An important factor ignored by most (as in 99.98% of users) is that each and every tube stage has a diode connected across the input (the grid to cathode diode), which both clips input signal assimetrically , and rectifies Audio creating DC which strongly changes biasing, operation point, gain, etc.
The waveform you show is exactly how a triode behaves when driven hard, and, as mentioned by PRR, an important part of the "flavour" which makes tube amps preferred by most Guitar players.
If you are annoyed by it, just go Solid State
An important factor ignored by most (as in 99.98% of users) is that each and every tube stage has a diode connected across the input (the grid to cathode diode), which both clips input signal assimetrically , and rectifies Audio creating DC which strongly changes biasing, operation point, gain, etc.
If you are annoyed by it, just go Solid State
No, I'm not. I like both.
I'm just trying to get a deeper knowledge of how tubes work in real circuits.
Don´t take it literally, that is just a generic transformer model pulled from the library, very much doubt a cheap SE transformer has that high primary inductance, specially considering the huge DC current component through it, but, hey, they had to write *some* inductance value or simulator won´t run he he
In any case it will give you an unreal low frequency extended simulated response, but at, say, 1 kHz results should be reasonably accurate.
In any case it will give you an unreal low frequency extended simulated response, but at, say, 1 kHz results should be reasonably accurate.
The topic is more relevant for guitar amps, the thread moved to the Instrument & Amps forum.Download this:
http://www.valvewizard.co.uk/Common_Gain_Stage.pdf
A few pages down toward the end see "Grid Current Clipping" , its is right after the "Cut Off Clipping" section (Sections 1.13 and 1.14).
Cheers,
Ian
http://www.valvewizard.co.uk/Common_Gain_Stage.pdf
A few pages down toward the end see "Grid Current Clipping" , its is right after the "Cut Off Clipping" section (Sections 1.13 and 1.14).
Cheers,
Ian
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.