My journey of building a tube amp (12AX7 and EL84 SE)

[zeh]

Hi there,

iam trying to build my first tube amp and iam having some issues.
if i only connect the power tube and plug in a test signal from my signal generator directly into the grid of the power tube, i hear output on the speaker, if i crank up the voltage to about 4-5V it starts distorting, i think this is normal behaviour since the powertube gets overdriven at this voltage.
if i only connect the input stage (1/2 12ax7, the unused pins of the 12ax7 are grounded and the heater is only using pin 9 and 5, so only the first half should be used) i can measure a clean amplified signal after my coupling capacitor.
The issue is, as soon as i connect the coupling cap output to the grid of the powertube and increase the supply voltage, at some point around 150V, the signal on the grid of the powertube starts 'ringing' with some high frequency and the output is kind of distorted and the output power is also not very high. i have tried to simulate the project as good as i can using LTSpice, but since in reality always everything is slightly different, the voltages will not 100% match, i guess this is due to the series resistance of the power transformer which differs in reality, i tried different values, also measured the DC resistance of the power transformers secondary.
any help will be much appreciated, since i never built a tube amp i dont know how to proceed now, maybe you can guide me into the right direction.

at the test point i can measure about 5V AC, so the input signal gets amplified correctly, but ONLY if i disconnect R3 from the grid of the output tube.

in my simulation actually every looks totally fine and its working correctly. i also tried exchanging input and output tube and its the same error..

thanks and kind regards
chris

 

[zeh]

usually in my simulation the power supply has a slightly higher voltage, so that the screen grid of the EL84 is about 250V and the anode is arround 300 or so. i changed the voltage in the simulation to get values that are more like in reality, in reality i measure more or less the voltages shown in the schematic... i dunno why it differs from reality and how to get a more accurate simulation. what do i have to do to make it more like in reality? simulate a real transformer and measure all values of the actual device like L and R and X and so on?

 

[zeh]

first of all, does the schematic itself make sense or is there anything wrong here? i did not pay too much attention on the excact values of all the components since i just wanna get it running first before i think about cut off frequencies and stuff like that..also i have only a limited amount of components available that match the specs, for example HV Caps.. i understand that the resistors and caps act like filters, but since iam not to picky about it at this stage, i think those values should be 'ok' for testing purposes? thanks for any help

 

[Khron]

What if you reverse the connections of the secondary of the output transformer? Might be a case of positive feedback...

Or actually, do that in simulation first (since it's easier) and see what happens there.

 

[zeh]

hi khron,
thanks for your reply. the NFB node is not connected at the moment.. i tried to switch the primary leads of the OT but this did not make a difference.. but, there is a difference then in the schematic, in reality, i have hooked up my speaker directly to the secondary of the OT, there is no connection to ground here, can this be the issue?

 

[zeh]

if i remove the ground node from the OT secondary, nothing really changes (in the simulation)

 

[zeh]

ah, you say 'positive' feedback, how can this happen? is this some kind of self inductance or something else related to the OT?

 

[zeh]

i added the ground connection at the secondary, did not make a difference. the thing is, if i have the circuit completely connected like in the schematic and only apply a small supply voltage of about 100V, it works. as soon as i start increasing the supply voltage, it starts to squeal and displays high frequency on the scope, also it sound totally distorted but not in any pleasing manner. i have a seperate supply for the heaters so i only increase and decrease the anode voltages..

 

[musipol]

What about your lead dress? Is the input to the triode grid anywhere close to the XFMR output? If so, this may be oscillating. May be helpful to include a picture of the circuit you have built.

 

[zeh]

ok, i will post some photos and maybe a video of whats happening. it could be totally due to all the cabling, since iam beginner and i tried to build also like a kind of test rig for trying out tube circuits.. i basically have a contact strip, kind of a old turret board and soldered all the tube outputs there and labeled all the pins, there on this 'patch bay' i have laid my resistors and stuff but i left lots of leads cause i might wanna reuse the components... i will post some pics to make it easier to understand...

 

[zeh]

 

[zeh]

this is not any final build so please ignore the burning marks and overall placement of wires and components, iam trying and learning, this is the first time iam trying to build a tube amplifier so please dont blame me. i know about safety and so on and i try to keep all the things in mind, basically i dont touch it at all while its powered up and i always discharge and whatch the power capacitor. iam very cautious. thanks for your help people

 

[zeh]

i just discovered that the OT also has a center tap on the secondary, is this supposed to be used for the screen grid?

 

[CJ]

No.
Check all your grounds with the ohm meter.

 

[zeh]

tomorrow i will rewire the whole thing, change the tube sockets so the preamp tube will not be located near the OT and will try to improve the overall cabling and overview and post some more pictures

 

[zeh]

its weird, i wanted to just get it running like in classroom experiment kind of manner and then improve it later.. i expected some hum and stuff, but this is not usable at all, like there is something really totally wrong.. however, thanks for you help so far, i will try tomorrow what i can achieve with rewiring and relocating things

 

[musipol]

Based on the pics, You might consider relocating the output xfmr away from the input jack and keeping wiring clear of the input triode circuitry.

 

[ruffrecords]

Just as an aside, I am impressed with how you get such neartlooking schematics out of LTspice with those nice thick lines reminiscent of old magazine articles. How do you do that?

Cheers

Ian

 

[zeh]

Hi Ian,
you just have to set the 'pen thickness' in the preferences of LTSpice, i also changed the colorscheme a bit..



Kind regards
Chris

 

[zeh]

Hi Guys,
thanks for your help, after rearranging tubes, moving heater cables out of the way, shortening cables, and most of all, completely redoing all connections did the trick. basically i removed all components and started from the beginning. i can hear my guitar coming out of the speaker now without any hum or other unpleasent sounds. i did not install the cathode bypass caps yet, but everthing else is the same like before. yaaay, it works