Help needed, please! Schematic review before build.

Discussion in 'Amp Tech Center' started by Ibanezgrind, Mar 26, 2020 at 4:15 PM.

  1. Ibanezgrind

    Ibanezgrind TDPRI Member

    Posts:
    6
    Joined:
    Nov 25, 2010
    Location:
    Tucson, Arizona
    Greetings!

    I have this same post on another fourm but I figured to try it here as well and get as much eyes on this before I get my component board built. Anyways, I built a 6L6WGB princeton reverb recently and sounds great! I recently tried to add a presence control to it but gave that idea up. I wanted to try something new and this is what I came up with.

    My idea is combining a mod I saw on Rob Robinette's 5E3 mod page that cascades one gain stage into the other. The channels are then separated by 2 mixing 220K resistors.

    What I am not sure of:

    Am I correct to place the mixing resistors in there? If so, are they in the correct spots?

    I currently have a Master Volume (represented in the schematic) that works great in my Princeton circuit . Now this cascade mod included a Master Volume for it's channel which is really the bright volume knob on a 5E3. The normal volume knob is now the gain level knob. I kept everything the same in regards to the schematic on his page but have no idea if having 2 master volumes, one for each channel is ok. Maybe one would be more simple but have no idea how that would look in my schematic. What would you guys suggest?

    I know too much overall amp gain can be a factor but I think what I have is ok since this proven mod is for a 5E3. However, my concern is for the V5A 12AX7 half in my schematic. I have the cathode tied to a 10K resistor for cold clipping. Not sure about that one either but thew it in. Maybe too much? Put far less resistance here? Maybe bypassed with a cap? I don't know..

    I'm pretty sure this schematic will work but I'd like to be more confident. If you see anything at all aside from my questions, please let me know. Feel free to make edits on the attachment if that is easier and apprecaite everyones knowledge. Thank you!

    Miguel
     

    Attached Files:

  2. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,287
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    Mixing resistors look ok
    That cold biased 5A seems a bit extreem...but if you like the sound...What are your voltages expected to be?
    You have all preamp stages marked as coming from the same supply node D, and none from B and C , not good, but i'm sure it's just a drawing mistake. also the 5E3 input jack is lacking a ground.
     
  3. Lowerleftcoast

    Lowerleftcoast Tele-Holic

    Posts:
    516
    Joined:
    Dec 30, 2019
    Location:
    california
    Concerning the 220k mixing resistors. One area I question is the 470k grid stopper on the PI. For that value of grid stopper the 220k mixing resistors might allow too much guitar signal to be bled off, but it wouldn't bleed off the lows. Using the 5e3 input(s) would be further impacted by the position of the "Princeton side" master volume.
     
  4. Ibanezgrind

    Ibanezgrind TDPRI Member

    Posts:
    6
    Joined:
    Nov 25, 2010
    Location:
    Tucson, Arizona
    Thanks for the reply, yes I agree it is extreme and will consider adjusting that for lower resistance. Maybe a more common value of 2.7K but it couldn't hurt to try 10K! Currently at all D power nodes, if I remember correctly is around 280V. I do have node B on pins 4 of the power tubes and also on the reverb transformer at the moment. Yes, node C isn't being used but thinking of moving node C to the phase inverter - especially after reading your wonderfully informing thread on the Cathodyne PI (thank you for the information btw!)

    The ground missing on the 5E3 jack is intentional according to the mod. Please check it out for more info - https://robrobinette.com/5e3_Modifications.htm#High_Gain_Channel
     
    Bendyha likes this.
  5. Ibanezgrind

    Ibanezgrind TDPRI Member

    Posts:
    6
    Joined:
    Nov 25, 2010
    Location:
    Tucson, Arizona
    Thanks for the reply! If this is the case upon hearing it for the first time, would you recommend a different value of grid stopper to pin 7?

    If I understand correctly, are you saying that using the 5E3 inputs will still interact with the Princeton side master volume despite the mixing resistors being there? Yeah, I don't want them impacting one another, no bueno! If this is the case, maybe eliminate that master volume altogether or place it elsewhere? Just leave the master turned up all the way?
     
  6. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,287
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    The the 5E3 input socket will not work if there is no ground connection. It is correct that a switching socket that self grounds when the jack is removed should not be used.
    But I see that you are right in that robrob disconected the ground completely in the schematic you are refering to. Why this is so he will hopefully relate.
     
    Last edited: Mar 27, 2020 at 4:31 PM
    Ibanezgrind likes this.
  7. Lowerleftcoast

    Lowerleftcoast Tele-Holic

    Posts:
    516
    Joined:
    Dec 30, 2019
    Location:
    california
    Instead of the 5e3 vol controls causing the effect it is the voltage divider of the 220k/470k/1m (acting as a fixed volume control.) causing the effect.

    What I envision is similar to how the 5e3 looses signal when it interacts with the other channel. The AC signal is allowed to form a RC filter through the coupling cap and the plate resistor to the B+3. The 5e3 has .1uF cap and 100k resistor which forms corner frequency of 15.9hz.

    For your design the Princeton input would bleed to the 5e3 side thru the .022uF cap and 100k resistor - corner frequency of 72.4hz. (So bass signal under 72hz remains.)

    The 5e3 input would bleed to the Princeton side thru two filters. The.022uF cap and the 1m pot corner fq would change with the position of the wiper. The next .022uF cap and the 100k resistor forms corner frequency of 72.4. (Signal hz would vary with pot position. With the MV on max it would be similar to it not being in the circuit.)

    IDK how prominent this effect would be. Upping the value of mixing resistors would help reduce the effect. Getting rid of the 470k grid stopper would eliminate most of the effect.

    I am pretty sure you included the grid stopper to get rid of blocking distortion common to the 5e3. (Some 5e3 builders claim replacing the .022uF with a .0022uF to .0033uF coupling cap gets rid of blocking distortion. I have not tried that yet.)
     
    Bendyha and Ibanezgrind like this.
  8. Ibanezgrind

    Ibanezgrind TDPRI Member

    Posts:
    6
    Joined:
    Nov 25, 2010
    Location:
    Tucson, Arizona

    I knew there was more to it but seems like there is a way around this. Trial and error time if it comes to that. I will keep your suggestions in mind, thank you! And yes you are correct, the 470K grid stopper resistor is there for that reason.
     
  9. Ten Over

    Ten Over Tele-Holic

    Posts:
    523
    Joined:
    May 13, 2015
    Location:
    Central California
    No matter how hard I try, I can't see a 220K/470K/1M voltage divider. From the point of view of the Princeton side (V4B), I see a second order high pass filter. Both capacitors are .022uF. Using 38K as the output impedance of V4B, the first resistor is 1038K. The second resistor couldn't possibly be any less than 220K in series with the parallel combination of 220K and 16.6M -- call the total second resistor 440K. Calculating for -6dB, I get 10.7Hz and converting that to -3dB, I get 16.6Hz. This is the worst case scenario where V5A and its coupling capacitor have zero impedance. Once you figure in the actual impedances at the V5A end, the corner frequency is even lower. So the 5E3 side has no effect on the audible frequencies of the Princeton side.
     
  10. Ten Over

    Ten Over Tele-Holic

    Posts:
    523
    Joined:
    May 13, 2015
    Location:
    Central California
    It's not a good idea to have a master volume within a negative feedback loop because the NFB changes as you change the MV.
     
  11. Lowerleftcoast

    Lowerleftcoast Tele-Holic

    Posts:
    516
    Joined:
    Dec 30, 2019
    Location:
    california
    @Ten Over I am no expert. I was considering the way the 5e3 circuit has losses when the volumes interact. From my understanding the 5e3 circuit looses signal through the .1uF coupling cap/100k plate resistor to the B+3 on the unused side.

    It appeared the signal would take a similar path in the proposed design. Am I totally off base?
     
  12. Ten Over

    Ten Over Tele-Holic

    Posts:
    523
    Joined:
    May 13, 2015
    Location:
    Central California
    I'm no expert, either.

    The output impedance of the typical Fender gain stage isn't 100K. It is 100K in parallel with the internal plate resistance. I decided long ago that the internal plate resistance at the typical Fender operating point is 62K, so the output impedance rounds off to 38K. This is the value you would use when doing frequency dependent calculations. A 0.1uF capacitor in series with a 38K resistor would have a corner frequency of 42Hz. This 42Hz is meaningless for the behavior of the amps under consideration because there are other components involved.

    I never paid any attention to the frequency dependent aspects of a 5E3 before, but it quickly became apparent that I could only calculate the corner frequencies with the Tone control removed and then only under two conditions: Both channels maxed and one channel maxed with the other off. However, I believe that the corner frequencies are always under 17.4Hz as far as interaction is concerned. For guitar frequencies, we could replace the 0.1uF capacitors with straight wires and then analyze the interaction of the two volume controls keeping in mind that each plate looks like a 38K resistor to ground.

    Now that I look at it some more, I can't calculate the corner frequency with both channels maxed, either, because the signal is being taken from the middle of a second order high pass filter. For some reason I feel that the corner frequency is lower at the middle than at the end of the filter and I am still confident that all corner frequencies are under 17.4Hz.
     
    Last edited: Mar 28, 2020 at 1:57 AM
  13. Lowerleftcoast

    Lowerleftcoast Tele-Holic

    Posts:
    516
    Joined:
    Dec 30, 2019
    Location:
    california
    This is thought provoking. I would like to follow along with the observations/calculations. What values are we looking at for this second order high pass filter? It may become obvious when I see the values, but currently I am lost with the statement "the corner frequency is lower at the middle than at the end of the filter".
     
  14. Ten Over

    Ten Over Tele-Holic

    Posts:
    523
    Joined:
    May 13, 2015
    Location:
    Central California
    Like this:

    5E3 Mixer.png

    The 5E3 mixed signal isn't taken from the second order filter output. The corner frequency at the second order filter output isn't the same as the corner frequency at the 5E3 mixed signal out. I think the corner frequency at the 5E3 mixed signal out is lower than the corner frequency at the second order filter out.

    Also, the 20K at the "in" side complicates things so that the equations for a simple second order high pass filter aren't readily applicable and I don't know how to alter the equivalent circuit to make them work. Still, I think the corner frequencies are going to be lower with that 20K in place as opposed to no 20K at the "In" side.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.