Which phase inverter for a deluxe

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I would use this one:
http://www.webphix.com/schematic heaven/www.schematicheaven.com/fenderamps/deluxe_5e3_schem.pdf

It has a Cathodyne Phase Splitter.

It is auto bias (cathode bias) so there is no bias adjustment. If you want to be able to use unmatched output tubes then:
- Rather than the common 250 Ohm cathode bias resistor on the 6V6es with a 25uF bypass capacitor, separate the cathodes and put a 470 Ohm 5 Watt resistor with a 47uF bypass capacitor on each 6V6.

If you expect to overdrive it I would change the 1500 Ohm grid stop resitors on the 6V6s to 22 KOhms.

My own (home built) guitar amp is like this except I use 6SL7 octal for the phase splitter.

Cheers,
Ian
 
Ian thanks! I was leaning towards the Cathodyne (5E3), but being only my 2nd build (beyond a 5F1) I'm needing to learn lots about PIs and biasing PPs. The non-PI half of the 12AX7 is a gain stage not in other deluxe designs? Does it actually increase overall gain or just compensate for the Cathodyne's lower gain (relative to other PIs)? And thanks, I will separate the common cathode biasing (seen it on early Fender preamp designs, and will utilize it), but then would you adjust the 470 ohm resistors, to dial in the plate dissipation? And are factory bias adjust pots rated at 5W? I know some only "balance" the bias (not really an adjustment), but they seem to be normal (1/4W). Jeff
 
Gain of the Cathodyne is 1 (actually about 0.95) so the Common cathode amplifier (gain stage) preceeding it in the power amp is used to get the sensitivity up such that a signal level of about 1V is enought to drive the power section to full output.

The suggested 470 Ohm cathode bias resistors will bias the 6V6 tubes very conservatively. You could take these down to 390 Ohms if you want to bias hotter or try say 2 off 820 Ohm 2 Watts in parallel (420 Ohms) for something in between. Biasing hotter gives a richer harmonic content but tubes don't last as long. Less than 390 Ohms per tube is a BAD idea.

If you are a real "shreader" then you could think about using a design with a LTP phase splitter. It responds best to metal head style playing, else the Cathodyne is perfectly adequate.

Cheers,
Ian
 
Sounds good! And since I'm wanting to experiment with PIs, I wonder how difficult it would be to have a separate eyelet board for the PI, so I could build one Cathodyne and one LTP, and swap them out to compare. Or.... install them both... and find a 6PDT switch.....
 
I see no reason that you could not do that.

I recall a post on one of the forums regarding a switchable Phase Inverter, hang on while I start a second session and do a search......
Ah! here it is:
http://www.diyaudio.com/forums/tubes-valves/207178-switchable-pi-paraphase-cathodyne-ltp.html

Cheers,
Ian
Way too complicated. A four pole switch is plenty. Cathodyne to LTP. Preamp switching changes from a 5E3 to semi 18 Watt config.

18Watt5E3simple01JPG_zps984012b4.jpg


Mind you it takes some liberties with the standard Fender values. Problem with doing something of this sort is that you have to be careful you do not get ultrasonic oscillations. Layout is important, not really for a beginner. Have schemed up a few other PI switching circuits, will have to try them all some day.
 
Thanks P2 for the update of your switched PI.

@Latubegeek
My early posts show signs of being done late afternoon with the brain in lurking till go home time mode.
When you put separate cathode bias R and C on each tube then you need to double the resistance (develope the same bias voltage from 1/2 the current) and to keep the same frequency response you need to halve the capacitance.
That means you replace a single 250 Ohm resistor with a 25uF bypass cap with a 500 Ohm resitor and a 12.5uF bypass cap on each tube.

470 Ohms and 22uF on each tube will be fine.

Cheers,
Ian
 
One more thing about the concertina PI; If you do hope to overdrive the 6V6's you'll need as high a voltage as possible on the PI to ensure it has enough headroom to not overdrive itself if the incoming signal is high. In my PPP 6V6 amp I've got the concertina PI running at the same level as the screens, filtered through a 7H choke (probably doesn't need to be that large, but it's a choke I had handy).

This is a good read.
 
STARTED MY HYBRID 5E3 DELUXE!

Thanks Dave! I wanna get started on something, so decided on the 5E3 Cathodyne. I am going to drop the 2nd pair of inputs, use both 1/2s of the preamp 12AX7 on one input (and the cathodyne PI gives me a 3rd stage of gain!), but use a single tone pot (ala AA764) to keep it simple. Attached is the schematic and the eyelet board on my breadboard. So far so good. Now to build the PI eyelet board and put them both together!
 

Attachments

  • 3-STG GAIN DESIGN.jpg
    3-STG GAIN DESIGN.jpg
    661.6 KB · Views: 168
  • IMAG1412.jpg
    IMAG1412.jpg
    788.4 KB · Views: 164
Must to add a grid (leak) resistor and perhaps a (grid) stopper resistor to the 3rd stage and PI.

Since you don’t have signal loss in the tone section, the 3rd stage may clip before the PI, unless some attenuation is placed before the 3rd stage. Although non “Fender” cathode resistor values might avoid that.

If the 3rd stage clips before the PI clips the output tubes will never reach clipping nor full output potential. All of this (preamp distortion) could be an intentional outcome if so desired. Using a pot instead of the grid leak in front of stage 3 could be a way to "dial in" the attenuation value needed for whatever result you want on this. Then you could replace the pot with resistors. Marshall uses that attenuation approach on some amps.

A “master volume” could be used on the input to the PI to keep speaker volume down while allowing the 3rd stage to clip via increasing the “normal” volume control.

Slamming that type of PI hard is also known to be an issue and is mitigated with a very high value grid stopper on it’s input.
 
Thanks Dave! I wanna get started on something, so decided on the 5E3 Cathodyne. I am going to drop the 2nd pair of inputs, use both 1/2s of the preamp 12AX7 on one input (and the cathodyne PI gives me a 3rd stage of gain!), but use a single tone pot (ala AA764) to keep it simple. Attached is the schematic and the eyelet board on my breadboard. So far so good. Now to build the PI eyelet board and put them both together!

Don't forget your anode and cathode bias resistors on the cathodyne (not shown on your schematic). Also, you'll want 22uF decoupling at each tube. I would also recommend at least 1K between each B+ (so with the 22uF and 1K's you'll have Pi filters along your high voltage power rail) to isolate each stage and keep it nice and clean. Read this for more details on power supply design. Getting the power supply right the first time will save you a lot of annoyance later. It's not as critical with a low gain amp such as this, but it will still help the amp be noise free.
 
Must to add a grid (leak) resistor and perhaps a (grid) stopper resistor to the 3rd stage and PI.

Since you don’t have signal loss in the tone section, the 3rd stage may clip before the PI, unless some attenuation is placed before the 3rd stage. Although non “Fender” cathode resistor values might avoid that.

+1

Definitely 1M to ground on the grid of each stage, plus at least 100K grid stoppers between each stage. You can experiment with grid stoppers for subtle variance in tone once you're amp is built.
 
5E3 BUILD UPDATE

Doozer - My hastily-drawn schematic was more for showing the preamp mods, and I plan to use the 5E3 for the PI and PP sections.... but I will probably turn the 1M grid leak resistor to the PI grid into a pot, to control the 3 stages of gain to it. Will give me a leak and a stopper in one pot? And keep the 0.1mfd de-coupling caps (per 5E3). Why 22ufd? And increase 1500R stoppers if needed on the 6V6s (suggestions from other posters THANKS ALL!). Am slowly assembling it on breadboard (see pic), with separate eyelet bd for PI, and more submodules for PI and power tubes. Tested preamp section already, works fine.
 

Attachments

  • IMAG1421.jpg
    IMAG1421.jpg
    714.8 KB · Views: 77
Doozer - My hastily-drawn schematic was more for showing the preamp mods, and I plan to use the 5E3 for the PI and PP sections.... but I will probably turn the 1M grid leak resistor to the PI grid into a pot, to control the 3 stages of gain to it. Will give me a leak and a stopper in one pot? And keep the 0.1mfd de-coupling caps (per 5E3). Why 22ufd? And increase 1500R stoppers if needed on the 6V6s (suggestions from other posters THANKS ALL!). Am slowly assembling it on breadboard (see pic), with separate eyelet bd for PI, and more submodules for PI and power tubes. Tested preamp section already, works fine.

When the pot is at full gain you will have no grid stopper. I have the same setup on my current build (master volume control) and haven't had any problems, but I think I'll add a fixed grid stopper anyway.

I meant power supply decoupling, not inter-stage signal decoupling.
 
OK, you caught me trying to short cut using one pot for both grid leak and stopper.....:) Might go with a fixed 500K leak and 500K pot for vol/stopper? Will keep input impedance up, but still have full volume control (and who knows what will happen when stopper goes to zero?). And when you say "power supply decoupling", you mean the filter caps? I am using 22ufd instead of the 5E3 values (16ufd), and expect to play with the power rail dropping resistors, to adjust the circuit plate (and screen) loads.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.