pentode phase inverter?

[ALBATROS1234]

i saw a you tube video the other day by the guitologist, he reviews a danelectro built montgomery wards amp model 25JDR8424. i think he mentions its push pull and was designed to compete with fender deluxe tweeds. it is a sweet sounding amp which uses 2 x 6SJ7 and 2 X 6V6 with a 6X5 recti. i can only assume that one 6SJ7 is a preamp tube and the other is a phase inverter. i am sure this is possible as a split load albeit rare because i havent yet seen any phase inverters that were not  triodes and i have poured over many schematics in my short time fooling with fire bottles. i find this amp and concept extremely interesting but can not for the life of me find a schematic for this amp or any amp that uses the tube compliment to see what makes this tick. anyone with the schematic for this amp please post it or tell me where i can find it. or if anyone knows of an amp model that uses a pentode as split load pi i would love this info. thank you all for any and all responses. below is a link to the video for those interested.

[sluckey]

Not that amp but I bet it's very close. Also has tremolo...

https://el34world.com/charts/Schematics/files/Danelectro/Dano_special.pdf

[tubeswell]

See attached schematic for 6SJ7 pentode PI (courtesy of David Jones)

[tubeswell]

Not that amp but I bet it's very close. Also has tremolo...

https://el34world.com/charts/Schematics/files/Danelectro/Dano_special.pdf

Similar to Wards Airline GDR8513A (albeit sans tremolo)

[ALBATROS1234]

thanks fellas i just checked out the one sluckey posted and it looks to be the same animal with a tremolo. i supposed to eliminate the tremolo i would just ground the line from the voltage divider which connects the tremolo?

[ALBATROS1234]

interesting in the one sluckey posted it looks like the pentode is run as a triode and made into a pi and the 1st one tubewell posts looks like pi comes from plate and screen grid. didnt realize that would work

[sluckey]

Notice the unusual way the first tube gets it's screen voltage.

To remove tremolo look at this pic...

[PRR]

Any pentode can be a triode. Sometimes pentodes were a glut on the market and this made sense. (It was a real thing in broadcasting. 1620 was a selected 6SJ7,/6J7 and when triode-strapped was same-as 6J5 and a fine audio triode. Because broadcasters have to have deep spares of every tube, using 1620 in all small holes was thriftier.) So it is a simple triode cathodyne.

The "Angie" in post #2 is totally different. Cathode current normally splits 80:20 plate:screen. But this rig is funny bias. The long-tail sets a fairly constant cathode current. G3 is way below K. The screen is taken as a second "plate" output. Apparently in this bias G1 will steer the constant cathode current to current to plate or G2, giving opposite outputs. There are several related schemes. Most are not stellar performers, but may be adequate for a low-cost guitar-amp.

[bmccowan]

Here is another all 6SJ7 preamp - Epiphone EH-150. I think the schematic is questionable, but you can get the idea. BTW, I have one of those Dano Special 12s waiting in the wings. It's a complete basket case. Chassis is rustier than a Chevy Vega and the pressboard cabinet is composting nicely. Everytime I think about working on it, I get sidetracked by a more attractive project.

[jjasilli]

PRR, thanks for that explanation re the Angie in Reply 2.  I still don't get how the 2 PI outputs are out of phase.

[ALBATROS1234]

Yea sluckey I saw that, 2.2 meg plate resistor  thank gents especially prr for his usual technical explanation. And I do know you can set any pentiode as a triode by tying the other grids to the plate giving them the Same attraction as the plate to those wayward electrons. So a 6j5 is electrically the same as a 6sj7 triode strapped. I have a ton of 6j5 tubes perhaps I will just try and adapt the circuit to use a 6j5. If I am not mistaken a 6j5 is half of a 6sn7 too right? I was also confused regarding the one where screen and plate provide reverse phase to the power tubes.

[tubeswell]

I was also confused regarding the one where screen and plate provide reverse phase to the power tubes.

Basically works off the principle that in a pentode, screen current and plate current are in anti-phase (in order to achieve 'constant current'). In a normal pentode, under signal conditions, the screen is held at a constant voltage and when the plate current dips down (when the plate voltage swing peaks), the screen current zooms up  - and vice-versa (all other things being equal).


In the above PI example, the screen is supplied with an independent current source (through the screen load resistor). The screen is not bypassed in any way, and the tube current is supplied via the 'constant current source'* by way of the tail resistor. So when the plate current dips, there is an alternative pathway for the tube current, through the screen, so the screen current rises - relative to the plate current (and vice versa when the plate current falls). Having a separate load resistances for the plate and for the screen, results in a differential in plate and screen voltage swings. But as PRR noted, this is not a very efficient phase inverter. This is because of the totally different physical (and electrical) properties of the plate vs the screen electrodes. (Note the different load resistances: 220k for plate and 120k for screen). Therefore, this PI typology would tend to suit some types of pentodes (ones which have higher-screen-current-relative-to-plate-current - and more robust screen grids) better than others (but even then it wouldn't be 'perfect')

*Re: the tail, the 'substantial' (22k) tail resistor at the cathode mimics a constant current supply for the pentode (like the tail in an LTP does), and the bootstrapping of the grid load resistor in combination with the bypassing of the cathode (which prevents changes in cathode voltage), helps to maintain the constant current function.

[kagliostro]

Any pentode can be a triode. Sometimes pentodes were a glut on the market and this made sense.

Note the double pentode ell80, triode strapped, used as PI in this old italian guitar amp

Franco

[jjasilli]

Tubeswell, thanks; but I still don't get it; even though you must be right.  Maybe my brain is stuck stuck in triode mode.     IOW, my understanding of how tube stages are out of phase is this:  it happens at G1.  AC signal voltage wobbling on G1 induces a mirror image of itself onto the stream of electrons flowing from K to A.  So the waveform impressed upon the internal tube current flow is 180 degrees out of phase with the incoming signal's waveform on G1.  The phase of the internal tube current seems to be the same everywhere inside the tube, including at the plate and at the screen.  The fact that the intensity of current flow in the plate & screen varies inversely, does not seem to have anything to do with it's phase.  I.e., we can vary the intensity of AC voltage and/or AC current without varying its phase.  Hence my continued confusion.

[ALBATROS1234]

very interesting, thank you tubeswell. i appreciate the explanation. i like that amp so at some point i am going to try to build one, i really like 6SJ7 tone, and pentode preamps in general. i know they are supposed to be horribly microphonic but i havent had too much of that yet, i have built a couple 6SJ7 types, an ef86 type, a 7199 and a couple 6U8a types and i did find one of my 4 6U8a types was a little microphonic but the rest are great. they are all separate chassis not built into a combo so maybe thats why i havent had trouble.

[shooter]

The fact that the intensity of current flow in the plate & screen varies inversely

I'm gonna butcher this, sorry 

I'm tangled, like you JJ.  Here's "how I can sleep", AC happens because volts increase n decrease based on current flow.  this thing is flowing I 2 places, if current is going to the plate (sinewave forming going neg)because V is dropping, G2 gets less I, more V. (sine wave forming positive)because V is increasing.
HOW G1 gets to pick is where my thought process falls apart 

[Tony Bones]

The fact that the intensity of current flow in the plate & screen varies inversely

[...]
HOW G1 gets to pick is where my thought process falls apart 

Me too! Staring at plate/screen curves doesn't help. (I realize that all the curves are at constant screen voltage, which just confuses things.)

If I were just a tiny bit more motivated I'd breadboard this ...

[shooter]

a tiny bit more motivated

you know just like the rest of use, it WILL produce 2 180 out-of-phase signals and you still won't KNOW why 

there are just some mysteries in life that should stay a mystery 

[sluckey]

Guys, that circuit is unproven. Tubeswell dropped that circuit on the forum six years ago. A couple people breadboarded the crazy PI circuit and proved that the screen and plate are indeed in phase, and different amplitudes. Kinda useless as a phase inverter. Read about it here...

https://el34world.com/Forum/index.php?topic=15872.msg154041#msg154041

[DummyLoad]

actually, it does work. just wonky out of balance with a 6sj7. you need a high gm pentode to balance properly. i did make it work and posted a few o-scope screenshots. screen and plate are 180deg. out of phase.


--pete

[shooter]

the crazy PI circuit

Nice memory!
Me, I'm filing it with the this PI, which I know works, under heading "don't bother a tranny is way simpler" 

[ALBATROS1234]

thanks shooter now i am all duh over what you posted. i am just gonna try to build the original concept with a 6j5 split load or perhaps just put a 6sn7 and use half as a gain stage after a simple tone stack and the other half as the pi. i just love me a 6SJ7 in v1.

[bmccowan]

Hey Albatros, I too love a 6SJ7 in V1. But I am curious as to why you are not going to follow the Dano circuit as Sluckey posted, using two 6Sj7s? I was look forward to that as I am working on an old Stromberg Carlson Muzak PA with all Pentode, and I have a Dan Special in waiting. I'm away for the holiday now, but would be happy to confirm that Dano schematic with my amp  when I get home.

[brewdude]

I built an amp based on the early Gibson G40 that uses two pentodes in the preamp with a tilt control between them. However, it has a triode cathodyne PI.


It sounds great.

[ALBATROS1234]

I could and I may. I was just thinking since a triode strapped 6sj7 is the same as a 6j5 which is half of a 6sn7 theoretically if I put a 6j5 which I have a ton of in multiple different brands and have been wanting to make use of. And was also thinking I could just pop in the 6sn7 making one triode identical to the pi but adding a gain stage giving me a tad more balls. I also have a bunch if 6sn7 types not being used but I think I only have a couple decent 6sj7 types one is a gt and for sure nice sounding and the other is metal and fair at best. I was mainly thinking of making use of what's on hand. I do however have a pair of metal 6j7 types which are predecessor to 6sj7 except have the grid connecting to a post coming from the top of tube. So that's a possibility. I actually have already drawn plans for a 6sj7 6sn7 double 6v6 circuit but just haven't built it. And I happened on the amp above and like it.

[PRR]

Reading the Sulzer article, it seems to need *large* bias on G3, and give only a small zone of out-phase outputs.

https://www.americanradiohistory.com/Archive-Communications-Magazine/1940s/Communications-1948-08.pdf

Tangential essay by Joe: https://www.radiomuseum.org/forum/dual_control_pentodes.html

[tubeswell]

Guys, that circuit is unproven.

See Sultzer article in PRR's post. (I was in the middle of writing my mid-life MSc thesis back in 2013 when I posted that in the other thread. I must've been procrastinating hard to ponder that'n'all)

[sluckey]

I stand corrected. That Sultzer article was an interesting read. I don't recall ever running into this circuit or even being aware of it. Learned something this morning. It's a good day! 

[bmccowan]

I was mainly thinking of making use of what's on hand.

I get it. I should do more of that myself. The 6SJ7 was so commonly used in the old PAs I used to buy up for cheap, that I ended up with a few of them. BTW, I have not found them to be microphonic, but then I do not try to squeeze every bit of gain from them. Merlin has a good chapter on pentodes which has helped me. One interesting amp I built used a Matchless Clubman type preamp with a pentode second gain stage. The early Clubman used a 6SH7, I used a 6SJ7 fed by a first stage 6AU6. Cool amp, it was a rebuild of a Bogen chassis that had one 7 pin socket and the rest were Octals.
Whatever you go with, please post the progress as this has been an intresting thread.

[ALBATROS1234]

ok , i am confused about the first preamp 6SJ7 in the schematic sluckey posted. i dont see how the screen is getting current. the feed appears to go to the plate of this same tube thru a 2.2meg plate resistor. the screen seems to be fed from ground? if the .01 coupling cap was on the other side of the 2.2meg which seems to be a ground reference to the second 6SJ7 grid 1. i dont see a way for the screen to get juice????????? what am i missing?

[sluckey]

The cathode of cathodyne phase inverter V3 has a big voltage on it. Gets developed across that 220K cathode resistor. The screen of V1 is connected thru a 2.2M to the top side of that 220K cathode resistor. That's the source of the B+.

[ALBATROS1234]

ok so the cathode of the pi has voltage on it, thanks, i am used to seeing the plate and screen of pentode preamp tubes being supplied from the same node, i could see clearly that coupling cap made that impossible and seemed it wasnt getting juice. thanks steve

[jjasilli]

I'm still struggling with the Sultzer article on the PI function.  The best sense I can make of it is in here:
www.audiodesignguide.com/New2A3/ETF06TS.pdf

G3 is disabled(??), due to ground connection; even though Sultzer seems to say G3 needs a negative bias voltage!?!  Or, @ nominally -0- V, G3 has a "charge" relative to the other inner tube elements.  But then, why isn't G3 a short circuit for inner tube current to ground? (Maybe because it's somehow repelling that current???)


Anyway, the pentode tube is wired as a triode (as PRR seems to suggest).  Somehow G2 is functioning as another control grid, per the article I referenced above.  So G2 remains out-of-phase with G1;  but somehow G2 reverses the phase of the electron stream passing by it to the plate.  So the plate is now back in-phase with G1.  (The output voltages of the Plate & G2 are balanced by the value of the plate resistor.)

This raises the musical question: are all triode-strapped pentodes non-phase inverting? And if not, why not?

[tubeswell]

..., the pentode tube is wired as a triode ...

Not in the Sultzer PI. The screen is connected to HT through its own load resistance, and is AC coupled to the following output tube grid. This means the screen is free to act independently from the plate.

If it was triode-connected, the screen would be connected directly to the plate (with or without a screen grid stopper), but this is not the case in the Sultzer PI.

[PRR]

The most shocking thing about the Sulzer (no T!) is that "the plate moves wrong".

As G1-K bias gets less negative, we expect plate voltage to fall. Happens in "all" tubes.

Here, it rises!

Also note that G2 current in the middle of the design range is 6X *larger* than plate current (normally 4X smaller).

The key is the extreme G3 bias, nearly 30V -negative- of cathode. Normally, near cathode potential, it offers no great effect on cathode electrons (while turning-back secondary electrons from plate which have acquired a large positive charge). But this VERY negative bias means it has large effect on "all" electrons from the cathode.

The voltages and electrode spacings mean that many cathode electrons turn-back at G3 and are acquired by G2.

As cathode current rises, G2 voltage falls due to the 100k.

As G2 falls, it steals more current from the G3-shielded plate. Plate current falls to zero at G1-K = 0. Over the useful "inverter" range the sum of Ip and Ig2 is nearly constant.

This is probably very critical of exact internal construction. Another tube type, or even another "works like type 6AC7", may not work the same in this very unusual operatating condition.

[ALBATROS1234]

Interesting,  I am glad I posted on this amp, I am learning a lot, most of all there's many ways to skin a cat

[bmccowan]

Albatros,
Although this thread went in multiple directions, I had said that when I returned home, I'd confirm that Dano Special schematic with an amp I had awaiting repair. I started looking at the amp sketching things out, and then I looked in the cab and there was the original schematic! Doh! With a few minor insignificant cap and resistor values its the same schematic.
I plan to replace the caps, get it working and then see what else it needs.

[tubeswell]

... and then I looked in the cab and there was the original schematic! Doh! With a few minor insignificant cap and resistor values its the same schematic.

Does the schematic state the model number (out of interest)?

[shooter]

"the plate moves wrong

I'm still struggling
see what happens when you eat negative apples 
guessing this circuit does have to operate under the -R rules 

[bmccowan]

Does the schematic state the model number (out of interest)?

It only states "Special Model" and in my head the Church Lady is saying, "Now, isn't that special."

[sluckey]

Does the schematic state the model number (out of interest)?

It only states "Special Model" and in my head the Church Lady is saying, "Now, isn't that special."

Is the schematic different than this one...

     https://el34world.com/charts/Schematics/files/Danelectro/Dano_special.pdf

If so, would you scan it or take a hi rez pic and post it? I'll make sure it gets into Hoffman's schematic library. THX

[bmccowan]

The only difference is things like 68K V1 grid stoppers instead of 75K. Is that enough to bother posting it as a separate schematic? If so, I'd be happy to do so.
B

[sluckey]

I think so. THX

[tubeswell]

The only difference is things like 68K V1 grid stoppers instead of 75K. Is that enough to bother posting it as a separate schematic? If so, I'd be happy to do so.
B

It adds to the special pool of obsolete schematics knowledge, which is highly valued by internet helpers on this forum.

[bmccowan]

OK - I'll get 'er done.

[bmccowan]

Danelectro Special schematic

[DummyLoad]

i transposed the schematic posted in reply 45.

i believe the schematic linked to in reply 40 may have a couple of errors. around the G2 supply of V1 and bootstrap of the cathodyne.

i omitted the .05uF mains filter cap.

--pete

[sluckey]

Danelectro Special schematic

I notified Doug. It should appear in the library shortly.

[sluckey]

Danelectro Special schematic

It's been added to the library...


     https://el34world.com/charts/Schematics/files/Danelectro/Danelectro_Schematics.htm