Allen Classic 10 Conversion to Tubenit's 5E3/18w Hybrid

[edthill]

Hi All,

Just wanted to say thanks to Doug and everyone on this forum for making this the best place to learn and get help building tube amps. After a long break from making music and amp tweaking, I'm back. I recently used the resources from this forum and some of Doug's fine parts to convert a broken down Traynor YBA-2B into a solid state rectified 5E3 tone monster. I don't ever remember my previous 5E3 build sounding this good. The solid state power supply adds a nice tightness to the amp.
Anyway, I have an Allen Classic 10 head I put together from kit. Even though the amp is a single-ended, solid state rectified, tweed Princeton, the power transformer can easily support another solid-state rectified 5E3 build. I was doing some digging last night and came across Tubenit's 5E3/18w hybrid schematic. It would be perfect for the Classic 10 Chassis as I wouldn't have to do any drilling or hacking. My question would be regarding the layout. I'm still not savvy enough to convert schematics into layouts. I can read and comprehend a schematic for the most part, I just don't know how to fully visualize all the components in a layout diagram. I attached the layout from the Allen Classic 10 and Tubenit's 5E3/18w hybrid. Any and all help is beyond appreciated.

Thanks,

Ed

[sluckey]

Tubenit has a layout for that 5E3/18W Hybrid schematic. You may even be able to adapt the Allen board but you'll probably need an alternate way to mount your filter caps or use a cap can.

     http://el34world.com/Forum/index.php?topic=17585.msg177130#msg177130

Keep in mind that you will still have to use the FWB rectifier. That PT will not work with a 5Y3 tube. Also, you'll need to get an appropriate OT since yours will not work with a push/pull amp.

[edthill]

Thank you, Sluckey,

I was thinking I might just ditch the caps on the board for a cap can. I am also really the digging FWR with the 5E3 circuit I recently built. It's got a tightness to it that I like but still retains the growl, and gradual breakup from the 5e3. I've used a bunch of the Classictone 20 Watt Deluxe style OT's as of late too, so it will be interesting to get a contrast with Heyboer 18Watt OT from Doug. I've got a Gibson Hawk single Ended EL84 amp project coming up that I might use the old Allen OT for.
Thanks again.

[edthill]

Is there a program for Mac that allows me to open up the .sch files. the free one? The one I found is only for windows. Unless there is a layout in gif or jpeg format. Thanks again for your help. I'm pumped to tear up my Classic 10!

[kagliostro]

http://dhost.info/jschem/

Download your OS Version

Franco

[edthill]

Thanks Franco,

No such luck with jschem, or Eagle by Autodesk.
It can't be this difficult, can it?

[tubenit]

CHECK for ERRORS!   Compare with schematic.  IF there is a discrepency, go with schematic.

IF you build this, let us know how you think it sounds, please

with respect, Tubenit

[edthill]

Thanks, Tubenit, I definitely plan to build this soon. I'm jonesing for a new project. I am going to get a parts list together and then order from Doug.
Thanks again for turning this into a gif. I will double check against the schematic as I have both printed out now.

[tubenit]

Here is another idea.  Would have 3 pots (volume, treble, bass & mid resistor).   Uses four tubes and a mosfet CF (which I have had great success with)

This would allow you to have sort of a minimalist HoSo56 using the 5879 V1 pentode and uses a concertina phase invertor instead of the long tailed phase invertor.

I think this would be a very touch sensitive amp with a beautiful overdrive to it.   It's a DRAFT idea, I haven't built it but would love to.  CHECK for ERRORS!  Compare schematic with layout.  If there is a discrepency, go with the schematic

With respect, Tubenit

[edthill]

Tubenit,

It's a sweet looking build, probably sweet sounding too. Honestly the Mosfet CF intimidates me. Perhaps it's just my fear of the unknown and the fact that I'm still getting used to building again. You've given me something to think about though. Where does one find the Mosfet CF? Thanks for all of your help and ideas.

Ed

[sluckey]

If you don't like the MOSFET you can always just use another 12AX7 tube. Looks like that Allen chassis has holes for three 9-pin tubes.

[tubenit]

I think you should build whatever you want and something you are comfortable with.  I have NO agenda for you ......... just tossing out ideas to consider.   

Doug Hoffman sells the mosfets and zener diodes, etc...........  and everything else you need for the mosfet CF in his store.

It is every bit as easy to work with as a tube is, IMO.

Having said that,  I certainly understand wanting to pass on something you aren't comfortable with. 

This photo of one of the builds I did with a mosfet CF matches the layout I drew for you.  It is that simple. 

With respect, Tubenit

[DummyLoad]

http://dhost.info/jschem/

Download your OS Version

Franco

i have read elsewhere on this forum of reports that mac OS sierra breaks jschem binary: probably related to apple changes to java libraries.

try a couple of work-arounds? 1) download the source & compile, or 2) install and run JAR version if you have a java run-time installed?

--pete

[edthill]

I think you should build whatever you want and something you are comfortable with.  I have NO agenda for you ......... just tossing out ideas to consider.   

Doug Hoffman sells the mosfets and zener diodes, etc...........  and everything else you need for the mosfet CF in his store.

It is every bit as easy to work with as a tube is, IMO.

Having said that,  I certainly understand wanting to pass on something you aren't comfortable with. 

This photo of one of the builds I did with a mosfet CF matches the layout I drew for you.  It is that simple. 

With respect, Tubenit

It will definitely be on the backburner for my future list of projects. Is there a Hexfred in that picture? Those I have worked with. It's just a diode :-)
I think I have a potential buyer for the 5E3/18watt hybrid when it's complete, my studio engineer friend quite enjoyed the tone of my home brew 5e3 I used today for the session. I do have a couple questions on the schematic vs layout for the 5e3/18W hybrid. First, I'm not seeing the screen resistors for the 6v6's on the layout, the 1K3W? I will manually draw those in. Also, what component/wire is running through the 4th turret? it meets at the intersection of the 100K's of V1? is that just the connection for Part D of the power supply?
Thanks again, guys,

Ed

[tubenit]

First, I'm not seeing the screen resistors for the 6v6's on the layout, the 1K3W? I will manually draw those in. Also, what component/wire is running through the 4th turret? it meets at the intersection of the 100K's of V1? is that just the connection for Part D of the power supply?

Yes add 2.2k/3W resistors to the 6V6's pin 4 screens. 

Yes, it is just a piece of wire to connect to the B+ rail.   See row #3, #10 & #14 on the turret board picture.  Just wire connecting to B+. (different amp obviously)

With respect, Tubenit

[sluckey]

Is there a Hexfred in that picture?

No. That's the MOSFET.

First, I'm not seeing the screen resistors for the 6v6's on the layout, the 1K3W?

The screen resistors are 2.2K, not 1K. They are shown connected directly to the tube sockets. Same with the 8.2K control grid resistors. Fender has a neat method of connecting those resistors. Connect the 8.2K between pins 1 and 5 on the socket. Keep the resistors down close to the tube base. Connect the 2.2K/3W between pins 4 and 6. Stand the resistors about 1/2" above the pins for heat dissipation. Look at this pic for an example...

     http://sluckeyamps.com/6v6plexi/P-6V6_05_big.jpg

[edthill]

OK, just to clarify, the 5E3/18W schematic that Tubenit has posted has those resistors at 1k, but I should use 2.2K instead? Lay them out like typical Fender octal sockets?
Thank you,

Ed

[sluckey]

OK, just to clarify, the 5E3/18W schematic that Tubenit has posted has those resistors at 1k, but I should use 2.2K instead?

Sorry, I was looking at the wrong schematic. The schematic in your original post shows 1K.

[edthill]

Just out of curiosity, What does increasing/decreasing the value of the screen grid resistor on a 6v6 do? Or any tube for that matter. I am also aware that this being a cathode biased circuit is going to affect that as well.
Thanks,
Ed

[tubenit]

My experience is that 2.2k has more compression then 1k  tonewise.   So, the 2.2k sometimes might sound "smoother whe overdriven, IMO. 


With respect, Tubenit

[sluckey]

And just FWIW... Fender used 470Ω/1W screen resistors on all their 6V6 amps that even had screen resistors.

[edthill]

Thanks again for all the tips, sluckey and Tubenit,
The build is underway and going well. I had a question about the layout where it says NFB, negative feedback. On the schematic this says to tie the 33K and .1uF cap to ground. Is it literally that simple or do I tie it to the speaker jack?
Thanks in advance,

Ed

[sluckey]

I had a question about the layout where it says NFB, negative feedback. On the schematic this says to tie the 33K and .1uF cap to ground.

CHECK for ERRORS!   Compare with schematic.  IF there is a discrepency, go with schematic.

I would believe the schematic.

[tubenit]

I would believe the schematic

I would too!   

[edthill]

Ha ha ha,  I'm just going off the schematic then.
The other thing I noticed was that the plate resistors on V2 on layout both say 100K and on the schematic one is 82K and the other is 100k. Again out of curiosity what does changing the value to 82K do? Since most preamp tube plate resistors tend to be 100K.
Thanks guys,

Ed

[sluckey]

Using a 100K and 82K will give a more balanced output from the LTP PI. That is, the plate signals for both triodes will be more equal in amplitude.

[Tony Bones]

It's interesting that Fender used the same 82k / 100k combination right from the first LTP applications in the 50's through at least to the end of the SF era (at least all the examples I've looked at.) It doesn't matter what size tail resistor or even whether 12AX7 or 12AT7.

Not sure what my point is; I just think it's interesting. 

[sluckey]

Also notice that Fender specified 5% tolerance for those resistors. Most other resistors were 10%.

[edthill]

The Classic 10 conversion to Tubenit's 5E3/18w Hybrid is now complete.

I loaded her up with GE 6v6GTA's, and some older 12AX7's I had kicking around.

Thank you Tubenit and Sluckey for helping me with the technicalities of the build.

The verdict. Sounds awesome!

It will need some tweaking as the volume seems to be over loading the preamp with massive amounts of gain at a very low setting.
I left the original V1 22uf Cathode cap from the original classic 10 build, I should probably use the 5uf per the schematic.
Any other gain lowering suggestions are welcome.
I posted a gut shot pic.
Thanks again to everyone for all the help.
What to build next is the million dollar question. Ideas? I l'm loving amps in the 20-30 watt range. Also looking for something with octals as I have a more than a few EL-84 based amps.

Thanks,

Ed

[Tony Bones]

That build looks great Ed! You do good work.

I wouldn't expect changing from 22u to 5u cathode bypass to have much effect at all, especially in the first stage. You might try removing it altogether. Remove the same part on V1b to reduce gain a little more.

If that's not enough, then think about splitting the plate load resistor on V1b into two resistors in series, then take the signal off the junction. It would be very easy to do if you used Tubenit's layout in post #6. I can markup the layout if you want.

EDIT: If your volume pot is linear taper it will seem hair-trigger. Switching to audio taper might make the gain more manageable.

[edthill]

Thanks Tony. I appreciate it.  Am I just removing the capacitors and keeping the 1.5k resistors on the cathodes of V1?
The 1meg pot is audio taper. It does seem to jump into crazy OD territory right away.
What value would I use to substitute for splitting the plate resistors?
Thanks again for the tips,
Ed

[tubenit]

Tony Bones gave you some good ideas.   The plate resistor split idea on V1-6 is a reasonable idea to try.  56k/47k

I would change your V1-3  22uf to 4.7uf or 5uf or 2.2uf.  Change the V1-8  4.7uf to 2.2uf.   You can change the V1-3  1.5k resistor to 2.2k.

Make sure you have a 1MA (audio not linear) volume pot.

Use a 5751 or 12AY7 in V1.     AND/or  ............. use a 12AT7 in the LTPI position.

With respect, Tubenit

[edthill]

Thanks, Tubenit,

I'm pretty sure I am using a 1M Audio Taper pot, just not entirely sure now. It is the same pot that came in the Allen Classic 10 kit, which worked well when it was in the amp in a different form. Is there a way to tell? I tried to look up the Alpha pot codes stamped on the side of the pot and nothing camp up for them.
For the split load plate resistors, how do I lay that out? Am I just reducing the values of the 100K's to 47/50K or I am doing something different?
I have a 12AT7 in the PI now. I don't have any 12Ay7's kicking around that I know of. I will have to order some just have kicking around.
Aside from the Volume Pot being too sensitive the amp sounds ballsy and chimey. I had her running through two different 2x12's and a 4x10. Sounds great through all of them. Really digging it. Thanks again for all the help.

Ed

[edthill]

Tony Bones gave you some good ideas.   The plate resistor split idea on V1-6 is a reasonable idea to try.  56k/47k

I would change your V1-3  22uf to 4.7uf or 5uf or 2.2uf.  Change the V1-8  4.7uf to 2.2uf.   You can change the V1-3  1.5k resistor to 2.2k.

Make sure you have a 1MA (audio not linear) volume pot.

Use a 5751 or 12AY7 in V1.     AND/or  ............. use a 12AT7 in the LTPI position.

With respect, Tubenit

So I tried all the mods to the circuit that were listed, with the exception of the split plate load resistors. I am just not sure how the layout for the split load plate resistors looks. I have gotten a tiny amount of play in the Volume pot in terms of head room, but past 1.5/2 on the dial the amp is in ridiculous high gain territory? Otherwise the amp sounds good and I have rehearsed with her and the band.
What else am I missing? Are the Voltages in my B+ rail too high? I have swapped numerous combinations of cathode resistors/and capacitor values on V1. I have two 12AT7's in the amp. I swapped the 1MA volume pot for another 1MA volume pot...just in case. I'd also like to point out that the master volume functions perfectly and is very linear. Any and all suggestions are appreciated.

[ac427v]

Sometimes I prefer the thoughtful approach to troubleshooting. That would start with posting the DC voltage from each tube pin to ground and studying the results to find wiring errors or inadvertent part substitutions.
The soldering iron approach would be to remove one end of each cathode capacitor and see how it sounds without the gain those add. For me, thoughtful often gets used only when the soldering iron fails.
--Craig

[edthill]

I agree, might as well do this properly. Voltages, I've got voltages. Volume and master volume down while measuring them...as much as I like the sound of hisses and popping 

V1
Pin 1: 200
Pin 2: 0
Pin 3: 1
Pin 4/5: 2.9
Pin 6: 197
Pin 7: 0
Pin 8: 1
Pin 9: 2.9

V2
Pin 1: 200
Pin 2: 29
Pin 3: 89
Pin 4/5: 2.9
Pin 6: 198
Pin 7: 32
Pin 8: 89
Pin 9: 2.9

V3
Pin 1: 0
Pin 2: 2.9
Pin 3: 370
Pin 4: 376
Pin 5: 0
Pin 6: 378
Pin 7: 2.9
Pin 8: 20

V4
Pin 1: 0
Pin 2: 3.0
Pin 3: 372
Pin 4: 377
Pin 5: 0
Pin 6: 379
Pin 7: 3.0
Pin 8: 21

[Tony Bones]

Ed,

In the end, this amp is just gonna have a lot of gain. The concertina PI in a stock 5E3 has no gain (well, it has gain = 1.) The long tail PI you used has about half the gain of an ordinary 12AX7 gain stage. That means you have a boost of 30x (30dB) over a 5E3, which is already pretty quick to break up. The 18W Marshall has only ¹/₂ 12AX7 before the long tail PI. You have another one, so you have a full 60x (36dB) gain over that circuit!

Yeah, the throttle is touchy. 

{I wonder, is there a way to wire up a dual audio taper pot so that it has an even steeper taper than audio?}

Anyway, here's a markup of Tubenit's layout from post #6. I made it awhile ago but didn't bother posting it when I noticed from your photo that you'd built it up differently. Maybe it'll still be useful.  Note that compared to Tubenit's layout, the 100k resistor with the red X is completely removed. The 47k resistor is added, and the 56k resistor replaces a jumper.

[edthill]

Thanks Tony,

I will give this a day in court.
The amp sounds great. I really do love its' inherent tone.
It's got a tweed/marshall vibe without the ear-bleed volume
I'd just love the Volume to be more useable. It literally takes the most precise movement to dial in my preamp volume. Otherwise the amp sounds/functions absolutely killer.

[Tony Bones]

Taking another look at the schematic I realized this has the funky 5E3 backwards volume control where the wiper is the input. Could it be that the way it's wired it's not audio taper but reverse audio taper, which means it's even more hair-trigger than linear taper? I don't think so, I think the taper is right the way it's wired - that is assuming that twisting the knob clockwise makes it louder.

In any case, you could try wiring it the normal way. I've always assumed that Fender chose the backwards way with the 5E3 because there are two channels and this arrangement makes the two volumes less interactive (which is ironic because they're known for having interactive volume controls.) It does end up loading down the first stage when the volume is low, so might contribute to the unique tone of these amps.

But, just as an experiment, try rewiring the volume control. Leave the left lug grounded (viewed from the rear), move the two wires on the right lug to the middle and move the wire currently on the middle lug to the one on the right.

[ac427v]

Thanks for the voltage information. I don't see anything unusual there. Certainly the power tubes should be in a happy place. I like Tony's idea of the split load resistor. Also, I'm struck with the similarity of the hybrid amp design with Doug Hoffman's Stout(page 6 of attachment).
He used EL84 power tubes but is otherwise very similar. That amp uses different part values to reduce gain: 2.2k cathode resistors in V1a and V1b, no cathode cap on V1b, 500ka volume pot with conventional wiring, an extra coupling cap with a 470k resistor to ground after Vib. The 470k resistor to ground could be replaced with a 470k pot to provide prePI volume control. Lots of tweaking opportunities!
--Craig

[edthill]

Taking another look at the schematic I realized this has the funky 5E3 backwards volume control where the wiper is the input. Could it be that the way it's wired it's not audio taper but reverse audio taper, which means it's even more hair-trigger than linear taper? I don't think so, I think the taper is right the way it's wired - that is assuming that twisting the knob clockwise makes it louder.

In any case, you could try wiring it the normal way. I've always assumed that Fender chose the backwards way with the 5E3 because there are two channels and this arrangement makes the two volumes less interactive (which is ironic because they're known for having interactive volume controls.) It does end up loading down the first stage when the volume is low, so might contribute to the unique tone of these amps.

But, just as an experiment, try rewiring the volume control. Leave the left lug grounded (viewed from the rear), move the two wires on the right lug to the middle and move the wire currently on the middle lug to the one on the right.

I will give that a shot, Tony,

Worth a try and easy to put back to schematic form.

The pot definitely works clockwise and is hair trigger.

Thanks again,

Ed

[edthill]

Thanks for the voltage information. I don't see anything unusual there. Certainly the power tubes should be in a happy place. I like Tony's idea of the split load resistor. Also, I'm struck with the similarity of the hybrid amp design with Doug Hoffman's Stout(page 6 of attachment).
He used EL84 power tubes but is otherwise very similar. That amp uses different part values to reduce gain: 2.2k cathode resistors in V1a and V1b, no cathode cap on V1b, 500ka volume pot with conventional wiring, an extra coupling cap with a 470k resistor to ground after Vib. The 470k resistor to ground could be replaced with a 470k pot to provide prePI volume control. Lots of tweaking opportunities!
--Craig

Thanks Craig,

I did quite a bit of playing with Cathode Cap/resistor values on V1a/V1b. I noticed the higher the value of resistor/lower value cap I put in just made the amp sound colder. It did very little in the way of getting any headroom on the volume pot. I have definitely considered changing out that volume pot. Thank you for the idea. Much appreciated.

Ed