AO-39 to Spitfire Variant conversion

[vampwizzard]

Hey folks! Back at it again with another hammond conversion.

In addition to restoring the EKO Duke on another thread, Ive come upon a good source for AO-39s at a pretty reasonable cost. I successfully converted an AO-29 to a 5E9-ish clone (also on this forum) and they are freaking sweet.

What to do this time? I went through a ton of EL84 sound clips and wanted to try a spitfire-ish circuit.

One input, dropped the input resistor to 33K to compensate. Decoupled the triodes and adjusted the resistor values to compensate for the gain. Thats my first source of concern.. does that look correct to maintain a similar gain structure?

Why only one triode? I want to steal the other triode for trem. I yanked it out of the Traynor circuit/Muchle$$ Liten Revibe. Makes a lot of sense, should work fine.

Other changes I'm looking for opinions on include removing the master volume control. Planning on inserting a tone cut control instead. Also curious about power filtering. Sluckey's AO-39 conversions use 4 caps for 4 B+ nodes while the revibe uses 5. Can I get away with sharing a node on phase inverter and trem triode?

Lastly, Goodsell mentioned that his Super 17 doesn't use a choke. How much of an impact would that make on the power filtering? Should I do something to compensate?

Big thanks to all the documentation Sluckey put out on the AO-39. Really useful. Glad to be working on amps again.

See my incomplete schematic below. Also willing to listen to critiques on my LTSpice skills. LOL 

[sluckey]

Circuit looks basically workable to me, except your trem circuit. I suggest the changes in the attached. I would power the trem from node B. I would build without a choke. Plenty of popular EL84 amps without one. You can easily add one later if you think it would help.

[vampwizzard]

Thanks! Which power supply should I mimic? Your 39 conversions or the spitfire?

[sluckey]

Since you are building a spitfire amp, build the spitfire power supply too, including the choke.

edit... I also suggest you use spitfire values for R3 and R4.

[vampwizzard]

Makes sense!

R3, R4.. I was confused what to do when you separate paralleled triodes, which is why I doubled the values compared to the stock amp. That’s still the suggestion, use the same R values on the separated triode?

Thanks Sluckey.

[vampwizzard]

Long time no talk. Got stuck in a bunch of gigs, went on a mini tour.. etc. Things finally wound down show wise for the year.

Update: Chassis stripped, holes for pots drilled. Phase inverter tube socket installed. Retaining rings for new cap cans installed. Output jack installed and wired up (broke the heck out of the original output board and it was in the way) Pics attached. I certainly messed up ordering my screen resistors so its time to re-evaluate and order more parts.

First, for enlarging holes while converting the filter caps to new cans.. I used a dremel to horrendous results. How do you all do that metal work? Punch it? Hole saw bit on a drill?

Next, still trying to wrap my head around how to fix the values on v1. The spitfire uses paralleled stages with a 220K load resistor and a 1.5K cathode resistor with 25u bypass cap. I want to use half the triode for the preamp and the other half for the trem. Per another thread on the subject https://el34world.com/Forum/index.php?topic=15125.0 you halve the values of the resistors to compensate when going from single triode to parallel.. meaning that I would need to double the values of a paralleled triode once converted to a single triode. Is that correct? In this case,



The 220K resistor would become approx 440K and the 1.5K would go to 3k. Is that correct thinking?

As always, thanks everyone. And for you NC folks, hope youve dug out alright. SWVA doesnt know how to run a snow plow and the school districts have been closed all week. HA!

 

[sluckey]

I used a punch to enlarge square holes for the cap cans. I might use a step bit if the holes were already round. You can't use a hole saw for enlarging an existing hole because the hole saw relies on the small drill bit to guide the hole saw.

I would leave the resistors on V1 just as they are. Maybe reduce the 220K to 100K.

[vampwizzard]

Thank you for the advice on the metal work.

As far as the resistors.. is there a explanation as to why to do that? I feel like im missing something important. The gain difference the same values on V1 is approximately 30% more gain in parallel vs single triode. Then its my understanding to get that gain back I'd have to DOUBLE the Ra resistor, not halve it. (220 to 440, etc). on the 18w marshalls, Ra is 100k for paralleled triodes. Do we just not really mind the lower gain? 


For what its worth, I think im stumbling on an old problem solved by a boutique builder. This triode feels like the make or break of the design.. especially because it appears any and all gut shots of the amp in question seem to have been censored from the webs. Im taking a guess at the design.

[sluckey]

I just told you what I would do with the resistors, but you can experiment with it and do what you want. I don't hear much difference between parallel triode and single triode so I assume any difference would be subtle. Better ears may say differently. I never put a scope on it to see the actual difference in gain of parallel v. single.

Look closely at the Marshall 18W since you mentioned it. You'll notice that the normal channel uses the two input jacks differently than most dual input preamps. The HIGH jack sends the guitar to both grids (parallel operation), but the LOW jack only sends the guitar to a single grid (single operation). I don't hear much if any difference between the two inputs. This is why I say I would leave the resistors as they are.

I also built a Lightning (same preamp as your spitfire). It is a gainer sounding amp than the 18W, but then there are other differences between the two amps. I really don't see that triode as a "make or break" of the design but maybe it is. If that's important to you, maybe just stay with the parallel triode. You can always just add another tube if tremolo is important.

[1blueheron]

I used a punch to enlarge square holes for the cap cans. I might use a step bit if the holes were already round. You can't use a hole saw for enlarging an existing hole because the hole saw relies on the small drill bit to guide the hole saw.

I would leave the resistors on V1 just as they are. Maybe reduce the 220K to 100K.

A tip for turning a square or round holes into a larger round hole if you don't have a hole punch:

Using the area that will be come waste in the flats of the square, drill a small screw hole on either side of the hole.  Use these holes to secure a small block of wood or peice of sheet metal on the underside with small screws.  Now you can drill your center pilot hole for using a hole saw to enlarge the hole.

[vampwizzard]

Thanks 1blueheron, that's a heck of an idea.

Sluckey, I made the assumption that one gain stage made it important lol. I may have been a little tired and frustrated with gain calculation spreadsheets. No one should be doing math after midnight. Thank you, again, for your patience. Experimentation looks like the way to go. resistors are cheap lol

[PRR]

> The gain difference the same values on V1 is approximately 30% more gain in parallel vs single triode.

For those values (anything like that) there is hardly any difference between single and double. The Idiot says 0.4dB which is not a difference you can be sure of hearing, certainly not important. And more than swamped by different tubes, resistor tolerance, and how many vitamins/spinach you eat.

Splitting as two consecutive gain stages is for-sure a WOW change. Probably too much until tamed with loss or heavy tone networks.

[vampwizzard]

Thanks for the info PRR, much appreciated.

[vampwizzard]

Update: She LIVES! At least as a single triode spitfire


I have to go back and look at the discussion about the trem circuit. May have ordered the wrong cap for part of the circuit.

Gigantic props to Sluckey for the documentation on his own 39 conversions. True P2P was a bit tougher to layout without a plan but I was able to figure it out.

I grounded the power caps to the rectifier socket along where the PT grounds were. I grounded the incoming power to a PT mounting screw, and I grounded the preamp to one newly dedicated terminal strip segment. I did not ground the pots via the preamp bus you see on a lot of layouts. My first voltage dropping resistor was the chassis mounted 200Ohm from the AO-39 and it got my voltages down pretty well. Going to try and get the pictures embedded correctly.

I do not like that ive got a couple of flying connections, especially on the grid leaks (they fly to the depth control, which i then grounded to make it a non-trem spitfire.

[vampwizzard]

More pics.

[vampwizzard]

sound demo is uploading but can be found on my youtube: https://www.youtube.com/watch?v=bg8NHOFZ_oA&feature=youtu.be

[silverfox]

What is amazing is what you don't hear- No AC hum... Untwisted filament leads and lots of wires hanging around but nice and quiet.
Hmmmm.
silverfox.

[vampwizzard]

yeah.. the only filament i didnt mess with was the initial run to the output tubes.. that was existing. I wouldve thought coming off the volume and control id get hammered but its a relatively short run.

Better lucky than good I suppose.

[vampwizzard]

Alright folks.. finished my first tolex covered cab in the matchless style. HOLY COW this was a pain in the butt. Learned a ton. Shes ugly but functional. Box finger jointed box, carolina blue tolex. I used all plywood for this.. i think 1x10 dimensional lumber will work better.

Should be complete with a tube set coming in the mail and the knobs. On to the next one.

[vampwizzard]

Went through and started testing the amp through various settings.. maxing out knobs, etc, and ran into a very strange problem.

Intermittently it sounds like im shorting to ground. I get the 60cycle hum thats unaffected by tone/volume. I can recreate this by tapping on the chassis with my chopstick. I can then also make it go away this way as well. I can make it go away by wiggling the input guitar cable. If i turn the amp off in this condition, it will dump some signal to the output and sound garbled. Typically restarts with


I've tried different cables and guitars. Replaced all tubes. Cleaned the sockets with contact cleaner. Removed preamp tube triode 2's cathode res/cap out as its unused without the trem, reflowed and rechecked ground connections. All seems in order. Very unsure what to make of this. It doesnt matter where I chopstick.. connection wise, chassis wise.. itll cut out. Any advice on where to look?

[shooter]

.. itll cut out.

bad solder connection somewhere

[vampwizzard]

The "somewhere" is what im trying to pin down. If i can chopstick any connection to recreate the problem, is it reasonable that its a connection to ground I should be worried about? Sorry, I was pretty frustrated yesterday and a little tired after work.

Im biased really high as well.. near 95% of the output rating so that will need to be addressed as well. Not enough time on brand new tubes to cause great damage concern but still need to get it dialed in.

[sluckey]

Im biased really high as well.. near 95% of the output rating so that will need to be addressed as well.

95% is fine for this amp.

[vampwizzard]

Thanks Sluckey. 

Reflowed all ground solder joints, really messed with the incoming ground connection.. and alas, problems solved. Had a coated lock washer preventing solid contact.

Sometimes its that easy.