UL output ("baby" Major) prob.

[Mikko Vainiala]

Hi guys,

I have a bit annoying problem with my Marshall "baby" Major build. I use Hammond 378CX PT and 1650R OT with pair of KT88's and B+ around 540-560V.

The prob is that I have removed the concertina and used that half of 12ax7 as an effects loop return stage. This means also that instead of the later paraphase boost stages I have used common type LTP (with 12BH7) which just seem not to give enough output/volume/swing for the UL.

Today I removed the LPT and (as a test) put in a direct-coupled gain stage and concertina splitter instead (plus added IRF820 cathode follower on the anode side of concertina). The result: I sure got all the volume back (and some more) but didn't quite like the sound. Especially the leads were kinda "dry" and too tight in a bad way without the sustain character which were there before. I found myself wondering if this drier tone was because it was cathode follower driven (grid current in action?) and would the sound be similar with the interstate transformer too?

Now I'm tinkering what should I try next. The original tone with 12BH7 parapahase boost stages was just right. Should I install one more 12ax7 (have already three channels with 8 12a_7's in there) only for concertina and wire it also as parallel-consertina to get both halves in use.

Other possibilities or chances without adding a tube? Could there be a transistor based paraphase amplifier before the 12BH7 boost stages? Or wiring the 12BH7 in parallel and use an interstage tranformer. (Any examples how to actually wire this things up?) I have a chanche to get Lundahl's
here. Triode Electronics sell also universal interstage transformer UIS-1, is this good stuff? http://store.triodestore.com/unintrui.html

Ok, hope to get some thought and tips from here, but please remember that I'm WAY more musician (and preamp tone tweaker) than deeply involved in electronics or engineering stuff.

Greets,

Mikko

[loudthud]

I'm not sure if I can help you with your problem but wanted to ask about the transformer set, I've been looking at the 1650R and KT88's for my next project. How much power are you getting? I would think more B+ would be the way I want to go.

It's not easy to drive KT88's which, IMHO, is why you don't see them used very much. Is it more hassle than they are worth?

My first guess would be to increase (maybe 2x) the the feedback resistor going from the speaker back to the cathode of the 12AX7. This will change how the tone controls work but loosen up things a bit. It sounds like you have changed the gain structure inside the feedback loop which will alter how the amp sounds and operates.

[Mikko Vainiala]

I've been looking at the 1650R and KT88's for my next project. How much power are you getting? I would think more B+ would be the way I want to go.

Specs say that with a pair of KT88 and B+ around 560V you should get 100W output with 4,5K plate to plate load, both ways, tetrode or UL connection. So this 1650R is just perfect. Its just that the UL needs more swing from the PI.

It's not easy to drive KT88's which, IMHO, is why you don't see them used very much. Is it more hassle than they are worth?

You can drive them easily with a common LTP in terode connection...the hassle comes out when you run out of (ideas and) tube sockets needed for the proper driver setup and still wanna keep the UL.

My first guess would be to increase (maybe 2x) the the feedback resistor going from the speaker back to the cathode of the 12AX7. This will change how the tone controls work but loosen up things a bit. It sounds like you have changed the gain structure inside the feedback loop which will alter how the amp sounds and operates.

Yes, in normal conditions. My last stage is same as Major has (where the nfb is connected) except I don't have tone stack there plus I use that for a fx return gain and the feedback works there nicely keeping things cool (so that the effected signal will not get distorted).

Well, that is it for now. A moment ago I wired up a "floating paraphase" (from Norman Crowhurst lessons) and already like it much better than a LPT or direct coupled gain stage and concertina. Let's see if I can get enough voltage to/from it...

Mikko

[Mikko Vainiala]

Update: Think I'm closer to the problem...

Hi,

Yesterday I added tube for concertina and changed LTP back to paraphase boost. I turned the amp on and no raise in overall volume level but after half minute mains fuse blowed. Replaced the fuse and removed the nfb loop and result: Enormous raise in volume! "Now this is a 100W amp..."

Important note:

I tried to connect the nfb loop to many places on the circuit and it works. If I happened to get positive feedback I just change the leads going to paraphase boost stage and everything works again...BUT here is the thing:
I can't connect the nfb loop to the same point as Marshall uses, otherwise power tube milliamperes goes sky rocketed and I must quickly turn the amp off. Btw, before that gain stage I have an virtual earth amplifier (tube stage for active eq).

Can someone explain what is happening and how to get this under control? Plus for the tonal reasons - its a must for me to connect the nfb loop to that "wrong" point.

Edit: One more notification - with the floating paraphase arrangement I got pretty loud volume levels ("almost there...") and connecting the nfb loop worked just fine.

Mikko

[Mikko Vainiala]

Solved!

Hi guys, sorry for hysterical and rapid postings. :-)

Well anyway, I used that gain stage where Marshall Major injects the nfb as a fx return gain stage. The solution is to add similar local feedback as Kevin O' Connors (Best All Tube Loop) has in return stage, which is 47nf and 1Meg resistor in series from plate to grid. After adding this everything is in control and you can use also the nfb from the speaker - but in this new situation I don't like it no more (it was good and "open" sounding before!) or perhaps I'll end using 100k there.

Hey loudthud, thanks for listening! And also thanks to a suggestions I got from a member with a private mail.



Mikko

[Satamax]

Terve Mikko.

Well, since you're into a testing mood, instead of your floating paraphase, why don't you try a proper high power PI

http://music-electronics-forum.com/showthread.php?t=354 Usual gain of 300 iirc, can go up to a thousand, or so i've read.

[Mikko Vainiala]

Terve Mikko.

Hello Satamax, lol! Long time, since we met last time in old ampage forums.

Btw, here is what I learned (and deserved for only tweaking preamps):

1. Whenever you add a gain stage (or concertina) to a circuit remember disconnect the nfb from speaker jack first, just to be sure.

2. If everything works connect the nfb from speaker jack AND keep your eye or multimeter on power tube current behaviour (rather than just waiting positive feedback squeal, which is normally solved by changing PI leads). If you're not fast enough with your current readings and turning the amp off your power tubes or (hopefully) fuses blow. To solve this problem try to inject some local feedback (cap and resistor in series) from anode to grid on that gain stage where the nfb from speakers were connected. If that works you can then try connect the nfb from the speakers again.

As an EDIT I'll add this:

3. If you don't like nested feedback loops in your amp you could also wire a cathode follower from the anode of the concertina - now I have a possibility to connect or disconnect whatever nfb combination I like only for tonal reasons plus the concertina output with two followers seems improve the overall tone.

Well, since you're into a testing mood, instead of your floating paraphase, why don't you try a proper high power PI

http://music-electronics-forum.com/showthread.php?t=354 Usual gain of 300 iirc, can go up to a thousand, or so i've read.

I have now the real deal (added tube for concertina) and the paraphase is not floating anymore. Today I've tested different tubes for the paraphase boost and I liked the 12au7 for the cleans best (its so clear and crispy) but 12BH7 and 6CG7 (even more) shine in the crunch/high gain territory - actually I measured it - a lot more Newton meters in tone with them.

But hey, I've read this super high gain pi thread before and its really tempting - already tonight?! I just can't believe it before trying or has anyone, even Ray, tried it yet? Think we have to lower those bias feed resistors to around 150K with KT88's too? If someone has tried this please tell your opinions in the thread linked above.

Mikko