AB2 guitar amp design questions

Hi,

I have posted this question in the thread 'how does class ab2 sound' but it probably wasnt the best place to post it.

I have designed a guitar amp using 2x 807's in AB2 push pull, they produce around 75W in this configuration and require a load of 4k2 - 5kohms.
My first question is are the 2 output transformers i am looking at good choices (price is a factor)?
they are 1650R and 1650N
Here is a link to the specs:
http://www.hammondmfg.com/1608.htm

Also i have 2 x 12ax7 gain stages followed by a 12ax7 phase inverter (LTP type) so in total only 2 preamp valves, i have calculated more than enough gain with them using 120k anode loads and 1k5 cathode bias resistors to produce the required 100v peak swing from 10mV peak input.
(the buffering after the LTP is via IRF820 mosfets to drive the 807's)
But all guitar amp schematics i have seen of this power use another 12ax7 in the gain stage, am i missing somthing? will i be alright using 2 preamp tubes?

Cheers
Craig
 
I'm not an expert at amp design, but here's some of what I've learned.

12AT7 might be a better choice for the LTP, since it has higher current ratings. I believe it is equivalent to the 6SN7 (and 12AX7 = 6SL7), although the octals are usually more linear, which is why they're common in HiFi.

The preamp is designed to present the grid voltage the output tubes want to see (usually a lot more in a guitar amp where output tube distortion is desirable). For 6V6 that is about 25 volts.

The least number of preamp gain stages I've seen in a guitar amp is one (plus the LTP gain) in a Matchless Spitfire or Vox AC15. Modern guitar amps use more gain stages along with master gain in order to get preamp distortion.

I've been experimenting with two gain stages (i.e. one 12AX7 preamp into a LTP, or three preamp gains into a cathodyne), which appears to be rare in guitar amps. Most have three, which I believe is to compensate for gain lost in reverb.

Here's a nice example of a simple 807 amp
http://www.tubes.mynetcologne.de/roehren/807pp/schematics_e.html
 
Craig405 said:
My first question is are the 2 output transformers i am looking at good choices (price is a factor)?
they are 1650R and 1650N

In a case like this, I'd go with the 1650R. The 1650N is rated at 60W, and your design will easily exceed this. From the Hammonds I've used, they won't go much over the rating. Trying to force too many watts through the OPT will cause core saturation, and that's horrible-sounding, even in a guitar amp.

Also i have 2 x 12ax7 gain stages followed by a 12ax7 phase inverter (LTP type) so in total only 2 preamp valves, i have calculated more than enough gain with them using 120k anode loads and 1k5 cathode bias resistors to produce the required 100v peak swing from 10mV peak input.
(the buffering after the LTP is via IRF820 mosfets to drive the 807's)
But all guitar amp schematics i have seen of this power use another 12ax7 in the gain stage, am i missing somthing? will i be alright using 2 preamp tubes?

You got a passive tone stack in there? That "extra" 12AX7 in most guitar amps is used for the tone stack.
 

ray_moth

Ex-Moderator
2004-01-27 8:55 am
Jakarta
A 12AX7 isn't up to the task of dealing with grid current in AB2.
Brett, it doesn't have to. As Craig says, the buffering after the LTP is via IRF820 mosfets to drive the 807's. This arrangement is fine - any tube could precede those MOSFET drivers.

I don't know whether to agree with Miles or not. After all, this is a guitar amp and an OPT designed for 60W might do the job very well (unless the amp is meant for a bass guitar, I suppose). Both OPTs have the same max. primary current rating but the advantage of the 1650N is that you save 4lb in weight. You also get a lower p-p loading with the 1650N, which also needs to be considered.
 

Brett

Member
2002-01-07 6:02 pm
ray_moth said:

Brett, it doesn't have to. As Craig says, the buffering after the LTP is via IRF820 mosfets to drive the 807's. This arrangement is fine - any tube could precede those MOSFET drivers.
Missed that, must'nt have been awake yet.

The IRF's are still going to have a goodly chunk of gate capacitance, even as SF's. Still not sure about the AX7.
 
Firstly thanks for replies all!

I checked out the Edcor page, transformers look fine i would definatly consider if i was in the U.S. but hammond are more convenient and probably cheaper on shipping within the UK.

The 1650R is looking more attractive, the load is slightly higher so i will not be getting max power transfer but is this likely to have a negative effect other than that?
Plus the transformers are the same price, how wierd.. a better watt/pound ratio cant be bad!

I am totally open to better transformer manufacturers in the UK, so please suggest if anyone knows of them.
Sowter are too costly for me though :(

Miles Prower, thats gotta be it, i dont have a tone stack at moment so i guess im safe until i decide to add one in.
Im hoping the amp will sound so fantastic i wont need to shape tone (HA! i wish).

edit.
That 200w 807 beast is cool, though wouldnt the 6SN7 have too high a Zo for 3 807's in ab2? good to see the hammond is taking a pummeling though.
 
Craig405 said:

That 200w 807 beast is cool, though wouldnt the 6SN7 have too high a Zo for 3 807's in ab2? good to see the hammond is taking a pummeling though.


Bear in mind AB2 power supply requirements. This is harsh. Some time ago I modded a 20W 6550 amp to ++50W AB2, and backed quiescent to 20mA per tube...however at full load was a wacker 180mA per tube. If you are using true AB2 for guitar power in overdrive then your tubes may have a short life under peak currents.
A compromise ? Once youv'e tried AB2 and heard it (I find it very bright and hard on the top end)... the temptation is to veer back to AB1 for extended tube life.
I found AB2 likes a 20% lower output Z transformer for best power transfer.

ps. Morgan Jones book mentions thrashing output transformers and hasn't destroyed one even under the worst stage conditions. Copy that.

richy
 
One of the better sounding guitar amps that I built (I built LOTS of them once upon a time) was a PP 6V6 with two 12AX7 configured as you describe and included a fairly standard Fender style tone stack. That amp had too much gain for my taste and I ended up attenuating about 10dB before the tone stack. The 6V6's were biased around 20V IIRC, and had only moderate NFB. (Sorry, I don't remember exactly how much.)

So, without the 10dB attenuation and similar NFB you should do OK provided you aren't looking for massive amounts of gain. The most I ever ask for from an amp is anything from bluesy distortion to a Rock and Roll bark. If you want Santana style sustain then you will either need more tubes or a stomp box in front of the amp.

Also, I agree with Ray Moth; a guitar amp needs to go down to only about 80Hz. For 100 watts the core only needs to be about 1/4 the size of one that goes all the way down to 20Hz. In the end, the load will affect the tone. In general, lower resistance tends to sound smooth and punchy while a higher resistance load will have more crunch and bite. It all depends on the sound you're looking for.

-- Dave
 
Rich, I wondered if running into AB2 would shorten tube life.. well its good to have the power in reserve even if i dont drive it into AB2. I guess i chose AB2 because it was a bit different. As you say, if it proves trouble/ tube killer then i can always back it down. the 5K OPT would also be a very good match for AB1 i think. Even if its a little high for AB2.

Dave,
NFB.. havent considered this yet hmm. Im going to run without then ill add some experimentally if im not happy with performance.
Your almost certainly right about the power bandwisth point, but i cant resist a more powerful transformer for the same price!

A question:
My layout requires the OPT to be right next to the choke, about 1 inch away mounted on the top of the chassis (its a partially shielded choke but the core isnt shielded).
Is this likely to cause me issues wich induction into my output transformer windings?