This is a Gibson Skylark in the “Crestline” series.  It is one of the many amps released with the GA-5T model number.   This was made sometime in the early 60′s as the later model “Whiteface” amps would be released in ’65.   This is not as popular as the true Class A SE Skylarks from the 50′s so it cost me $202.50 with shipping a couple of years back.  It’s not working right now and I’ve decided to refurb it.

When it worked (which it actually did at some point) it had the best tremolo I had ever heard.   It is so weird that Gibson tremolo is the shame of high end amp world these days.  They used to do it brilliantly.   However, I don’t like trem, so it’s going out..  in it’s place is a Master Volume.   I’ve read all the gasbags saying “It’s not the same!  You won’t get .. “  but the truth is I don’t care.  It makes a good sound.   If I get a different sound from 11 on both dials, great!  That’s just more versatility to me.

Here is the plan in general :

• Use the original  Iron, Speaker, Cabinet and Chassis as much as possible
• Mod back to original GA-5 specs, removing tremolo circuit.   except as noted below.
• Replace all discretes using original bone stock values
• Add Standby switch (hopefully I can do this with a rotary switch like they have now)
• Use multiple parallel caps for filtering ( a la TUT5)
• Add Master Volume before the Phase Splitter
• Add Pentode/Triode/Other switch
  • Pentode – Standard operation except provide 1k flameproof resistor for each screen
  • Triode – Put the screen in the plate loop (keeping the 1k resistors)
  • Other – Lower the voltage supplied to the screen (should be something between the other two in terms of breakup)
• “Galaxy” grounding, again, a la TUT5
• Careful layout and use of shielded wire on the signal path.

Should really be doable in a weekend if I have the parts, but I really want to nail this build, so I’ll take my time.

So where are we ?   I built out my McTube clone and it actually worked.   Here are a couple of pictures :

So yeah, she’s nice looking ..  I’m a fan of the clean look.   The back has the instructions / names of the controls.   The three banana plugs on the top are for heater and plate voltages (recall that my power supply is separate).

Here are the guts.. quite a bit less nice looking :

So what happened here is I didn’t carefully lay out the components before hand.   I ended up with a lot of bare wire crossing a lot of other bare wire and naturually got some interference.  In fact.  I got an AM radio station and had to tie some ground points together before I got a good sound.

How did it sound?  Well, I plugged it into a crappy Frontman 15B (that I got for 13 bucks) and it sounded like a tube amp.   Not a great tube amp, but a tube amp nonetheless.  It would be very fun to A/B with this setup and a good solid state amp to really get a chance to play with the ‘sound’ of tubes.

For a distortion box, I would tweak it I suppose, because I don’t have anything that can give you that barely hanging on sort of distortion and this doesn’t get there, although it gets pretty close.

If I were to use this more often, it would need a complete rewire.   I will probably consider it though because as always.. I have a plan

]]> http://www.paleoelectronics.com/blog/2009/10/the-real-mctube-wrap-up/feed/ 2 http://www.paleoelectronics.com/blog/2009/04/a-game-changer-the-gibson-ga-20-rvt-initial-review/ http://www.paleoelectronics.com/blog/2009/04/a-game-changer-the-gibson-ga-20-rvt-initial-review/#comments Fri, 24 Apr 2009 05:56:38 +0000 Dilapidus http://www.paleoelectronics.com/blog/?p=129

I have purchased the best amp I have ever heard.    This is a 2006 Gibson GA 20 RVT.   It’s not perfect, I might have to mod it, but I am definitely in love.   My strategies about building the Songbird and other things are all up in the air now.

Let see why..

Once again.. this is the best amp I have ever heard.  I suspect that other people get better tone (T-Bone Burnett playing for Plant/Krause comes to mind) but I have never heard anything personally that even comes close.   In the first 20 minutes after starting this amp for the first time, I got a bit concerned.   It is voiced very dark and I’ve backed my Strat pickups off thestrings already to moderate Fender shrillness.  However, I very quickly came upon settings that were just magnificent.

Specs

Here are the specs from Gibson (with comments):

15W Class A   -   12AX7 pre-amps,  6V6 x 2 Output

The circuit really looks like it’s Class A.   No phase splitter.    12AX7 to 6V6, just like the old 50′s Champs and Skylarks.   The two output tubes are in parallel, but not push-pull (very rare .. last seen in the BR-9 I beleive), so I wonder if you couldn’t disable a tube for even more low volume fun.   I haven’t really pushed it very hard yet, so I don’t know how clean it would stay in a club situation yet.    Rumor is they use Sovtek or Chinese tubes in production.   I’ll replace with quality when the time comes.  It can only help.

Do I still need the Songbird?   Hmm..

Hand Built, Point to Point

I’m really not a purist when it comes to these things.    The Fender Hot Rod / Blue Deluxes are circuit board and they sound great.   My concern was that I would not be able to repair them myself.    Any fool can beleive he can repair a PTP amp.

2 channels plus mix input

Looking forward to trying the other channels someday :-)

Tube Reverb,  Tube Tremolo

The reverb is the most un-Fenderish reverb you can imagine.   It’s got a deep reasonant hall sound to it.. not bright and springy, but very very filling.   Hard to describe.. I’ll work on sound samples this weekend.  It’s a great sound and it’s nice to not be making the same sound everyone else does.

There is a tremolo but it could not really suck more.    This, to me, is amazing because the 50′s and 60′s  Gibson Tremolos sound fantastic.    They had an oscillator varying the power tube bias and were great.

Because of the fact that this has both reverb and tremolo, I’ll bet a lot of people bought it for either surf or very bayou swampy sounds and they were probably very disappointed in that.  Too bad for all concerned.   Gibson loses some luster with such a crappy tremolo (taking up two knobs and a full tube).

Again, the reverb is different but outstanding.

Pentode/Triode Switch

This is nice.   Cuts power and lets you reach a higher level of preamp saturation before you wake the kids.    Some tone change, but nothing untoward.

12??? Eminence Legend Speaker, Asymmetrically Placed

I am very interested in how this responds to other speakers.   It has various outputs so I’m wondering how it will sound against a classic Jensen Ultrasonic or the new Weber Alnico.    I’ve heard that this is a good first mod as these are probably not the made in Britain speakers.

Metal Film Resistors

Note :  How is that a feature?   Metal film is widely believed to cause sterility in amps.  This is balanced against the fact that the amp is silent even when dimed if you don’t play something.

DC Powered Filaments

No hum.. it must be working but it seems to be AC to the power tube filaments.    I wonder why.

Hand Wired Turret Board

Again, I’m not a purist, but this should make repairs/mods easier.

Two-Tone Vintage Brown and Tan Tolex

It does look good, but you can’t use right angle cables because the case is too thick and too close.

Comments

This thing streeted for 1k in 2006 when it was made.   Honestly, I wasn’t even shopping for an amp, but I saw that come up on ebay for 300 to start, I bid 300 and until the last few hours, I was the only bidder.   I got it for $406.01.  For a hand built PTP amp, that is a phenomenal deal.

So far I have only barely touched the “crunch” and mix channels and really haven’t taken it out of Triode mode much either.  I literally stop the backing tracks all the time now and just listen to the amp play the notes.   I’ve never enjoyed hearing a guitar more!

As I said, it is voiced very dark.   You have to shape your tone a fair bit (which I am not really used to doing) but it’s glorious once you get there.

Drawbacks

• Tremolo is completely useless to me.
• Very loud pop when switching to standby.
• No master volume.
• Only one control for reverb.
• The channels share pre-amp tubes, so you cannot swap tubes in a channel

Nothing end of the world here but unsurprisingly, I am starting to think about some …

Mods

• Yank the Tremolo and add a master volume.   You could get one per channel since the Trem takes two knobs.
• See if there is a place in the reverb circuit to adjust the dwell.   Pop a pot in there and bring to front if I only use one master volume.
• revamp the pre-amp topology so that I can use a 12A?7 to fiddle with the sound on channels

Truth be told.. I might never do any of those mods.     Maybe I’ll just leave this pristine and wrap up the McTube and the Songbird.

That is the way it’s always done on the old amps and I’m surprised in retrospect that he didn’t remember the simpler trick.   Better amps would place a 100 Ω potentiometer between the two sides of the 12 V tap from the PT.   Of course, I’ve just remembered that too now that my rectified filament is done.    Oh well.

So the way to determine quality with rectification is to ask how close you got to the desired voltages and how much ripple (AC in your DC) remains in the signal.   The only way you are gonna see this is on a scope or a dedicated tool for this.   You might get it measured with a DMM, but I never got that to work.  If the ripple is complex, the DMM will not settle on a value so you will be seeing snapshots that make no real sense.   Since the DMM is reading RMS,  the highest number you see will be about 2/3 rds the total ripple.

Unfortunately, I didn’t take a picture of the ripple in the Skylark when I had it open.   Suffice it to say that it was plainly visible.   With a scope, it’s easy to see.   You set the scope to AC (and you don’t have to worry about the DC voltage spec within reason) and you slowly dial down the volts/div.

For the McTube, on both the B+ and Filament voltages, there was practically no ripple.   At 0.01 volts/div at the slowest sweep, I saw about 0.005 swing that took a bit over a second.  My scope wouldn’t trigger on it but it was plain enough.   Normally you expect to see something at 60 Hz at least and I saw nothing.

So I guess it cost me some extra parts, but that is a clean power supply.

Next time I get an amp opened, I’ll shoot a pic of the ripple.

So the B+ voltage is expected to be around 140V with load and 150V unloaded.  I got 173V unloaded so I might be running a bit hot, but that is ok, as this tube is rated to 250 typical and 300 max.  As we shall see when we do the calculations, this will actually make it just a bit easier to chart.

In fact, in the Champs, Fender ran this tube at 150V and Gibson ran it at 165V in the Skylarks, so we will be operating right smack in the late 50′s!

For the filament, I got 16.6V unloaded, which is close enough to the expected 15-16V that I am happy with the power supply.   I will button it up separately from the rest of the McTube stuff as this is stage one of the Tube Amp Computer!

More soon.

I’m not gonna do a full analysis either, but I want to cover some of the basics, and then measure once we are built.   I’ve done some back of the envelope already and I’m pretty sure I don’t know what I’m doing.  

12AX7

The tube is 12AX7, a very popular pre-amp tube, so knowing how it sounds and how it used in a circuit will be very handy.   This is a dual triode.   It houses two triodes in one glass bottle.

Power

Each triode is rated for 1 watt of power.    Power is    P = VI, so we can draw that curve on the standard datasheet just to keep ourselves safe.    The datasheet shows voltages from 0 to 500 and currents (in mA) from 0 to 4.0.    All we need to do here is pick a number, do the math and plot the point.

I’ll start with 250 V for no reason at all and find the current that creates 1 watt of power.  Please see the datasheet below

1 W = 250 V  * I  ->  I = 1/250 A = .004 A  = 4.0 mA.   Isn’t that handy?  Right at the top of the graph.  Wow, what luck!

Lets do some more

So just draw those points on the graph and connect them with lines to get a good idea of the safe ranges for this tube :

12ax7 data sheet with max power curve

Click on the sheet for a better picture, but there is no way we are gonna get close to the max power rating, it’s not even in charted territory.    Recall from the real McTube article that the max voltage is 140 V.    That would be over 7 mA of current and this is a stomp box meant to go in front of your amp.   I don’t think we need to worry.

Next up.. where on this chart are we?

I was talking about the Hammond 269 before (not overly cheap, to be sure) but as luck would have it, I found a better choice for this application.   Antique Electronic Supply has a PT called the  P-T442 .  Which I got for about 16 bucks.   Here are the specs :

Primary: 117 volt, 50/60 Hz
Secondary: 125 V, 15 mA
Filament winding: 12.6 volt, 0.3 amp
Mounting centers: 2″, U-bracket
Weight: 0.7 lbs.

Way more secondary current than we need (the doc says 500μA) and since we will have DC for the 12A*7 heaters, this thing can push twice the heater current required.   (Hmm..  perhaps a second tube .. two different circuits side by side .. switchable or stereo…  hmm..)

Where were we?  Oh yes.   The transformer is on the way, and cheep.   Time to start crunching numbers and see if I really am getting any of this.   More later.

Now many of the options that I was considering are simply to try them out, with all other things being equal.   I have no real idea what effect they would have…  I want to hear them for myself.

Here is a quick list of things I am considering :

Input

• Fender vs. Gibson style grid stoppers,  Old school cap and grid.

Pre-Amp Stage 1

• Switchable bypass cap, Tweakable bias voltage,  Bias negative warning light, The Dave Hunter small cap mod, Other plate voltages

Pre-Amp Stage 2

• Variable NFB, Switchable “Tone Expander Cap”

Output Stage

• Switchable Bias Cap, Master Volume Control, Triode/Pentode switch

Power

• Standby Switch, Hum balance vs. rectified heaters, choke vs smoothing resistor.

That’s a lot of options and I’m pretty sure I’ve missed some.   So my plan is now somewhat different.   I am going to build the amp as an analog computer first.   The person who gave this blog it’s name calls it bread boarding, but I think of that as using the board with all the little squares.

I will create component boxes for all the pieces I need.  I’ll put banana jacks on them for connections and wire them up as need to try things out.   I’ll have generic parts, like decade boxes, alligator clip leads,  octal and nine pin tube sockets, pots etc. and I’ll make up custom stuff like switch sets, and transformer holders.

I’d like to have them all sitting on something stable.   Maybe I’ll put velcro on the bottom of all these boxes.   I think I’ll do the Real McTube II like this first.   Woo hooo.. this sounds like fun

Tube Biasing – 1st Stage

As I understand it from my readings, the standard way of biasing, self biasing, is a complex reaction.   Essentially as the power drawn from a tube increases, the bias goes more negative, somewhat thwarting the increase in power.    This could be a very good thing.   To my mind, this sounds like a plausible explanation for why compression happens.   Now, as you all know, my understanding of what’s going on in these amps is mild, at best.    My thoughts for biasing are actually three fold.   Here is the first stage of the GA-5 :

The Fender Champs (5E1 and 5F1) and original Skylarks biased at 1.5V.  Interestingly, the schematics from Crestline Skylarks (GA-5 and GA-5T) are annotated with voltages and despite the fact that these are exactly the same circuits up through the second stage, the ’5 shows a bias voltage of 1V while the ’5T shows 1.5V

I want to use a potentiometer here, to play with various values.   Note that setting the bias is not really an issue for 12AX7 / 6EU7s in this case.  The circuit is self biasing.    I just want to see if fiddling here would change the tone.

Additionally, I want to be able to switch the bypass cap on and off.   The 5E1 Champ had it but the 5F1 champ did not.  Finally.. and here’s the kicker,  I am considering a straight bias option.  Sending a fixed voltage to the cathode.   This is how tubes used to be biased, with a battery or other source directly supplying the right voltage.  So this would be a complex little bit of business, just so I could play around with biasing options.   Knowing me I’ll find a setting I like and never change it.   Dave Hunter’s Two Stroke amp has a three way switch with two different caps and a no cap selection.

But still, that’s not complex enough!   Next up an LED/diode combo to tell me when the grid voltage is higher than the bias voltage  (Actually when Vg + Vb >= 0).   No point in buying a lot of extra tubes if I can give myself a warning before I cathode strip them.

More Coming…

It uses a 12AX7 as a two stage pre-amp in much the same way that many other guitar amps do. The ouput of the first stage drives the output of the second stage with pot in between.

You can find the original article here and I recommend printing it or better yet storing a PDF locally.
As designed, it should be fairly cheap to build. That was a large part of the intention. Transformers are usually the most expensive part of any amp (outside of the cabinet) and this idea uses two standard ‘wall wart’ adapters (7-10 bucks) to power the heaters in a pretty clever two stage deal.

(Click on image for a better look) As you can see, the first transformer steps down from 120 VAC to 12 VAC. The heater power is tapped off and a second transformer steps back up from 12 to 120. A couple of bridges and some filtering and you get clean power. This ends up with about 140V or so at the tube plates. A pretty good value when you consider that the old GA-5 Only ran 165 to the plates of it’s 12AX7. Hotter is not more distorted, btw, in fact, the reverse may be true under some conditions.

So I went for the power section first. Cramming some components into a tag board and then dry testing. I had just about the voltages expected without load so I started soldering it all together. Unfortunately, I left the power on while I soldered the damn connection. Luckily for me, all I did was blow both of the cheap transformers. DOH! That’s a major fuckup though. I probably could have been hurt.

Forget the price. I’m gonna buy a hammond, maybe a 269BX (150 – 0 – 150) or EX (190 – 0 – 190) so that I can play around a little with higher plate values. Throw in a solid potentiometer and I could probably learn a lot about plate voltages and sound.

More as soon as I get the part.