The 6550 Experience

His guitar slung across his back, his dusty boots is his cadillac.

Moderators: VelvetGeorge, BUG

Post Reply
jazz1
New Member
Posts: 5
Joined: Wed Nov 08, 2017 5:08 am
Just the numbers in order: 13492

Re: The 6550 Experience

Post by jazz1 » Tue Jan 16, 2018 1:43 pm

daveweyer wrote:
Tue Jan 09, 2018 2:27 pm
Something to keep in mind; the lowered screen voltage means the screens will not draw nearly as much current during full power situations, therefore the resistor on the screens has much less current limiting effect. Bottom line is that any size will work, and will lower the voltage at the screen grid only slightly at full output power. It is handy to have a resistor there to act as a fuse should a tube short, and even though the tendencies for oscillations are drastically lowered with the reduced screen voltage, it is probably wise to have a stopper there. Also remember that the voltage developed as AC across the screen resistor is current feedback, and lowers the gain of the tube accordingly. You could probably get by with 100 ohm resistors.
In situations where the plate voltage was only 580, I usually derived the preamp supply from the HV--this is a personal taste thing, but higher preamp supply voltages give a different distortion characteristic downstream.
I'll be sharing the split supply Marshall Major mod I just completed here on the forum in just a few days, maybe there will be more West Coast ideas that you can play with, possibly even some I haven't mentioned yet.
Cheers to the mods!
Dave

In that mode of operation not only screen grids draw less current but power tubes plates to ,which indirectly results in lower cathodes emission and cathodes wear, that Significantly Extend power tubes life ,
my praxis is also to add 10 ohm /1W from each power tube cathode to the ground , except that those two small values resistors improve dynamic mutual current sharing ,they in the same time makes very good test points for biasing control ,
btw , I collect parts for one Marshall clone with four 6L6GC ( RCA- NOS ) and three ECC83( EI-NIS -NOS) , OPS will work with 500V/250V , and B+ supply for preamp stage will be derived from 500V line .

Best Regards

shakti
Senior Member
Posts: 2053
Joined: Tue Apr 05, 2005 9:06 am
Just the numbers in order: 7
Location: Ramnes, Norway

Re: The 6550 Experience

Post by shakti » Wed Jan 17, 2018 2:44 pm

I'm learning a lot here, thank you all!

Dave; as long as I make sure the preamp has higher voltages, is there any great advantage to having 560V plate voltage (and 280V screen voltage) as opposed to 530 and 265? I understand that the higher voltages yield greater power output from the output stage, but that was originally intended for loud stage amps through multiple 4x12s. These days we are all trying to get "the same tone" at lower volume. In a situation where we are using attenuators/re-ampers/post phase inverter master volumes to get a reasonable volume level, do we need the extra few volts? As long as the preamp (including phase inverter) is working at the higher voltages, can I expect any noticeable extra dynamic range from bumping the plate voltage further?
JTM45 RS OT, 1973 18W, JTM45/100, JTM50, JMP50 1986, JMP100 "West Coast", AC15, AC30, BF Super Reverb, Boogie Mk 1, Hiwatt CP103, DR103

Tek465b
Senior Member
Posts: 207
Joined: Sat Jan 10, 2015 9:43 am
Just the numbers in order: 13492
Contact:

Re: The 6550 Experience

Post by Tek465b » Wed Jan 17, 2018 5:47 pm

i use 1 Ohm cathode resistor, since 1ma = 1mv i can read bias current in millivolt directly.

If you want to experiment with B+ voltage you can try a VVR(variable voltage regulator)
Its a type of Power scaling.
Here is a schematic for a marshall fixed-bias type amp.
You must insulate the fet from the amp chassis.
Image

revolver1
Senior Member
Posts: 511
Joined: Thu Jul 28, 2011 1:37 am
Just the numbers in order: 7

Re: The 6550 Experience

Post by revolver1 » Fri Jan 19, 2018 10:03 am

Hey fellas I've been following the threads going over it and 're reading alot.... I'm learning all the time but by my own admission am not the most technically knowledgeable guy.

Fantastic info all thanks to Dave.

I really want to build my own West Coast modded amp but have a couple of questions.

First one is for the separate screens supply what type and value choke would be used?

Then what value filter capacitor would follow?

I'll post the questions one at a time to make it easier to get an answer.

revolver1
Senior Member
Posts: 511
Joined: Thu Jul 28, 2011 1:37 am
Just the numbers in order: 7

Re: The 6550 Experience

Post by revolver1 » Fri Jan 19, 2018 10:07 am

Sorry, but please answer Shakti first.

I'm interested in the response as well....

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Sat Jan 27, 2018 1:44 pm

Shakti, probably not much. Everything will work fine with the voltages you have. If you watch the scope of the output from the ECC83 input stage, you'll see it can handle anything a guitar can output without slamming into the supply. So, higher voltage on the supply in this case doesn't really change much. This is of course the reason for adding gain in front of the first stage, to get it to start clipping.

Now I'll be posting here soon, the rebuild of the Marshal Major, which uses a wave compressor to generate the distortion you want, all without needing to run the amp at more than 5 watts output. It's the WeyerPower solution to getting the output clipping sound, without actually clipping the output stage.

In this upgrade I'll also show the mosfet screen regulator implementation that will give you a reliable screen supply you can set at any voltage you need to get the output power desired, and desired output tube sensitivity.

In many instances, half of the plate supply is insufficient for optimum power tube operation, determined largely by the screen sensitivity of the tubes you are using. In the instance of KT120s, the screens prefer 400 volts, and draw the ideal idle current (18ma) at -56 volts on grid one. This set of voltages, 610 on the plates and 408 on the screens allows the tubes to output 239 watts, with a rather comfortable margin of safety. But as I said, you don't need all that power to get a great sound.

In looking at some of the other comments, there is also a question of what choke to use; I have used models from 2 to 15 henries with good results, but you have to make sure the screen current is not too much for the DCR of the choke. KT120s for example draw 160ma at full power. In the end, a mosfet is cheaper and does not add any weight to the chassis or require chassis real estate on top.
The cap after the choke will determine the ripple on the DC for the screens and anything else you run on that line. Screens are grids, and they do amplify AC, in some tubes more than others, so don't hold back--I often use 400uf at that spot. A mosfet in that application will have zero ripple, and the screen voltage will not fall when the amp is run into the rails. This feature gives more of the sound that having the screens at full plate voltage offers, but still protects them at square wave saturation. As stated before, smaller screen resistors are fine with this system.

Some small resistors on the cathodes, even 1 ohm, will protect the tubes from destruction if something goes wrong, help balance the tubes, and allow you to set the bias voltage easily by measuring the idle voltage across the resistor. In the Major rebuild I used 3 ohm units, and summed those together into a 2 amp fuse. The HV fuse never blows on the Marshalls because it is between electrolytic capacitors, so using a cathode fuse is a good idea for protecting the output tubes.

Should I post the Marshall upgrades on this thread, or on the West Coast thread?

User avatar
Xplorer
Senior Member
Posts: 2470
Joined: Sun Apr 19, 2009 5:27 pm
Just the numbers in order: 7

Re: The 6550 Experience

Post by Xplorer » Sat Jan 27, 2018 5:55 pm

Here ! From Dave, a beautiful Marshall Major, what a clean work !! I love it



https://imgur.com/Nte5WNp

https://imgur.com/kVOaZK4

Upgrade schematic :

https://imgur.com/lwBeWMP

i wish i could make them visible here on the page but the forum keeps telling me : "It was not possible to determine the dimensions of the image. Please verify that the URL you entered is correct."
even if i decrease down to 800 pixels large. Can't attach the files either.

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Sun Jan 28, 2018 1:00 am

If you have the time to check out the links Xplorer posted you can see a couple pics of the Marshall Major, including its old West Coast sticker, and you can see the entire schematic too. This was an original mod project from back in 1969, but the amp had been through a number of changes since then, and apparently a blowup too, as there was quite a bit of carbon and blackened tube sockets. This model must have been from late '67, because it had the 250 ohm screen resistors, the polarity switch, and dual cathode resistors on V1, all the things you see on that Marshall schematic floating around on the net and in Aspen Pittman's tube book that show the hand-written notes on various parts of the plan. This is the oldest Major schematic currently available, unless any of you know of an older one.

The new upgrade took some of the West Coast concepts and added to them with things learned throughout the years of amplifier building, and with things not available back in the '60s, like a 1000 volt mosfet.

Briefly, so I don't get to droning on about it to a vacuum, there are some extra stages: Input is to the conventional ECC83 input stage, but following that stage there is a 6KZ8 pentode/triode tube, the purpose being to get the gain way up to drive a wave compressor circuit. The first two stages give a gain of about 14,000, so there is 350 AC volts to drive the wave compressor. Out of the compressor there are only a few volts available, so a dual 12AT7 stage is used to amplify by 40 in order to drive the tone stack. The tone stack has been moved to right before the PI, and the power amp section has been reduced to a 12AT7 gain and inverter stage (called the floating paraphase system), a 5687 cathode follower power amp to drive the output tubes and drastically lower the input impedance of the output tubes, and of course lastly, the power tubes. The driver of the tone stack is a feedback amplifier set up so that part of its feedback may be shunted to ground by a capacitor in order to put a broadband high end "sheen" on the signal. This substitutes for the presence control which was eliminated, at least electronically. The other 12AT7 is the wave compressor, and it does a number on the waves passing through in a process called instantaneous wave compression (unlike a regular compressor, there is no time element, every instantaneous rise in voltage is met with an instantaneous amplitude dependent reduction in voltage, the result being a pretty good simulation of a power amp in full overdrive). The master volume control is right before the 12AT7 feedback amp, allowing super crunch at any volume. The crunch control adjusts the amount of the signal subjected to the wave compression process, done in six steps with a six position dual rotary switch. The other side of the switch is a six position attenuator which keeps the level between positions even as you go from low crunch to high crunch.
The screens are supplied by a passive mosfet regulator which holds the voltage at 408 under all conditions. The power tubes (KT120s) are biased at -56 volts. The mosfet circuitry can be seen at the very end of the amp chassis, the mosfet using the chassis as a heat sink, but electrically insulated from it. The original screen taps on the output transformer were disconnected from the screens, obviously, since that seems to be the biggest complaint on the sound of the original amp.
There is a hum balance control to eliminate filament buzz from slightly imbalanced heater current, and there is also an 85 volt bias on all the tube heaters to keep the heater/cathode insulation rating in range on the cathode follower stages.
The output tubes are protected with 3 ohm resistors which sum into a 2 amp fuse. (Right up front on the underside chassis pic.)
If you see something you like, or have any questions, please fire away. I hope to put up a video on youtube demonstrating the amp and the various sounds it can make, so if that comes together, i'll also post a link to that.

Tek465b
Senior Member
Posts: 207
Joined: Sat Jan 10, 2015 9:43 am
Just the numbers in order: 13492
Contact:

Re: The 6550 Experience

Post by Tek465b » Sun Jan 28, 2018 12:09 pm

Wow, that is nice engineering.
I like the Screen regulator/supply, something i want to try definatly.
Do you have more detail about the wave compressor module?
I really like the DC elevated heater with hum balance. Am thinking of adding it to my amp too :)

niknik
Senior Member
Posts: 82
Joined: Sun Jan 14, 2007 11:54 am

Re: The 6550 Experience

Post by niknik » Mon Jan 29, 2018 11:53 am

Dave,
This is a really awesome project!
I've always been intrigued by Marshall Major models, but I never had a real one.
I just finished a mini Major, with just 2 kt88, an original Dynaco A-431-70 ot, and a new P782 S Power Transformer from Triode.
I'm about to start to study the schematic Xplorer posted, hoping to be able to understand.
What transformers do you suggest?
Thanx, this is a wonderful place! :D
Sammy who?!

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Mon Jan 29, 2018 4:15 pm

The main thing about the Major is to get away from the screen taps and run the output tubes in pure pentode mode. Also to get rid of the tone stack in the feedback loop of the power amp, you wouldn't believe what weird things that does to the signal when the amp is under stress, pretty scary, and unstable.
I think Marshall had it right when they followed the 59 Fender Bassman, and put the tone controls right before the power amp section--all preamp distortion gets "toned", for much more versatile sounds.
If you have some bucks, Souter makes a 250 watt output tranny, and some smaller ones as well.

And while I'm here, I have my doubts about a post PI master volume, because that puts it in the feedback loop of the power amp. Just add one more stage before the PI, and put the master volume there. Once again, the feedback loop of the power amp section is important for stability reasons, that's why I eliminated the presence control--you should see the ringing on the square waves when the presence is turned up, which is what happens when you are shorting out the upper feedback frequencies. One more stage is so simple to add, in fact you can use the second input stage.
Also, if possible, drive the tone stack with a lower impedance, like a cathode follower, as Fender originally did. I used a 12AT7 with both sections connected in parallel, and then added a feedback loop around that to lower the impedance. The tone controls are most effective when driven by a low impedance.

One more hint, preamp distortion sounds best I think, when it is done in push-pull, like the output stage. Among other things, there is a reason that power tube distortion is so desirable. If you've played much with Mesa amps and their extra preamp stage distortion, you would know the difference pretty well.

Clear as mud I am sure........

niknik
Senior Member
Posts: 82
Joined: Sun Jan 14, 2007 11:54 am

Re: The 6550 Experience

Post by niknik » Mon Jan 29, 2018 7:51 pm

Is this Sowter type U071 ot, correct for this project? It's a push pull Output transformer for 2 pairs KT88 Fixed bias.
2250 ohms Rla-a.
200W at 20 Hz. For 2 parallel pairs KT88 or 6550 GEC circuit. 250 mA per side. 10 Hz to 60 kHz. Size Q.
Is it good also with 4 kt120?
Looking at the schematic, v1 is pretty stock bass version, shared chatode bypassed by a smaller capacitance cap, 100uf, then two100k res plate, two .022 coupling caps, and a 500p in parallel with 470k mixer res in channel 1, touching also the input of the volume pot.
Tone stack and eq section is classic bass spec, plus 2 series zeners facing on, bypassing a 1m res going from wiper of the treble control to ground. Tone stack is platen driven from a 12at7 with plates and cathodes shared.
This 12at7 is just after the master volume, coming from this wave compressor, which is also conected to the cathode of second stage 6kz8 tube (is it in a cathode follower arrangiament?). And so many things that I don't understand completely.. :oops:
What rfc is? I'd to learn more about this wave compressor.
It's a diffucult schematic for me, I not a tech, but I think, with patience, a lot of passion and determination, and also with some help, I could do it! Maybe.. :D
I'm really into this great project,
Thanx to you, dave, and thanx to you all
Sammy who?!

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Mon Jan 29, 2018 9:26 pm

The 6KZ8 pentode runs directly into the triode section which is used as a cathode follower to drive the wave compressor. Once again, driving things requires some power, and a cathode follower is an extremely simple way to get some driving power. Because of that it can output more voltage than the plate of the tube would be able to on its own--it gets loaded down by the impedance of what it is driving.
The back to back zeners on the PI/gain stage are to limit the input voltage to the power amp section to about 8 volts, otherwise the KT120s can output over 400 watts into an 8 ohm load, not good for speakers, or other components.
The dual 12AT7 is a feedback amp, used to get both gain and power, enough to drive the tone stack and still have 7 volts left after the network to run the power amp section to 230 watts. The feedback gives you the option of the presence control, just short out some of the HF in the feedback signal and you get an HF boost across a wide band; this lets you keep the power amp feedback path intact under all conditions, helping to control snivets and ringing.
Driving the tone stack with a little power, and a lower impedance makes the controls work better, with more treble, mid and low range. I cut the low range pretty sharply in this amp, because guitars don't really need that much low frequency information, it actually gets muddy and causes some "ghosting", to use the term presented here several times.
The Souter transformers will work great with KT120s; they appear to be a wonderful replacement for the KT88, but even sturdier by some margin. 2Kohms is fine for (4) KT120s.

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Mon Jan 29, 2018 9:38 pm

And by the way, an RFC is a radio frequency choke; it is used to control high frequency oscillations, which can easily occur if you are using high gain tubes which were designed for radio frequency operation, the 6KZ8 generally used in television tuners as a mixer oscillator. The reason I use it is because the microphonics are very controlled in the pentode (better than a 12AX7), and it sounds wonderful for audio--plus you can get a gain of 300 easily with a 150K plate resistor.
There are a couple of chokes in the plate circuit of the outer KT120s; these are used to control snivets and parasitic oscillations, and, to cause the outer tubes to have a different reactance profile than the inner tubes.

niknik
Senior Member
Posts: 82
Joined: Sun Jan 14, 2007 11:54 am

Re: The 6550 Experience

Post by niknik » Tue Jan 30, 2018 7:24 pm

Thanx Dave,
It's a little more clear to me. I need this Sowter ot, and a 610v b+ PT, with quite enough mA to handle filaments of 4 kt120 plus 6 preamp tubes. Any suggest? Mercury Magnetics Major pt?
It's an expensive project, so I have to procede little by little, buying components one at a time. I'm really into this wave compressor module, is it an original stuff by you, Dave?
Sammy who?!

Post Reply