Lowering the noise floor in a higher gain amp

[Tone Junkie]

I was just wondering if there were any tricks I could use to lower the noise floor without robbing the sparkle off my top end. I understand how important good lead dress and and proper grounding are I have those covered fairly well. I just wondered if there were any other ways to lower the noise floor on these. I built the YJM mod as they called it on the metro or sloclone forum I forget which. And I have a quickrod clone I built to compare it to. the Quickrod is one less gain stage . and its only a hair quieter. I do have elevated heaters . but mayby I possibly should have used DC heaters instead of AC I dont know Im just wondering I know this isnt neccessarily a high gain forum but you guys never flame me for dumb questions or just ignore me because Im such a newbie.
Thanks Bill   

[JayB]

It's really about lead dress and proper lay out. I build high gainers and they are just as noise free as any other amp.

Another thing is to use the highest wattage plate resistors that makes sense. I use 3watt metal oxides (some may think is blasphemy).

DC heaters and elevating the heaters is to much of a hassle in my opinion. Done right, there is no need for it.

[phsyconoodler]

Level 2 eh? You newbie!

  I ran across that issue in a Soldano amp.They have so much gain on tap it's hard to control the noise.The trick is not too much gain per stage.They cold bias the stages and add more stages to get the amount of gain.
  The lead dress is extremely important.DC heaters do little to tame the hiss you get from cranking the gain.Hum yes,hiss no.
  The large plate resistors are a really good idea.
I avoid all that by not using high gain amps.I use a pedal to get my hard rock tones.(too lazy)

[Tone Junkie]

JayB thanks for the high wattage resister on the plates I had never heard of that before. I tend to always at least now elevate my heaters because of the problem we have had with some of the new production 12 ax7,s out there. phsyconoodler I just ask a lot of questions cus Im dumb so I have a lot of posts, but if I keep hanging out with you guys that wont be the case for long.
                                                       
Thanks Bill

[labb]

If you are using carbon com pots give the conductive plastic element pots a try. More expensive but quieter. I have used the Vishay/Dale RN series low noise resistors for the plate resistors with some luck..If not then the higher the wattage rating the better.

[Tone Junkie]

Labb whats a good brand for those pots or a mouser part # would be even better.
 Im all for using better stuff if i know there is an actual need that is worth some extra cost . so are they quieter or what this is something else i havnt seen talked about. Thanks Bill

[DummyLoad]

http://www.newark.com/honeywell-s-c-clarostat/rv4laysa105a/pot-cond-plastic-1mohm-10-2w/dp/01F2665?Ntt=RV4LAYSA105A

part # is RV4LAYSA105A  105A is 1Mohm Audio taper.  10 + 5 zeros = 1M

so then part # RV4LAYSA254A   254A would be a 250K Audio taper.  25 + 4 zeros = 250K

http://www.newark.com/honeywell-s-c-clarostat/rv4laysa254a/pot-cond-plastic-250kohm-10-2w/dp/01F2663?Ntt=RV4LAYSA254A

both in stock - not a bad price for a 2W conductive plastic element, mil-grade pot.

--DL

[labb]

for a good read on resistor noise go to www.aikenamps.com     open tech info, under technical papers go to advanced, go down to " resistor types-does it matter".

this statement is from that paper...Don't forget that potentiometers are also resistive elements, and are almost always carbon composition, and generally are large values (such as 1MEG for the volume control).  These can be a major source of noise in a guitar amplifier.  For absolute lowest noise, conductive plastic element pots should be used, again, the lowest practical value, and the largest practical power rating.

Weber, www.tedweber.com  has them, not sure what brand. I usually go with the Honeywell from Newark.

[tubenit]

Layout and lead dress is important. I prefer quieter metal film resistors over carbon comp or carbon film. I realize this is totally a preference thing.

I owned a Fender '76 Princeton Reverb for a few yrs. So that will be the reference standard.

The Tweed Overdrive Special and the Tweed BluezMeister using 6V6's or 6L6's (I can use either in either amp) has NO more noise at idle then the PR did when the volume of the TOS and TBM is set comparable to the volume of the PR. And that is with either the clean or OD channel engaged.

In a high gain amp there will always be some floor noise with the OD engaged. However, I could always hear floor noise with the Princeton Reverb and even the Fender VibroChamp amp I owned.  

Many of the scratch built amps I've done were so quiet I could not hear them at all at idle & when I first cranked them up that first time ........ I thought they weren't working until I strummed the guitar. I can remember checking the speaker connections on several thinking a wire had come loose when it hadn't.

With respect, Tubenit

[tubeswell]

Anything you can do to lower the noise:signal ratio, especially at V1 is good. Getting a good strong signal at V1 is important. Apart from what the others have said about resistor size and type, going with parallel stages and using half-size plate and cathode resistors, lowers SN but keeps the gain the same (and also lowers the stage's output impedance).

[jerrydyer]

on the CCV 's by Mark Cameron ( i was trying to help him for a while) instead of going straight to ground off the 1meg gain and volume pots etcand a few other locations,  he filtered it thru a .22 cap. So all thru the circuit you would see these .22 caps to ground.

Also using like an old hum balance control like on the old fenders helps on the heater circuit in higher gain amps. ( iahve a cool mod but i have to remember who gave it to me and ask his permission. I think its a hoffman guy here) ref heaters to b + thru similar circuit) 

once again the layout. Ive had to change my entire chassis layout from my first design. Sheilded cable, lots of sheilded cable. 

[Tone Junkie]

Jerry if your talking about the one that goes from what you would normally ground on the humdinger to your negative B+ referance I have a schematic with that on it I just havnt tried. I have an even more interesting one were they made the front 3 tubes DC then referanced it through the humdinger to B+ for a higher dc referance on dc which thoroughly confused me but they said it was an old High Fi amp trick. havnt tried that one either.
 I kind of like to research new radical things in my world , Im new enough people could be be messing with me  I burnt up a PT when really green because I would just try what ever people told me . I could see in my meager understanding please correct me if Im wrong that the caps would filter some high end hash off of the top maybe. In my latest high gainer I used merlins elevated grounding schem. I can say its quieter than the quickrod I built with one extra gain stage. Im sure my lead dress and other things could be better but Ive tried getting better on each one. I have a friend with a 5150 and his is as noisy or more so than these but there has to be a better way thats why I asked. I will be happy when I can make one as quiete as my TBM or the TOS was and yes I understand 2 totally differant beasts but I can dream.
Thanks Bill

[jojokeo]

Rather than posting my own ideas on this one I'd rather post something from someone that's respected but is also MIA or I'd ask him. But, since it's already in the public domain it should be okay anyway. I liked most of what I read as I've done or agree on many of the same things.
Here's part 1:

My builds look like rat's nests inside, but are silent as death... like all the old HI-FI gear that was even messier, and even quieter. Radiated fields will kill you... squeals galore. Shielded wire will help, but it is not a cure all. Everything is exacerbated with higher gains and more stages, so you have to be much more deliberate in your designs.

My tips...
+It starts with a good power supply: -Get the ripple out! I usually regulate, and then bypass the supply nodes close to the pins with a .1uF CERAMIC capacitor to shunt any noise picked up along the wire run to ground.
-Make sure your rectifier diodes are not radiating RF
-use fast, soft recovery diodes and snub them, or use a tube rectifier (I do in almost all mu builds now).
-In addition to regulating, use a CLC power supply stage first. Just do it. Capacitors' effectiveness decreases as the impedance increases... that's the ESR problem, but chokes are the opposite, so they compliment each other and work better across the board. They also by their nature, convert sine AC ripple to a more saw tooth shape from their ramping up/down action, so the regulator will more effectively cancel out the ripple.
-MIND YOUR CURRENT LOOPS! Don't use crappy multi-section can caps. Use separate capacitors, and place them near the stage that they are filtering and keep the wiring in the loop short. I use 1 cap per triode if they are opposite in phase, if you have two cascading sections that are the same in phase, then separate their supplies.
-Keep the AC input and rectifying part of the supply HELL AND GONE from the preamp circuit... all the way in the opposite corner of the chassis. Keep the loop around the first filter cap as short as possible so that it does not become a quarter-wave antenna.

+Filaments:
-Elevating AC filaments works really well for getting the ripple noise at a different potential than the cathode, so you are not shooting AC noise into your circuit, but you still have AC traveling around your circuit, which will radiate crap for your sensitive grid nodes to pick up. A better solution is to use a good REGULATED DC supply, with a choke filter and then also elevate it, AND use shielded twisted pair wiring. This is not necessary, but it will eliminate completely the chance of filament interference. BUT I must stress that it has to be a GOOD regulated supply, or else you will be only making things worse.
-Keep your filament wiring as short as possible, and use solid wire so that you can make all your bends at 90º angles... loops are RF antennas! So keep the bends sharp. Keep the filament circuit away from the other wiring. I always send the signal wires in on one side and the filaments on the other. I also use 12v filaments so that I can only have to approach the tube socket from one side, rather than wrapping around to the other side to hook up to pin #9... and as a bonus, you'll need less current. And lay the filaments against the chassis for extra shielding. -Even with DC filaments, bypass each tube's filament to ground right on the tube pin with a .1uF CERAMIC capacitor to kill any picked up stray noise in the heater wiring.

[jojokeo]

Part 2:

+Grids:
-Always always use a grid stopper, and always put it right on the pin. Always make sure that the leads are copper, not copper clad steel, and make sure that the resistor end caps are copper and not zinc... and make sure that the composition of the resistor is not inductive. I use good carbon comp stopper resistors... always...always. If you don't think it is important to use a magnetically inert, non inductive resistor, then just wire up a circuit with a crappy resistor and then put a magnet close to the grid resistor, then do the same with a nice carbon comp... you'll see what I mean. With all the stray fields and cell phone interference around, it becomes very important.
-Make sure that the grid connection is away from ANY other signal carrying wires, or power wires, and especially AC carrying power wires. Make your layout so that NOTHING has to come anywhere near these. I keep the grid wires running to the socket at a radial angle that is perpendicular to the tangent of the heater wires on the opposite side to completely eliminate the possibility of coupling.
-If you do have to cross the grid wires for any reason, keep as much space as possible between the wires and only cross them at right angles

+Transformer layout:
-Orient your transformers in a 3-dimensional plane. Look at the radiating field from a standard E-I core tranny, and orient your three trannies on 3 axis...X, Y and Z, then there can be no coupling. You can keep some distance between them, but if they are oriented properly, I have snuggled them up pretty close with no coupling. More important than keeping them away from each other are to keep them away from the preamp and phase inverter! Put them away in that corner where your rectifier circuit is, and then space them out according to the room you have available.

+Grounding:
-I run a buss/star ground that keeps the plate supply nodes, grid nodes and cathode nodes in loops around each circuit. It is a star ground that is kind of stretched out. Then reference the preamp buss to ground at the input jack. The rest of the circuit should practically lay itself out at this point.

+Other considerations:
-Don't use more gain than you need. Why generate uber gain from a section, and then dump it away with a voltage divider? Re-work your operating points so that you get the clipping that you want without having to crank up the gain so much.
-Keep the grid resistance low. Big resistors on the grid will only increase noise and raise the impedance of that node, making it more susceptible to RF and hum. You can bleed off treble in ways other than miller capacitance roll-off with HUGE series resistance. Just do the math to get those huge resistors out of the signal path. I usually add just enough series resistance to get the highs to roll off from Miller effect just outside the audio band, and then use plate capacitors or grid capacitors to tailor the high end.

[Fresh_Start]

Anything you can do to lower the noise:signal ratio, especially at V1 is good. Getting a good strong signal at V1 is important. Apart from what the others have said about resistor size and type, going with parallel stages and using half-size plate and cathode resistors, lowers SN but keeps the gain the same (and also lowers the stage's output impedance).

This may be painfully obvious, but noise at the start of the preamp is much more audible than noise later in the signal path.  Parallel triodes for the first gain stage probably helps lower the noise floor. 

I'm not about to spend over $10 on each and every pot in an amp; however, using a conductive plastic pot(s) for the first gain control may have a good cost/benefit ratio.

Don't know who Keo was quoting above but my limited experience leads me to agree with most of it.  I got thoroughly flamed years ago on the 18-Watt forum for simply asking a question about using shielded wire for heaters.  It makes sense to me and I'm trying an experiment in my next build.  Elevated DC reference for the virtual center tap made a noticeable difference in my single-ended build - and it wasn't even high gain.  I don't understand the details of the ceramic caps in the power supply stuff.  Also, a C-L-C filter before the plate node on the power rail seems like overkill for any push-pull power amp.  But what do I know.

I've been happy with the noise reduction using 1 watt metal film resistors compared to other types.  YMMV.

Remember the first rule of AX84:

Don't Bozo the layout!

Cheers,

Chip

[Tone Junkie]

Does any one have 12 volt dc referanced schematic so I could see how to do it. I dont quite understand how you hook it up.
Thanks Bill

PS. I now how to do a dc referanced type but dont now how to do regulated 12 volt dc
I will have to reread that one above a few times to understand the .01 cap to ground on the pins of the tube

[TIMBO]

Hi guys, It seems to be a common problem that layout ,lead dress and heaters are the buggbear of amp builders and everyone has their own ideas.I too have my own ideas but i may have dodge the problems up to my last build.After finding the problem by elevating the power supply for the relay the problem fixed.I also elevated the heaters and as for lead dress, layout and grounding i left it much the way i originally had it.The amp is very quiet the only noise i get is when the OD is at max i get a kind of white noise (this i can live with).

If i may, their are a lot of you guys who are much more experienced than i and have better ideas on this subject and if we were to have a group contribution of ideas this would give us new guys a much need point in the right direction.In the construction industry their is a saying "if your foundations are correct the rest will follow".

Could we start with a schem and add to it bit-by-bit and possibly agree on the best ways to achieve the goal of "QUIET"

The schem below is "TEST SUBJECT" if we were to agree on the first step to build this and treating it as a HI GAIN build what would be the first thing we would do?

The parts are run of the mill OT, PT , chokes etc.
The chassis is not much bigger than the board shown.(being the main problem of problems)

I hope this might spark some interest and build a platform for everyone to benefit from ( also reduce the same old same old questions that come up all the time.

A lot of the time i find it hard to understand the technical info you guys supply and find it far easier to follow when it is drawn as a plan, so by having a MUD MAP we can see what you are saying.

So  that the layout is a good place to start (tube sockets,transformer placement and orientation etc.)

Look at the schem as if you were looking at the board inside the chassis.

Thanks



 

[overtone]

Being a visual guy I think this is a great idea and now at last understand the short ground connection between the input to the V1 cathode resistor!

I have a suggestion: it will really help everyone if we name and date our revisions. Later on it would be good practice to mark wrong versions accordingly, or as superseded.

Where I am still unclear, in this kind of a layout, is where the filter cap -ve poles (grounds) are tied into the two busses. I tried drawing the filter caps on the board and as I am ill in bed I just messed it up. (I cannot undo on a MAC and I made a clumsy click and drag only to end up with a spaghetti layout.)

Here is my stab at it. I have made my additions and questions in green for clarity and a more schematic way of drawing the send return jacks. I started drawing up the other filter and reservoir caps, the speaker ground etc, but it all became  tagliatelle. I'm calling it a day, back to bed with a hot tea, over to you all with a parting question:
I use isolated jack sockets and run the ground lugs of the loop send return to the negative pole of the C stage filter cap because the loop is after the 5879 coupling cap. Is that OK? Please also comment on my filter cap negative pole tie-ins to the ground busses,


best, tony

[Fresh_Start]

Timbo & Overtone - I applaud your efforts. 

However, there is no one "correct" approach to an amp's layout.  Hoffman's layout approach and grounding methodology (see Library) clearly are proven as known-good but they're not the only way to get a quiet amp.  Leo Fender figured out layouts that worked and could be mass produced - they aren't anywhere near as random as they looked to me at first.  At the other extreme, some of the high-end, point-to-point builds I've seen photos of approach art to my eyes. 

You could take that sample circuit and implement point-to-point, turret/eyelet board or PCB construction and get a great result.  One builder might use a ground bus, another star grounding and the third a ground plane and all three could work.  Some guys use tons of shielded leads and others figure out how to minimize shielding.  DC heaters?  Seems to make sense on a PCB build, at least for the first 1 or 2 preamp stages, but is it really necessary on a well-done turret board build?  I don't know.

Bottom line:  copy a "known good" layout.  Figure out why it works.  Chopstick and diddle with wire arrangement until you get a feel for what works and what doesn't.  Look at as many gut shots and layout diagrams as you can find.  IME layout is part of the art/craft of amp building.  Sure there are scientific principles, but there's also an element of ingenuity and craftsmanship.  That's part of what makes building amps fun for me.

Hope this helps,

Chip 

[TIMBO]

I understand that every one has their ideas and procedures. This is a way of cutting through the mountains of info on the net ,pictures, builds and so on. Its to get thoughts on paper to make it easier to construct these amps (when shown the right way, not that there is really a right or wrong way but there is a better way) This is just away of addressing different parts of the amps construction.I do not believe that we will agree on some aspects of the build but it will be food for thought.

I was thinking of making a list of the MAJOR steps that are taken when planing to build.(Steps in the order that you would do when constructing)

1) Type of amp to build - Schem provided
2) BOM                      - Hoffman Supplies
3) Chassis Type           - Type of amp to be constructed

I know that this might take some of the fun out of building these great machines (there is learning when you have a problem and you have to try and find it) but there is also frustration and bad practices.

Fresh Start is right, some parts will be left of the builder to decide what method they prefer to adopt ( Turrets, eyelets, busswires,star ground etc.)but at least this will show that there are options.

Please feel free to add to the list.Thanks

[Fresh_Start]

I'm not trying to be difficult, but some research here and on other amp boards (AMPAGE, AX84, etc.) will yield better results than trying to re-invent the wheel.  Some authors like Kevin O'Connor (TUT 3) already have discussed build details at length and systematically. 

This is cooking, not chemistry.

Sample layouts:
AX84 Single Ended Lead

Ceriatone 2204

Trinity Amps Tri-Watt

Brown Note's D'Lite

Hoffman board building instructions

A modified AB763 using my take on Hoffman's layout & grounding scheme (preamp ground bus on board & distributed power supply filtering)


A big problem with all of these drawings is that they are, by definition, 2 dimensional.  It took me a while to understand that we have THREE dimensions to work with when laying out wiring, etc.  For example, it helps to put high voltage plate leads down on the chassis as soon as they leave the board.  That gives you more room to keep sensitive, low voltage grid leads isolated and in the air.

A gut shot of a really cool point-to-point build is attached.

Respectfully,

Chip

[jojokeo]

I really dislike the way all of those yellow zip ties are laid out

[Fresh_Start]

I really dislike the way all of those yellow zip ties are laid out

IIRC that amp builder is color blind so you'll have to forgive him!

Chip

[jojokeo]

You don't happen to remember who he was do you? Did he like aluminum foil perhaps?

[Willabe]

Hey, I think I remember that guy, but it was only Blue colored tin foil on his head!       


      Brad           

[Tone Junkie]

Hey you shouldnt make fun of my hats like that they keep the aliens from taking over my thoughts dont you know, and the blue tin foil works the best.

[Fresh_Start]

You don't happen to remember who he was do you? Did he like aluminum foil perhaps?

Yes, I do remember who he is.  He is color blind.  That's why he likes those jumbo brown resistors with their values printed on them that you see in the preamp - he can't read color bands on resistors.

That layout style is called a Gothik Ring, if that rings any bells...

Chip

[jojokeo]

Hmm, my initial thought was of a belly dancer w/ a penchant for Ouzo but I think I may have an idea?

[Willabe]

Hey, jojo,

Is that you in the pic sitting with a Flying V in your arms? Way cool!


      Brad     

[jojokeo]

Na, that would be the man of our subject, but I do have a really special custom flying V of my own straight from a guy w/ the initials EVH.

[TIMBO]

Flying V, love those guitars  Thats not a flying V. Anyway, Fresh Start those are great drawings and are great on paper.When it comes to do the build it becomes a nightmare. Its what you have in your head (things that become second nature to you)that we want in front of us when we do our builds.

A couple of things, those layouts are known builds that work,but its the info that you can't see is what i am getting at.We have said that the LAYOUT is the foundation.Its how we get the foundation, do we start with the circuit board and arrange the components in such a way that gives a direct connection to the relative tube and avoid wires crossing each other and are able to be earth and as well as not being too close to other components that might react to it. 

Or punch a heap of holes in a chassis and then try to design a board that will fit and can service the arranged tubes as well be not too far from things like the OT etc. 

I was told that some of the problem with the Silver Jubilee was lead dress and layout.The layout is pretty much standard but by the time i added the reverb and a few other things it became a rats nest and most was unavoidable as space became tight.I looked at your take on the AB763 again it looks great on the layout but you would still find that things could get out of control in the wiring 

So this leads me back of how do we start these builds if we do not know where to start ( scratch are even harder)

Sorry for the RANT. Thanks

[38Super]

I don't understand the details of the ceramic caps in the power supply stuff.

short answer: A 0.1 or 0.01 high quality cap across each of your filter caps supresses high frequencies in your power supply, because the electrolytics are imperfect due to ESR and ESL (can look inductive at higher frequencies).

long answer: All capacitors have ESR, the effective series resistance of the device, and ESL (series inductance) due to their physical construction.  When you build a lowpass RC filter, there is a pole (corner) created where the frequency response is -3dB with respect to the passband frequency.  It occurs where the capacitor's impedance 1/(2*pi*C) is equal to the series resistance that is "sees", e.g., an actual series resistor or impedance of a tube rectifier.

The wrinkle in this scheme is the filter is not perfect, primarily due to the ESR in the filter cap: this series impedance creates a zero (opposite of a pole or corner) which causes the frequency response to break flat again - because it counters the pole created by RC filter.  Think of very high frequencies where the capacitance is near zero, the filter series R and the ESR create a simple voltage divider, which creates an fixed attenuation, rather than an ideal slope.

So the 0.01 to 0.01u cap that you place across each electrolytic bypasses the ESR, because at higher frequencies, 0.1u impedance is less than the electrolytic's ESR.  That is how it is supposed to work, anyway.  Effective series inductance has effects similar to ESR.

Bypassing your filter caps with 0.1uFs is good practice, used extensively in commercial power supplies, but for some reason, not so much in factory made guitar amps.


cheers,

rob

cheers,

rob

[jojokeo]

Nice Explanation 38Super, thanks.

Timbo, it takes me longer to design everything from wants and needs, then to schematic, then finally to layout in that order - than the actual building phase. Once all is in order then the building (the really fun part) goes real fast. It depends on my amp design but I'll start w/ more of a high gainer since grounding and layout are more critical. Not only do I do the normal schematic and actual real layout designing but I also may do the actual schematic and layout for the entire grounding system separately. I highly suggest this for those that really want to take all precautions and get it right - the first time when new to building. Don't wait until you've started building, reading the schem and layout as you go...then forget or maybe not even know where or what and make corrections afterwards? My layout design has filter caps closest to all of their respective stages, the resvoir cap's ground has the PT's CT going straight to the neg end w/ short lead length of heavy guage wire, etc...this stuff is all accounted for in the regular layout and the grounding layout also(if necessary).
When deciding where and how the actual layout of the circuit is going to go, you are correct in your thinking that the entire circuit should flow from one direction to the other, from input to output w/out anything crossing over each other if possible. IF it does keep distances or use shielded wire. There's times where multi-strand shielded wire can be used also. Don't limit yourself or your creativity.
I actually decide my chassis size by the number of pots and tubes used w/ care of tranny amounts and sizes too but don't forget the filter caps here as well.
Now, I'll either place my PT & OT on opposite sides of the chassis ala Mesa Boogie and many others or if they need to be closer I use a simple "hearing amp" to know where and how to place my tranniies to each other. I've learned that this helps rather than guesswork and thinking that simply placing them 90 degrees to each other is simply okay - it's not (if you care about the least amount of floor noise possible).
Now, I decide if what I'm working with is going to be eyelet, turret, or tag strip. I actually draw on paper if necessary and it's a tight fit, all components as close to actual size as possible. I'll cut out the drawn tag board and place it inside the chassis along w/ the trannies and whatever else to get as close to an actual mock up as possible. Have you ever seen American Chopper or similar? I build my amps a bit in the same way fitting everything together as I drill and mount all of the components as they warrant it in this process. I account for all tube sockets, local grounds, tag strips, jacks, test/bias jacks, you name it so that all will fit nicely and everything is accounted for.
After all of this, I disassemble what needs to be, all holes are cut, drilled, sanded, reamed, etc. Then it's time to paint or powdercoat the chassis put it all together for real and it all fits like a glove w/ no guesswork what so ever. This is where all of the time consuming and painstaking planning and designing pays off and makes the actual build easy and enjoyable.

I know I've missed a few items, got a little off track w/ the building phase info, and not gone into too great of detail on some stuff but these are the kinds of things I plan for and think of. I am just to trying to help give one person's way of doing things that goes into this process of combining everything with the ultimate goal of not only working loud and proud but also being very quiet when not playing too.

Hope this helps? Keo

[Tone Junkie]

There is some really good stuff here I love the info in this Im in the process of working these things out and printing some out to go on my wall of important info. so lets look at one of the most important things GROUNDING what are your thoughts. the second amp I built was a plexi 50 watt it was kind of noisy then I changed the grounding to what was at metro known as the larry grounding schem, novosiber was the gentlman who infented it for his marshall builds it helped my noise level greatly. In this last one I incorpurated the ground schem shown to us by are great friend Merlin Blencoe I hope I spelled that right who has given us some great books to read lately. what are your guys thoughts on this subject . Oh I dont want to leave out Hoffmans grounding schem that helped me greatly in an ab763 build. but here is the way i set up the board on this last one using Merlins scheme.
Thanks Bill
                                                             

[Tone Junkie]

I should have added that my grounding was the 2 turrets on each side of my cap so I can ground the differant stages as close to each cap as possible.

[TIMBO]

Good stuff jojokeo, What you have said makes a lot of sence. This is the sort of stuff i was looking for and i'm sure the Level 1s will lap this up because there is a lot that goes into building these things.

I can relate to what you have said, my procedures are much the same (unfortunatly the early days of my building saw a lot of failures)but i still think that there is always things to learn and better ways to build.

Tone, board looks great.Thanks 

[Tone Junkie]

jojokeo could we talk about some of the things in that great paper you put up. Like how do we snub our rectifier diodes is that were we put a cap in parallel to them, and what is a clc power supply stage, he also says to bypass the supply nodes close to the pins with a .1uf ceramic cap. these are things I dont quite understand Im going to aiken to read up right now.
Thanks Bill

[Fresh_Start]

Bill  - a C-L-C filter is cap/choke/cap.  C-R-C is cap/resistor/cap.

Personally, putting snubbing caps on rectifier diodes seems like overkill in a guitar amp.  If you want something "better" than the ubiquitous 1N4007 diodes, try a UF5408 ultra-fast recovery diode, 3 amps, 1,000 volts.  My guess is that amp builders who obsess over rectifier switching noise spend too much time looking at their oscilloscopes and not enough playing guitar.  However, I could be all wrong about that...

FWIW Merlin's grounding scheme is very close to Hoffman's and Kevin O'Connor's "galactic ground", at least from a functional standpoint.

Hope that helps,

Chip

[Tone Junkie]

You know I thought that would be the case I bought a bunch of the uf5408 and uf4007,s last time i ran out of the others, just so i wouldnt have to snub them.
I guess I already use a clc power section. so what about putting those caps on the B+ node close to the pins as he said, i just got done reading at Aikens site again Im glad I seem to understand a little more everytime I go there I remember the first time everything was over my head, It was interesting what Jerry said about Mark Cameron using the .22 caps to ground as his ground from all the gain and volume pots any thoughts and Ideas on this, and i have to say all my gain pots on the front end will be conductive plastic from now on, on any high gain amps anyway. Bill  

[jjasilli]

A bit late to the party, but what about a step-up transformer at the amp's input, say 1 : 3.  That would boost 200mV to 600mV; 800mV to 2.4mV.  It would only have to handle about 100 - 10,000Hz.  Not sure how to do impedance matching, but am guessing 500K input impedance; with the secondary impedance determined by the necessary turns ratio.  This would noiselessly boost the guitar signal to line level.

[Tone Junkie]

jjasilli now your way over my head but sounds cool, I dont know how to impliment it but gives me another thing to study.
Thanks Bill

[jjasilli]

Well this might be a bit pricey, but who cares if we're eliminating noise!:  http://www.jensen-transformers.com/mic_in.html

guitar signal > JT-11K8-APC wired backwards; the 1.8K secondary, used as a primary, should impedance match the guitar output well enough > JT-115K-E > amp preamp tube stage

[Tone Junkie]

Cool so this will like give me 3 times the normal signal with 0 noise floor gain .
I had to read it a couple more times to try and understand what you were doing i could still be wrong thats why I asked.
Thanks Bill

[jjasilli]

Yes, using a tranny we could step-up the guitar signal noiselessly; but we need to do impedance matching on both sides of the tranny. 

[jjasilli]

Or: 

*  guitar > lo z phono plug tranny > balanced line > step-up input tranny in amp

*  guitar > DI box > balanced line > step-up input tranny in amp

This keeps noise out of the guitar cable; and gives the noiseless boost to guitar signal.  However it requires special equipment and cables outside the amp.

[jjasilli]

Decided to read TUT1 by KOC again. This time I darn near almost understand it!  Anyway, Ch 2 is a must read for noise reduction: separate B+ filter nodes for ea preamp gain stage; regulated B+ supply for preamp; methods of filament noise reduction. 

[PRR]

> what about a step-up transformer

Higher winding impedance, less bandwidth.

The guitar pickup is ALREADY "wound up" as high as is reasonable. DCR near 5K and audio impedance over 100K at 4KHz, then steep fall-off. What you could do with a transformer, the pickup winder already did by adding more turns of finer wire until treble loss happens.

Guitar should be loaded with over 100K and really more like 1Meg. You don't find 1Meg audio transformers, and 100K is unusual.

> 1.8K secondary, used as a primary

It is sold step-up. You propose to work it step-down? This works OK as guitar-to-mike-line, but gives voltage LOSS, not gain.

Published chart for JT-11K8-APC gives 14H for the "150" side. It will barely be over 7K impedance at the bottom of the guitar band. Secondary would be 88K to the guitar, but that would be step-down.

Note that JT-11K8-APC has the least treble of any part in that series: only 80KHz when used AT design impedance.

[jjasilli]

Clearly impedance matching is not my forte!  So a pure tranny arrangement between guitar and amp appears to be not feasible. That leaves the alternative of: guitar > buffer (e.g., DI box) > amp with internal step-up tranny as 1st stage.  However, with an active buffer, we can get a signal boost anyway, which may render the step-up tranny superfluous.  Perhaps this explains why nobody is doing it!