Legacy fixation

[demym]

Hi,

some time ago i did the jfet CarvinLegacy, based upon the schem called Leegazzy:



The jfet versions sound very good, don't know how much as the real Legacy, but to me it's similar..

Now i am exploring tube stuff, and i'm trying to recreate that schem using 12AU7 (but i also have some 12AT7 and 12AX7 to use, but at the moment i'm stucked with 12AU7, as i've tried valvecasters and so on);

so i did the Rick Holt multiplier based on MAX1044 (but i use LT1054 as my dc voltage is 12V); i have, so, 80V for the plates..

I'm onto the second stage of the amp; but (hey yeah, there's a BUT)... at second stage i don't get that crunch that was instead evident using jfets (and that i think the real amp should have); the sound remains as clean as the first stage (maybe just a little bit more clipping); i know that using 80V at plates change a lot from the real amp that runs those stages at 400V with 12AX7.

does anyone know what changes i should make in cathode and anode resistor to adapt that design to a 80V one ? I'm talking about the second stage (the first works like a charm, giving me a clean that crunches a bit upon hard dynamic picking)..

At the moment i'm using the same exact values from the Legacy official schematic (220k at plate, 1.5K at cathode, for the first stage i used 220k at plate and 10k at cathode)...

Is this a possible dream ? Or is it a mission impossible one... just wanted to be sure before wiring the next stages...

Would it be different using submini 6111 instead of regular 12AU7 ? I've seen that they have 100V max voltage plate, so maybe 80V is pushing them to the (right) limit ?

I'm doing all of this mainly because don't wanna take any risk with 'true' HV...so newbie here


Also, i've noticed that on the original legacy schem, the first stage (which acts as clean or as first stage of lead channel), they wire pins 6,7,8 of the tubes, and then, on next stage, 1, 2, 3... can the two triode pins be used in any order or there are difference in it ? I mean, to me it would be logical to have 1,2,3 and then 6,7,8...

Thanks in advance and sorry for the long questions

[deadastronaut]

sorry to not answer any questions..i have no idea on tubes..

but do you have a soundclip of your fet leegazzy....? would like to hear it very much...
rob.

[DougH]

You should use 12ax7's instead of 12au7's. 12au7's have a much lower plate resistance and will not produce the gain (transconductance) of a 12ax7. At the same time, your 80v power rail is going to kill any headroom you might have had with the 12au7, and the ridiculous gain will probably produce a lot of fizz at that voltage.

You can't just sub parts using the same resistor values and expect it to work or sound right at whatever voltage supply level you feel like. Plate resistors are there to provide the correct voltage at the plates, and with the cathode resistor, establish the correct current flow for the plate and cathode, and to establish the operating point of the tube stage. At the same time, plate resistors (along with grid resistors of the next stage) determine gain parameters for the given stage. Everything depends on everything else.

The JFET circuits are fun to play with and in some instances the same resistor values as the amps will work with JFETs at 9v. Consider that a happy coincidence (or happy accident as JFETs vary all over the map and may or may not bias correctly for a given circuit configuration). But the metaphor only stretches so far and falls apart when you are trying to build a (presumably) low-voltage starved-plate tube circuit. You're going to have to do some research into setting up tubes to run at low voltage, figure out the gains you want out of each stage, and scale components accordingly to get the same frequency response and gain characteristics that are in the original amp. IOW- you've got a lot of reading to do.

You up to that? If not- enjoy your JFET circuit, or take another tack and learn about building amps instead.

[Earthscum]

http://www.freewebs.com/valvewizard/

That site helped me more than any other about valves. Go through it all, and compare parts of the original schematic to the section you are reading up on, and you should start to see exactly what Doug is talking about.

You should use 12ax7's instead of 12au7's. 12au7's have a much lower plate resistance and will not produce the gain (transconductance) of a 12ax7. At the same time, your 80v power rail is going to kill any headroom you might have had with the 12au7, and the ridiculous gain will probably produce a lot of fizz at that voltage.

You can't just sub parts using the same resistor values and expect it to work or sound right at whatever voltage supply level you feel like. Plate resistors are there to provide the correct voltage at the plates, and with the cathode resistor, establish the correct current flow for the plate and cathode, and to establish the operating point of the tube stage. At the same time, plate resistors (along with grid resistors of the next stage) determine gain parameters for the given stage. Everything depends on everything else.

...

I learned enough from that site to use a different tube (12AX7 LP) and make it work and sound nice as a preamp using 100V. The LP really isn't much different, but it does have some obvious gain changes vs. voltage, etc. I did the "leap" and bought a tube before learning enough about it... originally wanted to do a McTube, but decided I just want the "tube compression" sound for my bass. I wish I would've found that site sooner, but oh well... the LP sounds great!.

[petemoore]

  Kind of looks like a nice clean Fender type preamp.
  The output amp of an amp driving a speaker is huge part of the distortion, Fr Response and compression.
  Not that overdriven preamps can't sound good, it is a completely different affair.
  Jfets @ 9volts can be hard to bias, as are low voltage preamp tubes [biasing either one is intermediately difficult, small voltage makes less room for error, limited clean headroom]...are primarily '1 parameter' affairs, ie signal peak gets to x level and something happens to round-ish over or chop off the peak, the process is naturally sharp and precise. Tends to be hard sounding compared to when a speaker/suspension/compressions of air and other factors are all influencing waveshapes, these 'events' occur at different phases in the cycle, in varying amounts depending many things...how hard the speaker is 'leaning' on the output section at that moment and maybe a fraction or two of a moment ago, speakers compress and distort and these have movement.

[anchovie]

 Kind of looks like a nice clean Fender type preamp.

With 4 gain stages, 3 of them with bypass caps on the source?! This is a gain monster!

[deadastronaut]

 Kind of looks like a nice clean Fender type preamp.

With 4 gain stages, 3 of them with bypass caps on the source?! This is a gain monster!

hmmm...sounds right up my street!...

if i could hear it...

[petemoore]

With 4 gain stages, 3 of them with bypass caps on the source?! This is a gain monster!
  There it is then, such a large size I could see the forest for the trees...
 

[deadastronaut]

With 4 gain stages, 3 of them with bypass caps on the source?! This is a gain monster!
  There it is then, such a large size I could see the forest for the trees...
 

pete do you do a lot of crossword puzzles..?..your ever so cryptic..but its always worth reading...

[petemoore]

  The design is for 400v, 80v is very much smaller.
 The second and subsequent tubes are not sent as hot a signal from the first gain stage, the concept being that each stage sees X amount of input and that distorts a little bit or however much [settings etc.]...repeating the stages adds ''gain'' [I get the idea from the phrase ''Gain Monster'' that this design is intended to create distortion.
 If at 80v, a redesign using low voltage tubes would seem to be a possible fit worth checking out, scaling the whole 'gain structure' [how much each stage is overdriven...the first valves output being multiplied many times by later stages. For example if the first stage is 'just under' the amount required to slam the second stage sufficiently, the multiplied gains further down the line won't push the rest of the preamp as hard as the original design...400v is exceeds max plate for 12ax7's, starting with a tube that 'says sort the same' at 80v would seem logical starting point if 400v is out of the question.
 Brings back the Jfets are easier/safer to work with scenario, they don't necessarily behave all that well at 9v, see Fetzer and Jfet bias articles, sort them by the amount of power supply they waste, or maybe give them a bit more power supply to: waste+ have leftover for gain -the voltage used to set up bias.
 I'm not sure at all which would be the better ''not 400v'' choice for tone, even if tubes were as easy to mess with as the Jfets.
 I can say that the voicing of the circuit looks cool/though through, so the topology and voicing of the circuit-map could probably remain very much the same, however 80v from 400v will require some redesigning, choosing what kinds of active components since the restructuring of the gain stages was fundamentally altered.
 Yet another approach that maybe shold be in a nother post..
 Start with 12v jfets or pick them from a batch, get 1 or 2 of them biased up nice, use that to drive the stages that use this higher input voltage [latter after being boosted once or twice] at higher supply voltages.
 @ 400v, the first 12ax7 stage is probably quite clean, used to bump the signal voltage to a 'workable' [distortable] level for all [or some? scope or probe to find out more] the later stages. Maybe just because I like my single Jfet boost about the same or different than a 12ax7 stage1, the input presented from guitar is teeny at the first stage.
 Without the 'enourmous' supply voltage it will start to look or sound and work very differently, if that is scaled to 80v, a tube which 'lights up' at that voltage would probably allow similar function.  
 Another reason I mentioned Jfets is..how does the current rating compare to the current draw of the plates, and bumping the signal to a 'workable' level @ low voltage works well and can be played, fiddled with without the HV worries..
 ''All Tube" is a big selling point, I'm not exactly sure how that equates to ''all goodness'', especially when input signal is near guitar output level.

[DougH]

Oh btw, I don't know where that schematic in the first post came from but I definitely did not "start the jfet emulations", but thanks anyway to whoever drew that up.  

[demym]

Hi folks, first of all, a big thankyou for all your thoughts and suggestions.

I have read the Valvewizard site, but it seems a little bit complex for me to understand, mainly because they always refer to HT 280V and so on, and that make me difficult to adapt things to my current voltage supply.. I know that using formulas could help me a lot (and i would do some maths with pleasure), but...

For example, i've noticed that in other LV designs (such as valvecasters and so on) the plate resistor is often 220K or 100K; does this mean that correct biasing is devolved to the cathode resistor and grid resistors ? In the first stage of the legacy, i had to use a 10k resistor, as that seemed to give me a more hot signal on the first stage... the sound coming out from the first stage is beautiful this way; as i've left the following cathode resistors as in the original legacy shem, maybe i could try to trimmer them (as in the valvecaster, say 50k) to find the right spot at my voltage ?

I enjoy my jfet builds, but just i wanted to go one step further with tubes, as this was maybe my first objective when all of this started.

Please don't see me as the one that would not like to read (in fact, i've read and still keep reading a lot in this year and a half of DIY).. I'm , though, much toward to a (basic) practical approach.

Do you think that a high gain preamp (as the legacy is) is possible with tubes and 80/100V of plate voltage ? If it is, i will be pleased to continue these experimentations (i have tons of gear, so i really don't have the need to build a pedal for using it today); but if all my efforts in this direction would be unuseful, i wouldn't like to waste time (even if this time is pleasing)..

Would the situation change using submini tubes ? I mean, giving the fact that a 6111 has 100V max plate voltage, would 80V be nearer to the limit as 400V is near to the limit for a HT project ?

And space charge tubes ? Would these, running at 12v, provide a behaviour that is much similar to their HV couterparts ?

Just freely thinking though, any further help and suggestion would be very appreciated.

I've tried several tube LV projects (valvecaster, tube star, marshavalve tube, ecc..), but none of them gave me the high gain i'm after, even with voltage at 80V; i'm definitely a high gain guy, and love sounds a la VH, Vai, Satriani, ecc..

I will record some samples of the jfet legacy asap, for you to listen.


Once again, thanks a lot to all of you !!!

[DougH]

Demym, hopefully I didn't give the impression that I was accusing you of not reading. My only point was that this stuff is not as simple as just doing simple substitutions, especially when you are changing the device type. 

I'm also not implying that you need to figure all this out perfectly ahead of time before trying things either. Sometimes the best way to learn is just roll up your sleeves and dig in.

I would suggest finding some low voltage circuits that are out there and experiment with modifying them to get the sound you want. If you find an existing circuit/project with a particular tube type and voltage level, start from there. Do it on a breadboard so you can experiment and move things around.

Ten years ago or so I modified a BK Butler tube drive circuit to use 6ak6 pentodes instead or the 12au7/12ax7 in the original design. I wired it up on a breadboard and experimented with it a lot. I had to tweak some things and move things around to get the sound I wanted. As a start, you generally want to get the plate voltage on the tube around 1/2 of the supply voltage, give or take. I also played with radio space-charge tubes on the breadboard too a few years back. All I can say is a lot of this depends on the tube you are trying to use at what voltage level, what kind of sound you are looking for, etc.

I would start with wiring stuff up on a breadboard, adding stages, adding frequency shaping, etc, etc. All the usual stuff. Be careful if you're playing with 80V on a breadboard- don't get bit.

[demym]

Thanks Doug, no problem at all, i said that just to explain that i've read a lot in this period, but for my noobism it's not yet sufficient.

I, instead, appreciate all your pointings.

Will try other combinations (and already have them on my breadboard)..

Thanks once again

[demym]

Deadastronaut,

here's a sample of the pedal:

http://www.youtube.com/watch?v=mPOmrRstUdw

Have a nice time

[deadastronaut]

Deadastronaut,

here's a sample of the pedal:

http://www.youtube.com/watch?v=mPOmrRstUdw

Have a nice time

thanks. nice to hear it....sounds good nice sustain, nice licks too!.cool....i love these emulation pedals...gwar!