Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 07:40:43 PM

Login with username, password and session length

Recent Posts

 

Hughes & Kettner Blues SS Clone

Started by joecool85, March 24, 2011, 11:47:09 AM

Previous topic - Next topic

joecool85

Quote from: tcio on March 24, 2011, 09:41:12 AM
Hello all!

I am having a heck of time trying to get a little circuit to work. I am trying to do a tube to JFET conversion based on the preamp section of a Hughes & Kettner Blues Master. I already know how to do these type of conversions but the Blues Master uses a NE5532 chip in a bipolar configuration. I have a variac with plenty of amps and built a couple of rectifiers which produced +-15v but the thing just makes a low hum. Any pointers or tips would surely be appreciated



Quote from: tcio on March 24, 2011, 10:24:39 AM
I tried many times to get my attached schematic pics up but they would not post in the original message. Now I cannot even open the message up to edit it with nonstop "Error 500" so I am adding this link to my schematics in hope that this will assist anyone out there that may be able to help me out:

http://www.freewebs.com/tcio/schematics.htm

Please help if you can - Thanks in advance  :o


Note sure what was going on, that one thread was damaged.
Life is what you make it.
Still rockin' the Dean Markley K-20X
thatraymond.com

tcio

#1
Yes. I have been having serious issues trying to access and modify my post. This is now the first time I have been able to log in and access this thread since the MANY attempts I made a few hours ago. Weird.......

Any help on this NE5532 Op Amp 15 +- problem would still certainly be appreciated  (hint hint) :)

J M Fahey

Youwould have to redesign each and every Op Amp stage from split supplu to single supply, which means:
1) All Pin4 now go to ground instead of -15V
2)  All pin 8 now go from +15V to +9V (I suppose your pedal is battery powered)
3) You will need to add a +4.5V source to bias all opamps
4) All "+" inputs now are not referred to ground but to new +4.5V bias supply
5) All inputs, outputs and feedback networks now are not referred to "old" ground but to "new" one, they sit at 4.5V DC from there, so any ground referenced point must have a capacitor to block DC if it didn't include one previously, and if an electrolytic, proper polarity must be considered.
Our friend joecool is still struggling to turn a good Dean Markley preamp into a pedal, because of all reasons stated above.
It *can* be done, just it's not easy, it involves major work.
6) Your Fets will have to be properly biased and feel happy with only +9V, that's all there is.
Personal opinion: what makes this pedal good (correct that: *very* good) is, precisely, that it uses *real* tubes, and uses them as God intended, with a macho +250V supply.
I fear that the conversion to 9V may work if done properly, but sound will be bland and gutless.
I suggest you build yours as-is and enjoy it fully.
Good luck.
PS: the "easy way out" is to use a split supply made out of two 9V batteries; it will save you 95% of the hassle.

joecool85

Quote from: tcio on March 24, 2011, 04:50:43 PM
Yes. I have been having serious issues trying to access and modify my post. This is now the first time I have been able to log in and access this thread since the MANY attempts I made a few hours ago. Weird.......

I apologize for that, you should be good to go now though.
Life is what you make it.
Still rockin' the Dean Markley K-20X
thatraymond.com

teemuk

#4
I think a bit of clarification about the goals of this project would help everyone.

Is this a conversion from +-15V dual supply to a 9V battery-powered single-supply? If not, then we can ignore that.

Do you want to power the FET stages from dual supply rails or from a single-supply rail? The answer naturally also depends on what is the answer to former question.

Why FETs? I know the FET+100K trimmer resistor at drain is the typical "easy" scheme, but really the conversion from tube to FET isn't as simple as just substituting the triode with a FET and dropping the power supply voltage lower - at least if you want to nail similar operation and characteristics as in the original design that used tubes. I could go on for a while why such a simple scheme doesn't really work sufficiently but I think that has been discussed extensively in various forums and a moment of Google searching and reading about the topic will prove to be more fruitful.

My personal opinion is that you achieve the triode stage emulation goal more easily - and more importantly, much more consistently - with a well-thought OpAmp+diodes -based clipping circuit. Practically, in circuits like these you want to simulate the frequency response of the stage, it's input impedance, it's output impedance, grid conduction/clipping phenomenon, and the saturation/clipping. Since the emulating circuit most likely will be operating from lower power supply voltages you will also find out that you need to scale the clipping points and stage gains to match that. And when you start scaling stuff it will affect all component values. All in all, you will end up with something far different than the original design using tubes and B+ of 250 volts, but that's where the magic of a proper design capturing the characteristics of the original lies.

IMO, all this is much easier to achieve with a gain block (OpAmp) that has more consistency than a FET with widely spread parameters and that lends itself much more easily to various different circuit configurations.

joecool85

Quote from: teemuk on March 25, 2011, 09:56:19 AM
IMO, all this is much easier to achieve with a gain block (OpAmp) that has more consistency than a FET with widely spread parameters and that lends itself much more easily to various different circuit configurations.

I would LOVE to see some tube emulation with some basic opamps (TL07x etc).  It would be much easier to build with less testing of components.  FETs require matching, biasing and all that, opamps don't which is nice.
Life is what you make it.
Still rockin' the Dean Markley K-20X
thatraymond.com

teemuk

#6
Here's one example. It's a design already presented in one of the threads here: http://www.ssguitar.com/index.php?topic=1372.0


In case of emulating a single triode stage only you really are interested in focusing only on a small section of the schematic, that is one of the OpAmp stages following the phase splitter stage:

- The diode from inverting input to ground simulates the grid conduction mechanism: Signal at grid becomes more positive than signal at cathode and current begins to flow from cathode to grid. This saturates the tube making it conduct fully.

Anyway, the grid conduction clips one of the input signal's half waves already at the grid, at the same time creating DC bias shift. This effect becomes enhanced if the coupling capacitance is quite low - like it is in many tube amps. The clipping and bias shift creates a lot of asymmetry.

A generic high gain tube preamp design usually pushes one tube to grid clipping, and then proceeds to overdrive the signal into "square wave" territory. The end signal is an asymmetrically and hardly clipped wave. You can perfectly nail that operation with ordinary diodes. Ironically, your greatest issue is likely replicating clipping that is "hard" enough. A LED gets to the ballpark, even an overdriven OpAmp. With that in mind, creating subtler overdriven effects with semiconductor diodes isn't overly difficult.

The characteristics of the clipping are somewhat tweakable with proper selection of resistor values and diodes. Additionally you can add some shunt capacitance to the grid to simulate the same effect existing in tubes.

- The following stage is configured for unity gain. It doesn't have to be but it worked for this particular design. What's really more important are the diodes following the stage: These simulate the cutoff mechanism, in which tube stops conducting current. Normally the diodes are forward biased but when the output signal's amplitude exceeds a certain threshold the diodes stop conducting and limit the output signal.

This clipping circuit can again be tweaked to ones liking, even configured around the feedback loop of the OpAmp. You can also tweak the frequency response of the gain stage to match that of a typical triode gain stage. Basically you can use all the various frequwncy response shaping methods an OpAmp stage lends itself into.

I used the particular DC-referenced clipping configuration so I could control the clipping threshold with a modulated DC reference signal, this to simulate power supply voltage sag under loading, which results into earlier clipping threshold. So, there you have that aspect as well. Other clipping schemes you can imagine likely work equally well - especially if the design goal is just a preamp emulator.

With proper balancing of "grid clipping" and the "cutoff" clipping you can "bias" the tube emulator to bend into various types of performances, similarly as designers do by varying biasing of triode stages. Want less asymmetry: configure the grid clipping emulation to clip at higher input signals. More asymmetry? Do the opposite. And so on. The great thing in basic circuit like this is that it's easily tweakable, much more than a FET with drain trimmer.

This is just one example of a basic principle you can use to replicate the generic triode stage -style clipping mechanisms and resulting bias shifts. With careful choice of components in relation to circuit you'll use them in it'll get you to the ballpark. The generic asymmetric diode clipping setup found from numerous stompboxes can nail the overall asymmetry but it won't really nail the resulting bias shifts. That characteristics was vital for nailing the crossover distortion effect.

Also, you actually want to subtly cascade few of these circuits and process the EQ while you're at it - like it's done in actual tube amps. A typical stompbox clipping stage just clips in one stage, a typical tube preamp usually clips the signal in several stages so you want to replicate that interaction.

Anyway, this post has gotten fairly long already. The main point is you can take a fairly simple and basic design like this OpAmp with clipping diodes in input and output and proceed to build from that.

joecool85

Quote from: teemuk on March 25, 2011, 11:58:45 AM
Here's one example. It's a design already presented in one of the threads here: http://www.ssguitar.com/index.php?topic=1372.0


Do you have an updated schematic?  You mentioned in the other thread that you re-worked it but didn't put in another schemo.  Have you worked on this circuit since then?
Life is what you make it.
Still rockin' the Dean Markley K-20X
thatraymond.com

J M Fahey

QuoteIMO, all this is much easier to achieve with a gain block (OpAmp) that has more consistency than a FET with widely spread parameters and that lends itself much more easily to various different circuit configurations.
+1
I *fully* agree and then some.
Fact is, Op Amps are "Universal Gain Blocks" and were invented specifically to generate or follow the strangest curves possible which Mathematicians could invent.
Before Digital Computers, existed Analog Computers, which were used to solve Military and Scientific problems, real time.
Problems such as Radar processing, anti aircraft gun pointing, Spaceship controls, flight controls, Industrial Process controls, you name it.
And what were they filled of?: you guessed right: Op Amps.(even tubed ones).
I remember most early Op Amp datasheets were full of differentiator and integrator circuits, plus log generators, scale extenders, voltage shifters and instrumentation circuits.
Fact is, Digital computers , an entirely different approach became possible, and Op Amps, by sheer chance, were usable in Audio circuits, plus being inexpensive.
I almost forgot: "Operational" means exactly "Mathematical" operation.
Some examples:
1) Loebe Julie's 1941 vacuum-tube operational amplifier with differential inputs.
2)  1942 M-IX Gun Director Operational Amplifier
As you see, it was a tube type one, yet still needed split supplies (+/- 350V !!!!)

3) 1950-era operational amplifier for Nike Missiles.

Incredible, huh??
Now you know why any Op Amp we use today has, say, 100x gain within the audio band ... but 1000000 gain at low frequencies or DC.
Yes, not a typo, 1 Million X gain.
Teemu, congratulations again. :tu: :tu:

tcio

#9
I know this is way too late to reply to any of these posts but I feel I should certainly take the time out to first off apologize for not getting back to reply until now but most of all thank everyone for taking all the time they have to go in depth and explain all of these things to me. I was basically going by what I have seen done many times from runoffgroove.com in regards to mimicking and converting tube amp circuits to solid state. I thought if I kept the 15± on the opamp which is what was applied to it in the original circuit the tube to JFET section would hopefully be somewhat straight forward (throwing some pots on the FET(s) to find the correct bias) but it appears I have even FAR more to learn than I ever imagined. I know it's ridiculously late for this but now that things seem to be working properly for me here is the schematic if anyone is still interested

Here's the original schematic:




Here's a stripped down section a little doctored up with the section I was working on (circled in red):