Hi folks!
I've been building my solid state guitar amp quite long and experimenting it pretty much. So for now I'm really stuck and I don't know how to continue. Maybe you can give some advice.
I've got my power amp ready (I built it on chip) and it works as one usual solid state power amp should work. The problem is with preamp. The aim was to have preamp with distortion and clean in the same channel. So if you increase gain the sound will turn from clean into more and more distorted sound. Clean sound is almost ok. Using JFET's, it gives a good sound. But how to make it with distortion? With JFET's, if they are going to distort, then the sound turns out farting on low frequencies and it's very metallic. If I'm cutting down the frequencies from about 100Hz it becomes better, but still it's very metallic. I believe, that I should do pre EQ before overdriving JFET. I was using Marshall JCM900 dual Reverb amp pre boost passive EQ (http://www.mif.pg.gda.pl/homepages/tom/files/JCM900dualRev.gif) before overdriving JFET. Still it doesn't have much effect. Probably wrong frequencies for my system. Maybe you have some suggestions what I should experiment.
Many thanks!
I've been building my solid state guitar amp quite long and experimenting it pretty much. So for now I'm really stuck and I don't know how to continue. Maybe you can give some advice.
I've got my power amp ready (I built it on chip) and it works as one usual solid state power amp should work. The problem is with preamp. The aim was to have preamp with distortion and clean in the same channel. So if you increase gain the sound will turn from clean into more and more distorted sound. Clean sound is almost ok. Using JFET's, it gives a good sound. But how to make it with distortion? With JFET's, if they are going to distort, then the sound turns out farting on low frequencies and it's very metallic. If I'm cutting down the frequencies from about 100Hz it becomes better, but still it's very metallic. I believe, that I should do pre EQ before overdriving JFET. I was using Marshall JCM900 dual Reverb amp pre boost passive EQ (http://www.mif.pg.gda.pl/homepages/tom/files/JCM900dualRev.gif) before overdriving JFET. Still it doesn't have much effect. Probably wrong frequencies for my system. Maybe you have some suggestions what I should experiment.
Many thanks!
Attachments
Pilt,
Don't offend, but on the first glance i noticed couple of flaws in your design: why use a buffer after the jfet input stage and an opamp stage in front of tonestack? The jfet input impedance is already high enough and if you feel the gain is not enough you could preceed the stage with another jfet stage that has an adjustable gain - therefore allowing you to ditch at least the buffer. If you really need buffers put one in front of the tonestack and one as the last stage in preamp. Also, if you want a real high gain distortion why overdrive jfets? It would seem to me that few simple diode clippers would do the job. You could put them either in opamp or jfet feedback loop: drain to gate - decoupled of course.
What kind of distortion you are after? A mild breakup, good for blues and hard rock, should be quite easy to achieve with only few overdriven jfet stages coupled with very small value capacitors. Runoffgroove website http://www.runoffgroove.com has some interesting circuits for that purpose. Huge part of great tone is actually due to right dynamics and picking techniques so circuits that seem to sound crappy at first can actually sound impressive with the right touch.
If you are aiming for modern metal music type tones you actually have to overdrive several stages and do some signal processing - pre and post - for each one. Simple high- and low passing isn't enough in this case and bear in mind that if you overdrive too hardly at any point you will loose a certain amount of detail, thus making the guitar sound like a synthesizer (especially on higher notes). Also, compressors tend to voice these circuits very personally. Few years ago i heavily modded some high gain distortion circuits and noticed that the compressor always gave them the final distinctive tone.
Few rules of thumb i have discovered:
Too much bass overdrive = blocking distortion, undetailed, muffled or farting tone.
Not enough bass overdrive = "clean", thin, 70's rock distortion tone.
Too much high frequency content = annoying "shredder"- like tone (the shredder not in good sense of it), synthesizer - like high notes.
Not enough high frequency content = muffled, undetailed tone.
Pre compression = good for modern metal with damped riffs etc.
You can get a good idea of what you are really after with couple of graphic equalizers that you stick in front and after the gain stages - you can crudely simulate the same thing with some audio software (like cooledit) too. The middle frequencies are very important (ie. scooped mids) and using two tonestacks to control them seems like a good idea.
Some tones require a lot of gain, which means lot of slightly overdriven gain stages and a lot of signal processing. Unfortunately this also means that these circuits tend to get very complex - it might not be such a good idea to try to implement one with the "clean" channel. In my opinion you really have more control if you use separate channels.
Teemu K
Don't offend, but on the first glance i noticed couple of flaws in your design: why use a buffer after the jfet input stage and an opamp stage in front of tonestack? The jfet input impedance is already high enough and if you feel the gain is not enough you could preceed the stage with another jfet stage that has an adjustable gain - therefore allowing you to ditch at least the buffer. If you really need buffers put one in front of the tonestack and one as the last stage in preamp. Also, if you want a real high gain distortion why overdrive jfets? It would seem to me that few simple diode clippers would do the job. You could put them either in opamp or jfet feedback loop: drain to gate - decoupled of course.
What kind of distortion you are after? A mild breakup, good for blues and hard rock, should be quite easy to achieve with only few overdriven jfet stages coupled with very small value capacitors. Runoffgroove website http://www.runoffgroove.com has some interesting circuits for that purpose. Huge part of great tone is actually due to right dynamics and picking techniques so circuits that seem to sound crappy at first can actually sound impressive with the right touch.
If you are aiming for modern metal music type tones you actually have to overdrive several stages and do some signal processing - pre and post - for each one. Simple high- and low passing isn't enough in this case and bear in mind that if you overdrive too hardly at any point you will loose a certain amount of detail, thus making the guitar sound like a synthesizer (especially on higher notes). Also, compressors tend to voice these circuits very personally. Few years ago i heavily modded some high gain distortion circuits and noticed that the compressor always gave them the final distinctive tone.
Few rules of thumb i have discovered:
Too much bass overdrive = blocking distortion, undetailed, muffled or farting tone.
Not enough bass overdrive = "clean", thin, 70's rock distortion tone.
Too much high frequency content = annoying "shredder"- like tone (the shredder not in good sense of it), synthesizer - like high notes.
Not enough high frequency content = muffled, undetailed tone.
Pre compression = good for modern metal with damped riffs etc.
You can get a good idea of what you are really after with couple of graphic equalizers that you stick in front and after the gain stages - you can crudely simulate the same thing with some audio software (like cooledit) too. The middle frequencies are very important (ie. scooped mids) and using two tonestacks to control them seems like a good idea.
Some tones require a lot of gain, which means lot of slightly overdriven gain stages and a lot of signal processing. Unfortunately this also means that these circuits tend to get very complex - it might not be such a good idea to try to implement one with the "clean" channel. In my opinion you really have more control if you use separate channels.
Teemu K
Hi!
Thank you for fast reply! Now I've been already experimenting with JFET circuits from http://www.runoffgroove.com page. Have you tried any of them from that page?
You were right with flaws in my design. Some of those stages are really not necessary. I was just copying my circuit from breadboard as it was after some experiments. Still as I said I'm now dealing with one circuit from that page. I'm after clean signal and distortion when the gain is increased. Distortion should be more for blues and rock playing. I've been trying also diode clipping in different circuits. Still as final result I use my stomp box (Super overdrive) for that. So from that page I choosed circuit called Professor Tweed and I kept from my previous design tone control stage after that. I choosed this circuit cause it's based on one of my ideas to mix overdrive and clean signal. Usually the solid state overdriven signal is dry, that mixing it with clean signal maybe shoud make it sound better. At least I have such kind of believe. Still I have some problems with that circuit sound. When the very first cliping starts, there are some disturbing high frequencies. So I can hear, rough clipping. Maybe it's kind of solid state thing. So again I believe that I should maybe use the EQ's to get rid of this certain frequency. Some notch filter.
For EQ's I've been using Cool Edit ones already with my earlier experiments. I recorded some licks, when I was playing guitar through some part of circuit. Then I was EQ'ing it and playing it through next part of circuit. Still I think it sucks a bit, cause I lose the touch between guitar and amplifier. I need realtime EQ's. I hope I'll not have to build those.
Gonna work a bit more with that circuit. I hope I'll not lose my patience.
Thank you for fast reply! Now I've been already experimenting with JFET circuits from http://www.runoffgroove.com page. Have you tried any of them from that page?
You were right with flaws in my design. Some of those stages are really not necessary. I was just copying my circuit from breadboard as it was after some experiments. Still as I said I'm now dealing with one circuit from that page. I'm after clean signal and distortion when the gain is increased. Distortion should be more for blues and rock playing. I've been trying also diode clipping in different circuits. Still as final result I use my stomp box (Super overdrive) for that. So from that page I choosed circuit called Professor Tweed and I kept from my previous design tone control stage after that. I choosed this circuit cause it's based on one of my ideas to mix overdrive and clean signal. Usually the solid state overdriven signal is dry, that mixing it with clean signal maybe shoud make it sound better. At least I have such kind of believe. Still I have some problems with that circuit sound. When the very first cliping starts, there are some disturbing high frequencies. So I can hear, rough clipping. Maybe it's kind of solid state thing. So again I believe that I should maybe use the EQ's to get rid of this certain frequency. Some notch filter.
For EQ's I've been using Cool Edit ones already with my earlier experiments. I recorded some licks, when I was playing guitar through some part of circuit. Then I was EQ'ing it and playing it through next part of circuit. Still I think it sucks a bit, cause I lose the touch between guitar and amplifier. I need realtime EQ's. I hope I'll not have to build those.
Gonna work a bit more with that circuit. I hope I'll not lose my patience.
I agree Teemu
To get something going quickly. Some things to try before you think to ditch your design. I think all the pieces are there, they might need a little rearrangement and tweeking.
1) Put aside the input Jfet for now.
2) Move the input buffer to the output.
3)Play with the gain on the new input buffer
4) once the gain is close then play with the EQs on remaning jfet stage for tone.
5) If you want more distortion stick the Jfet stage back in before the volume. Then spend the time tweeking this stage.
To get something going quickly. Some things to try before you think to ditch your design. I think all the pieces are there, they might need a little rearrangement and tweeking.
1) Put aside the input Jfet for now.
2) Move the input buffer to the output.
3)Play with the gain on the new input buffer
4) once the gain is close then play with the EQs on remaning jfet stage for tone.
5) If you want more distortion stick the Jfet stage back in before the volume. Then spend the time tweeking this stage.
pilt said:
As is, i haven't. However, most of the circuits in there are very basic design. I used similar circuit parts before i even found about the website. I have built a jfet booster circuit similar to fetzer/valve. It had different transistor and component values but the circuit is really nothing but a schoolbook example of common-source amplifier. I also built a practice amp around LM386. If you ever heard about Smokey amps they are based on this little halfwatt IC. The circuit turned out to be surprisingly similar to Ruby, except that the buffer was replaced with a gain stage and preceeded with a simple tonestack. The power amp was a almost a carbon copy from National's datasheet. So, basically i have built "modded" versions from some of the circuits.
However, note this: I have never believed that jfets would be the exact key to great tone by themselves - overdrive them too much and they still sound quite crappy. I guess that the idea that they sound like tubes comes from the fact that you can use tube-like circuits with them: Low capacitance coupling etc. With some good filtering they tend to have a little less harsh sound than plain opamps though.
And speaking of filtering, it is quite tricky so i can relate to you Pilt. So far i haven't completely nailed the tone i'm after with a circuit that would have stayed simple enough to be worthwhile to build. With some jfets and small coupling values a fairly OK "blues" distortion should come quite naturally. The guitar signal makes a big difference too: i have never been able to produce a good blues tone with a humbucker-type tail pickup. Instead, a single coil neck pickup sings. I actually found out that a simple jfet booster that i overdrive slightly is good enough for that type of tone. If i really want it to "punch" i put a compressor or limiter in front of it. With that combination you really start to get a tube-like tone.
I think a stage that would split the signal to three bands - bass, middle and high - should be the first stage in a distortion circuit: It would allow controlling of gain for each frequency band individually. Usually the bass has to be cut drastically so that it won't mess up the signal so the distortions tend to either be fuzzy, synth-like on high notes or lack the bottom end. I think mixing the distorted signal with clean bass signal would yield quite impressive results.
The high frequencies are usually a problem too: They should be attenuated when the gain rises but if the attenuation is constant it will surely ruin the clean sound - unfortunately this tends to mean one extra knob more.
I'm more a fan of very high gain distortion myself: Death- and black metal - type tones are much closer to my heart so i rather just stick with some digital units. What i really miss is control of each effect, so from a performers perspective, you might be happier if you make the gain level switchable. For example, have two gain settings, one for "clean" and one for "distortion" and a make the circuit footswitchable to select between them.
pilt said:
Yep, the touch is quite important and even some good circuits will sound boring when they just play a monotonous lick. But hey, at least the EQ gives some idea. I started designing with LTspice few years ago and i have been happy ever since: It allows you to feed the circuit model with a .wav file and get the output as .wav file too. I loose touch, of course, but at least it saves me from the trouble of either guessing how the circuit might sound like or from building a crappy circuit as a proto. If results from LTspice are ok i usually find that the proto will sound quite good too.
Don't loose patience. You can always take a break and design something else for a while. It might actually give you new ideas. Happy DIY:ing!
Teemu K
From the LTSpice docs -
Syntax: Vxxx n+ n- wavefile=<filename> [chan=<nnn>]
This allows a .wav file to be used as an input to LTspice. <filename> is either a full, absolute path for the .wav file or a relative path computed from the directory containing the simulation schematic or netlist. Double quotes may be used to specify a path containing spaces. The .wav file may contain up to 65536 channels, numbered 0 to 65535. Chan may be set to specify which channel is used. By default, the first channel, number 0, is used. The .wav file is interpreted as having a full scale range from -1V to 1V.
For saving, there is the .WAVE command. Again, check the docs
Syntax: Vxxx n+ n- wavefile=<filename> [chan=<nnn>]
This allows a .wav file to be used as an input to LTspice. <filename> is either a full, absolute path for the .wav file or a relative path computed from the directory containing the simulation schematic or netlist. Double quotes may be used to specify a path containing spaces. The .wav file may contain up to 65536 channels, numbered 0 to 65535. Chan may be set to specify which channel is used. By default, the first channel, number 0, is used. The .wav file is interpreted as having a full scale range from -1V to 1V.
For saving, there is the .WAVE command. Again, check the docs
Hi Teemuk,
are you still active? (This forum seems to be a bit outdated) I just fluttered into it.
Reason : Trying to simulatie a fuzz box design (generalguitargadgets) but I don't have a LTspice pickup model. That's the first question: is there a LTSpice model available? I mean, do you know that?
The other question is different: What guitar effect is Alan „Blind Owl“ Wilson using on the Canned Heat " Let's work together" slide guitar solo. I'm asking everyone but haven't got any positive answer...
I'm a hardware design engineer, I don't know if it is permissible to ask like that in forum like this, but anyhow. I try it
are you still active? (This forum seems to be a bit outdated) I just fluttered into it.
Reason : Trying to simulatie a fuzz box design (generalguitargadgets) but I don't have a LTspice pickup model. That's the first question: is there a LTSpice model available? I mean, do you know that?
The other question is different: What guitar effect is Alan „Blind Owl“ Wilson using on the Canned Heat " Let's work together" slide guitar solo. I'm asking everyone but haven't got any positive answer...
I'm a hardware design engineer, I don't know if it is permissible to ask like that in forum like this, but anyhow. I try it
You mean a guitar pickup? As far as I know there is no ltspice model for one but this should get you pretty far in making one: BuildYourGuitar.com :: The Secrets of Electric Guitar Pickups
Though if you plan to use that waveform input discussed earlier the pickup's response is already in the recorded media and there is no need to simulate one.
The effect? Back in the days I acquired the exact effect by plugging the guitar to cheap mic preamp of 1980's stereo and overdriving the heck out of it. So its likely some kinda of a basic fuzz circuit, like fuzz face, fuzz-tone, or whatever there were. There were hunreds of those designs made and most of them should get you there. It's just the guitar overdriven without any extra tricks and that's why it sounds so "filthy" and "chokes" on lower notes.
Though if you plan to use that waveform input discussed earlier the pickup's response is already in the recorded media and there is no need to simulate one.
The effect? Back in the days I acquired the exact effect by plugging the guitar to cheap mic preamp of 1980's stereo and overdriving the heck out of it. So its likely some kinda of a basic fuzz circuit, like fuzz face, fuzz-tone, or whatever there were. There were hunreds of those designs made and most of them should get you there. It's just the guitar overdriven without any extra tricks and that's why it sounds so "filthy" and "chokes" on lower notes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.