New GxCreamMachine.lv2
Moderators: MattKingUSA, khz
-
- Established Member
- Posts: 2347
- Joined: Mon Jul 01, 2013 8:13 am
- Has thanked: 9 times
- Been thanked: 466 times
New GxCreamMachine.lv2
Hi
A couple of people suggest me to simulate this or that, that's nice, thanks a lot therefore.
However, some suggestions going along to simulate a power amp.
After some work and research I decided to create my own circuit, surly hardly based on known schematics.
So I come up with the CreamMachine. It's based on a input stage driven by two 12AX7 tubes, here is the schematic:
The two 12AX7 tubes nicely compress and saturate the input signal, so that the input becomes "creamy" without spikes in the signal.
A tonestack with bass and treble control allow to fine-tune your signal, when needed.
The input stage is followed by a power amp stage with two EL34 tubes:
On the input of the power amp stage is a lowpass with 12kHz.
The two EL34 pentodes drive the output transformer to the final volume control.
and flow finally into the cabinet simulation:
The result is ( i found) a nice sounding Power Amp simulation with a creamy sound.
Here it is:
Even if it ain't looks like, be aware that this is a power amp, not a small pedal like it looks.
Here is a plot of the power amp stage:
You can get it here:
https://github.com/brummer10/GxCreamMachine.lv2
A couple of people suggest me to simulate this or that, that's nice, thanks a lot therefore.
However, some suggestions going along to simulate a power amp.
After some work and research I decided to create my own circuit, surly hardly based on known schematics.
So I come up with the CreamMachine. It's based on a input stage driven by two 12AX7 tubes, here is the schematic:
The two 12AX7 tubes nicely compress and saturate the input signal, so that the input becomes "creamy" without spikes in the signal.
A tonestack with bass and treble control allow to fine-tune your signal, when needed.
The input stage is followed by a power amp stage with two EL34 tubes:
On the input of the power amp stage is a lowpass with 12kHz.
The two EL34 pentodes drive the output transformer to the final volume control.
and flow finally into the cabinet simulation:
The result is ( i found) a nice sounding Power Amp simulation with a creamy sound.
Here it is:
Even if it ain't looks like, be aware that this is a power amp, not a small pedal like it looks.
Here is a plot of the power amp stage:
You can get it here:
https://github.com/brummer10/GxCreamMachine.lv2
On the road again.
- funkmuscle
- Established Member
- Posts: 2806
- Joined: Mon Jun 02, 2008 2:30 pm
- Has thanked: 131 times
- Been thanked: 33 times
Re: New GxCreamMachine.lv2
I am starting thinking that this plugin-from-schematics thing is literally the best thing since sliced bread. Gosh I still have to give it a try...
- funkmuscle
- Established Member
- Posts: 2806
- Joined: Mon Jun 02, 2008 2:30 pm
- Has thanked: 131 times
- Been thanked: 33 times
Re: New GxCreamMachine.lv2
Perfect meme for that plugging also. Run one of those fuzz plugins in front of the cream machine and then use your neck pickup on your guitar and just try the guitar solo from Sunshine of Your Love from cream and you'll see you will almost nail the tone.CrocoDuck wrote:I am starting thinking that this plugin-from-schematics thing is literally the best thing since sliced bread. Gosh I still have to give it a try...
-
- Established Member
- Posts: 363
- Joined: Sat Nov 17, 2012 3:36 pm
Re: New GxCreamMachine.lv2
+1 This is what has got me interested in the project too! I have tried circuit simulation on guitar signals with Spice, which does a similar thing, but in a highly batch-oriented way, not suited for music. The cool thing about this project is that the simulation is done in real-time, which is absolutely incredible. It opens the possibility to build custom boards and simulations of rare music machines etc. which is much, much more interesting than what most other simulations do. Remember, all musicians like to have a few special items that give them an edge soundwise. This project has the framework to provide digital versions of these items.CrocoDuck wrote:I am starting thinking that this plugin-from-schematics thing is literally the best thing since sliced bread. Gosh I still have to give it a try...
Vox, Selmer, Yamaha and Leslie amplifiers. Rickenbacker, Epiphone, Ibanez, Washburn, Segovia, Yamaha and Fender guitars. Hammond, Moog, Roland, Korg, Yamaha, Crumar, Ensoniq and Mellotron keyboards. Xubuntu+KXStudio recording setup.
-
- Established Member
- Posts: 564
- Joined: Thu Mar 12, 2015 8:41 am
- Has thanked: 44 times
- Been thanked: 8 times
-
- Established Member
- Posts: 2080
- Joined: Mon Sep 28, 2015 8:06 pm
- Location: Here, of course!
- Has thanked: 227 times
- Been thanked: 400 times
- Contact:
Re: New GxCreamMachine.lv2
Indeed!gimmeapill wrote:Damn, so much to catch up with & so little time!
Good stuff Tramp
Now about that 'spare' time I was supposed to have when I went part retired...
The Yoshimi guy {apparently now an 'elderly'}
-
- Established Member
- Posts: 2347
- Joined: Mon Jul 01, 2013 8:13 am
- Has thanked: 9 times
- Been thanked: 466 times
Re: New GxCreamMachine.lv2
Thanks for the kind words.
Well, I need to refine the fuzz plugs for what I've learned in the mean time. So, for example this is how a fuzz face response in the simulation run
To archive that, the plug needs more then 80% CPU power here, unacceptable.
Currently the fuzz plugs I've published only use the first quarter of the response table, which reduce the used CPU power below 2%. Currently I work on a solution to use the full sine wave response on a low level CPU consumption. Duno, when I archive that, but then, we'll be there.
Well, working with the DK simulator means a lot of fun (for me, at least), and a lot of learning about analogue circuit's and the responding dsp expression. A lot of fun, if that is what interested you.
I know, you work yourself on a system to simulate non-linear responses, how far you are?
I must admit that we are indeed be able to simulate complete circuit's with there non-linear response, just, like pointed out above, this leads to a non acceptable CPU consumption. This may be different when we use a embed system like kemper does.
My current way is to split the response into strict linear parts, and implement the non-linear part with the help of a interpolation tables, between the linear responses of the IIR filters. The tables could be generated from the circuit simulation run, in non-realtime.
You've experienced with fir filters as far I know, do you come a step forward??
However, when ever you feel you like to give the DK simulator a run, and need any help to get started, let me know and I'll try to explain a bit better what we've done so far.
This invention is true for anyone here.
regards
hermann
Seems like we've a similar taste for music.funkmuscle wrote:Run one of those fuzz plugins in front of the cream machine and then use your neck pickup on your guitar and just try the guitar solo from Sunshine of Your Love from cream and you'll see you will almost nail the tone.
Well, I need to refine the fuzz plugs for what I've learned in the mean time. So, for example this is how a fuzz face response in the simulation run
To archive that, the plug needs more then 80% CPU power here, unacceptable.
Currently the fuzz plugs I've published only use the first quarter of the response table, which reduce the used CPU power below 2%. Currently I work on a solution to use the full sine wave response on a low level CPU consumption. Duno, when I archive that, but then, we'll be there.
A good slice of bread, with a taste of garlic and onion, some cream fresh and a slice of Serrano ham, oh, wow,CrocoDuck wrote:I am starting thinking that this plugin-from-schematics thing is literally the best thing since sliced bread. Gosh I still have to give it a try...
Well, working with the DK simulator means a lot of fun (for me, at least), and a lot of learning about analogue circuit's and the responding dsp expression. A lot of fun, if that is what interested you.
I know, you work yourself on a system to simulate non-linear responses, how far you are?
I must admit that we are indeed be able to simulate complete circuit's with there non-linear response, just, like pointed out above, this leads to a non acceptable CPU consumption. This may be different when we use a embed system like kemper does.
My current way is to split the response into strict linear parts, and implement the non-linear part with the help of a interpolation tables, between the linear responses of the IIR filters. The tables could be generated from the circuit simulation run, in non-realtime.
You've experienced with fir filters as far I know, do you come a step forward??
However, when ever you feel you like to give the DK simulator a run, and need any help to get started, let me know and I'll try to explain a bit better what we've done so far.
This invention is true for anyone here.
regards
hermann
On the road again.
Re: New GxCreamMachine.lv2
Experimenting, which is great fun. I am trying to clone a distortion pedal as of now. Got not so very good MISO Hammerstein FIR implementation which doesn't sound great. But I plan to be playing with some more in the weekend.tramp wrote:I know, you work yourself on a system to simulate non-linear responses, how far you are?
I will surely do as soon as I free up the time that this deserves!tramp wrote:However, when ever you feel you like to give the DK simulator a run, and need any help to get started, let me know and I'll try to explain a bit better what we've done so far.
-
- Established Member
- Posts: 363
- Joined: Sat Nov 17, 2012 3:36 pm
Re: New GxCreamMachine.lv2
I must congratulate you on the basic sound of this simulation. It sounds really good. Lots of harmonics and top. With this simulation, I can run a line of effects in Carla, using the CreamMachine as the last stage before the interface out.
I took the CreamMachine to the RPi. It compiles and runs beautifully, and latency is still very low. RPi CPU usage is between 25 to 50%. I put the RPi in front of a solid state Yamaha G100 amp on a 4x12" closed back cab,so the RPi basically functions as a complete modelling preamp and effects unit for a real amp. And the sound of the CreamMachine makes it inspiring to play the RPi setup.
About the amp schematics. I couldn't find any similar schematics on the web. Most amps use three valve stages before the power amp. Which amp were you inspired by? Another question. Am I right that there is no feedback loop between power amp and cab sim?
I took the CreamMachine to the RPi. It compiles and runs beautifully, and latency is still very low. RPi CPU usage is between 25 to 50%. I put the RPi in front of a solid state Yamaha G100 amp on a 4x12" closed back cab,so the RPi basically functions as a complete modelling preamp and effects unit for a real amp. And the sound of the CreamMachine makes it inspiring to play the RPi setup.
About the amp schematics. I couldn't find any similar schematics on the web. Most amps use three valve stages before the power amp. Which amp were you inspired by? Another question. Am I right that there is no feedback loop between power amp and cab sim?
Vox, Selmer, Yamaha and Leslie amplifiers. Rickenbacker, Epiphone, Ibanez, Washburn, Segovia, Yamaha and Fender guitars. Hammond, Moog, Roland, Korg, Yamaha, Crumar, Ensoniq and Mellotron keyboards. Xubuntu+KXStudio recording setup.
-
- Established Member
- Posts: 2080
- Joined: Mon Sep 28, 2015 8:06 pm
- Location: Here, of course!
- Has thanked: 227 times
- Been thanked: 400 times
- Contact:
Re: New GxCreamMachine.lv2
FWIW very few guitar amps have significant negative feedback, so that output stage is quite normal. Also the characteristic sound of these amps tends to be defined by two stages.
The output stage gives some compression and mostly odd-order distortion. it is heavily influenced by the O/P transformer and the cab loading.
The input stage (usually a triode) also gives some compression, but also adds quite a bit of even-order distortion which tends to 'thicken' the sound.
Any intermediate stages, phase splitters etc. add very little (if anything) to the character.
The output stage gives some compression and mostly odd-order distortion. it is heavily influenced by the O/P transformer and the cab loading.
The input stage (usually a triode) also gives some compression, but also adds quite a bit of even-order distortion which tends to 'thicken' the sound.
Any intermediate stages, phase splitters etc. add very little (if anything) to the character.
The Yoshimi guy {apparently now an 'elderly'}
-
- Established Member
- Posts: 363
- Joined: Sat Nov 17, 2012 3:36 pm
Re: New GxCreamMachine.lv2
I believe the majority of Fender amps use negative feedback to achieve the clean signature Fender sound. Marshall used it too, while Vox AC30 amps so do not. It's one of those things that are important to the final amp sound. On Vox' original blue series Valvetronix, the negative feedback loop was variable, depending on the amp model, which meant feedback on Fender models, no feedback on Vox models etc., and this made the amp models feel very realistic.
Vox, Selmer, Yamaha and Leslie amplifiers. Rickenbacker, Epiphone, Ibanez, Washburn, Segovia, Yamaha and Fender guitars. Hammond, Moog, Roland, Korg, Yamaha, Crumar, Ensoniq and Mellotron keyboards. Xubuntu+KXStudio recording setup.
-
- Established Member
- Posts: 363
- Joined: Sat Nov 17, 2012 3:36 pm
Re: New GxCreamMachine.lv2
Small bug report. At 96kHz it crashes Carla.
Vox, Selmer, Yamaha and Leslie amplifiers. Rickenbacker, Epiphone, Ibanez, Washburn, Segovia, Yamaha and Fender guitars. Hammond, Moog, Roland, Korg, Yamaha, Crumar, Ensoniq and Mellotron keyboards. Xubuntu+KXStudio recording setup.
-
- Established Member
- Posts: 2347
- Joined: Mon Jul 01, 2013 8:13 am
- Has thanked: 9 times
- Been thanked: 466 times
Re: New GxCreamMachine.lv2
Ups, was a typo. Fixed now in git.Frank Carvalho wrote:Small bug report. At 96kHz it crashes Carla.
The input stage is inspired by the Silvertone 1484 Preamp.Frank Carvalho wrote:About the amp schematics. I couldn't find any similar schematics on the web. Most amps use three valve stages before the power amp. Which amp were you inspired by?
The power stage I've created from several circuit's were I most ain't know now from what I've take what.
The cabinet stage is the Hughes & Kettner Cream Machine cab out circuit.
Yes, there isn't.Frank Carvalho wrote:Another question. Am I right that there is no feedback loop between power amp and cab sim?
On the road again.
-
- Established Member
- Posts: 2347
- Joined: Mon Jul 01, 2013 8:13 am
- Has thanked: 9 times
- Been thanked: 466 times
Re: New GxCreamMachine.lv2
. . and I forgot to thank you for confirm the quality of the simulation.Frank Carvalho wrote: must congratulate you on the basic sound of this simulation. It sounds really good. Lots of harmonics and top.
I had it running on my MOD, were it takes 6% CPU and it makes it hard to me, to stop playing.
On the road again.