Hello people,
this is my first projekt, that I post here. I want to build a very robust guitaramp. In the past I have built some hifi amplifier, so I am not a bloody beginner 😉
The headline says it. I plan to use irfp240/9240 Mosfets for a 50watt Amp. And I like to give my projekts names, so I call it 'Gimos50'.
There are some conditions for Gimos50:
- it should be oversized to have some reserves
- soft clipping
- stable work
With LT-Spice is this schematic caused. Gimos50 has a very low open-loop-gain. There are high lokal currentfeedback in the voltagegain. The negative-feedback-factor is very week. The result is a stable circuit, it don't need great frequency compensation.
The z-diodes in the negative feedback are limit the output voltage. About 25V should be enough for 50watt at 8ohm.
The overcurrent protection limit the output current to 2A per Mosfet. There should be enough reserves for continuous load.
this is my first projekt, that I post here. I want to build a very robust guitaramp. In the past I have built some hifi amplifier, so I am not a bloody beginner 😉
The headline says it. I plan to use irfp240/9240 Mosfets for a 50watt Amp. And I like to give my projekts names, so I call it 'Gimos50'.
There are some conditions for Gimos50:
- it should be oversized to have some reserves
- soft clipping
- stable work
With LT-Spice is this schematic caused. Gimos50 has a very low open-loop-gain. There are high lokal currentfeedback in the voltagegain. The negative-feedback-factor is very week. The result is a stable circuit, it don't need great frequency compensation.
The z-diodes in the negative feedback are limit the output voltage. About 25V should be enough for 50watt at 8ohm.
The overcurrent protection limit the output current to 2A per Mosfet. There should be enough reserves for continuous load.

Last edited:
i have made a FFT with LT-Spice. This pictures shows 1kHz, 1watt and 50watt:
This simulation shows the cut-off-behavior:


This simulation shows the cut-off-behavior:

strange 😕
I can see it.
This is the url:
http://www.diyaudio.com/forums/gallery/uploads/86962/Gimos50.jpg
I can see it.
This is the url:
http://www.diyaudio.com/forums/gallery/uploads/86962/Gimos50.jpg
strange 😕
I can see it.
This is the url:
http://www.diyaudio.com/forums/gallery/uploads/86962/Gimos50.jpg
I get webpage not found.
Ok, I uploaded it on other sever :
Hostarea Gimos50.jpg – Kostenloses Bilder Hosting
An I hope it work...
edit: it can't be true 🙁
Hostarea Gimos50.jpg – Kostenloses Bilder Hosting
An I hope it work...
edit: it can't be true 🙁
This is the layout of the first board(100x160mm). I work with kicad, it is open source and I like it.
The space is not filled with components, so I consider to shrink it.
The space is not filled with components, so I consider to shrink it.
An externally hosted image should be here but it was not working when we last tested it.
The new layout has not gotten smaller, but better.
The key tracks are shorter.
There are fewer bridges.
The differential amplifier has now a thermal coupling.
To protect the acid I have closed the open areas as much as possible.
The key tracks are shorter.
There are fewer bridges.
The differential amplifier has now a thermal coupling.
To protect the acid I have closed the open areas as much as possible.
An externally hosted image should be here but it was not working when we last tested it.
This is the prototype. I built and tested it with different Signals.
I have shorted the output at full level, but the amplifier has no problem with it.
25Vss - 1kHz - 8ohm:
15Vss - 1kHz - 4ohm:
12Vss - 10kHz - 8ohm parallel 1uF:
As you can see, the circuit works very stable. But the short-circuit protection is limited to early. It's easy to customize, so that the amplifier brings the required 50 watts.
And on the triangle signal can be nice to see the various distortions. The peaks are nice rounded and the edges are slightly bent.
That was not so hard. Now comes the jfet preamp and this will be a hard nut.
I have shorted the output at full level, but the amplifier has no problem with it.
An externally hosted image should be here but it was not working when we last tested it.
25Vss - 1kHz - 8ohm:
An externally hosted image should be here but it was not working when we last tested it.
15Vss - 1kHz - 4ohm:
An externally hosted image should be here but it was not working when we last tested it.
12Vss - 10kHz - 8ohm parallel 1uF:
An externally hosted image should be here but it was not working when we last tested it.
As you can see, the circuit works very stable. But the short-circuit protection is limited to early. It's easy to customize, so that the amplifier brings the required 50 watts.
And on the triangle signal can be nice to see the various distortions. The peaks are nice rounded and the edges are slightly bent.
That was not so hard. Now comes the jfet preamp and this will be a hard nut.
Since the preamp is based on JFETs and JFET are subjected very high tolerances, do I have to set each step with trimmers.
This is probably the reason why commercial manufacturers avoid JFETs.
In the design I will find my way close to tube circuits. And there can be observed that in the gain stage frequencies strongly emphasized to 10kHz. Exactly what will make my circuit as well.
I already have a first idea, but before I post here a half-baked plan circuit, I will test it first. Should something useful come out , I will report back.
This is probably the reason why commercial manufacturers avoid JFETs.
In the design I will find my way close to tube circuits. And there can be observed that in the gain stage frequencies strongly emphasized to 10kHz. Exactly what will make my circuit as well.
I already have a first idea, but before I post here a half-baked plan circuit, I will test it first. Should something useful come out , I will report back.
I have made a lot of tests with different Jfets, but the results were not really good. So I went back to the diodes. My vavorit circuit have a very soft limit in the first stage. The limit have a window of +/-12V, there is enough headroom for clean play. The second stage is harder, it cuts the signal above 700mV and mix some additional overtones, like a tube. The effect of this circuit is similar like an oktaver. I call it 'tubemachine' and 'snot'. After the preamp comes a 5-band-equalizer, discret and low-noise. With a footswitch you can switch from clean/crunch to hard/heavy. This is my test constuction:
This is the circuit diagram from the preamp:
And this is the equalizer:
I have also made some sound samples, you can find it on my site at the bottom.
Patrick's kleine Homepage
When I finished the Amplifier, I will make a video on youtube. I hope a friend of mine will play guitar for me.
An externally hosted image should be here but it was not working when we last tested it.
This is the circuit diagram from the preamp:

And this is the equalizer:

I have also made some sound samples, you can find it on my site at the bottom.
Patrick's kleine Homepage
When I finished the Amplifier, I will make a video on youtube. I hope a friend of mine will play guitar for me.
As long as the power supply voltages are under ±50V, I would try using the IRFP9140. The Rds on is a better match with the IRFP240.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Solid State
- Gimos50 - a robust guitaramp with IRF240/IRFP9240