I use a PIC -- same difference.....Not quite. With an arduino, you've got the admiration of the duct-tape wielding, POV loving, Maker hipsters.
If you're geek enough you can do both. Arduino software and hardware compatible with a 32 bit 80 MHz PIC chip:
Digilent Inc. - Digital Design Engineer's Source
Maybe they'll bring back Schaefer, Rheingold and Schlitz.
When we see the return of Red White and Blue or Falstaff then we know the end is near. We had the untimate in stink beer, Regal, brewed in downtown Miami, but I'm not sure it ever got distributed outside of south Florida. Disgusting stuff.
I did a project for a Microchip / Circuit Cellar magazine contest back in 2007 that used a dsPIC chip to measure and control the bias, allow several choices of operating points, modulate the B+ on the output tubes in real time to improve efficiency (Class H operation) and even test the tubes. The project won a prize, and was published in the magazine. Sadly, I never took the project much further than the basic functionallity needed for the contest, and it has since been robbed for parts. The publication is no longer on CC's web site, so I will eventually put it on my own site.
Schematics of the controller and amp are attached. Feel free to copy whatever you can use. My code sucks too bad to publish. Real programmers...even Arduino hackers laugh at me.....and I have a degree in computer engineering!
Page 1 is the dsPIC chip and EEPROM for saving settings.
Page 2 is bias controller. SPI commands are sent to the dual DACs which create a voltage from 0 to +5V. The opamps scale that into a 0 to -150 volt bias. There are 6 duplicate circuits. There were 4 output tubes and two drivers under PIC control.
Page 3 is the modulated power supply. The dsPIC samples the two audio inputs and generates a PWM pulse train with a duty cycle equal to the instantaneous audio amp. This circuit uses those pulse trains to control two buck converters that step down the +450 volt supply to the range of +150 to +420 volts in step with the audio. Think of it as a mega SE class D amp. This voltage is the B+ for the aUdio output tubes. The output tubes get only enough B+ to keep from saturating at any given moment. This whole system operates at a 500KHz to 1 MHz rate.
Page 4 is the sensing circuitry. I used an external ADC for monitoring the output tube current, B+ voltage, B+ current, and driver tube plate voltage. I needed the internal dsPIC converters full attention to digitizing the audio stream for the PWM generators. The fet in the lower left is for soft start or B+ shut down if things go wrong. It is tied to the CT on the power transformers HV winding.
The schematic for the tube amp being controlled is shown. There are two conventional stages, abd a cathode follower output stage. A CF output stage was used since it has a very high PSRR. The ability to reject power supply ripple is needed to remove any DSP artifacts that remained on the modulated power supply.
The amp and controller PCB pictures are included.
A few years after publication I received correspondence from the owner of a major tube amp manufacturer regarding an European patent application for a tube amp operated by a microcontroller. Several key claims were taken directly from my original contest submission (including the wording), which had been posted on the internet by CC. I discussed this with the owner, but could not get involved due to the employment contract with my current employer. I don't know the final outcome, but I don't see how those claims could stand.
I have started down the road to building a new design that will be completely controlled by the large ChipKit board mentioned above. I have been using these boards at work for automated IC testing, so it makes sense to use one here. This project will be done in small steps. The first step is bias monitoring and control, and a few of the features that I never got competely ironed out in the 2007 project. Ultimately I plan for a tube stereo power amp of 100+ WPC, a phono preamp (probably a chip) and a tube guitar preamp all controlled by the ChipKit. I will be starting a thread about it as it progresses, but things will be slow for now, and the controller won't be started until the amp is built.
Will the Arduino be accurate enough alone or do i need an amplifier between?
The Arduino can directly measure the drop across a small resistor in the output tube's cathode. I used 10 ohm. You need a RC low pass filter to remove the audio, and a 5.1 volt zener diode clamp to protect the microcontroller from a tube short.
Providing for Arduino based bias control is a bit harder. Some Arduino's provide a PWM based output that could be used to generate a variable voltage, but you would need to keep the 3.3 volt square waves out of the audio. I choose to use SPI to communicate with a multi channel DAC chip. The contest required the use of a Microchip DAC, but I choose TI today. Don't have the part number handy....I am at work.
You still need an opamp circuit to use the variable positive voltage from the PWM or DAC to control a negative voltage for the output tubes grid. Other functions can be controlled with relays or opto-isolators. I use LDR's for controlling the functions in the guitar preamp. Most are specced to 60 or 100 volts, but I have used them in 300 volt circuits without issue at low currents. Soft Pots can be used for low level controls but most are limited to 5 volts or less.
