DIY Hammond Novachord

Hi all. Haven’t posted here in a while again!

I’ve been wanting to do this for years. I had an old piano in the back room that wasn’t doing anything so I took the keyboard out of it and put the rest in the garden. It was a junk piano, not worth anything.

One of these boards has two Phantastron oscillators (no divide down here) and two differential pairs as VCAs (each board has two notes). The sets of pads off to the far right are at the bottom of the board, and will have nickel strips soldered to them to become leaf switches. These switches will be switched by the keyboard and trigger the VCA, passing the oscillator’s signal to the main amplifier. I’ll need 44 of these for the whole keyboard.

They all connect via DB25 connectors to a backplane. Needless to say, the filament current will be pretty high (132A as a preliminary estimate) so there’s a heavy eyelet screw terminal on each segment, which will connect to an individual transformer to power that section.

The schematic for each section. This is for two notes so I’ll ignore the lower half. The left pentode is the oscillator and is fine tuned with RV1 to the correct frequency, and all notes are connected to a reference voltage to give a global tuning control. The envelope generator is borrowed heavily from the original Novachord circuit.

Here’s a description of how it works:

There are two voltage references - A and C - which are connected to the same pot. When the pot is in the Slow position, A is at 0V and C is at -100V. When it’s in the Fast position A is at -100V and C is at 0V.

Normally, J1 (cutoff) is closed and J2 (attack) has 2/3 closed. The Cutoff voltage is about 20V, so that is what C4 is charged to, preventing current from flowing through U1 (a 6J6 dual triode) and therefore no sound output from the oscillator.

If the attack pot is set to slow, C5 will be charged to 0V. When the key is pressed, J1 opens and J2 connects pins 1 and 2. C4 and C5 are connected and the -100V on the C voltage cause them to slowly rise towards that voltage.

If the attack pot is set to fast, C5 will be charged to -100V. When the key is pressed, J1 opens and J2 connects pins 1 and 2, as before. C4 will rapidly charge to C5’s voltage (and peak at about half that voltage) and slowly decay to the C voltage, which is at 0V.

I’ve not drawn a KiCad schematic of the output amplifier because I don’t need to make a PCB for it. However, I’ve got a diagram of it in LTspice. The signal bus, SIG+ and SIG- exit to the left and connect to all of the notes. The 6DJ8’s are there to keep the voltage constant on the bus, and to amplify the current being sunk into it. The difference in the current on the bus is the signal we want. The differential pair underneath converts the differential signal back into a single ended signal and removes as much common mode signal as possible. The output goes to the right, off to a filter or a line output to an amplfiier.

Here are a few sound samples from LTspice:
https://www.dropbox.com/s/lms9uxyk453pb0b/oscvca%202%20.wav?dl=0

I have no idea if I’ll be able to finish it, but here’s hoping!

10 Likes

Any ideas yet how to implement that?

Yep. Each segment of 8 modules will have its own 10A 6.3V transformer.

Use only in winter :slight_smile:

1 Like