lacuna vrm

Description: A Doepfer LMK3+ keyboard was purchased from a seller in Dunedin through TradeMe, a well-known auction site in New Zealand. The keyboard is equipped with a Fatar keybed resembling a traditional piano, housed in a robust flight case. The unit was acquired at a lower cost due to the absence of a MIDI interface, which the previous owner removed and placed in a separate box after switching to a different keyboard. The flight case lacks labels, preventing confirmation that it is indeed an LMK3+ model rather than the original LMK3 version. Communication with Doepfer revealed that production of the LMK3+ ceased in 1993 and that they do not have compatible parts available. Therefore, the construction of a custom interface for the keybed is necessary. The keybed has been removed from its case and cleaned, revealing a label that likely contains the serial number and assembly date, but no part number for technical specifications was found. The key switches operate by breaking contact with one jumper and making contact with another when a key is pressed, with the timing of these actions determining key velocity. However, the mapping of the 88 key switches onto a 40-pin connector and the complete circuit design remain unknown. The aftertouch sensors are believed to function on a piezoelectric principle, but the reason for the presence of two strips and the circuitry required for their integration is unclear. An update indicated that the keybed may be an earlier version of the TP/10MDF, and a request for specifications has been submitted to the manufacturer. Despite a lack of response from both Forte Music and Fatar, an alternative solution was discovered through Hinton Instruments in the UK, which offers an FPGA capable of velocity-sensitive keyboard scanning. Although not as comprehensive as other solutions, an evaluation board from Hinton Instruments could serve as a good starting point for building a custom interface, potentially allowing for keyboard splitting and various MIDI effects. Information from Doepfer regarding the connection of FATAR TP/7, TP/8, and TP/9 keyboards, including schematics and pinouts, has been reviewed. These models utilize two normally-open switches per key, indicating that the LMK3+ may utilize a "change-over" type switch, which aligns with the observed "break-before-make" operation. If the keyboard employs a diode matrix, the available schematics may provide sufficient information to determine the pinouts on the IDC connector. Additionally, it was noted that the aftertouch sensor operates as a force-sensing resistor.

The Doepfer LMK3+ keyboard employs a sophisticated mechanism for key detection and aftertouch sensing. The keybed utilizes a break-before-make switching principle, which is essential for accurately capturing key velocity. Each key is equipped with a dual-switch configuration, ensuring that the pressing action is translated into a digital signal that can be processed for velocity sensitivity. This switching mechanism is likely integrated into a matrix configuration, which allows for efficient scanning of the 88 keys through a limited number of connections.

The absence of the MIDI interface necessitates the design of a custom interface to facilitate communication between the keybed and external MIDI devices. This involves mapping the key switch outputs to a 40-pin connector, which may require the use of diodes to prevent ghosting effects during simultaneous key presses. The integration of the aftertouch sensors, which operate on a piezoelectric principle, will also need to be considered in the design, as they provide additional expressive control over the performance.

The exploration of the KV64 FPGA from Hinton Instruments presents a viable option for implementing the required functionality. This FPGA can handle velocity-sensitive key scanning and can be adapted to work with the unique characteristics of the LMK3+ keybed. The evaluation board available from Hinton Instruments can serve as a foundational platform, allowing for further development of additional features such as MIDI effects, keyboard splitting, and the incorporation of control knobs and buttons.

In summary, the project involves reverse engineering the LMK3+ keyboard's keybed and developing a custom interface that meets the performance requirements of modern MIDI applications. The integration of existing schematics from related FATAR models will aid in understanding the necessary connections and configurations, ultimately leading to a successful implementation of a fully functional MIDI interface.I bought a Doepfer LMK3+ keyboard from a very pleasant chap in Dunedin, via TradeMe, a popular New Zealand auction site. The keyboard features a nice, piano-like Fatar keybed in a rugged flight case. Here`s a photo of another LMK3+, taken from a recent auction on German eBay : I didn`t pay very much for mine, because it was missing the MIDI interface.

The previous owner liked the knobs and wheels so much that, after changing to a different keyboard, he had the interface removed and placed in a stand-alone box. So, the end of mine looks like this: Also, the flight case is devoid of labels, so I have no way of confirming that it`s actually an LMK3+ (with the uprated keybed) rather than an original model LMK3.

I have emailed Doepfer, and they told me that they stopped producing the LMK3+ in 1993 and that they have no suitable parts for it. Consequently, it looks like I will need to build my own interface to the keybed. (I`m assuming that if the previous owner was prepared to part with the interface, he would`ve put it back in the keyboard and sold it as a complete working unit for a lot more money.

) After taking the keybed out of the flight case and cleaning it up, I was able to find the label with (I assume) the serial number and the date of construction ("montaggio" apparently means "assembly"): I was hoping to find a part number that I could use in my quest for technical specifications on the keybed interface, but I haven`t seen anything like that. I can see how the switching for the keys works: when a key is pressed, a piece of wire breaks a contact with one jumper and then makes contact with a second jumper.

(This is hidden behind the leftmost aftertouch sensor in the photo above. ) The timing between the break and the make gives you the key velocity. What I don`t know is how the 88 key switches are mapped onto a 40-pin connector, or what the complete key switch circuit looks like. I believe that the aftertouch sensors operate on some sort of piezoelectric principle, but again I don`t know why there are two strips or what sort of circuit would be needed to incorporate them.

Update: After looking around the FATAR website (which I think is currently being constructed ” certainly I couldn`t find it last week), I think that my keybed is an earlier version of the TP/10MDF (see below). So I have filled in their contact form and asked for some specs for the interface. We`ll see what they say. Update 13 March: Well, no reply from anyone at Forte Music. Or Fatar, for that matter. But I`ve found (via the synth-diy mailing list) Hinton Instruments in the UK, who have the KV64 FPGA that does velocity-sensitive keyboard scanning.

It isn`t as complete a solution as the FMI-MK3, but they do have a nice-looking evaluation board that would be a pretty good start. And I`d be happy to build the rest of the interface myself; all sorts of cool keyboard splitting & "MIDI effects" ideas are possible, not to mention adding lots of knobs and buttons.

Actually, looking at the Doepfer MKE page again, they have information on connecting FATAR TP/7, TP/8 and TP/9 keyboards, including schematics with pinouts. And apparently these keyboards have two normally-open switches per key, and are thus specifically not the "change-over" type (which I guess means one normally-closed switch and one normally-open switch).

Which would in turn imply that my keyboard is a "change-over" type, since I can clearly see the "break-before-make" operation on the left-most key. And since the KV64 is apparently designed for "changeover contacts", it might indeed be compatible with my keybed.

If my keyboard is wired as a diode matrix (and there are a lot of diodes visible on the circuit boards), then there`s probably enough information in those schematics for me to nut out some of the pinouts on the IDC connector. Luc Hondareyte informed me that the aftertouch sensor is in fact a force-sensing resi