I just upgraded the firmware in my newly bought MKS-70 from 1.03 to 1.08. I’m not actually sure what the differences are, but since I have the possibility to program EPROMs there was no reason not to.
However, the engineers at Roland decided to put the EPROM in a position that makes it impossible to pull it straight out. The EPROM is located at the mainboard in the bottom of the MKS-70 and marked with an A. The two voice cards (same as used for the JX-10) are stacked upon that. It is possible to fold away those cards, but a lot of cables has to be detached. I should mention that there each voice card also has an EPROM, these are didn’t have to be upgraded.
The TOP853 couldn’t burn the TC57256D-20 EPROM that are used in the MKS-70 and JX-10. My advice is not to buy the TOP853 if your’re going to use it for programming. The GQ-4X did it after I added a custom line to a new text file called customdevices.txt
I know it’s a bit insane, but I got a very good deal and bought an MKS-70. The MKS-70 is the rack version of JX-10. The JX-10 contains two JX-8P. So at the moment, I have a total of five JX-8P engines!
It seems to be fully working, has a few scratches on top and the front panel looks very nice. However, when booting it and pressing the VALUE button the screen read “Ver 1.03 FINAL JX”, which is not the last version of the firmware. Regarding “FINAL JX”, it’s kind of strange because it was not the last JX synth produced (JX-1 was the last one, though it’s not analog) and it was not the last firmware version either. Just like on the JX-10, the firmware was very buggy on the MKS-70, but they got it all sorted out with version 1.08. I’ll have to open the MKS-70 and check what kind of EPROM it uses.
I just wanted to share how the SCI Drumtraks creates the closed hihat (ch) sound from the sample that is actually an open hihat (oh).
The hihat chip in the Drumtraks is a single 2764 containing a whopping 8 kb of samples. Remember that electronics where much more expensive back then, and costs must be cut. One method was to use the same sample for the oh and cc. This was the case on the Linn LM-1, Oberheim DMX / DX, Drumtraks and probably a lot of other digital drum machines from this era.
The trick is to use the last part of the oh sample and add some envelopes to it. Below is the Drumtraks hihat chip loaded in Adobe Audition.
The red marker marks sample number 4096, which is exactly half of the total 8192 samples. It’s approximately here the ch starts. But, when recording the ch, it doesn’t look exactly like that. Take a look at the recorded audio below:
To make the ch fade out quickly, an envelope is added by the Drumtraks. If you don’t believe me, compare this image to the first one, the peaks are easy to identify.
What about the oh, does it only play until sample 4096? The answer is no, the oh plays the whole 8192 bytes. Take a look at the screenshot below:
The first part is the oh recorded from the Drumtraks. Last part is the data from the chip. As you can see, they are ‘identical’. This means that when the oh is played back from the Drumtraks, the whole sample is used.
Notice how different it sounds when processed by the Drumtraks, and that’s the reason for having a Drumtraks. 🙂
Then after a few uses it went all bananas again. At this point I still suspected some kind of “software error” because of some faulty CPU or memory. I disconnected everything, part by part. Still didn’t work. Then I disconnected the lower main board (the DPX-1 has two main boards). The upper board contains the CPU, memory and midi. The lower board contains all audio specific stuff like filters and outputs. As soon as I disconnected the power to the lower board, the DPX-1 booted perfectly. Quite fun that the upper board can work without the lower one powered up.
Anyway, as soon as I connected the lower board again the DPX-1 wouldn’t start. If I booted without the lower board and then connected it, the DPX-1 would hang. So there was something fishy about the lower board. I measured all voltages and they were stable.
So I disconnected the power to the 8-output and cd-rom expansion card and the DPX-1 booted just fine. At this point I was very happy, but that didn’t last long since I had no output. Nothing but very low distorted audio came out of the original single output.
Next thing to try was to remove the expansion card and then I discovered that the cable going from the lower board to the expansion had to be connected to a connector (J8) on the main board below the expansion card. So now there’s a big hole in the back of the DPX-1 – but I can live with that. I don’t think the 8-output expansion is usable anyway. The cd-rom is very rare and only reads cds that also are very rare. The eight outputs work the same way as on the Emax. Each voice has its own output, so every time you hit a key that sample is heard in the next free output. Eight voices = eight outputs. I’ve heard that there are some EII disks with drumkits and stuff that actually maps one particular sample to use one particular output, but I haven’t found one yet.
A couple of months ago I scored three M-16C on German eBay. My intention was to do the famous M-16C to M-64C conversion, since an M-64C costs more than the three M-16C together. Plus it’s good soldering practice.
When I got the JX-10, it had no factory sounds – the internal memory was all messed up. To solve this two things are required:
A Roland M-64C cartridge. The JX-10 can only take full dumps, and those can only be done to a cartridge. Once on the cartridge however, they can be copied to the internal memory.
Installing the new firmware is the easiest part. The harder part is to modify an M-16C. So if you choose the easier and more expensive path, just buy an M-64C.
I’ve found two guides the original one and another one based on the original one. The original guide has a nice description, but very low-res photos. The other one has good pictures but a not much text.
The first thing to do is to desolder the old memory chip from the M-16C and for me this was the hardest part. It took about an hour and I used a solder sucker.
Another thing was to actually understand where to solder each lead. Some are easy to see, some are not. One that shouldn’t be missed is the one that you have to solder beneath the chip before soldering the chip. Here’s a photo from one of the guides, it’s the red lead. It’s connected to the second pin in this photo, very hard to spot, but if you look carefully it’s visible between the blue ones.
You also have to bend a few pins upwards, this is quite clear in the text in the original guide, in the other guide you can see it in this photo.
The final thing to do is to modify the casing, one of the guides recommends a Dremel, for me a filet knife did the job.
There are some instructions scattered all over the internet, sosummarized it here: