|Capacitance probably is the issue. At one time Intel produced a DMA Controller that had the same problem. Customers complained it didn't work|
and the Intel engineers said it did. However they had never run it without a scope on a certain pin. Eventually they figured it out and changed the
tech note to include having a small capacitor on that pin. Congrats for intuiting the issue.
On Sep 4, 2012, at 9:46 AM, norwe...@aol.com wrote:
I am sending this email with great apprehension (expecting a lot of criticism), but maybe it will help someone else? ..... I finally got my dual IDE card working. It was a struggle. First, the adapters that John mentioned (from NewEgg) are the ones to use. I tried some from DealExtreme, but they don't seem to work with this board, although they work fine elsewhere. Second, the diode, D16, is a mystery to me. I tried several different kinds (I've never seen any specific type specified), and whatever I put in the card seemed to starve the TIL311's for power. One digit would sorta' work, but when I stuffed more than one digit into the board, most of them just displayed a "-". I finally just installed a wire jumper. That did the trick. Now all 6 digits light up as they ought to. Did I do something wrong? And lastly, the CF select circuit was a major headache. I put a 'scope on the output of the 74LS02 (pin 1) to see what was going on, and suddenly it started working. Without the scope, any attempt to select the "B" drive wouldn't work. The two inputs to that gate (WR* and SEL_SECOND*) looked OK to me, so I didn't know what to do. Finally I figured that the capacitance of the 'scope probe was the answer, so I put a small cap. from pin 7 to pin 1. That seems to work OK, but I'm not feeling really confident about it. I'm not an EE type, so I'm fumbling in the dark. Another strange thing that happened to me is that at first I used a couple of 128 MB SanDisk compact flash modules (cosmetically identical). One worked fine with John's diagnostic, but the others I tried (I have a number of them) errored out (sector not found) with attempts to address sectors beyond 20h. These CF's have 8 heads, 32 sectors, and 980 cylinders, so maybe that is to be expected? But, why did one work?