I don't have the PCBs yet, but I did order a round of parts to make sure all the mechanicals were correct. Everything seemed like it is going to fit and function correctly with one exception. I haven't found a good solution for the JTAG cable yet. I ordered a pre-formed flat cable and IDC connector, but the gauge was a bit off and it would involve a lot of exact-o knife trimming to get right. All I need is a way to connect a 6 pin in-line .1" row of PCB holes to an identical set on the other board. For the JTAG end, the wire can just be soldered straight into the holes. For the sidecar end, I was thinking a pin-strip header, but digging through Mouser and Digikey gets confusing with a couple hundred thousands parts in those categories. If anyone has a suggestion on specific cable and pin-sleeve part numbers or has an alternate suggestion, I'm all ears.
I've seen discrete single wire jumpers with pin-sleeves on both ends at Fry's. I'm going to pick some of those up this weekend and solder in break-away pin-strips on both ends to begin with. Not the cheapest solution, but one that would work out of the gate for the first boards. Ideally the need to JTAG will diminish greatly after the first few weeks.
XT-IDE on PCjr
Re: XT-IDE on PCjr
PCBs arrived today. I already ordered 4 parts kits to send out to folks too. However I've already found several issues. One pretty severe. The drill size on the holes for the side car connector (~1mm) are a bit too small. Once the assembled board comes out of the dishwasher and I recharge my camera, I'll take a pic. The pins slide back in, but the friction locks are sticking to the point the pins don't seat by about 1.5mm. They stick out slightly from the socket. I've tried squeezing down the flares with needle nose pliers, but I'm afraid shy of respinning the boards, the only way to fix the issue is carefully filing down one of the two flares on all the pins. I'm pretty busy at work this week and possibly this weekend. I'd like to rework my board and do some basic hardware validation before I send out the other boards and kits - basically ensure the PLDs can be reflashed with the dongle cables.
Another similar issue involves the drills for the IDE connector. They were sized from the right angle female headers. Turns out the male right angle and male vertical alternates have thicker pins. That may kill the viability of this spin for folks (unless you're good at filing some more). I was planning on installing a 2.5" HD anyway but a cable or industrial flash module may be difficult with this prototype.
On the positive side, the boards look absolutely amazing. Everything fits mechanically in the side car box. The non-plated drills line up absolutely perfectly. The HD doesn't stick out as far as I anticipated. Some of the silkscreen font size got screwy but it's pretty minor.
Another similar issue involves the drills for the IDE connector. They were sized from the right angle female headers. Turns out the male right angle and male vertical alternates have thicker pins. That may kill the viability of this spin for folks (unless you're good at filing some more). I was planning on installing a 2.5" HD anyway but a cable or industrial flash module may be difficult with this prototype.
On the positive side, the boards look absolutely amazing. Everything fits mechanically in the side car box. The non-plated drills line up absolutely perfectly. The HD doesn't stick out as far as I anticipated. Some of the silkscreen font size got screwy but it's pretty minor.
Re: XT-IDE on PCjr
Nice!
My donor sidecar is ready to go. And yes, the connector was hard to get apart ..
(I still need to post my version of the BIOS for the original XT-IDE. That is coming.)
Mike
My donor sidecar is ready to go. And yes, the connector was hard to get apart ..
(I still need to post my version of the BIOS for the original XT-IDE. That is coming.)
Mike
Re: XT-IDE on PCjr
Wow, that looks good! I can't wait to see that booting up a hard drive. 
Re: XT-IDE on PCjr
I had some free time tonight to grind the pins down with a Dremel and finish assembling the board. I plugged it into my Jr just to see how things fit. The board connector probably need to shift over to the left a few hundredths, but I was able to get it plugged in and the holes registered in the stand-offs.
Even though the PLDs are not yet programmed, I decided to test power the Jr on with the board plugged in to make sure no magic smoke came out. And to my shock the damn thing counted up to 640K! I guess the RAM/ROM CE lines are registering low when the PLD isn't programmed causing all three to go active. And since the ROM is definitely going to have all the bits set, the RAM actually works (though both RAM chips will be mapped at every address in the 1MB range). I was happy though that at least the board appears to be working out of the gate.
I've got some things to take care of tomorrow (Thurs) night but I should be able to program the PLDs Friday night with my gang programmer and do some basic register access testing. Hopefully I can get the parallel port dongle tested by Monday and get the boards and kits out to folks by Tuesday at the latest.
Looking good so far.
Even though the PLDs are not yet programmed, I decided to test power the Jr on with the board plugged in to make sure no magic smoke came out. And to my shock the damn thing counted up to 640K! I guess the RAM/ROM CE lines are registering low when the PLD isn't programmed causing all three to go active. And since the ROM is definitely going to have all the bits set, the RAM actually works (though both RAM chips will be mapped at every address in the 1MB range). I was happy though that at least the board appears to be working out of the gate.
I've got some things to take care of tomorrow (Thurs) night but I should be able to program the PLDs Friday night with my gang programmer and do some basic register access testing. Hopefully I can get the parallel port dongle tested by Monday and get the boards and kits out to folks by Tuesday at the latest.
Looking good so far.
Re: XT-IDE on PCjr
I programmed the PLDs and the POST display came up and worked. Beyond that I need to reorganize my bench to get my JR and a 386 w/ a 360K drive next to each other in arms reach for further testing. Good old floppy net. At this point you will have to grind pins down to use these PCBs and if there is a major fault with this spin, those pins are wasted. So I'm not going to send these boards out until I know everything works 100%. And then I'd like to get some feedback on exactly what to do about the connector hole sizes. I could order a small board with some sample drills this weekend, but it would take another 2-3 weeks for them to arrive. I'm going to check around at work for some really really accurate calipers to see if I can make a better drill estimate. But different board houses have different drill steps so we may be screwed any way we go.
I'm not sure how much time I'm going to have tomorrow to work on this. If I can confirm the PLD programming dongles work (not using them atm), program the main PLD for I/O address IDE operation compatible with XT-IDE mk1 at a fixed address, program the existing XT-IDE image into the flash part with an external programmer, and the hard drive comes up and works, I'm going to declare victory and and leave it up to you guys. I'll keep you posted.
I fine with eating the cost of this board spin if we want to do a round 2 as it has several problems. But if I get my prototype board working with some rework, there will be less incentive to front a second spin - especially with the pin friction issue lingering. I'm open to suggestions.
I'm not sure how much time I'm going to have tomorrow to work on this. If I can confirm the PLD programming dongles work (not using them atm), program the main PLD for I/O address IDE operation compatible with XT-IDE mk1 at a fixed address, program the existing XT-IDE image into the flash part with an external programmer, and the hard drive comes up and works, I'm going to declare victory and and leave it up to you guys. I'll keep you posted.
I fine with eating the cost of this board spin if we want to do a round 2 as it has several problems. But if I get my prototype board working with some rework, there will be less incentive to front a second spin - especially with the pin friction issue lingering. I'm open to suggestions.
Last edited by alanh on Sat Jul 23, 2011 9:15 pm, edited 1 time in total.
Re: XT-IDE on PCjr
I have an idea. You could order a set of small drill bits and drill sample holes in a piece of scrap to fit the pins in. Then you'd know the proper size of hole from the size of drill bit used to drill the hole.
Re: XT-IDE on PCjr
Alan,
The sidecar connector pins were a real pain in the rear to separate originally. The pins are a standard spacing, and I've seen IDC connectors used as a cheap substitute.
I'm going to look in my DigiKey catalog, but if we could get a female IDE connector and stand it off the board far enough we could avoid reusing these awful AMP connectors.
The sidecar connector pins were a real pain in the rear to separate originally. The pins are a standard spacing, and I've seen IDC connectors used as a cheap substitute.
I'm going to look in my DigiKey catalog, but if we could get a female IDE connector and stand it off the board far enough we could avoid reusing these awful AMP connectors.
Re: XT-IDE on PCjr
Another status update, though it isn't much of one. I've been having difficulty finding time to actively work on this lately. I'll have some time this week, but the next two weekends are definitely out as I'm moving and have to make an out of state trip. I did scratch a few hours out this past weekend and have some good news and a lot of bad news. The good news is I programmed the flash and got the XT-IDE BIOS up. The bad news is a disastrous layout error. I have the IDE connector layout flipped. I was able to get a hard drive plugged in and connected properly with a cable I just happened to have - however the drive still wasn't properly recognized and I have some doubts about the cable integrity. Bottom line it means given all the problems with this spin of boards, they are pretty much junk. I plan on ordering a second revision, however I just don't want to do that until I get other things proven out. And that's what is going to take time. Here are some things that I think should happen before a second ordering of boards:
1) I need to ohm out the drive cable I have and make sure there are no connection problems with it.
2) Do some basic poking around on IDE registers to see if I'm really hitting the drive
3) Validate the RTC
4) Validate in-circuit programming
5) Find an explanation to why D7 is pulled low on the XT-IDE schematic & resolve other differences in the XT-IDE layout I believe are incorrect (and why)
6) Investigate why DASP isn't driving the hard drive LED on power up. It should at least blink on drive power-up. I get nothing on the scope.
7) Rework the IDE connector layout for the proper drive orientation
8) Misc board layout changes: IDE header drill sizes, more ground coupling, JTAG header change (BB.II compatible).
9) I need help reproducing a XT-IDE BIOS build environment - where is the latest code, what compiler, compiler settings, etc
10) Have one or more other people review my schematic and board layout in detail (desperately need another pair of eyes here)
11) Have one or more other people review my code in detail (desperately need another pair of eyes here too)
12) Figure out a accurate way of estimating the side car IDC connector drill diameter or find a connector alternative.
The last four items are the most critical - especially #12. Any help here would be appreciated. I think the rest I can have done by the end of the week in-time for the next dorkbot order deadline. I'm only going to order 3 boards this time until I have an actual Jr booting off a hard-drive. However this means another 3 weeks (from next weekend) turn time. So this, unfortunately, is progressing slower than I had hoped.
-Alan
1) I need to ohm out the drive cable I have and make sure there are no connection problems with it.
2) Do some basic poking around on IDE registers to see if I'm really hitting the drive
3) Validate the RTC
4) Validate in-circuit programming
5) Find an explanation to why D7 is pulled low on the XT-IDE schematic & resolve other differences in the XT-IDE layout I believe are incorrect (and why)
6) Investigate why DASP isn't driving the hard drive LED on power up. It should at least blink on drive power-up. I get nothing on the scope.
7) Rework the IDE connector layout for the proper drive orientation
8) Misc board layout changes: IDE header drill sizes, more ground coupling, JTAG header change (BB.II compatible).
9) I need help reproducing a XT-IDE BIOS build environment - where is the latest code, what compiler, compiler settings, etc
10) Have one or more other people review my schematic and board layout in detail (desperately need another pair of eyes here)
11) Have one or more other people review my code in detail (desperately need another pair of eyes here too)
12) Figure out a accurate way of estimating the side car IDC connector drill diameter or find a connector alternative.
The last four items are the most critical - especially #12. Any help here would be appreciated. I think the rest I can have done by the end of the week in-time for the next dorkbot order deadline. I'm only going to order 3 boards this time until I have an actual Jr booting off a hard-drive. However this means another 3 weeks (from next weekend) turn time. So this, unfortunately, is progressing slower than I had hoped.
-Alan