keenerb wrote:

Also, Sergey seems to think it's possibly an issue with the flash ROM as well. I may move my ROM to a lo-tech ROM board.

thats interesting. I need to see if i have a smaller rom. I think I could try putting the xtcflite + bios rom onto the xtcflite board and removing the rom from the xi8088. wonder if that would work.

whats your SW2 memory configuration switch settings on your xi8088?

SW2 is all-off right now. The xi8088 rom needs to be at a particular address though, my xt-ide card can't be configured for that address. I'm looking at picking up one of James' dedicated ROM boards but they seem to be out of stock.

So, the two memory ICs have memr/memw as ACTIVE LOW. When I pull that chip, isn't that potentially causing an ACTIVE signal, or an indeterminate signal on that line? I don't think there's a pullup resistor/capacitor/whatever on that line, is there?

Is this possibly why the 1mb ram board is having problems?

Or are those inputs only?

*edit*

Never mind, those are inputs to the chip.

Whoa, I'm lucky I say down and sanity checked it this evening. I had the buffer wired backwards! 🙁

Here is a schematic I ended up with:

The board is live on Osh Park at https://oshpark.com/shared_projects/kRd1DfRG and a set of three can be ordered for just over six bucks.

I wish to stress in the strongest possible way that this is untested, very lightly reviewed, has tight clearances between components and has a possible range of outcomes that vary from fixing everything to doing absolutely nothing to causing your xi 8088 to explode. I'm going to order a set and see what happens.

Scott

I"m aware. I"m going to compare it to James' design and give it a quick once-over.

Note that the above board is for JSOP36 SRAM, like the AS7C4096. I picked those because there were 10ns parts available, and it seemed like another variable to try.

We could also invent a board that used STSOP32 or TSOPII-32 parts like the AS6C4008 SMD variants. Wouldn't be too hard to do, and it would have the advantage of being the same SRAM that we are using now.

Scott

I should be able to breadboard this, right? I can pull the existing chip and use a spare ACT245 and a bunch of jumper wires. Think that will throw off timing or anything?
What should the OE and CE lines be connected to on the 245 IC? Should they simply be spliced into the OE/CE lines into the memory IC?

keenerb wrote:

I should be able to breadboard this, right? I can pull the existing chip and use a spare ACT245 and a bunch of jumper wires. Think that will throw off timing or anything?
What should the OE and CE lines be connected to on the 245 IC? Should they simply be spliced into the OE/CE lines into the memory IC?

Yes, you ought to be able to breadboard it. Assuming you're mating one 74ACT245 to one SRAM chip, then connecting the OE and CE as you describe I think should work.

It's going to be a lot of jumper wires. 😁

Tried again to get my 8ch logic analyzer on my Rhode & Schwarz scope going, but it must add too much impedence to some lines. if I attach to clk or osc, system wont boot. If I'm on MEMR/AEN/ some others, I cant get past bios. Think I'm just going to have to get my Agilent 1670G out and try that with its 132 channels.

It's weird you're seeing such issues with loading on the lines. I don't have any problems on my X1 or X10 probes, and I've been able to attach resistors to most the lines, to +5V and/or to ground. The only one that gave me any grief was IRQ6, which seems to cause the boot sequence to halt after emitting the boot music.

yeah I think my board must have something else going on that your guys boards dont. I may see if I can switch my 74F tranceivers to 74HCT or something.

smbaker wrote:

It's weird you're seeing such issues with loading on the lines. I don't have any problems on my X1 or X10 probes, and I've been able to attach resistors to most the lines, to +5V and/or to ground. The only one that gave me any grief was IRQ6, which seems to cause the boot sequence to halt after emitting the boot music.

It's interesting you mention this.

My xi8088 NEVER boots the first time; it only boots after I hit the "RESET" button.

It plays the boot music and stops responding, never initializes video. Once I hit RESET it's all good.

BloodyCactus wrote:

yeah I think my board must have something else going on that your guys boards dont. I may see if I can switch my 74F tranceivers to 74HCT or something.

I remember you mentioned you jumpered 5V to ground through the speaker connector, just like I did.

That caused me MASSIVE issues with random lockups and weird behaviour; when I replaced my 00 and 32 series ICs the problem went away.

You can find that back on page 8 or so of this thread.

I'd highly recommend doing the same, it's entirely possible that you damaged one of those chips just like I did...

keenerb wrote:

It's interesting you mention this.

My xi8088 NEVER boots the first time; it only boots after I hit the "RESET" button.

It plays the boot music and stops responding, never initializes video. Once I hit RESET it's all good.

This is interesting. Do you have a POST display? When I was messing with that IRQ line and it froze after the music, it would always get stuck on POST code 06.

Other than that, I've not had any recurring problem with power-on boot.

I haven't tried in my Malinov backplane lately. I'll do that. 06 seems familiar.

keenerb wrote:

I remember you mentioned you jumpered 5V to ground through the speaker connector, just like I did. […]
Show full quote

BloodyCactus wrote:

yeah I think my board must have something else going on that your guys boards dont. I may see if I can switch my 74F tranceivers to 74HCT or something.

I remember you mentioned you jumpered 5V to ground through the speaker connector, just like I did.

That caused me MASSIVE issues with random lockups and weird behaviour; when I replaced my 00 and 32 series ICs the problem went away.

You can find that back on page 8 or so of this thread.

I'd highly recommend doing the same, it's entirely possible that you damaged one of those chips just like I did...

this has been niggling me ever since it happened, I just dont have a bunch of chips spare to test it. I should do this before anything else really. sigh

BloodyCactus wrote:

keenerb wrote:

I remember you mentioned you jumpered 5V to ground through the speaker connector, just like I did. […]
Show full quote

BloodyCactus wrote:

yeah I think my board must have something else going on that your guys boards dont. I may see if I can switch my 74F tranceivers to 74HCT or something.

I remember you mentioned you jumpered 5V to ground through the speaker connector, just like I did.

That caused me MASSIVE issues with random lockups and weird behaviour; when I replaced my 00 and 32 series ICs the problem went away.

You can find that back on page 8 or so of this thread.

I'd highly recommend doing the same, it's entirely possible that you damaged one of those chips just like I did...

this has been niggling me ever since it happened, I just dont have a bunch of chips spare to test it. I should do this before anything else really. sigh

I really would; all you have to do is go back and look at the problems I experienced to get pretty suspicious...

I remember coming across this thread before I built my Xi 8088. I made sure to pull one of the header pins out of my speaker header, just to ensure that could not happen to me.

I've found a new-in-box 640k XT memory card in my stash. Ordering some 41256 ram right now to populate it, I've been meaning to test it out anyway. The manual DEFINITELY states it'll fill from 0k - 640k on an xt system, so we'll see what happens.

If it still fails I'll breadboard the memory buffer this weekend, so we should have a final answer on both the buffer chip design and whether a memory expansion card should be expected to work with the xi8088...

I also received my 12mhz DMA controllers in the mail, so that'll be added to the list as well.

http://wiki.osdev.org/ISA_DMA

According to this, DMA will run at 4.77mhz, "NO EXCEPTIONS" on an ISA system, so this was probably a waste of $16, but we'll try anyway.