Thanks 640K!enough,

Appreciate you taking a look. Agreed, let's get the perfect fit. I'm of the same mindset. I'll be standing by.

The Tariffs are an unfortunate turn of events beyond our control. It reminds me of that scene in 2010: The Year We Make Contact where doctor Floyd and the Russian Commander have an exchange:

"Listen, just because our governments are behaving like asses doesn't mean we have to. We're supposed to be scientists, not politicians."

I'm sure we can all find the answer.

I forgot one major point: we're missing the stamped/etched jack labels; that's a big part of the reason for doing this. I will edit my previous list so it's all in one place.

ARGUS logo would be cool too, wouldn't it?

hard1k wrote:

ARGUS logo would be cool too, wouldn't it?

It might be nice, but I don't think we even want to know what the price would be. It would also be a little difficult to get the alignment right for all cases; it would have to be pretty small.

matze79 wrote:

it does not honor legacy mode.
disabled isa pnp for isa slot, and now it seems to stay on my Settings.

Can you put the pnp setting back as it was and try "iwinit -v9 -legacy > iwinit.txt", then post both IW.INI and IWINIT.TXT (or send them to me privately, if you prefer)? If you're willing to help with this, it would be nice to figure out where this is coming from, and maybe put together a modified IWINIT to get around the problem. I don't have a machine that has this problem, so testing may be somewhat tedious if the problem isn't immediately obvious.

matze79 wrote:

its annoying that i need 2 IRQ's. So i can't use LPT Port anymore if i want to use SB Emulation 😀

Often enough, the LPT IRQ isn't used. Also, I don't think it will cause too many problems, especially if you don't usually print (or use other parallel-port devices that need the IRQ) while IWSBOS/Mega-Em is loaded.

Unless there is some condition that requires a separate IRQ for the emulation, you can usually use the same value for both IRQ settings.

matze79 wrote:

Well, Wolfenstein 3D sounds very different 😳 on megaem, on sbos it just freezes.

In most cases, IWSBOS produces better sound than Mega-Em for FM emulation. Of course, that usually just means that it is a little less likely to make your ears bleed. It is not anything that can be described as authentic OPL2 sound.

vittek wrote:

Appreciate you taking a look. Agreed, let's get the perfect fit. I'm of the same mindset. I'll be standing by. […]
Show full quote

Appreciate you taking a look. Agreed, let's get the perfect fit. I'm of the same mindset. I'll be standing by.

The Tariffs are an unfortunate turn of events beyond our control. It reminds me of that scene in 2010: The Year We Make Contact where doctor Floyd and the Russian Commander have an exchange:

"Listen, just because our governments are behaving like asses doesn't mean we have to. We're supposed to be scientists, not politicians."

I'm sure we can all find the answer.

It's just a question of finding the most economical option (and not paying the tariffs if we don't have to). Even without the tariffs, if shock__ agrees, and wants to ship them with the cards, it would likely be less expensive (from the standpoint of brokerage, import fees, etc.) to have them made in an E.U. country. It could also be that production in China is so much cheaper that, even with the fees, it makes sense to do it there.

There are a few European companies that do that sort of work. If they don't charge exorbitant tooling/engineering fees, that could be worth looking into. There is even a German company, with production in Germany, so maybe someone who speaks the language fluently can give them a poke at some point, to see what their rates are like. I won't mention them by name until/unless we have a fluent volunteer.

Of course, it could also be that shock__ will want absolutely no part of this. Then we'll have to figure out another way to handle distribution of the brackets.

Guess who's back 😀
I'm currently still a bit busy with another high priority project for the Vintage Computing Festival Berlin 2018, but should be able to pick up work on the ARGUS by the mid of next week.

Hey shock__ ! Glad you are back 😀

shock__ wrote:

Guess who's back 😀

It's good to see you back.

I've had a bit of a hard disk mishap, so my efforts will be stalled for a little while. Fortunately, I managed to salvage my data (the source code most importantly), but will have to re-build the rest of my usual work environment on a temporary, smaller hard disk for the time being.

I'm really looking forward to seeing where this goes. I currently run a GUS in my 486 but wouldn't mind an additional one or two for other builds I'm in the middle of/planning on. Keep up the great work, guys!

Okay, the other high priority project is finished for now (can you guess what it is? https://i.imgur.com/xnLt8bm.png)
Work on the ARGUS will resume tomorrow.

^ what is it? 😁

Hint: Top are IR diodes and IR phototransistors ... where would you see rows of 12? 😜 Just don't try to decypher which machine it's for ... it's a reproduction of a part for a learning aid "computer" (8 opcodes, 7 bit data!) from 1974. Supposedly only 2 prototypes were produced (or at least are still known to exist) and the whole thing seems quite philosophical rather than a full blown functional machine ... also all the documentation for trainers/trainees apparently has been lost (or was never written).

Just started on keeping up with developments and suggestions. I can say this so far:
- non-PNP modifications will be implemented, even tho I don't see how that would work yet. I guess it would fetch the enviroment settings from iw.ini? A friend of mine has a non-PNP 386 mainboard, so I'm quite excited to test this.
- JP4 / JP5 will get a cleanup as proposed by 640k!Enough
- socketed Op-Amps: Nope sorry, that area is already quite crammed. If they need changing: cut them off, clean up the pads and solder in a new one 😉
- uncramming C725, C718, C721 & C718 - done already
- additional audio in: Nopes, there's already AUX and Line-In
- headers for Line Out, Mic, Line In jacks - maybe, as said that area is quite crammed and I don't want to reroute most of the analog out at this point. Sorry.
- thermal relief for ground pads - gotta have to look into that, I think the options on thermals in my version of eagle are extremely basic, manually reworking might be possible.
- add receptacles for X3 and the header option of U31 - to be implemented. Good suggestion.
- add support for other SIMM slots - likely to be implemented
- additional 0.1uF capacitor for CFILT - to be implemented

Brackets: I'll get back to that ... I already have a printout of a proposed layout contributed by vittek, but that needs some minor adjustments (spacing between DSUB and audio jacks are fine, but the overall bracket is off). About distributing them ... I have to admit I have rather little interest in that ... we'll see at a later point, but don't expect miracles.

Paper tape reader?

ab0tj wrote:

Paper tape reader?

Almost ... punched cards. Fun fact: SOL1 & SOL2 are for a solenoid ... "transport" of the cards is done by gravity by falling through the slot til the solenoid (relais with an additional wire soldered to one of the contacts) closes and holds it in the slot.

shock__ wrote:

- non-PNP modifications will be implemented, even tho I don't see how that would work yet. I guess it would fetch the enviroment settings from iw.ini? A friend of mine has a non-PNP 386 mainboard, so I'm quite excited to test this.

The idea is that the board would then be initialised by software at start-up (or at some other convenient time). There are a few approaches to getting this to happen, but we shouldn't need to worry about this, as it would ideally be appropriately handled by iwinit or the corresponding Windows drivers. I have tested it, and it works, for the most part, but since I don't have a non-PnP machine, there were a few hiccups that I am unable to diagnose. Also, there have been some reports of iwinit not being able to assign the requested resources (using the legagy option under DOS, with PnP enabled), but, again, since I don't have access to the unco-operative hardware, I can't investigate.

There is another way to activate the non-PnP mode: with a specially-crafted EEPROM. This would mean that no hardware modifications are needed, but I haven't tried this yet, and have no idea if/how it works.

shock__ wrote:

- add receptacles for X3 and the header option of U31 - to be implemented. Good suggestion.

I think the easiest approach here would be to convert X3 into a through-hole receptacle, if that doesn't create a routing mess on the other side of the board. The problem is a lack of support with the tiny pads that are on the current prototype, making unfortunate accidents more likely. I'll wait for revised measurements before designing a new ROM module.

shock__ wrote:

- additional 0.1uF capacitor for CFILT - to be implemented

Further reading seems to indicate that this should be a ceramic capacitor. There is no reason not to use a 0805 SMD part, like so many others on the board; C0G/NP0 are likely most suitable.

For ceramics in analog side you should only ever use C0G/NP0 types, others are completely not suitable.

640K!enough wrote:

shock__ wrote:

- add receptacles for X3 and the header option of U31 - to be implemented. Good suggestion.

I think the easiest approach here would be to convert X3 into a through-hole receptacle, if that doesn't create a routing mess on the other side of the board. The problem is a lack of support with the tiny pads that are on the current prototype, making unfortunate accidents more likely. I'll wait for revised measurements before designing a new ROM module.

Wouldn't that collide with the height of the 1.27mm grid headers for U31? IIRC through hole pinheaders are considerably higher than those with a smaller grid. I'll check into that tho.

CFILT capacitor has been added to the layout and one of my prototypes modified to allow easy switching between PnP and non-PnP operation. Gonna test it in a few hours at a friends place.

shock__ wrote:

Wouldn't that collide with the height of the 1.27mm grid headers for U31? IIRC through hole pinheaders are considerably higher than those with a smaller grid. I'll check into that tho.

I meant a 1.27mm-pitch through-hole receptacle; I should have specified. I have had them sitting around for a few months, but there are parts that match perfectly with the low-profile headers I used on U31: Digi-Key's 609-3753-ND, for instance. They are just plain through-hole receptacles, with no additional board guides or other unnecessary features. At the time, they were also less than $1 each, I think.

Now that I look at the differences, however, moving the extra lines to U31 might actually be a better idea. The reason is that 50-pin, 1.27mm-pitch, 2-row SMD receptacles are much more common, and much less expensive, than their 44-pin counterparts. When I first started the design, Digi-Key had some 44-pin parts that were only $2.39 each, which wasn't so bad. Now, those are non-stocking items, and the least expensive comparable part is over $10 😲.

If you're even considering the conversion, in order to get everything to fit into a 50-pad U31, I guess we could sacrifice RAHLD#. In such a case, whether you leave X3 in place, for simplicity's sake, or scrap it is up to you.

Just a short roundup of yesterday's experiences:
Non-PnP operation generally works, but various pieces of software fail along the way (setup.exe and play.exe both can't find the card and don't allow forcing legacy mode, therefore the edits to autoexec.bat must be done manually) - luckily iwinit.exe works. Demos/games usually seem to work fine.
I'll likely add the non-PnP mode as an "experimental" feature in hardware via a 2 position resistor (instead of a [solder] jumper)

Other findings:
- On my friends DX4-100 somehow the card forces the computer to hang with a memory fail beep on ctrl-alt-delete
- On his DX-40 midi did not work for some reason (even tho that may have been a configuration issue ... didn't dig too deep into that as my main focus was the PnP operation [digital sounds worked fine] and we were repairing another machine along the way).

Proposed new design for former JP5:
https://i.imgur.com/rM5NUhp.jpg

Is that clear enough to understand?