Jump to content

wileyc's Photo

wileyc

Member Since 22 Dec 2012
OFFLINE Last Active Yesterday, 3:35 PM

#4164969 CollectorVision Phoenix Pre-Order is now live!

Posted by wileyc on Tue Nov 27, 2018 3:23 AM

I'll also keep an eye on the Yahoo Japan auctions.  Controllers seem to come up pretty frequently for between ¥1000 and ¥1500; I'll grab 'em as they come around.




#4164934 CollectorVision Phoenix Pre-Order is now live!

Posted by wileyc on Tue Nov 27, 2018 1:24 AM

Bad news, folks ...

 

... I hit both Super Potato and that smaller place out towards Ueno on Chuo-dori, and neither of them had any of the NTT controllers in stock (although they had some controllers with additional sliders that I had never seen before).  They didn't know when they'd be back in stock.

 

I'll keep checking every week or so, buy them when they appear, and let you all know when that happens.

 

Sorry again ...




#4163436 CollectorVision Phoenix Pre-Order is now live!

Posted by wileyc on Sat Nov 24, 2018 9:13 AM

I live in Tokyo, not too far from Akihabara (and thus Super Potato).  I don't have a problem with grabbing a few controllers and shipping them over to y'all.




#4159813 N.Y. Times 1983: The Coming Crisis in Home Computers

Posted by wileyc on Sun Nov 18, 2018 7:35 AM

Well, I'm certainly on the fence now about buying a TI.  Even though I love Bill Cosby, who was the best father figure a kid could want :)




#4157900 OT: I screwed up!

Posted by wileyc on Thu Nov 15, 2018 12:57 PM

This is appropriate for a TI-99/4A forum?  Shame on you.




#4149015 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Fri Nov 2, 2018 11:30 PM

The regulator output of 5v is correct and appropriate. 

 

... which puts me on a design tangent track.

 

The voltage output as measured from the output pin against board ground is actually +5.7VDC (on the HRD+ at least, IIRC the other variants too).  This is done to compensate for the voltage drop across the battery backup protection diode, and is done by connecting the 7805 ground pin to ground through a same-spec diode.

 

It's a neat trick to make a non-adjustable VR output a custom voltage, but drop-in switching replacements for the 7805 don't like it (definitely the CUI, maybe the Murata too).  The Right Thing To Do ™ is to replace the 7805 with, say, a switching LM317 workalike and set it for 5.7.  That would require reworking the board, hence my silence when Opry99er mentioned the immediate heat concern.

 

If folks want, I can speak to the germanium diode wired-OR mess in a future post.

 

End design tangent :)




#4148306 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Fri Nov 2, 2018 7:40 AM

I figured you were sleeping when you mentioned a TL866...lol...I Know nothing!!! (Schultz)... about it..so, you're suggesting we need to check the condition of these..or as in ?? What

 

*laugh* It's only 2130 here.  Not time to sleep yet.

 

A TL866 is a Chinese EPROM programmer that also can perform dead-chip tests on most 74xx logic (most notably omitting the 74xx245).  If you're getting into the TI, I strongly recommend getting one.  They're less than fifty bucks on eBay.

 

Get one of those, plug it into a Windows machine, install the Chinglish software, and do a logic test on all of your chip pulls (including memory).  Replace the bad stuff with either HCT or ACT chips (power saving, rail-to-rail, yadaa yadaa yadaa).  Enjoy the goodness thereby.




#4148297 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Fri Nov 2, 2018 7:16 AM

So I'm trying to understand a little here, is the 1 Meg upgrade the same as calling the board a Rambo board, and using the same os? Or would that be a completely different board altogether? I'd love to do the 1 mb option with this board, if it can handle the logic. I'm not sure which ram to keep or does all of it need replacing?

 

So, to use my one gig scheme, you'd pull all of the memory chips except for the ROS 6264, as well as the '154, and run what used to a) be the chip-enable for the '154s (plural) from the 259 stack to the daughterboard, and the synthesized address lines from the same stack to the same board.  We're basically taking advantage of the board originally being designed to handle memory in 512k chunks; plug in 512k/8 SRAM and the 154 nonsense goes away.

 

All that is for later, though.  Let's get your board matching the HRD+ spec and working before we Frankenstein it.  I'd really recommend picking up a cheapo TL866CS (and crossflashing it to the -A variant if you're up to it) to see what all you have to replace.




#4148283 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Fri Nov 2, 2018 6:56 AM

I think I they are from old Russian production, so they're clones of the official chips.

*bwahahahaha* Then they will *never* succumb to silicon rot.




#4148241 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Fri Nov 2, 2018 5:50 AM

Welcome back Chris.

 

*chuckle* Thanks for the welcome.  I feel like I owe a few people apologies; I was not a pleasant person the last time around and I've taken steps to correct that.




#4148232 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Fri Nov 2, 2018 5:25 AM

For what it's worth, here's a picture of my hacked up HRD+ after I restored it to documented specs.  That's one meg in two chips; theoretical max is four.
 

HRD



#4148160 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Thu Nov 1, 2018 11:45 PM

I see from this post in 2016 they were discussing the three stack 138's and two stack 259's and 154 dilemma
https://www.disavowe...hip-stacks.html

That's my site. Go ahead and pull the schematics from there.


#4147579 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Thu Nov 1, 2018 8:51 AM

I have plenty of new-manufactured 154s. Rochester Electronics still produces them on demand, and I bought about 1,500 of them (the minimum order) a couple of years ago. Mine are the .300 width chips, not the wider, .600 style.

There's a really neat way to take the 154s out of the equation entirely. Basically what you do is get a couple of the right size SRAMs, wire the 154 chip select from the 259 stack into the CE of each SRAM, and run the synthesized address lines in. I'll post schematics when I return from this deployment.


#4147434 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Thu Nov 1, 2018 3:14 AM

Yep, that's the board that's covered by the HRD+ doc I put up. If you've got a TL866, you can pull each chip except for the soldered stacks and do a logic test for everything except the '245s I think.

What I ended up doing was desoldering everything except the RAM, replacing all the DIP sockets with machined sockets, and swapping the dead '154 with a small circuit board that incorporated a SMD CMOS '154 (the DIP is only NOS AFAIK).

That was the first iteration. When I return from this deployment, I'll visit the ex, grab the board, and send photos. All of my stack replacements and RAM plugin are available on OSHPark.


#4145956 HRD (Horizon Ramdisk) 2018

Posted by wileyc on Tue Oct 30, 2018 9:14 AM

There's a lot more that needs to be rewired (thinking specifically of the 259 stack, but the 138s and the 254s are going to be wrong too), and there's probably a cut trace that needs to be restored between the fingers and the rest of the card.

Seriously, please take it from one who fixed a HRD that matched no known wiring diagram into a stable HRD+ testbed that I used for prototyping OSHPark boards to eliminate stacking and bring it up to a minimum of 2 meg. Make it look like the instructions; there's a reason they switched from HRD wiring to HRD+. Treat it like a fresh build with all the parts present and you're just touching up bad solder joints and flywires icon_smile.gif