HRD (Horizon Ramdisk) 2018

[GDMike]

I'm seeking schematics other than online schematics posted at whl and forums. I have a 3000 I think..all i can say is

I have 1 ea 154 chip, I have 3 ea 138 chips stacked.. I am able to power up the card (main red LED) says so. And in explorer I can turn on address 1000 (data LED blue) says so. And I can turn it off. When I place data at 4000 I get zeroes... when I place data at 5000 I get zeroes. The only schematic I can find that's close to mine shows different connections to the 3 ea 138 chips than what my original 138 chips were wired with under working conditions a while back, as it did work at one time. Help, what wire goes where?

Edited September 20, 2018 by GDMike

[RickyDean]

I'm seeking schematics other than online schematics posted at whl and forums. I have a 3000 I think..all i can say is

I have 1 ea 154 chip, I have 3 ea 138 chips stacked.. I am able to power up the card (main red LED) says so. And in explorer I can turn on address 1000 (data LED blue) says so. And I can turn it off. When I place data at 4000 I get zeroes... when I place data at 5000 I get zeroes. The only schematic I can find that's close to mine shows different connections to the 3 ea 138 chips than what my original 138 chips were wired with under working conditions a while back, as it did work at one time. Help, what wire goes where?

I do not believe there are any clearer schematics for the Horizons released into the wild. The ones for my 198+ model and the 3000 and 4000 models are pretty much what Bud released to the hands of those who chose to build their own. I built three of them using the included construction guide and they worked (at least the two I got complete,d before I stored everything away 20 years ago. Yep, just looking at them last night as I was going to try to find you something, till I found that Whtech had the same, brought back a lot of memories, good and bad). If the wires are not shown or talked about, maybe it was one of the various hacks for the Horizon, and I know nothing about those. Sorry if I didn't give you anything concrete to work with.

[apersson850]

I'm not at home, so I can't check. But when I discovered that the later DSR software on the Horizon isn't according to TI's specifications, and thus doesn't work with the p-code card, I did enough research to write my own DSR for the Horizon disk. So at that time, I did figure out how to control the bank switching to get the right memory in the right place.

Thinking about it, didn't I publish that driver here? Maybe not...

[GDMike]

At this point I'm in need of good schematics! I've got 3 stacked 138 chips, and a couple more support chips, like stacked 259s..ive also got an expansion Jack and a U1 which appears to be another junction port like the expansion Jack that runs by wire over to another junction port U10..all this has got me looped! Lol. I've got a schematic for a HRD 3000 and the layout for the 3 ea 138s is close but not the same as the way this boards 138 stack is when it was working. In other words, things are close but not exact. But the 3000 board shows exactly the same chip count and ID as my board. So it's confusing. Thx

[RickyDean]

I got my Horizon 4000 RD working last night(had a bad VR and Diode). It checks out in Megtest and the last option on the menu(can't remember, believe it has to do with checking all ramdisk), but am seeing an issue.

When I try to load the configuration file it reads from the drive and locks up or goes blank screen, without doing anything. I previously used the same files to check and configure my HD 3000, I repaired last week. I had saved that configuration on this disk, would that create a conflict that would cause a lockup here? 3000 versus 4000? Now I am using a disk image on a HxC, so it's not a faulty disk issue.

[RickyDean]

Well, last night I finally found that the 8k dsr sram chip had 17 errors show up on the U11 test, this would explain the locking up condition. After I swapped the sram with the one from the Horizon 3000 board, the 4000 works, but needs battery work and the 3000 works, but needs a new 8k sram. One of my, I guess '2000' series, works, but needs a new CRU switch and battery holder work. Progress is being made, one diode at time.

[+retroclouds]

Well, last night I finally found that the 8k dsr sram chip had 17 errors show up on the U11 test, this would explain the locking up condition. After I swapped the sram with the one from the Horizon 3000 board, the 4000 works, but needs battery work and the 3000 works, but needs a new 8k sram. One of my, I guess '2000' series, works, but needs a new CRU switch and battery holder work. Progress is being made, one diode at time.

 

Can you elaborate on how you tested the 8k dsr sram chip? I obtained a HRD 3000 and it immediately locks up on the title screen.

 

Already did multiple tests and this leads me to believe, either the sram or one of the logic chips has died

• Batteries are ok, tested outside and when installed in the ramdisk as well-
• dip-switch settings checked and ok as well.
• If I do manage to get pass the title screen and run CFG834, it either locks up or shows bogus information.

I've now gone so far to get new logic chips and an SRAM so I can replace one by one if required.

 

Edited October 4, 2018 by retroclouds

[+retroclouds]

I read on Thierry's page, that there are 2 utilities available for testing the HRD after building. They are called

MEMTEST and MEGTEST

are these available somewhere ?

[RickyDean]

I read on Thierry's page, that there are 2 utilities available for testing the HRD after building. They are called

MEMTEST and MEGTEST

are these available somewhere ?

If you have the ability to access disk images, here is one that would have it.

http://atariage.com/forums/topic/281651-horizon-ram-disk/page-1 post# 21

 

 

I tested it by putting another onto the 4000 and it worked correctly and accepted the ROS. Then on the 2000 series Horizon, (The last one that I have ever had worked on by Bud Mills in the mid 90's) I plugged into the PEB and ran the u11 test(didn't even realize this was there till I plugged this in and went into the Megtest selection). It ran and reported 17 errors in the 8k chip. But then last night I tested the same chip in my MiniPro and it tested normal, don't know about that yet. I will be trying to test it again in the 4000 tonight.

Edited October 4, 2018 by RickyDean

[RickyDean]

Well the 8k and another, that I dug up in a quick look through my stuff, are definitely bad. I checked them in three different version of the Horizons and had errors show up in each, though not the same amount of errors, two showed 18 errors and 1 showed 4 errors. But the strange thing is, my MiniPro eprommer, tested both these chips as normal. What would the difference be attributed to?

But i do have 3 8k's that are on three of the Ramdisks and they all check good, transferring them between the 3 working disks, I now have, but working only with power, I still have to fix the battery issue's on each. Original battery holders came off in storage, or were too damaged by leakage, or never mounted in the case of 1. Progress though.

[+retroclouds]

RickyDean, would you mind taking a short video of the Horizon Ramdisk 3000 in action? In particular I'm interested in seeing when the the LED located at CR1 is on and when it is off.

 

I have a still non-working HRD 3000.

The behaviour I'm seeing is that the CR1 LED flickers in irregular intervals as soon as I boot up the TI-99/4a, which I don't think should be happening.

If the switch in location SW1 is turned off, the TI-99/4A works as expected and there is no conflict with the Disk Controller, but ofcourse CFG834 doesn't recognize the HRD.

 

By now I have replaced the following components

U24 - SN74LS245N Motorolla I8514B           ok
U23 - 74LS244 PC F 8534 KOREA               ok
U22 - 74LS244 PC F 8534 KOREA               ok
U21 - SN74LS156N Motorolla I8543            ok
U20 - SN74LS138N Motorolla I8049            ok
U19 - SN74LS138N Motorolla I8049            ok
U18 - SN74LS138N Motorolla I8049            ok
U11 - JAPAN 8703 HM6264LP-12 U0422YY0       ok
U10 - SN74LS259N Motorolla FFXG8945         ok
U9  - SN74LS259N Motorolla FFXG8945         ok
U2A - AD +B9024 G MM74HC154N                ok
U1  - SN74LS138N Motorolla I8049            ok

After replacing these components, the card still doesn't work. Only difference being that the TI-99/4a no longer locks up on boot up.

It does lock up while loading anything from disk drive 1 (e.g. CFG834). So by now I wonder if there is some electrical leakage that messes with one of the busses causing CR1 to flicker like a christmas tree.

 

 

 

EDIT: The HRD 3000 is in slot 4.

Edited October 11, 2018 by retroclouds

[RickyDean]

Yes I can take a small video. But I do have one question, switch 1 is cru 1000, do you have any conflicting cards with that cru?

[+retroclouds]

Yes I can take a small video. But I do have one question, switch 1 is cru 1000, do you have any conflicting cards with that cru?

 

No, no conflicting cards. Also tried removing all cards except 32K and Disk Controller, no difference. Tried different CRU's already, no difference.

[RickyDean]

Now one other thing that you might try is using another switch.If it's doing the same thing in either position, the switch might be non-functional, have seen it happen. Meanwhile I will get a video tonight.

[RickyDean]

Well the video didn't happen, sorry, on Wednesday night, waiting out the hurricane I was trying to get my HDX, that I had built plugged into the PEB, so I could load the DSR, and try it out. It is tigt in that bow and I didn't see that an aluminum brcket that was mounted to the RS232 was bending in while I was plugging it in. The result took out a couple of components as well as the PEB. It still powers but it will not light up any cards, though those same cards plugged into another PEB do light. My PEBs have been stored a long time and I have to work out bugs. I do have a switched PS in one and will jumper regs to see if I can video this for you, over the weekend.

 

Have you tried running Megtest on it? With the HRD814F disk you auto load with XB, then click on the 3rd item on the list and go into megtest then it will ask you how many 32k chips you have, common HR 3000 Ramdisks usually have 12 and if you are using 128k chips, they each = (4) 32kchips. If you don't get the this far check the transistors first. I think there are two on the 3000, but would have to look at the construction manual. Also, I don't have any knowledge of CFG834, have only used 814F, but you might try the config option, # 1 on the menu and see if it show the cards cru address, if not this would also point to a hardware issue, such as a bad switch.

Edited October 12, 2018 by RickyDean

[+retroclouds]

Now one other thing that you might try is using another switch.If it's doing the same thing in either position, the switch might be non-functional, have seen it happen. Meanwhile I will get a video tonight.

 

I tested the switch with my multimeter and it's looking ok so far. Don't expecting any issues there.

Did order a few switches as this one is not in a so good condition, waiting for these to arrive, just to be safe.

[+retroclouds]

Well the video didn't happen, sorry, on Wednesday night, waiting out the hurricane I was trying to get my HDX, that I had built plugged into the PEB, so I could load the DSR, and try it out. It is tigt in that bow and I didn't see that an aluminum brcket that was mounted to the RS232 was bending in while I was plugging it in. The result took out a couple of components as well as the PEB. It still powers but it will not light up any cards, though those same cards plugged into another PEB do light. My PEBs have been stored a long time and I have to work out bugs. I do have a switched PS in one and will jumper regs to see if I can video this for you, over the weekend.

 

Have you tried running Megtest on it? With the HRD814F disk you auto load with XB, then click on the 3rd item on the list and go into megtest then it will ask you how many 32k chips you have, common HR 3000 Ramdisks usually have 12 and if you are using 128k chips, they each = (4) 32kchips. If you don't get the this far check the transistors first. I think there are two on the 3000, but would have to look at the construction manual. Also, I don't have any knowledge of CFG834, have only used 814F, but you might try the config option, # 1 on the menu and see if it show the cards cru address, if not this would also point to a hardware issue, such as a bad switch.

 

Hope your PEB and cards are ok doing well. As far as my HRD is concerned, did some measurements with my multimeter on the diodes and I believe to have an issue with both CR4 and CR5.

I repeated my measurements multiple times to be sure. CR4 and CR5 seem to be conducting in both ways. Also checked CR6, CR7, CR3, CR8, CR9; no issues there.

 

On whtech I found the horizon 3000 construction guide. At some point it says:

 

We have found the germanium diodes CR4-CR7 can be the problem. The diodes can be replaced with Radio Shack Cat. No. 276-1123

 

So my question is, what is a good replacement for these germanium diodes in 2018 ?

[+Ksarul]

The Radio Shack part is a 1N34A. You can probably substitute with a 1N914 or a 1N4148. . .

[Stuart]

The Radio Shack part is a 1N34A. You can probably substitute with a 1N914 or a 1N4148. . .

 

1N914 and 1N4148 are silicon diodes with a larger forward voltage drop than the germanium 1N34A. May be significant if they're in the battery backup circuit.

[Shift838]

1N34A Germanium diodes can still be found on ebay at a fair price. If you want them quick then Jameco carries a line of them HERE

[RickyDean]

 

Hope your PEB and cards are ok doing well. As far as my HRD is concerned, did some measurements with my multimeter on the diodes and I believe to have an issue with both CR4 and CR5.

I repeated my measurements multiple times to be sure. CR4 and CR5 seem to be conducting in both ways. Also checked CR6, CR7, CR3, CR8, CR9; no issues there.

 

Well, the PEB PS will need to fixed or replaced, and in the process of trying another P-box, I had my SAMs card blow a couple chips, it was regulated so, it must have been feeding on the ground plain or somethng, especially the 612 chip. But in that process I saw my SAMS acting like you video, so it begs the question, does your HRD 3000 have the 32k mod on it? that may explain the flicking light.

[+retroclouds]

Well, the PEB PS will need to fixed or replaced, and in the process of trying another P-box, I had my SAMs card blow a couple chips, it was regulated so, it must have been feeding on the ground plain or somethng, especially the 612 chip. But in that process I saw my SAMS acting like you video, so it begs the question, does your HRD 3000 have the 32k mod on it? that may explain the flicking light.

 

no, I do not have the 32k mod.

[RickyDean]

Well, I do have the one regulated P-Box, so I will try to find my jumper-ed flex card and jumper the 3000 and see if it exhibits anything like this behavior, but the other day it was also plugged into the box that ate my SAMS, but the Sams is the only one that exhibited a ozone smell, at the tie it was not flickering as the Sams. Though I do believe it also blew a minor chip or transistor on my Corcomp DSDD controller. Oh the pain....the pain!!!

[RickyDean]

I'm seeking schematics other than online schematics posted at whl and forums. I have a 3000 I think..all i can say is

I have 1 ea 154 chip, I have 3 ea 138 chips stacked.. I am able to power up the card (main red LED) says so. And in explorer I can turn on address 1000 (data LED blue) says so. And I can turn it off. When I place data at 4000 I get zeroes... when I place data at 5000 I get zeroes. The only schematic I can find that's close to mine shows different connections to the 3 ea 138 chips than what my original 138 chips were wired with under working conditions a while back, as it did work at one time. Help, what wire goes where?

GDMike, I just realized this, you can tell if it's a 3000, by looking at the card with the regulator down at the lower right. If you look at the right side of the card it will have the number 3000 or 3000B printed on the board,as well as some other info as well as the 4000.

[+retroclouds]

 

Hope your PEB and cards are ok doing well. As far as my HRD is concerned, did some measurements with my multimeter on the diodes and I believe to have an issue with both CR4 and CR5.

I repeated my measurements multiple times to be sure. CR4 and CR5 seem to be conducting in both ways. Also checked CR6, CR7, CR3, CR8, CR9; no issues there.

 

On whtech I found the horizon 3000 construction guide. At some point it says:

 

We have found the germanium diodes CR4-CR7 can be the problem. The diodes can be replaced with Radio Shack Cat. No. 276-1123

 

So my question is, what is a good replacement for these germanium diodes in 2018 ?

 

 

The card is working fine now. I replaced CR4 and CR5, that alone did not do the trick. Same behaviour as before.

I checked CR9 with the multimeter again and it was acting strange. I desoldered it, tested it isolated with the multimeter and it appeared fine.

As I did not have replacement 1N914 diodes, I soldered the same 1N914 back on the board again, and now the card is working. Still going to get some replacement 1N914's though.

Quite happy that it is working now, took me quite some effort, but that's part of the fun.