VOGONS


Octek hippo vl+ 3.02 no POST no Beep no Code

Topic actions

Reply 40 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

Yeah good point, this is isa clock that is showing on the card. Well with 8mhz I can confirm the clock on the cpu pin so this is good enough to post. Have the fix the post issue first before looking at cpu clock gen again.

I ordered the card but from China will take a while.

As for the cpus yeah I tried a few different ones that I use for testing motherboards so they are ok

Reply 41 of 62, by rasz_pl

User metadata
Rank l33t
Rank
l33t

just to be sure remove everything but the cpu and post card, that means no keyboard, no ram, no graphic card. You can even pull keyboard controller chip off

https://github.com/raszpl/FIC-486-GAC-2-Cache-Module for AT&T Globalyst
https://github.com/raszpl/386RC-16 memory board
https://github.com/raszpl/440BX Reference Design adapted to Kicad
https://github.com/raszpl/Zenith_ZBIOS MFM-300 Monitor

Reply 42 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie
rasz_pl wrote on 2023-03-11, 00:51:

just to be sure remove everything but the cpu and post card, that means no keyboard, no ram, no graphic card. You can even pull keyboard controller chip off

Just tried it, nothing

Reply 43 of 62, by Chkcpu

User metadata
Rank Member
Rank
Member
treeman wrote on 2023-03-10, 10:49:

I took the bios chip off and read it with my tl866 and it shows data, keyboard controller is showing voltages and no shorts, can't verify signals without a scope.

Please put the BIOS dump you made with the TL866 up here.
If it is an AMI or Award BIOS, I should be able to tell you if it is valid or corrupted by checking the BIOS checksums.

Jan

CPU Identification utility
The Unofficial K6-2+ / K6-III+ page

Reply 44 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

It is a mr bios but I am attaching it anyway. Inside the file I can see the mr bios heading and version so it is a good sign it works.

I also tried the Ami bios I found for this motherboard but I think its not for the 3.02 version but lower versions. Flashing it to a erasable chip I have also did not boot the motherboard.

Attachments

Reply 45 of 62, by Chkcpu

User metadata
Rank Member
Rank
Member

Thanks for the BIOS.
An MR BIOS, allright!
I did various checksum calculations on this 64KB file and the regular 8-bit checksum produced a nice 'zero' result.
Then I put the BIOS in 86Box and this is what I got:

Hippo VL+ Rev3 MR BIOS.png
Filename
Hippo VL+ Rev3 MR BIOS.png
File size
125.25 KiB
Views
1234 views
File comment
Octek HIPPO VL+ Rev 3 MR BIOS Setup
File license
Public domain

I also got POST codes, running through the numbers during POST and memory count, so this BIOS looks okay! 😀

Hopefully the Logic Analyzer will tell you more.
Jan

CPU Identification utility
The Unofficial K6-2+ / K6-III+ page

Reply 47 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

I was just doing some thinking it could be some missing communication with the cpu. I randomly probed a few pins on the cpu socket and a few had no continuity to ground. Well this doesn't necessarily mean anything but I did a quick comparison to a working board.

I didn't do all the pins, just found one as comparison that stands out. J17 PCD shows 0 continuity to ground.

On a working motherboard same pin shows 0.600~ to ground.

I have a feeling this is not the only pin like that

Edit: I did a comparison and these are the pins with no connection compared to a working 486 cpu socket :
N15 "lock"
L15 "opt"
C12 "smi act"
J17 "pcd"
B14 "tms"
C13 "other"

Reply 48 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

So I did a bit of digging around. I looked up what sounds like a serious pin to me
C12 "smi act"
System Management Interrupt Active. Indicates that processor is in system management mode.

On my working working motherboard (different model but also vlb 486) I traced this pin to a resistor then to the super io chip. I also got some resistance readings on alot of pins in the chipset.

On the octek the c12 pin I couldn't find a hint of any resistance to anything I probed. All open circuit on the chipset. This is not a good sign, the line from this pin is not connected. I don't know the function smi but sounds like it might be important

Reply 49 of 62, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie
treeman wrote on 2023-03-13, 01:30:
Edit: I did a comparison and these are the pins with no connection compared to a working 486 cpu socket : N15 "lock" L15 "opt […]
Show full quote

Edit: I did a comparison and these are the pins with no connection compared to a working 486 cpu socket :
N15 "lock"
L15 "opt"
C12 "smi act"
J17 "pcd"
B14 "tms"
C13 "other"

N15 - LOCK# - important
L15 - PWT - important
C12 - ignore this
J17 - PCD - important
B14 - ignore this
C13 - ignore this

LOCK# is required for atomic operations to work. Though I would think that BIOS should be able to output some codes even if it wasn't connected.
PWT and PCD control external cache behaviour according to page bits in the CPU. This is pretty important but again I would expect the CPU to boot and output some codes before this has a chance to break anything. But that depends on the chipset too I guess.

These are CPU outputs. Chances are the inputs in the chipset are high-impedance CMOS and you will not see any resistante (and there's no need for pull-ups). You should be able to detect input protection diodes though, both to GND and +5V, in che chipset when measuring empty CPU socket. No diodes would lead me to conclude the connection is broken somewhere along the PCB.

TMS is for JTAG, ignore it - though it should have a pull-up but I guess it's not strictly required since 486 has internal one. It's not even implemented on later 486 chips.
SMI ACT is not implemented in non-S (power-saving) CPUs so can be ignored, it's not a part of early 486 spec so the pin might very well be not connected, just like it's supposed to.

Reply 50 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

Thanks for your reply. This motherboard uses the dynamic cache technology so no cache from what I understand, perhaps why the cache pins are not connected.

Reply 51 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

I wanted to update this thread, I have received the usb board some time ago. Have been very busy with personal life and haven't had a chance yet to do anything. Slowly I am getting back into it. I installed pulse view and the drivers recently but having trouble detecting the board in pulse view.

I will get there but short on time. Just wanted to update to say I haven't dissapeared or given up

Reply 52 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

Ok so I finally got back to this project.

I wired up the controller like in the links rasz provided me. Ground to the ground on the bios chip and a few of the PB probes to the data lines on the bios.

The result wasn't exciting at all, I got some activity on pulse view on power up but thats all.

Rasz maybe you can have a look and tell me if I need to change any settings like the range or time outs. Also what do you think of the result.

Here is some pictures of the pulseview log and how I wired it up

test2.jpg
PXL-20230608-125945412.jpg
PXL-20230608-125939939.jpg

Reply 53 of 62, by rasz_pl

User metadata
Rank l33t
Rank
l33t

at least A0 A1 A2 address lines on the eprom, and pins 20 & 22 to see if cpu tries to fetch first couple bytes

https://github.com/raszpl/FIC-486-GAC-2-Cache-Module for AT&T Globalyst
https://github.com/raszpl/386RC-16 memory board
https://github.com/raszpl/440BX Reference Design adapted to Kicad
https://github.com/raszpl/Zenith_ZBIOS MFM-300 Monitor

Reply 55 of 62, by rasz_pl

User metadata
Rank l33t
Rank
l33t

bios chip

https://github.com/raszpl/FIC-486-GAC-2-Cache-Module for AT&T Globalyst
https://github.com/raszpl/386RC-16 memory board
https://github.com/raszpl/440BX Reference Design adapted to Kicad
https://github.com/raszpl/Zenith_ZBIOS MFM-300 Monitor

Reply 57 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

normal.png
D5 = pin 20
D4 = pin 22
Others are address lines
reset.png
This is when I pressed reset 2 times

So if there is no signal to chip enable basically the bios is not powering on from what I understand

Reply 58 of 62, by rasz_pl

User metadata
Rank l33t
Rank
l33t

just noticed right now you set pulseview to 20KHz sampling, that means mobo might be working and you just missed it 🙁 set it to 12MHz (or 24MHz with only 8 probes enabled, this is ~max this board can do) and do bios address/enable pins again

Next you can do the clock chip. What was it hmm, https://www.digchip.com/datasheets/parts/data … 107C-03-pdf.php
You can do this with or without CPU. Without will validate our guesses about clock generator and if its being enabled, with CPU will make sure its still able to generate clock under load. You wont be able to read real clock values due to sampling speed, but we will see if clock is running or not.
You have enough pins to wire all of them and cpu input 😀 https://www.retrobrewcomputers.org/lib/exe/fe … am486_16244.pdf (https://www.retrobrewcomputers.org/doku.php?i … linux-sbc:start) cpu CLK pin C-3
try different jumper clock combinations from the clock table (8,16,25,33), lets see if your multimeter was right about erratic output 😀

https://github.com/raszpl/FIC-486-GAC-2-Cache-Module for AT&T Globalyst
https://github.com/raszpl/386RC-16 memory board
https://github.com/raszpl/440BX Reference Design adapted to Kicad
https://github.com/raszpl/Zenith_ZBIOS MFM-300 Monitor

Reply 59 of 62, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

on 12mhz im getting a similar picture
6th.png

Anything over 12mhz pulse view crashes, I think the driver can't handle the faster speed from what I googled