H8-RTC/Network

[Norberto Collado]

H8-RTC:

 

I received the new H8-RTC board, assembled one, and it is working fine. See pictures below!

 

The only problem is that when I power-cycle the H8, then the RTC time and date are set to zero.  The prototype board does not have this issue.

 

The only change to the PCB was to wire the CS1 to the H8 /RESET signal following the same wiring as shown in the GIDE schematics. Before, it was connected to VCC (+5V). Besides this issue everything is working fine. Terry any suggestions?

 

H8-Network:

 

On the Network side, somehow the PCB board applied a solder mask on the signals, power, and ground pins. The proto board did not have this issue. I used my X-ACTO knife to clean then up but with caution. I will show proper operation in the assembly guide.

 

Once I can figure out the issue with the power sequence with the RTC-72421 IC, then I will start the distribution. This IC has been challenging since the beginning of this project. Plan B will be to enable the second RTC DS1307.

 

Thanks,

 

Norberto

 

 

 

 

[Glenn Roberts]

Nice work, as always Norberto.

 

In your instructions can you indicate how to build just the RTC portion? 

 

Thanks.

 

-          Glenn

--
You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.
To post to this group, send email to se...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Norberto Collado]

Yes, I will do that. The network is going to take more time due to all the settings to get it configured properly.

 

The RTC will be first.

 

Norberto

[Terry Gulczynski]

Norberto,

I've had a lot of trouble with the very same circuit you are.  The problem manifests itself during the first few milliseconds after power on when the various control signals are not stable.  Allowing CS1 to go high too soon has often caused the exact problem you're seeing.

The solution I've found that works perfectly every time is to tie CS1 high through a 10K resistor, and at the junction of CS1 and the resistor, add a 1uf electrolytic cap tied to ground.  This forms an R-C network that delays CS1 going high for somewhere around 20 mSec after power on.

(1 Time Constant is 10 mSec in this circuit.  Assume ~2 T/Cs to reach a logic 1 [70% x 5v] before CS1 goes high, allowing regular access to the RTC chip.  2 T/Cs = 20mSec.)

Applying this circuit has completely eliminated the problem you're describing in every system I've used the 72421 RTC.  Will it work for you?  I dunno, but you might should give it a try.

(BTW: I've also tied CS1 directly to +5 as you did earlier and I still had the problem.  It didn't happen every power cycle, but it was common enough to seriously irritate me.)


Terry


On 5/11/2013 4:08 PM, Norberto Collado wrote:

H8-RTC:

 

The only change to the PCB was to wire the CS1 to the H8 /RESET signal following the same wiring as shown in the GIDE schematics. Before, it was connected to VCC (+5V). Besides this issue everything is working fine. Terry any suggestions?

 

Norberto

 

[Norberto Collado]

Hello Terry,

 

I will make such change right now to the PCB board to find out the outcome.

 

Thanks for the feedback,

 

Norberto

--

[Norberto Collado]

It just worked for a few power cycles and then back to same issue. I spend so much time, effort and money on this board that I’m frustrated with it. The reset signal should take care of the sequence but that is not the case. My proto board works fine for some odd reason and the production board is not that happy when power cycling the system. Not sure what else to do here!!!!

 

At least Stanly RTC HDOS application detects such abnormally. See below!

 

Norberto

 

 

 

From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of Norberto Collado
Sent: Saturday, May 11, 2013 7:34 PM
To: se...@googlegroups.com
Subject: RE: [sebhc] H8-RTC/Network

 

Hello Terry,

[Norberto Collado]

I found that the RTC-72421 registers gets corrupted when powering down the H8 system and not on power-up. Based on the scope signals, when the H8 is losing power, the RTC controller enters into write mode as the /WR signal is transitioning to the low state. It is caused by feedback voltage from the backup battery into the IC /WR circuitry. I manually created a write-protect condition after setting the date/time and manually power-cycled about 20 times and the date/time was preserved. That is the good news. The bad news it is on how to implement such fix on the existing boards!!!

 

Very strange behavior and ready to celebrate Mother Days with my wife for all her support on the hobby!

 

J

[Glenn Roberts]

Any idea why this behavior didn’t show up on  your prototype?

[Norberto Collado]

I tested the proto-board and the time/date are preserved. No matter how many power cycles I do, it preserves the data. I will probe the new board against the proto to find out if any layout issues. Also this process will allow me to recheck the hand-made schematics to be correct.

 

Norberto L

[Norberto Collado]

See attached preliminary schematic. The first page is the proto I/O decoding. The second page is the Production I/O decoding. On the production board I ratified the Read/Write signals with the I/O port decoder which I do not do on the proto. These are the main differences on each board. The rest of the schematics are the same.

 

Let me know If I’m missing something obvious and why on power-down, the registers are corrupted.

 

Norberto

[Norberto Collado]

Latest & preliminary schematics attached!

[Rob Doyle]

On 5/12/2013 11:51 AM, Norberto Collado wrote:
> See attached preliminary schematic. The first page is the proto I/O
> decoding. The second page is the Production I/O decoding. On the
> production board I ratified the Read/Write signals with the I/O port
> decoder which I do not do on the proto. These are the main differences
> on each board. The rest of the schematics are the same.

Are you missing a ground on the negative side of the battery?

The schematic looks pretty suspicious right there...

Rob.

[Norberto Collado]

Thanks for the feedback! I just checked it out on the new board to be sure
that ground is connected properly to the battery and to the RTC IC. I fixed
the schematic and updated file attached. Also I have both boards (proto &
new) configured the same and only the proto works fine during power cycles.
A simple board is becoming a nightmare and I just need to put my finger on
the issue to get it resolved. I will put it to sleep for a week and
eventually we will figure it out.

Thanks,

Norberto

-----Original Message-----
From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
Rob Doyle
Sent: Sunday, May 12, 2013 7:15 PM
To: se...@googlegroups.com
Subject: Re: [sebhc] H8-RTC/Network

[Rob Doyle]

On 5/12/2013 12:54 AM, Norberto Collado wrote:

It is a pretty standard problem with battery-backed peripherals that the
Chip Select and Write signals go active low during power-down.

> I found that the RTC-72421 registers gets corrupted when powering down
> the H8 system and not on power-up. Based on the scope signals, when the
> H8 is losing power, the RTC controller enters into write mode as the /WR
> signal is transitioning to the low state. It is caused by feedback
> voltage from the backup battery into the IC /WR circuitry.

This could explain how /one/ register is corrupted. It doesn't explain
how more that one register is corrupted. That would required that the
address lines are cycled.

You have CS1 of the RTC-72421 connected to /RESET. Is RESET/ asserted
low before the power goes down?

Rob.

[Norberto Collado]

Hello Rob,

GREAT feedback!!!!!!!

A big NO. /RESET is not asserted low when power goes down. The Z80 CPU board does not have a Power Management Supervisor IC (TL 7705) as part of the reset signal, just a 555 timer. There is one on the GIDE section of the Z80 CPU board but is not enable to drive the /RESET signal to the H8 bus. 

I just went back to my H8 and using the control panel, I kept the /RESET signal asserted and executed about 20 power cycles and then released it on power-on and the date/time was preserved. This fixes the issue, so I guess that I will need to add the TL 7705 to the new board!!! Any suggestions?

Somehow I got lucky with the proto board!!!!

Thanks,

Norberto

 

[Rob Doyle]

On 5/12/2013 10:47 PM, Norberto Collado wrote:
> This fixes the issue, so I guess that I will need to add the TL 7705 to
> the new board!!! Any suggestions?

The TL7705A has a threshold of 4.55V +/- 0.05 which is good choice.

The trick is that you don't want to run the TL7705 output through any
logic gates because you can't guarantee how the gates will behave as the
power goes down. TTL gates aren't guaranteed to operate properly below
4.5 volts. The TL7705 is guaranteed to be activated before the voltage
drops to 4.5 volts which is why it is a good choice for this
application. If you pull the CS1 input to ground with the TL7705, then
the RTC chip will be protected when the power goes down.

You can probably hook the RESET/ bus signal to the RESIN/ input of the
TL7705 and hook the output of the TL7705 RESET/ output to the CS1 input
of the RTC chip. You'll need a pullup on the TL7705 RESET/ pin.

You need to ensure that the CS1 pin of the RTC chip is negated (low)
before the power goes down and all the other pins start wiggling.

Rob.

[Norberto Collado]

Due to space limitation I decided to go with the TL7757 Voltage Supervisor.
I ordered some to test it out, and also I'm getting some TL7705 in case.
Attached are latest schematics.

Norberto

-----Original Message-----
From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
Rob Doyle
Sent: Monday, May 13, 2013 12:07 AM
To: se...@googlegroups.com
Subject: Re: [sebhc] H8-RTC/Network

[Norberto Collado]

Status Update:

Just received the TL7757ILP SUPPLY VOLTAGE SUPERVISOR and now the board is
fully operational. Did a lot of AC cycling and I cannot longer duplicate the
issue. I added all the feedback received as much as possible due to space
limitations.

Thanks for all your feedback and I will re-order new PCB boards during the
week. Attached is the updated schematic.

Norberto

-----Original Message-----
From: Norberto Collado [mailto:norberto...@koyado.com]
Sent: Tuesday, May 14, 2013 10:49 PM
To: 'se...@googlegroups.com'
Subject: RE: [sebhc] H8-RTC/Network

Due to space limitation I decided to go with the TL7757 Voltage Supervisor.
I ordered some to test it out, and also I'm getting some TL7705 in case.
Attached are latest schematics.

Norberto

-----Original Message-----
From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
Rob Doyle
Sent: Monday, May 13, 2013 12:07 AM
To: se...@googlegroups.com
Subject: Re: [sebhc] H8-RTC/Network

[Rob Doyle]

On 5/18/2013 4:42 PM, Norberto Collado wrote:
> Status Update:
>
> Just received the TL7757ILP SUPPLY VOLTAGE SUPERVISOR and now the board is
> fully operational. Did a lot of AC cycling and I cannot longer duplicate the
> issue. I added all the feedback received as much as possible due to space
> limitations.
>
> Thanks for all your feedback and I will re-order new PCB boards during the
> week. Attached is the updated schematic.
>
> Norberto

Nice!

It is so neat that you are adding new life into these designs.

I put a WIFLY module in my PDP8 so that I could telnet to it wirelessly...

https://www.sparkfun.com/products/11049

Old dogs with new tricks.

Rob.

[Norberto Collado]

Cool WIFLY module device! It is nice to bring some life to these old
computers and the ease to transfer files between the PC and the H89/H8
systems.

-----Original Message-----
From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
Rob Doyle
Sent: Saturday, May 18, 2013 11:37 PM
To: se...@googlegroups.com
Subject: Re: [sebhc] H8-RTC/Network