Power-On Reset and Automatic Screen Clear for the Briel Replica One Plus

5 posts / 0 new
Last post
Offline
Last seen: 1 week 1 day ago
Joined: May 12 2023 - 09:04
Posts: 35
Power-On Reset and Automatic Screen Clear for the Briel Replica One Plus

 

I recently added an automatic power-on reset and screen-clear circuit to my Briel Replica One Plus. The goal was to have the computer start reliably with a cleared screen, without having to manually press the RESET and CLEAR buttons after powering it on.

The original idea comes from Jeff Tranter, who published a NE555-based power-on reset circuit for the Replica I TEback in 2012.

 

Original project:https://jefftranter.blogspot.com/2012/03/adding-power-on-reset-circuit.html

For my Replica One Plus, I adapted the circuit, added an automatic screen-clear function, and optimized the RC timing based on practical testing.

Modifications to the Original Circuit

1. Automatic Screen Clear

In addition to generating the reset signal, the circuit also activates the CLEAR input during power-up.

This is accomplished by connecting the NE555 output to the CLEAR input of the Replica One Plus through a diode.

As a result, both RESET and CLEAR are automatically activated when the computer is powered on. The diode prevents the manually operated CLEAR button from feeding voltage back into the NE555 output.

2. Determining the Reset Duration Through Practical Testing

An important part of this project was finding a suitable reset duration for the Replica One Plus.

During my initial experiments, I discovered that short reset delays did not reliably initialize the computer after a complete cold start.

Interestingly, the reset worked perfectly when the circuit was activated again while the computer was already powered on.

This suggested that the problem was not the reset function itself, but rather the timing relationship between the power supply, video circuitry, and CPU startup.

I therefore experimented with several resistor and capacitor combinations.

A relatively long reset delay of approximately 11 seconds worked reliably, but was unnecessarily long for everyday operation. Shorter delays in the range of approximately 0.5 to 1 second proved insufficient.

After several experiments, I settled on the following component values:

  • R = 300 kΩ

  • C = 10 µF (106)

  • Theoretical reset duration ≈ 3.3 seconds

This combination has proven reliable during my cold-start tests so far.

The 3.3-second delay was therefore not chosen arbitrarily. It is the result of a series of practical experiments aimed at finding a reasonably short reset delay that still allows the computer to start reliably.

Another consideration is that the video circuitry of my Replica One Plus takes approximately two seconds after power-up before an image becomes visible. Consequently, the additional waiting time before the CPU is released from reset is hardly noticeable.

3. Electrical Optimization of the RC Network

In addition to adjusting the timing, I also changed the component values of the RC network.

Instead of using a relatively large resistor with a smaller capacitor, I chose a lower resistance of 300 kΩ combined with a larger 10 µF capacitor.

The higher charging current reduces the relative influence of capacitor leakage currents and the NE555's input currents on the timing behavior.

This is particularly beneficial when using a conventional bipolar 555 timer.

Of course, the actual delay will still depend on component tolerances and the power supply's startup characteristics.

4. Adaptation for the Replica One Plus

Since Jeff Tranter's original circuit was designed for the Replica I TE, I adapted the connections and circuit configuration to the Briel Replica One Plus.

The modification essentially consists of an NE555 timer, a 74LS05 open-collector inverter, the RC timing components, and a diode for the CLEAR signal.

 

How the Circuit Works

When power is applied, the timing capacitor is initially discharged.

The NE555 output switches HIGH, activating the CLEAR signal. At the same time, the 74LS05 inverter generates the active-low RESET signal for the 6502 CPU.

While the capacitor charges, the CPU remains in its reset state.

Once the capacitor voltage reaches the NE555's internal threshold of two-thirds of the supply voltage, the timer output switches LOW.

This deactivates CLEAR and releases RESET, allowing the 6502 to begin executing instructions.

With the selected RC values, the theoretical duration of this power-on sequence is approximately 3.3 seconds.

 

 

 

Results

With the 300 kΩ resistor and 10 µF capacitor, my Replica One Plus has consistently started with a cleared screen during testing so far.

The modification requires only a handful of inexpensive components and can be integrated directly into the computer without the need for an additional PCB.

 

 

I also suspect that this circuit could potentially be adapted for use with an original Apple I. Since the basic principles of reset and screen control are comparable, such an adaptation seems technically plausible. However, I have only tested the modification on my Briel Replica One Plus. Whether it would work reliably on an original Apple I, and whether additional changes would be required, remains to be verified.

I will also share the adapted schematic and a timing diagram for reference.

Finally, many thanks to Jeff Tranter for his original idea and for publishing the circuit that served as the starting point for this modification.

Offline
Last seen: 15 hours 6 min ago
Joined: Apr 1 2020 - 16:46
Posts: 1405
Is it proper nostalgia to add POR / screen clear to Apple-1 ?

In post #1, 'Appleny' wrote:

 

" I also suspect that this circuit could potentially be adapted for use with an original Apple I. Since the basic principles of reset and screen control are comparable, such an adaptation seems technically plausible. "

 

Uncle Bernie comments:

 

While theoretically possible to add this circuit to the Apple-1, I would not do it. IMHO, part of the "charm" of the Apple-1 is the ritual of properly starting it up with the CLEAR SCREEN and RESET buttons. This is one of the "lovely quirks" of the Apple-1, which other than some of its more nasty quirks is harmless as to its robustness and fault free performance.

 

Did you ever see one of the High Priests / Magicians otherwise called "Computer Operators" booting up a 1960s era mainframe computer ? An unfathomable ritual with obscure procedure steps totally unknown to normal people. Or, as another example, toggling in the bootstrap routine of a PDP-7. Which depended on the mass storage device to boot from.

 

Why should we want to "hide" that truth ?

 

For the Apple-1, I  have done power on reset and screen clear in my YAAK keyboards with YAAKEN-1 encoder, without adding any hardware, and removed that software feature later because I didn't like it. I missed the @_@_@_@ ... "welcome" screen the Apple-1 greets us with. It just did not feel right anymore.

 

This may be "Apple-1 purism" in its worst form, but like anything where nostalgia is involved, personal preferences reign supreme.

 

But there is also a real benefit of not having automatic screen clear and reset on the real Apple-1: these machines may be finicky (depending on who built them and from which components) and sometimes they refuse to start up the Wozmon.

The @_@_@_ screen upon power up will tell us that the clocks are properly running and any issues with the on-motherboard crystal oscillator could be seen (such as jitter coming from a crystal dying due to being overloaded by this rotten oscillator circuit) without attaching an oscilloscope. The clear screen button working proves that the 'terminal section' is alive and well. Unless no cursor appears.  And then, upon hitting the RESET button, and Wozmon NOT starting up and giving us a prompt, we know something is wrong with the 'processor section'.

 

So for a real Apple-1, these two buttons are a crude but useful diagnostic tool.

 

For the Replica-1 however, all this does not matter and adding the automatic clear screen and RESET only "hides" all the extra  development work which went into it to show the "correct" Apple-1 power up behavior. Even developers of Apple-1 emulators like POM-1 put a lot of effort into properly implementing the lack of power on reset and screen clear seen in the real Apple-1. Otherwise they would not show the @_@_@_@_ !

 

(Hope y'all get the irony of this).

 

- Uncle Bernie

Offline
Last seen: 1 week 4 days ago
Joined: Jul 31 2021 - 08:38
Posts: 46
 I can completely understand
 
I can completely understand that and fully agree.  A few weeks ago, I modified an A-One with a simple RC-based power-on reset circuit; we have the unit on display in a museum to give visitors an idea of ​​how such a device from that era was operated.
On the A-One, the reset also triggers a screen clear. Operating it via the keyboard would have required modifying the keyboard itself, so the RC reset was the simplest solution.
I could have provided an authentic replica, but the focus here is purely on the operation, and the A-One is definitely easier to maintain. In any case, I personally find that the power-on sequence with a screen clear and reset is something I miss whenever it’s absent; without it, things just feel somehow "incomplete".
;-)

 

Offline
Last seen: 1 week 1 day ago
Joined: May 12 2023 - 09:04
Posts: 35
There’s often more than one way to look at things

Hi guys,thanks for your honest feedback.Of course, a circuit like this isn’t for die-hard nostalgics. That’s why I didn’t even try to implement it on my Apple I.

I also think it’s important to note that even on an emulator or a modern board, the startup sequence is recreated with quite a bit of effort to replicate the original startup behavior. Respect for this approach and the development effort. 

Bernie is, of course, right in pointing out that the startup screen or reset acts as a quality check and immediately reveals if something is wrong.

However, there are two points that were important to me in this matter.

I. To add another piece to the puzzle of a complete documentation of all Apple I-related projects in our great forum.

II. There may also be a goal other than simply recreating an Apple I 1:1.

My approach is to use my Replica One Plus as a hardware and software development and testing environment,

in order to also protect my expensive original Apple I board.

It’s really annoying to have to keep performing resets and clears after a cold boot

—which is necessary whenever you swap out an expansion card. My goal is therefore to get back to the monitor as quickly as possible. 

So we’ll leave it up to the forum readers to decide whether that makes sense or not. 

-Appleny

Online
Last seen: 24 min 35 sec ago
Joined: Apr 9 2021 - 04:31
Posts: 211
Purpose of Prototyping Area

I think the purpose of the prototyping area was exactly the one to customize the Apple I to your liking.

I guess I would have done same, but would not "wasted" so much space with a 1/6 used LS05 and would have use a transistor instead.

 

Log in or register to post comments