So now, I feel like I'm really getting close to my ideal Apple-1 system. I've got the whole keyboard thing at a point I'm happy with. I'm making progress on my case, but can't go as fast as I would like. I've been thinking about that expansion slot. Right now there's an ACI sitting in it, and I have been thinking about the future and possibly adding a CFFA-1, especially with the thought that I could play Zork on my Apple 1! I plan on making my case large enough to fit the CFFA in the future, if I want to add it.
I've also been thinking about how I can jazz-up my build, and came up with one conclusion - it needs LEDs! I dearly love LEDs and love them when they're blinking. I added some power-line LEDs to my Mimeo at the top, so I can monitor all 4 power lines (+5, -5, +12, -12), but it's not enough! I was looking through the Apple-1 Operations Manual, and came across discussion of "Address Display" on page 12. Hmmm. It looks like a series of LEDs that can display the state of the address lines. I searched online, and found that speedyG has already built this here:
I'd love to build this and have it incorporated into my enclosure, and showing the activity of my Mimeo. I'd love to have it on all the time, but since I've got the ACI in the expansion slot, am I out of luck? I then looked to the side of the expansion slot, and saw the Mimeo's edge connectors sticking out the side, and wondered if I could tap those, build my "step-diagnostic-address-display-card" to plug onto that card edge. It looks like it's just parallel to the expansion card edge connector, and I was wondering I could just use the card edge and this way leave my ACI where it is.
The only thing I have seen that uses the card edge like this is Wendell's 3-slot expansion board here:
Now looking at his design, he uses logic to buffer address and data lines. Now I'm nowhere close to understanding this, but my vague feeling is that this has to do with switching control between the 3 cards on the expansion. If I built my board with just the address display and step/diagnostic switches, would I need to do anything like this, and still keep functionality of the ACI in the expansion slot already?
And another thing I'm getting a bit worried about is if I am over-taxing the power limits of the Mimeo. In my build, I could conceivably have the PS/2-serial adapter (with its microcontroller), the keyboard multiplexor (just a few ICs), the Apple II keyboard, the ACI/CFFA1, and possibly this address-display-card (with about 40 LEDs and some logic chips). Has anyone ever hit the limits of the power-supply of the Apple 1 with their peripherals? I'm guess if you attach the 3-slot expansion board and add 3 other cards, and if the Apple 1 could support that, I should be okay.
I would guess the duty-cycle of the address display LEDs should be somewhat less than 25-50%, and likely even lower. And if I limit them to 10mA each, that would put the estimated power draw of the extra LEDs alone at about 100-200mA.
Any help or insights by anyone on this would be a great help. Thanks!
yozr probably searching for something like this:
Thanks speedy, and wow, you've put a lot of work into those expansion options! Very impressive.
I think this is a bit more than I had planned. I was mainly interested in hanging the address/data/debug board (with all the LEDs) off the card edge of the Mimeo, and hope that I have enough power to run them all together with either the ACI or the CFFA attached to the Apple 1 expansion socket and the PS/2-serial board and keyboard multiplexor running as well.
Wendell's website has a link to a youtube video where he runs 3 of his expansion boards in addition to his ACI board all at the same time on his Apple 1. I didn't see any auxiliary power cables in that demo, so I presume there's enough juice from the Apple 1 to run all this!
Here's the video:
I'm guessing I should have enough power to do what I want. Perhaps I'll just give it a go and try it out, and see what happens with it. If anyone sees a major issue with what I'm trying, please let me know.
Wendell has another attemt to the topic than me......
My attempt is trying to work with chips also availiable in the days of the apple - 1
Wendells attempt is to make cards as sophisticated as possible....
so his cards use a lot less power than my cards because i´m using old chips
and he is using less power consuming chips and parts with far later technology with SMDs ....
besides bear in mind that at the connector only unregulated +12 V an -12 Volt is availiable
and at the caps and the connector can be up to +/- 18 Volt to 20 Volt availiable depending to
the used transformer !
second bear in mind that at power on the consumption jumps high and then drifts down to regular consumption....
specially at the + 5 Volt rail....
not to talk about the temperature at the cooling sink of that +5 volt regulator....
it´s quite high at the normal use, but expanding will probably drive temperature fast out of range.....
Thats the reason i added the option to add additional external +5 Volt power to the expansion port...
some points to think about, before rocketing the +5 Volt power to "BBQ heaven"....
BBQ heaven - lol!
Hey speedy, did your diagnostic card work okay? It looks like it did, and it looks like it ran off the 5v line (pin 22 on the expansion connector). So the 5v line is active, right? I get that I may be taxing the 5v regulator. I will be putting an 80mm Noctua fan right at the heatsink (which is running on its own 12v power supply) in my final build, so it should keep it cool. Hope it doesn't blow the fuse, though, as I've read Mike W. write about that startup surge as well.
yep - its running and yes it uses only 5 Volt but from slotexpansion and
the slotexpansioncard with supply from external additional 5 Volt source.
i used it while building my replicas and getting them up running....
it helped me solve one problem by detecting a specific lockup off a dataline
in a bad RAM chip at one of the boards and while design of the other
So you added auxiliary power to it? OK - did you run into any actual problems before doing this, or was it mainly for added security?
I'm not sure I'll ever use the stepping capability beyond initially trying it out, so I may skip that whole part and just do the indicator LEDs since that's what I'm mainly interested in - an indicator panel.
I will probably build just the LED indicator part of this, and try it out. If I run into power problems, I may look into switching off unnecessary components (like the PS/2-serial adapter). I wonder if I can put a small switch in to simply cut power to the PS/2-serial adapter so I can turn it on only when I need it, but still leave it plugged into the keyboard multiplexor so I don't have to unplug anything when I want to use it. I wonder what the state of the microcontroller pins are when it is powered off, or if there will be other problems with this approach.
I could also go back to my rotary switch keyboard selector solution, since that only powers on the PS/2-serial adapter when you need it, but then lose the added functionality of simultaneously using BOTH the PS/2-serial connection and the internal mechanical keyboard.
You should review Wendell's notes on expansion. They are available on his site. He ended up buffering data and address bus. Since +12 and -12 volts are unregulated on the peripheral connector, you can use them as sources for a local power supply and it shouldn't affect motherboard power to any great extent.
If I remember right, someone (John Calende?) created a simple bus LED display that plugged into the motherboard edge connector. Maybe there are some comments about this in the archives of this forum. I don't think that it took any special effort to do this. Unless you also implement some kind of single step functionality, actual usefullness is pretty limited.
to tell the truth - i´m no fan of try and error....
therefor i just did some math and dicovered it to be a good idea to use external power......
and besides because i also intended to add other cards too - i immediatly added the external
supply option and never did test in fact, if i could really stress the internal + 5 Volt that far.....
And the LED option varies from small one only for data and adressbits ( approx. 250 Milliampere )
or also display of the driving communication signals ( approx. 320 to 35 Milliampere ) and the
stepping only affords 25 to 3 milliampere.... the main cost of the display is related to the amount
of LED´s ( nearly 20 Milliampere each and 60 to 80 Milliampere for the IC´s )
and it depends - if you only have 4 KB RAM on board, or if you have populated the entire 8 kB of RAM.
why try and risk damage of 15 bucks, if at the other hand the external 1,5 Ampere / + 5 Volt is
availiable only at cost of additional 4 to 5 bucks ?
Mike: Thanks for the info. I found John's page discussing his simple address display:
From what I can gather from the page and pictures, he's driving the LEDs directly off the bus, and each through its own 300-ohm resistor. I didn't see any other ICs on his board, but he is taking the signals off the card edge, like I plan to. Interesting idea about using the unregulated 12V line to drive this indicator board.
speedyG: Thanks for the input. I guess my main resistance to putting in auxiliary power is I don't want to add another plug on my setup if I can avoid it, because I'm dealing with some confined quarters in my case. Another reason is, I don't really know how to do it! Power delivery is something I'm still quite fuzzy on, having only recently learned about bridge rectifiers on this Mimeo project.
So let me ask you this. If I were to put auxiliary power on this thing, how would I do it? I would prefer to avoid adding another AC plug in my setup. So first some details on my setup. My Mimeo is running with the standard transformers, and I've got a small wall-wart type power supply (which I plan on tucking away in my enclosure) delivering 12V to the 80mm fan which I am using to cool the PEM and will also power the small amplifier board which I have connected to a small 8-ohm speaker to amplify the output from the ACI.
Mike had the interesting idea to use the 12V line on the Mimeo and felt this should not affect the functioning of the Mimeo.
So could I just use a 5V linear regulator or small switching converter to take the 12V unregulated input on the card edge, to generate the 5V to drive my five 74LS373 chips? And correct me if I'm wrong here, but in your design for your debug board where you use the 74LS373's to drive your LEDs through 100-ohm resistors, the power that gets sent to the LEDs comes from the 74LS373's input Vcc, right? Also, you're using 100-ohm resistors due to the VOH from the 373 being around 3.1V, correct?
I'm looking for the simplest and least involved way to do this. I don't plan on adding any other major peripherals (no RAM cards, ROM cards, etc.)
I'm open to any suggestions on my setup, and part-numbers if you are able and willing to go that far. I'm nervous about picking my own parts because I don't really know how to read the datasheets beyond the very basic parameters, and will probably mess it up.
I would welcome any thoughts on this - thanks!
lets just walk along the line of options in the order you have given:
If you have allready a additionl transformator running to supply the fan
then there would be the option just replacing that transformer by another
transformer instead that offers 2 voltages instead... optimum would be one
that offers 12 Volt 1 Ampere and 6 Volt 1,5 Ampere but you should insist
in a transformer with seperated coilwindings i.e. not a transformer that has
only 1 coilwinding with a split at 6 Volt or only 2 x 6 volt coils.
In first case the current you take away from 12 volt will be the missing current at the
6 volt splitoff and in second case the load from 6 volt coil will have far more
load than the use of that part where both coils are used together as 12 volt coil.
Therefor its real a better idea to search for a transformer with one 12 volt coil
and one seperate 6 volt coil.
Up to the next point taking of the voltage at the positive capacitor ( i.e. at the connector )
then followed by a smaller electrolytic Capacitor ( some value near
1000µF to 1500µF and then a 7805 IC for 1 Ampere to regulate the voltage. As explained
one problem is the fact that at the large capacitor on the mainboard and the connector
there is a positive voltage availiable up to 18 to 20 volts !
Lets just do some small math:
assuming that the LEDs and the circuits will use in "small configuration" with maximum of
24 LEDs and 3 x LS373 then total consumption will be approx. 280 Milliampere and
with difference from 20 Volt ( worst case scenario for safety ) the disaption ( transforming
difference of input voltage to regulated voltage will be 15 Volts x 0,28 Ampere resulting
to a power consumption of 4,2 Watt. Thats rather bad if the 7805 is used without heatsink!
It´s therefor recommended to use at least a heatsink like this one:
If you want to protect the 7805 Regulator from getting "grilled".
Yes the resistors are calculated on the output voltage of the LS373 and for smallest current
in the LEDs 1k ( that are metalfilm 2 % resistors - resulting to 5 ring code ! ) Ohm will
be O.K. with current of only 12 Milliampere per LED, but if the shine
of the LEDs is too dark then you may drop the resistors to 220 Ohms but in that case each LED
would use up to 35 Milliampere and the LS 373 might get quite warm - but not hot.
Resuming this facts and your aims i´d recommend to really only use the setup like displayed by John´s page.
I hope this help´s you up with your decission.
It´s based on your claim not to add up more expansions to the MIMEO. If you cross the line of
adding up more features or expansions to the MIMEO - it will bring you back to this point and to
review options of adding up external additional power to supply the added expansions....
Thank you for the in-depth discussion of this and for your suggestions. Based on this, I think I should probably look for a regulated 5V-12V power supply to drive the fan and the address display board.
Yeah, the linear regulator option sounds like it would create a lot of heat. How about a switching buck converter? I think in the end it's probably easier to just get an additional power supply to supply the extra 12V and 5V I need.
I'm a bit confused about your resistance calculations, though. The LS373 outputs at 3.1V, and the voltage drop across an LED array I'm looking at is 1.85V. If I run these LEDs at about 15 mA, then that comes to a resistor value of about 100 ohms, just like you have on your board. Is that right? I think your calculations of 1K ohms looks like it's based on a source voltage of 12V. I may be missing something, though.
And another question - with John's setup, aren't those LEDs drawing extra current through the address lines directly? Wouldn't that be a less-preferred way of doing this? Or am I not understanding this completely? I was thinking he mainly build up his display panel to be used only intermittently, not long-term, which is why he did it this way. I think he planned to eventually use a hex digit display.
well .... your calculation are ruled by data of simple "normal" LED´s....
There are different kinds of LED bars availiable and most of them
have / permit 35 mA per LED. And in general the normal ones work with emmitionvoltage of 1,2 Volt
and some with high lumen ignition have a emmitionvoltage of 1,5 Volt while normal LEDs have
emmition voltage of 2,7 Volts.....
and output of the LS 373 may be at "high"-level up to 4,7 Volts.....
so that´s the difference in the calulation database....
and like you see in my pictures my LEDs at the bar units are a bit "dimmed" - i.e. not
running at maximum power / brightness - so that´s reason for larger values of the resistors.....
reason: the brighter they shine the shorter their lifetime....
second: green LEDs have less emmitionvoltage than red LEDs - i.e. the red ones need higher voltage
thats the reaason the red ones are strong dimmed while the green ones glow stronger....
if you want "ovar all same brightness" - you´d have to use a bit higher values at the green ones
and less value at the red ones....
OK. Again, thank you for your ideas and insights.
I've heard as long as you don't stress LEDs near or beyond their limits, they should be pretty stable in their lifetime, so I'm going to shoot for the middle of their operating range of current, maybe a little above, especially since I'll probably add an external power source based on your estimated numbers. I'll keep you posted on the results.
I also just remembered I have a variable power supply which gives me a reading on the current draw, so I can test it with that to see what I'm dealing with before making any final decisions on the power source. I also found a cheap adjustable buck transformer I got off eBay for about $1 lying around here, so I may try it to convert that 12V unregulated line to 5V.
Thanks all for the ideas and info!
One more question before I go. If I add an external power source, what should I do with the ground lines? If I add a wall-plug powers supply to provide the 5V to this address display board, should I connect the ground from this to the Mimeo ground on the card edge connector? From my limited delving into the Mimeo power supply in earlier postings, I recall the Mimeo's ground is "floating". I guess my added wall-plug power supply uses a similar method of generating DC from AC, and thus would also be floating. I would guess I would want to connect these two grounds together.
Also, I would NOT want to connect my external power supply 5V line to the Mimeo's edge connector 5V line, right? I'm guessing no since it's main purpose is to drive the address display board.
Just want to make sure I'm not violating some fundamental rule of electronics design.
external power supply requests to take care of 2 rules...
Yes - the Ground at both MIMEO and external powersupply must be connected to ensure
that both circuits deal with same reerence and avoid drifting apart from each other.
Yes - you may not connect the 2 different + 5 Volt supplies together they must stay appart
otherwise you will make a breach in the integrity of the function. Both IC´s that regulate
the +5 Volt have sense at the voltage output and if one supply drifts apart - it will also
affect the sensing of the other regulation and pull it away from correct value.
I´d recommend to keep sure that both power sources are switched at power on at the same time.
OK. Thank you very much! That is very helpful information.