A3000 WarpEngine HDD LED conundrum

[StompinSteve]

Oi Fellas,

Got me WarpEngine 3040 working in me three thousand now and of-course, now me HDD LED no more workie work. Warpies own SCSI is after all not in any way connected to anything "activity LED wise".

I already soldered a LED header on the ZuluSCSI and now im pondering how to connect it to the original HDD LED.
The HDD LED lives together with the power LED on that cute little circuit-board (I still have the original one, made it great again as it was crusty...). The board has a 3-wire connection to the Motherboard.
I guess 1 wire = Power-LED +
1 Wire = HDD-LED +
1 Wire = common ground.

As I don't want to molest the LED board or its cable, I want to "put something in between" the connector and the 3-pin header on the motherboard so neither the board or the Motherboard need any modifications.
Has anyone else done something like this? Any tips?

[Jope]

Get some female -> male jumper wires to go in between?

[StompinSteve]

Yeah something of that nature. I'll leave the HDD-LED + from the motherboard disconnected and the LED is then fed by the + coming from the ZuluSCSI.
The common ground pin will go back to the motherboard but is also connected to it's corresponding pin from the Zulu. That should work.

What I'm unsure about is how much Volt is on the + Pin on the Zulu's LED Header and how much Volt the 3000's LED's are. I don't want to blow up the LED by sending to much to it.

In my A1200 in CheckMate 1500 Plus case project, I power the HDD led directly and completely from the ZuluSCSI' LED header but that is a modern LED, not something build 30 years ago. I don't know the specs of the 3000D's LED's. I have no idea what they used back in the day.

[StompinSteve]

Somewhat confused now. I talked about each LED having it's own + and the third wire is the common ground (I've read this assumption elsewhere too) but it seems to be incorrect.
If I poke a multimeter on pin 1 and 3, or look at the wiring diagram, pin 1, the red wire, is the +5V on the schematic (see 3rd attachment) and two black wires are ground for the PWR LED (pin 3) and ground for the HDD aka SCSI-LED (pin 2) respectively.
This can also be seen on the 1st attachment where the red wire clearly is a common trace towards both LED's.

To answer my own question about voltages: measuring the voltage using pin 1 and pin 3, I noticed the PWR LED is getting 4V under normal circumstances. Assuming the HDD LED is the same spec, it won't blow it up with the 3.0 V that is on the ZuluSCSI LED-Header (verified with ZuluSCSI developer-team).

So how does this work? +5V is on the LED-board all the time (actually 4V). The 7407 chip* connects the "SCSI LED"-ground (thus completing the circuit) only when there is activity on the SCSI BUS (making the HDD-LED light up). Is that how it works?

If it does work like that, then how am I going to power the HDD LED with the + and ground coming from the ZuluSCSI ?
I can interrupt the ground wire of the HDD LED coming from the Motherboard connector, by ensuring (using a small adapter made out of a Noctua 3-pin fan extension-cable) that pin 2 is permanently disconnected. So only Pin 1 on the MoBo connector (+5V) is live and so is pin 3 (PWR LED Ground).
But then what???

* Disclaimer: I hardly understand what a 7407 Chip is and does, so don't get the impression that I know what i'm babbling about...

[Jope]

You figured it right. The inputs prepended with an underscore are active low. Most Amiga signals are like this.

What did the ZuluSCSI folks say? Are they driving their LED high or low? You could add a little inverter circuit if they drive the LED signal pin high.

[giantclam]

Looks like highside drive to me on the zuluscsi (LEDs are connected to ground)

Depending on what you want to do, you could rope in the spare NOT gate on U711 to do the inversion.

[StompinSteve]

Originally Posted by giantclam

Depending on what you want to do, you could rope in the spare NOT gate on U711 to do the inversion.

Sorry I don't even understand what that means. Way over my head

[giantclam]

Originally Posted by StompinSteve

Sorry I don't even understand what that means. Way over my head

It just saves you having to do something like this ...





(active-high on input == active_low on output)

[StompinSteve]

Still don't understand. I'm not an electrical engineer. I know a bit about watts and volts and ohms and that's it. I can also solder simple stuff (replace a pass-trough cap, replace the microswitches in a Tank-Mouse. Basic knowledge).

In this context, I can "plug things in".
In my CheckMate 1500 Plus case (where my A1200 Mobo lives), I simply plugged the two wires from the ZuluSCSI's + and ground directly into the LED's header and the LED works. That is my skill-level.

[StompinSteve]

Come on guys 'n gals. I'm most certainly not the first WarpEngine 3040 owner that has this question. How did you folks get a working LED in the 3000?

[giantclam]

Ok then...then perhaps you know enough to understand this ...lets go back to your statement about not understanding how the 7407 gates work and what the LED layout is on the A3K...

For a start, pin1 (5v) is common to both LED anodes(+) at the LED board, and the separate anodes(-) come back to pins2,3 at the mainboard connector.



The scsi LED anode comes back to R300 (limit LED current), and is connected to a pair of the 7407 IC's gates. Each gate can only 'sink' around 20milliamp, so they connect 2 in parallel to achieve the current required to drive the LED...and you just think of the gates as switches to -power- ground...

The -logic- level signal SCSI_LED is at 5v (logic 1) when no scsi activity is present, and is pulled to 0v (logic 0) by the scsi controller IC. These logic level signals can't drive the LED, but the 7407 can, which responds by closing the ground switch for the LED, whenever the SCSI_LED signal goes logic 0 ..ie; zero volts,(digital ground not power ground).

From what I can see, the zuluscsi controller IC itself is driving the scsi activity LED ~ this is normal these days, thanks to a few decades of electronics advancements since C= had to do it like above, and this is mostly due to Amiga having to use logic I/O + driver circuit, whereas on the zuluscsi controller it'll be using a GPIO pin which can drive(supply 5v) or sink(provide a path to ground) up to a few tens of milliamps ...so you only need the current limiting resistor & LED itself to get the job done, by just turning the 5v on/off at the GPIO pin....

...this would infer, there might be a different way to write/compile the zuluscsi firmware, so it's LED pin switches to ground rather than power, and this would allow a 1-wire patch to achieve the ends of maintaining current scsi LED circuit at mainboard level..and everything working....hmmm....failing that..

...now that I think about it, the one device route would be 4pin optocoupler... small, cheap, will sink 50milliamps by themselves, typically tough and reliable, the zuluscsi LED output is just feeding a (duh =) LED..



Connect pins1,2 to zuluscsi LED out (observing polarity), connect pin3 to Amiga ground and pin4 to the output (LED) side of the 7407 gates (at the IC end of R300 would be a good place)...if you can solder wires to a 4pin DIL thru hole package and hide the connection with heat-shrink tubing, connected as described.. et_voila, you've just created your own custom zuluscsi LED output to A3K LED custom active cable hack, become rich&famous, etc etc B)

HTH

[StompinSteve]

Originally Posted by giantclam

lets go back to your statement about not understanding how the 7407 gates work

Sorry that statement was meant to communicate that I will merely "glaze over" when I read posts like yours

On a very high level, I understand what you are saying. I understand the concept and I understand what "high and low" mean in this context but gates? I only know Bill Gates...

I don't know how to physically implement such a solution other than "plugging in a thing into another thing".

You mean well, but you are overestimating my abilities and so far, I fail miserably to communicate my skill-level.
I'm not going to be rich&famous i'm afraid.

[giantclam]

That's fine ~ it's probably best to view it like this -- you pretty much now have the instructions, and with that the need for someone to read those and build the cable for you ; any electronics /hobbyist/repairer can read what I posted, understand the optocoupler usage, and make it happen -- you don't have ppl around you into electronics?

Big picture stuff...you'd think someone on this forum, close to your locale, could do it for you and just post it to you (for reasonable costs ofc =) ...but as is anyone anywhere could do the same thing. You just need to find such a person 8)

[StompinSteve]

I do know someone that is able to create a solution. I already asked him.

But I am somewhat astonished that I cannot find a ready-solution somewhere. There are (were) so many WarpEngines out there, it is hard to imagine that all those 3000 and 4000 owners had no working HDD-LED after installing their WarpEngines and did nothing to solve it.

(come to think of it, 4000D owners might be able to simply disconnect the IDE LED and hook it up directly to the ZuluSCSI, not sure though, my 4000D left my house 25 years ago).

[giantclam]

No, it's not like that - warpengine & zuluscsi are 2 different things...ie;



That's actually correct ~ most all 50pin SCSI hard-drives around back then had a 2pin jumper header to connect to an 'activity/remote' LED. I had a couple of single height 5.25" drive carriers, and the LED was on the front (connected to the drive itself).

[Jope]

Yep, the A4000 LEDs are just LEDs and wire. I have mine removed from the motherboard and hooked up to a SCSIDE board.

[StompinSteve]

The ZuluSCSI board does not come with a LED header but it does have the two "holes" prepared though so in all my Zulu's, I soldered-on a simple jumper-header that I de-soldered from some old motherboard.

In my A1200 (in CheckMate 1500plus case), two wires (+ and ground) come off the Zulu with standard jumper-wires and each wire splits into two wires (like a Y ). One pair goes to the case HDD LED and the other to the Beleths Drum SCSI HDD Sound Emulator board (for that good 'ol sound that them old Quantum LPS and Fireball drives used to make :-)

[StompinSteve]

Bonjour community,

I decided to not pursue a solution based on optocouplers and the like. Too complex for me.

On the LED board, I cut the trace of the HDD LED's "shared +" right at the edge of the solder-point (and made sure there was no more continuity). In then de-soldered the HDD LED's ground-wire (which goes to the motherboard).
Next I soldered a wire directly to the + pin of the LED (now isolated from the trace) and a new wire to where the former groundwire was.
Did continuity testing, made sure that the HDD LED's + is connected to the Zulu LED + and ditto for the ground wire. All good.

Hooked it up to the ZuluSCSI and.... nothing. I know the Zulu LD Header works because I "Y split" both wires coming from the Zulu to a Beleth's Drum + and - which generates HDD sounds (kinda emulates a Quantum LPS SCSI Drive) and it happily rattles away on the beat of the Zulu's onboard activity LED, so I know (I hear) that the Zulu LED Header works fine.
But the HDD LED does nothing.

The yellow HDD LED worked before I started this "operation": The LED blinks very dimly a few times when the 3000D boots and scans the internal SCSI-Bus (which has nothing on it). So at this point, the HDD LED still worked.

Could it be that these LED's have a minimum (forward) voltage under which they don't work at all? I measured the Power LED and it reads 3.82 V and 4.04 V depending on the state of the low-pass filter.
The Zulu LED Header outputs 3V. if these orginal "Commodore" LED's need, say, 3.3 V to start doing anything at all, that would explain why nothing happens at the moment.

[StompinSteve]

Update: I attached a "lab power-supply" to the LED. The PSU can supply 0 - 15V up to 3 AMP very granularly and with this, I confirm my own suspicion. These original LED's need a starter Voltage of 3.65 V.

At 3.65 V, the LED starts to shine very faintly. At 3.8 V it is a bright as the power LED if in "low". At 4.0 V the HDD LED reaches normal brightness.

So the HDD LED is good (pfew..) but I need to feed it 4.0 V to work and the ZuluSCSI only provides 3.0 V.

I know how to reduce a voltage by placing a resistor inline on the + wire. But I have no idea how to **safely** increase the voltage from 3 to 4 V without damaging the Zulu.

With my limited knowledge of electronics, I found these small Adjustable Step Up DC Power converters that can do this, but are they safe, as most of these units, take the + and - from the source, step it up and output a higher Voltage. They only have 4 pins: 2x "in" and 2x "out". But I guess this puts a strain on the source (higher Amps? I dunno).

Examples:
https://www.techtonics.in/dc-dc-3v-3...oost-converter
https://www.ebay.ch/itm/174959053725...Bk9SR5TBpfmVYg

[giantclam]

I know how to reduce a voltage by placing a resistor inline on the + wire.

Resistors in series do not change/lower voltage, they limit current. Resistors part of a voltage divider do change voltage (but the circuit isn't this)

LED output from the zuluscsi card will be limited to around 40ma (by current limiting resistor). This output will not driver those converters.

But I have no idea how to **safely** increase the voltage from 3 to 4 V

Same answer as before ~ opto-coupler ... easiest, lowest parts count (1 opto + 1 resistor), and protects the zuluscsi output.