Solid State Power Box

[posplayr]

Put me down for one too. I think its a great idea, and would go well with Rick's stator and the SH-775.

OK great; I will put you down.

[posplayr]

I had a quick look there and just the following:



I think having an alarm indication even with ignition off is going to cause problems. Some people may park the bike for a long time in disgust and with the intention of getting to it some time. Will the unit have any current draw when the ignition is switched off? Any current draw with ignition off in normal or fault mode should be avoided even if only a few mA.

If the alarm is going off then then there is an issue that should be attended to. I did not measure the thermal duty cycle , but the switch limits at 12 amps for about 200 uSec and then shuts off. It is probably something like 1% duty cycle so that alone is 120 mAmps draw for a shorted input.

I actually have more concern about not hearing the buzzer than it being annoying. I doubt you will hear it if you are sitting on the bike with the engine running and a helmet on. If you have reasonable hearing, then sitting on the bike with helmet but engine off you should be able to hear it. If the warning is due to anything other than the unswitched/Batt power the buzzer will probably go off with the ignition switch because power will be cut to those circuits. If it is really annoying you can always disable the buzzer, but then if you are going to ride like that I would recommend mounting a flashing warning LED on the dash.

I should note that this is a small buzzer not something like a theft alarm deterent.

The quiescent current draw will be less that 1 mAmp, probably much lkess which is far less than the self discharge rate of a typcial lead acid battery

.001 amps/14 AmpHr is C=-0.01% this is extremely low and is essentially going to be imperceptible. I'm not even sure how to measure the draw, but the spec sheets say it will be much less than 1 mAmp.

OK a LOT of people have issues with buzzers (that irritating ding dong in some cars!) on motorcycles, so rather make the buzzer only work at ignition switchon and time out after 10 seconds. The circuit if not attended to, will stay tripped out like a blown or pulled fuse and the rider could probably use the bike as he would have done before, with just a quick reminder at startup only.

This is not a key buzzer; it is a fault indicator. Generally if it is buzzing you have a short and you need to attend to it, most times this happens I could not even see continuing to ride unless it was to simply limp home. Since all circuits are protected and there is a status available, I plan to make the warning active on any fault key on or off. It may seem an irritation but from a safety and reliability stand point any fault should be dealt with immediately. There is a note in the device data sheet about repeated faults and that the device is not intended to indefinitely protect against a short like what would occur if you shorted your battery and left the bike sit for months. I don't know what the limits here are but it is one of the tests I will have to do. If draining the battery with a 30 mAmp buzzer/LED is the only way to stop the short current then that is not a bad thing.

I presume the Description and Operation note at the moment is just temporary. In you final operation/user manual I sugets you ad a section called Optional Enhancements or something like that. Here you can add the details for adding/converting to the Ignition Relay modification, headlamp cut out relay modification and the switched accessory jumper configuration.
This part as it is now included will cause a lot of confusion and make your unit seem unnecessary complicated. Reading and seeing it separately as an option will make the basic written guide seem simpler and easier.

This is not an instruction manual but a more technical/functional description of how to the device works. In fact I came to the idea of using this style of interconnect (equivalent Bosche relay) after reading both Koolaid kid's and your system relay mod write ups. When I get a unit produced I will go through the install creating descriptions with picture as I go through the process on my bike.

I like the idea of the fuses in series with the solid state fuse as this gives an option to isolate a circuit manually for faulting. I assume your electronic fuse set at 10A will trip faster than the 10A standard fuse? Or you may have a 15A rather as the secondary backup to the 10A electronic one?

These are fast blow solid state fuses intended as a fail safe in case:

1. the owner fails to use a Battery inline fuse and a short occurs
2. one of the smart FET's shorts high side closed and then subsequently there is a short to ground in the circuit that failed.

Otherwise these should remain in place and cant be removed without doing "brain surgery" with a soldering iron.

Good work you are doing here, very interesting, best of luck.

Thanks; it is coming along

Put me down for one too. I think its a great idea, and would go well with Rick's stator and the SH-775.

[posplayr]

Hi. I am not here. I am studying for my new job. Just looked quickly at the extra pages since I last logged in. I am interested in two units. But I would need to know the initial batch cost. Daniel

I'm still looking at $150 per unit, the length of the warranty is TBD at the moment. I'm toying with lifetime (repair or replace) warranty for original owner but don't really know how badly this can be abused.

The device should be very robust and not subject to failures under normal or moderately abused care. A direct lighting strike would probably kill it. Many of the other devices (Ignitors, R/R's) would fail first

Bump; I provided an update to the first page of this thread along with a downloadable functional diagram, Comments welcome. I'm getting pretty close to sending out to have the PCB's made so now is the time for any comments and changes.

I had a quick look there and just the following:

The SSPB will have two LED's and a Buzzer to provide feedback to you about what is going on inside. When the Ignition Switch is powered on you will see the GREEN LED come on. This is indicating that the unit is receiving power and the IGNITION switch signal. In the event that there is a FAULT, then a RED LED will flash and a Buzzer will sound indicating the presence of the fault. This will ocuur even if the Ignition switch is off and the GREEN LED is off. Once you are aware of the FAULT you should fix the problem. But it will be possible to open the SSPB cover and disable the buzzer. You will not be able to disable the RED LED however. There will also be a provision for connecting a handlebar mounted WARNING LED if you like.

I think having an alarm indication even with ignition off is going to cause problems. Some people may park the bike for a long time in disgust and with the intention of getting to it some time. Will the unit have any current draw when the ignition is switched off? Any current draw with ignition off in normal or fault mode should be avoided even if only a few mA.

OK a LOT of people have issues with buzzers (that irritating ding dong in some cars!) on motorcycles, so rather make the buzzer only work at ignition switchon and time out after 10 seconds. The circuit if not attended to, will stay tripped out like a blown or pulled fuse and the rider could probably use the bike as he would have done before, with just a quick reminder at startup only.

I presume the Description and Operation note at the moment is just temporary. In you final operation/user manual I sugets you ad a section called Optional Enhancements or something like that. Here you can add the details for adding/converting to the Ignition Relay modification, headlamp cut out relay modification and the switched accessory jumper configuration.
This part as it is now included will cause a lot of confusion and make your unit seem unnecessary complicated. Reading and seeing it separately as an option will make the basic written guide seem simpler and easier.

I like the idea of the fuses in series with the solid state fuse as this gives an option to isolate a circuit manually for faulting. I assume your electronic fuse set at 10A will trip faster than the 10A standard fuse? Or you may have a 15A rather as the secondary backup to the 10A electronic one?

Good work you are doing here, very interesting, best of luck.

[posplayr]

Bump; I provided an update to the first page of this thread along with a downloadable functional diagram, Comments welcome. I'm getting pretty close to sending out to have the PCB's made so now is the time for any comments and changes.

http://www.thegsresources.com/_forum/showpost.php?p=1887048&postcount=1

[Sci85]

This has been great to follow along. I wish I knew more about the technical specifics but am learning a few things anyway. I am eagerly awaiting the final product.

[posplayr]

Hi Jim,
Thanks for the feedback. The issue around the size or footprint and the hardware (enclosure, mounting and connectors) can sometimes be bigger challenge than the actual electronic design!

Yep the connector is one of the biggest compromises. The contacts are rated at 10 amps and as long as the main current comes in on the discrete wires it is OK. Will need a little fiddling to come in on the bottom but workable,

The status led and audible warning will eventually determine how user friendly your unit will be, regardless of the sophistication of the insides.
Your first clients will probably buy it as they understand what you are doing, others later on will not care as long as it does the job, which will include indicating clearly what is wrong if something goes wrong.

The RED LED flashes and that in concert with a two tone beep should be enough to detect a n issue when the engine is off but side over on. I might add a through hole pads in case someone wants to run an LED off the dash purely as an option . I was not planning on supplying an extra LED or wire for that through.

Individual circuit led's should indicate which circuit is faulty and the audible should not be annoying otherwise it will just be permanently disabled. Consider a slow beep or slow ticking sound. I suggest a miniature led for each circuit, let it flash with the thermal cycle of the Fet.

It is a small place and I don't want to make it too technical specially with automatic fault isolation. It is easy enough to determine of the engine will not crank, start , headlamps not on or tail lights blinker not operating without needing LED under the side cover. So for no I have just settled on one power LED, green and one status LED RED flashing when there is a fault detected.

You could consider backing up the circuits with miniature auto fuses, which can override a failed open Fet when plugged in. If you use a 3 way socket per fuse one way could be fuse in serial and pulled for disabling a cicuit manually and the other position in parallel to bridge the Fet...? If you bridge the Fet the individual status led should also come on to indicate this.

I am concerned amount either having a failure and causing a short in the harness; that just is not acceptable. Leaving someone stranded in the case of an internal fault, is not really an option. Many problems that will originate in the electrical system, will have to be dealt with bu the owner. The SSPB can only provide a heads up that something is going on. Putting blade fuses into the design would make it much larger and begs the question of what is the point if getting rid of them was the original concept in the first place.?

I assume that the ignition switch will not carry heavy current anymore and just switch on the Fets. Thus your highest calculation for voltage drop being 0.25V should be a major improvement on what we have now via the mechanical parts.

There should be only a few milli amps through the ignition switch , except in a "wakling wounded mode" , where you can revert back to using the ignition switch to power the bike in the event of an partial of full failure of the internal SSPB FETs.

I like your idea! How about designing one for fault finding? It would be great tool to have something like this you can just plug into a fuse position to find a short circuit or even an intermittent short blowing a fuse, without having a handful of blown fuses at the end.

Yes you could easily design a tested light for detecting shorts, but the not sure how much better it would be that disconnecting the battery and using an ohm meter.

More later

Sorry just saw your post.

You are correct in needing some feedback . I tested the prototype and with bike running and all lights on I grounded the switched ACC output and nothing outwardly appears to happen. The smart FET detects the condition within 200 us limiting current and then going into a cyclic thermal shutdown recycle.

The smart FETs have a status indication which I have tied to the flashing red LED. Given the SSPB will be added under a side cover I looking at adding a small buzzer that can be disabled by a jumper under the cover.

I've put together a FMEA (Failure Modes and Effects Analysis) matrix for various faults/failures am still working through fail safes and " walking wounded" modes for getting you back home without needing a soldering iron in the event a FET fails open. I've added additional SMD fuses (non replaceable) as a back up to the smart FET controls in case of a fail short. So there should be full protection against any type of fault with or without the external inline fuse.

I realize the importance of such a device to the safety and reliability of your motor cycle perhaps more critical than any other. The PCB will be we'll protected beyond what is already built in for a robust automotive grade parts.

I don't see any particular sensitivities that the unit will have and don't see anything that would give an abnormal failure rate. It should at at least S reliable as any MOSFET or SERIES r/r realizing that most all r/r fail due to poor connections. A big part of this design is go eliminate connections that can make you susceptible to corrosion . Preliminary calculations are that the power dissipation should be low (less than 10 watts at full capacity ). There should be no more than about 0.25V drop at 14.5V 20 amps

One of the things I'm working on at the moment is how to easily revert back to a
Running off the ignition switch with fused circuits if FET failed open. This the least likely failure mode but not a situation you want to be faced with without some limp home strategy.

Should be able to selectively override each FET circuit which I think also satisfies your diagnostic criteria.

Hi Jim,
Thanks for the feedback. The issue around the size or footprint and the hardware (enclosure, mounting and connectors) can sometimes be bigger challenge than the actual electronic design!

The status led and audible warning will eventually determine how user friendly your unit will be, regardless of the sophistication of the insides.
Your first clients will probably buy it as they understand what you are doing, others later on will not care as long as it does the job, which will include indicating clearly what is wrong if something goes wrong.

Individual circuit led's should indicate which circuit is faulty and the audible should not be annoying otherwise it will just be permanently disabled. Consider a slow beep or slow ticking sound. I suggest a miniature led for each circuit, let it flash with the thermal cycle of the Fet.

You could consider backing up the circuits with miniature auto fuses, which can override a failed open Fet when plugged in. If you use a 3 way socket per fuse one way could be fuse in serial and pulled for disabling a cicuit manually and the other position in parallel to bridge the Fet...? If you bridge the Fet the individual status led should also come on to indicate this.

I assume that the ignition switch will not carry heavy current anymore and just switch on the Fets. Thus your highest calculation for voltage drop being 0.25V should be a major improvement on what we have now via the mechanical parts.

I like your idea! How about designing one for fault finding? It would be great tool to have something like this you can just plug into a fuse position to find a short circuit or even an intermittent short blowing a fuse, without having a handful of blown fuses at the end.

[posplayr]

Very interesting and a brilliant idea!

Just a question or two:

Will your unit have any sort of warning when a circuit has been automatically cut due to a short?
I had a Harley here with a short to ground in the handlebar switch and the audible ticking of the 15A auto fuse was an immediate clue.

Lets say you have installed it in place of a 5 fuse unit, will there be an indication as to which circuit is switched out during a fault?

Will you have any manual method of isolating a "fused" circuit to allow easy fault finding? Based on pulling a fuse.

Sorry just saw your post.

You are correct in needing some feedback . I tested the prototype and with bike running and all lights on I grounded the switched ACC output and nothing outwardly appears to happen. The smart FET detects the condition within 200 us limiting current and then going into a cyclic thermal shutdown recycle.

The smart FETs have a status indication which I have tied to the flashing red LED. Given the SSPB will be added under a side cover I looking at adding a small buzzer that can be disabled by a jumper under the cover.

I've put together a FMEA (Failure Modes and Effects Analysis) matrix for various faults/failures am still working through fail safes and " walking wounded" modes for getting you back home without needing a soldering iron in the event a FET fails open. I've added additional SMD fuses (non replaceable) as a back up to the smart FET controls in case of a fail short. So there should be full protection against any type of fault with or without the external inline fuse.

I realize the importance of such a device to the safety and reliability of your motor cycle perhaps more critical than any other. The PCB will be we'll protected beyond what is already built in for a robust automotive grade parts.

I don't see any particular sensitivities that the unit will have and don't see anything that would give an abnormal failure rate. It should at at least S reliable as any MOSFET or SERIES r/r realizing that most all r/r fail due to poor connections. A big part of this design is go eliminate connections that can make you susceptible to corrosion . Preliminary calculations are that the power dissipation should be low (less than 10 watts at full capacity ). There should be no more than about 0.25V drop at 14.5V 20 amps

One of the things I'm working on at the moment is how to easily revert back to a
Running off the ignition switch with fused circuits if FET failed open. This the least likely failure mode but not a situation you want to be faced with without some limp home strategy.

Should be able to selectively override each FET circuit which I think also satisfies your diagnostic criteria.

Excellent work Jim, I am in for one or maybe two.

[posplayr]

The glass style off a 78 1000 is:
40mm wide
58mm long
35mm tall with the cover
58mm center to center at the mounting holes
72.5mm total width @ mount ears
Hope that helps

Thanks that help

The glass style off a 78 1000 is:
40mm wide
58mm long
35mm tall with the cover
58mm center to center at the mounting holes
72.5mm total width @ mount ears
Hope that helps

[salty_monk]

I can probably do the other one... looks like the GS1000 one. I even have a spare you could borrow if it would help.

[posplayr]

The one in top photo is looks like mine

L - 3.525
W - 1.535
C/C - 2.270

Thanks, that means the principle widths are the same as the later spade type fuse box and I'm good with the box I'm working with . It will be a little tight and need allen head screws to mount but look like it will fit in the OEM locations without having to relocate holes (width of the SSPB will be about 1.75").