Bike warms, inconsistent idle, blows fuses

[member14827]

So here's one for you...

I'm nearing the end of my '80 XS400G project (original post here) and am having an issue that only rears its head when the bike warms up and has been ridden a mile or two. First symptom is the idle: with the clutch in or while stopped at an intersection, the idle will start to drop to the point where the bike seems to want to stall. I can keep it from stalling either by revving as needed or using the throttle screw on the bottom of the carbs. Then sometimes it'll shoot into high idle and doesn't want to fall back down until I back that same screw out or put the bike in gear. It's all subtle at first, but gets progressively worse as I continue to ride (typically within the first 10-15 minutes). It really sucks to be at a stop light or in the middle of a low speed turn and have bike stall out because the idle dropped too low. Then -- this has happened three times now -- I lose electrical power due to a blown fuse, once the 10A "turn signal" circuit and twice the 20A "main" circuit.

The first time this happened, my main fuse blew. I traced it out at home and found that both my license plate light and speedo lamp had burnt out AND shorted. I don't understand why, per design, these lights (and the running light) aren't segregated to one of the other circuits. To correct this, I wired them into the headlight circuit to pull 12VDC from the R/Y wire. Everything is LED, so no need to increase fuse size. Anyway, I suspected that the blown lights were symptomatic of a larger issue, over-voltage maybe at high RPM.

I was still seeing ~500 Ohm from Main 12VDC to ground, measured at the fuse box. This is mostly a combination of the TCI box (~1k to ground from the red/white lead) and the voltage regulator (~1k to ground from the brown lead). I actually have a second TCI box and it measured the same, so I presumed that was within spec. This left me with the voltage regulator (original) which was already cause for suspicion. So... I replaced it with a BWD Select R296 per threads here and here. (FWIW, the R296 measures ~3k to ground from the same lead) After getting it wired in, I started the bike and tested the voltage to the battery at high RPM: about 14.7VDC, no higher. Seemed pretty good.

I test drove the bike shortly thereafter. Within five minutes, idling is rough as described above. Then, bam... my turn signal fuse blew (brown wire). Ignition was still good, albeit rough, so I limped the bike home and put her away in the garage.

I didn't find a short, so, for experimentation, I popped a 15A fuse in (I know, I know ...but... science), started the bike and took her for a test ride. Everything's good until about ten or fifteen minutes in -- same idle problems, worsening, then, bam... main fuse blew. This time I walked the bike home.

This does not appear to be RPM related. I can rev the bike pretty high after startup and have no issues. It's only after I've been riding for ten or fifteen minutes, at normal RPM, that things start to haywire.

Does this sound immediately familiar to any of you? I'm pretty familiar with electronics, but this really has me scratching my head. I'll post my updated circuit diagram to show the changes I've made and the things I've removed (starter, starter motor, horn, front turn signals, headlight relay...).

 

[member14827]

 

[member14827]

Some additional information:
- Both ignition coils tested good (primary at 3 ohms apiece); I have not tested the secondary
- The rectifier tested good (diode test), prior to all of this.

It sounds to me that, as the bike warms up, something is progressively pulling more current, starving the ignition system, and eventually pulling so much that it blows a fuse. The question is... what?

Addendum:
I just had a chance to check the circuit again. No short on the main circuit, or at least nothing shorted permanently. I plugged in a new main fuse and everything worked again. I still suspect the "turn signal" circuit this last time, and that 15A fuse I plugged in temporarily allowed for the temporary short to bypass that circuit and blow the main fuse.

On a related note, the "turn signal" circuit is kind of misnamed. There are a lot of components that are fed 12VDC from that brown wire circuit, both downstream of the fuse ("turn circuit") and upstream (technically "main" circuit):

Brown wire main circuit:
- Oil or Neutral switch
- Voltage regulator
- main switch (key)

Brown wire "turn signal circuit":
- rear brake light switch
- front brake light switch
- flasher relay
- oil and neutral indicator lights

 

[buztabuzt]

I'll be interested what others say but curious your timing setup and TCI

 

[member14827]

The pick-up sensor and TCI box are both stock. I happen to have an extra TCI box that I purchased back toward the beginning of my build when I was chasing down other gremlins.

 

[NewHavenMike]

Major post lol

Hows the battery? What are you using?

Check your grounds, definitely need a couple of those. Ive had frayed wires cross and cause a short too.

My 360 engine had similar running issues, which has an on/off switch for lights. The extra draw from the lights(when I used them) would cause a low idle that made me nervous. Without lights=high idle. I probably should’ve adjusted the idle with the lights on constantly. I use traditional bulbs=more draw.

My bike doesnt have a TCI, but it could be on its way out.

I wouldnt use a fuse thats rated higher than the circuit, thats how a fire is created.

 

[member14827]

Hows the battery? What are you using?

This is the battery I'm using: https://www.amazon.com/gp/product/B00KC399SY/ref=ppx_yo_dt_b_asin_title_o04_s01?ie=UTF8&psc=1
Bit of an experiment, but it's actually performed quite well thus far. I know it's for alarm system/UPS applications, but given that I've removed the bike starter, it's sized appropriately I think. It's actually held up better under higher current draw than a more traditional, miniature/auto lead acid battery I've used for comparison. It doesn't seem likely to me that it would cause the issue I've described above, but it is a bit of an unproven component in the build as I've not seen anyone using one of these before.

Check your grounds, definitely need a couple of those. Ive had frayed wires cross and cause a short too.

I've been very diligent about my grounds. Most of the individual component black wires connect to two "star" grounds on the frame, one forward and one aft of the battery box.

My 360 engine had similar running issues, which has an on/off switch for lights. The extra draw from the lights(when I used them) would cause a low idle that made me nervous. Without lights=high idle. I probably should’ve adjusted the idle with the lights on constantly. I use traditional bulbs=more draw.

It was my plan to convert everything to LED. However, my rear turn signal lights and speedo light are still traditional. Soon to change.

My bike doesnt have a TCI, but it could be on its way out.

It's a black box, literally and figuratively, with dozens of components in the circuit. It's cause for suspicion, but I'd think if that were pulling too much current then my "ignition" fuse would blow, which hasn't been the case.

I wouldnt use a fuse thats rated higher than the circuit, thats how a fire is created.

Generally I'd agree with you. In this case, I presumed that if the current draw on the turn signal circuit was greater than 10A (stock fuse) but less than 15A (what I plugged in as an experiment), the short would still be caught by the main circuit fuse (20A), which was what happened. The only components I put at risk by doing this were the brake switches, the flasher relay, and the oil & neutral indicator lights. It was a measured risk

 

[member14827]

Addendum II:
Regarding the usage of the alarm battery, I'm not special. Check this old thread. Apparently it has been done successfully.

 

[member14827]

Addendum III:
So far, I've been operating under the presumption that the idle issue is related to the electrical issue. This may or may not be the case, but if anything, the idle issue is the first problem.

I tested the bike last night on the center stand for about fifteen minutes, running with the multimeter hooked up, trying to see if I could induce any kind of voltage change by unplugging components, using the turn signals, etc. Results were inconclusive. Eventually the bike did idle at lower and lower RPM until it finally died and I couldn't get it started again (remember kick start only).

I took it for a test drive this morning without the turn signal relay plugged in, as an experiment. No fuses blew but after fifteen minutes of riding the bike idled lower and lower until it finally died and I couldn't get it started again. I rolled it back into my driveway and noticed that the bottom of my left fuel filter (Uni - 6") was wet with fuel. This is another problem I've been dealing with on and off since I replaced my petcock with the non-vacuum type from MikesXS. I pulled my plugs and the left plug cylinder was both black (rich) and wet (gas). Then it hit me...DUH.

So now I'm pretty sure my idling issues after warmup aren't electrical. It's just the carb bowls overflowing slowly and the engine start running progressively richer until it floods.

I understand that, with the non-vacuum type petcock, I need to ensure that my needle jets are sealing and that my floats are set to the correct height (I thought I'd already had this all taken care of, and I won't really get into detail here, but apparently I still need to address it). So that's one problem.

Second, is there any way that this running progressively richer problem is actually CAUSING the fuses to blow in the way that I've described? Or am I likely dealing with two separate problems?

 

[NewHavenMike]

Aside from other carb issues. Running rich when warm is a sign of main jets being too big.

 

[member14827]

Main jets are stock, but air filters are not. I'm using the stock 'H" intake boot with 6" Uni filters, per XSChris' recommendation. Not sure if jet resizing is needed with that setup.