Tuesday, May 11, 2021

MGB - Ignition Complete

My last couple of posts have spanned 3 weeks, and, honestly, most of the work has as well. Today's post ends with a running car, and covers the final step in the ignition journey.
Best Laid Plans
I thought my plan was so simple. I would just pull out a box of bolt-in ignition parts and I would go from a running-crummy car to a magical sled. As you have witnessed from the last couple of posts, it was not that simple. First, the battery was dead, then the fuel pump was bad. Now, we are finally ready to set the timing, but there doesn't appear to be any spark coming from this new ignition. Why?

Lucas, Lord of Darkness
First, that's not my joke; I borrowed it from Hal. Thanks, Hal. Years later, its still funny, and in so many cases, its still true.
Lucas CEI ignition wiring diagram

I figured that the red and black wires for the ignition signal in the new distributor could just re-use the red and black wires that had been feeding the old distributor. That was my mistake, because I did not recognize this was not your standard distributor. No... this was a Lucas CEI ignition system. While I imagine these were great in their day, this set up was a little more complicated. The 2-wire signal going to/from the distributor does not go directly to the coil. It routes through an amplifier box. This amplifier box appears to only work with the Lucas stuff, though. I believe that either the replacement hi-power coil, the new distributor or both are not compatible with the amplifier box.

The amplifier does not make things much more complicated: it only adds 2 wires to the coil, but the wiring around the coil is already a little confusing. Add in some extras and some grease shmears and trying to figure out what wire is for what starts to get more difficult, especially when the wires are not in the standard diagram. The image above is correct, but it is not in the standard "advance autowire" wiring diagrams, nor can I find references to the Lucas CEI in the MGB shop manual that I have. Still, this research did tell me that there is no ballast in the circuit from the coil to the distributor. There is a ballast between the coil and the fusebox (and that remained), but I figured I should be able to direct-wire the red/black wires to the +/- on the coil directly. The picture below should help show what a "ballast" coil set up looks like. See that "distributor resistor" in the drawing below? It is not in the drawing above so I don't have that. So, we start over. Kinda.
Final Install
standard 1978 ignition wiring diagram
I started by removing the distributor again. I had moved the 2-pin plug from the Lucas distributor onto the AccuSpark so I undid those solder joints and soldered on 2 1-foot lengths of 16ga wire (black and red for negative and positive, respectively). I heat-shrink'd the solder points. To the ends of these wires, I added female wire disconnects and heat-shrink'd those as well.

Next, I disconnected the wires going to and from the amplifier box and then removed the box itself. I re-installed the coil and the 2 things that shared the mount-points with 2 shorter bolts. Then, I slid the distributor back in, and re-set the mounting bolt to hold it firm. I had found TDC earlier, as noted in my previous post, so I oriented the distributor visually into the same place (took a picture before removing). In theory, everything should be fairly well set, at least well enough to start, albeit roughly. I optimistically pulled out my timing light just in case.

On an aside, the MGB battery is behind the passenger seat. This makes use of a timing light challenging. To remedy, I sat an old battery on the floor by the front bumper on the passenger side and clipped my positive and negative leads for the timing light onto it. 

It's Alive!
tidied up
I double-checked my under-hood connections, turned off the KBOO and then returned to the driver seat. I turned the key to run and heard the click-then-hum of the fuel pump as well as a click from the engine compartment. Thinking that was the coil leaping to life, I turned the key to start and Oliver started immediately. I mean immediately. No ba-ba-ba beforehand, no choke needed. He just started, and then dropped into a lope-y idle.

Setting Timing
Thrilled, I clipped my timing light sensor on spark lead #1 and pointed it at the main pulley. It flashed on the timing mark, indicating I was around 10* before top dead center (BTDC), which was in the ballpark (suggested 6-14 degrees advanced from TDC by AccuSpark). I tried rotating the distributor housing a little bit to advance or retard the timing, and found that Oliver was most content around 12* BTDC. My light is a cheapy, and does not have a knob to check full advance. So, instead, I used my ears, revving the engine up and down and then revving it up and quick-pulling my foot off the pedal. This last test was when Oliver used to backfire. There was none of that now. Of course, he is not under load right now either. Undeterred, I went back up front and checked his timing again: he was still settled at 12*.

I turned off Oliver's engine and tightened down the timing adjust set bolt (5/16") so the distributor would not rotate on it's own, and his timing would be locked in. I was done. Wow. I hadn't expected that. So, I spent a few minutes cleaning up some of the wiring that I had earlier cut loose from the taped loom by encasing it in some basic wire wrap.

leads now
I included the before picture of my ignition below (blue leads) and a picture of the ignition now on the right here (red/orange leads), taken from about the same spot. If you look closely, you will notice that the newly installed distributor is oriented at least 45* anti-clockwise compared to how it was before. So, either I had moved the dizzy a little before taking that before picture or the engine was running so incredibly off it is kind of a miracle it was running at all. Perhaps the backfires on deceleration could be attributed to the timing being so far off that the dizzy could not get back to a fire-at-the-right-time state when I took my foot off the pedal.

At this point, I'm not sure it was ever firing at the right time. Based on how quickly he started, and how well he ran after I revved him a few times, I am confident the mark I used for TDC and his timing are correct. The lesson here, I think, is to give your new-to-you car a tune-up as soon after you get it home as you can. I have been given this lesson before (See Bay City Blues), but I guess I need multiple examples before I learn sometimes. Bare minimum, I will start looking at my spark plugs during oil changes so I have a regular insight into the combustion. Might be worth a quick validation of the timing when I do that.

Oliver is pretty much ready to go again. He hard-started a few days after I did the timing I mentioned above. So, I still need some carb adjustment time, and he needs to go through DEQ, of course. An oil change is overdue as well. Otherwise, he is practically ready for Summer. As am I. Just thinking about a top-down drive with Boo for a picnic at a park sounds just about perfect. I'll post on the carb tuning when I get to it.

leads before
I'm not sure what I will be working on next, but nice weather is right around the corner so returning to the paint on Zed is inevitable. Some of him sat outside all winter, so I imagine there is quite a bit of rework coming before I can think about laying paint. Meanwhile, T is due back in the Pacific NorthWest to fight forest fires somewhat soon, and he will probably want/need Nemo on his off days. So, maybe I need to get that rear end sorted. Of course, Nemo won't start now or even put a signal to the OBD2 plug, so he is going from bad to worse fast.

Regardless, there's always something to do at ShadeTree, like helping a local kid swap out his muffler. Oops, I guess I forgot to post on that. We had fun and he learned something: a fart can muffler on an otherwise stock exhaust doesn't really do anything. For a meaningful change in tone and performance, he will need to remove at least one of the other resonators. He is saving up for a cat-back; that will be fun.

Thanks, as always, for following along.

Tuesday, May 4, 2021

MGB - Fuel Pump 2 Step

Picking up where we left off with Oliver, the 1978 MGB, we had completed the ignition swap out and we were ready to start the engine and set the timing. Until the battery went flat.

To Start Or Not To Start
should be an image from AutoZone
AutoZone product picture
At this point, I was ready to test-start it. Or, so I thought. Last Fall, I thought Oliver's battery was going bad. It later turned out to be the clock sucking all the juice while left sitting for weeks (see MGB Battery Monitoring). Before I learned that, however, I swapped the battery out for the battery from the donorZed. At this point in my electronic ignition install, I discovered that the donorZed battery actually was bad. I mean bad. Like, I would put it on the charger overnight and then I could put a voltmeter on it and watch it discharge down to 0. Fortunately, I had left Oliver's prior battery on the trickle-charger, because I guess I was not 100% convinced it was bad. So, I swapped his old battery back in. I turned the key, heard the fuel pump click once, and then heard nothing. In the past, I have had this happen when the fuel pressure had not let off since it was last running. Since the engine had not run since October, that was unlikely. Perhaps the inertia switch got nervous and is preventing the fuel pump? Well, then I would not have heard the one click. Still, I pushed the reset button. No change. Then, I jumpered around the inertia switch, running the wire disconnects through a 3A bladed fuse instead of the inertia switch. Same single click.

Fuel Not Pumping
what arrived
I concluded that the fuel pump, which had not been behaving consistently (sometimes just not working at start up), needed replacing. I slid under Oliver's backend and recalled that the original fuel pump had been replaced by the prior owner for what looks like a universal pump. I can't judge. Quite the contrary, I was not going to invest in a $200US+ original MG fuel pump on the assumption that the fuel pump is the problem either. So, I ordered a Spectra fuel pump from AutoZone for $48US delivered next-day-ish. This pump meets or exceeds the requirements for pressure (3.8psi) and delivery volume (18gph) for an otherwise stock MG engine. I sourced the requirements from this page on the MG Experience. The Spectra offers 2.5 to 5 pounds of pressure and 21gph, has reviews on Azn from MG owners saying they work great and even AutoZone shows these are a direct replacement for the Midget.

It is possible to over-power your carb with too much pressure, so that's one for the back of the mind while trying to diagnose what's going on with the fuel flow through combustion. I may introduce a pressure regulator, but one could argue that adding a cheap pressure regulator on top of a cheap fuel pump starts approaching the cost of a good stock-style replacement pump. While that's very true, I am not running a stock carb either. The prior owner installed a SK Racing (now known as OER) side-draft carb, which seems like a combination of the Mikuni PHH and the Weber DCOE, but this blog post has some great information about them. My knowledge of carbs is extremely limited, so I am accepting this as a learning opportunity.

Fuel Pump 2-Step
Holley Mity Mite
The pump that arrived was not as pictured. Lovely. This one was equipped with a 3/8" barb on one end and a 3/8" NPT female on the other. Into the 3/8" NPT, the consumer was expected to thread in the included mini-filter that has a 3/8" barb on the end. While I am sure a filter is needed, that one they included is tiny and virtually irreplaceable. Perhaps these pumps don't last past a fuel filter change. Since the pictures and the spec's I could see did not match, I was not going to go forward with it; I already have more than enough mysteries with the fuel system. I returned it unopened and got a Holley 12-434 Mighty Mite Electric Fuel Pump. These top-out at 4psi, push 28gph and actually ship with 5/16" hose barbs. Between the delays from CoViD and possible delays from that cargo ship that was blocking the Suez Canal for a week, the arrival of the fuel pump was not exactly immediate. Still, through the power of split-posts, it arrived in time to keep this blog active.
Fuel Pump Swap
The pump arrived mid-week, so I started my Saturday planning for the fuel pump swap. Over breakfast, I prepped the wires and assembled the pump and pre-filter. While the Spectra above also delivered with a filter like this, I know that Holley supports replacement pre-filters (I verified on their website). Either way, I figure I have one of those in-line clear filters just upstream of the pump, so I don't expect it to require service too terribly often. For the wire prep, I attached a male wire disconnect to the ground (black) wire and a male disconnect to the positive (red) wire.

ready for install
So, with the pump in one piece, and wire disconnects attached to the wires, I slid under the rear end of Oliver, bringing a collection of tools, a drain pan and a flashlight with me. I closed off the fuel supply from the tank with a small pair of vice-grips. I then moved from that hose forward through the fuel pump, disconnecting hoses and draining the contents into the pan. I only got as far as the pump before I didn't need to drain anymore. Clearly that old pump was not working anymore.

The MityMite is actually larger and heavier than the universal pump I had removed. So, I needed to adjust some fuel line lengths, but otherwise things just fit together. I reused the bolt hole which was holding the old pump to the front edge of the trunk, for example.Once all together, I wired the black wire to the grounding wire that was previously attached to the ground tab on the universal pump. Into the female disconnect coming from the pump, I put one blade from a 3Amp fuse. I put the other blade into the female disconnect which is ultimately wired to the inertia switch. I worked the disconnects with a pair of pliers to get the fuse to stay firmly in place. I wrapped the disconnects and the fuse with electrical tape. Last, I spooled up the extra wire and zip-tied it to the pump pre-filter so the connections would not shake loose.

Pump Test Fire
With everything as I thought it should be, I removed the vice grips and my tools from underneath. I turned the key to "RUN" and the pump fired right up. I left the key in "RUN" and checked the fuel lines for leaks. None appeared, so I concluded the pump was fine. Hoping for the best, I returned to the driver seat and tried to start the engine. Oliver would not start, but the engine would turn. That's a start. I removed a spark plug and I could smell gas on the tip, so We now have fuel and electricity. Next, we need to get electricity to the plug... at the right time.

Recall from my last post that my initial effort to find TDC were wrong, and that I needed to pull the valve cover to figure out true TDC? Well, I did that here. Still, even with true TDC found, Oliver would not start. I did figure it out, but that's my next post.

Thanks, as always, for following along--

Tuesday, April 27, 2021

MGB - Electronic Ignition

It was an unseasonably cold weekend recently. Since Oliver is in a nice warm-ish garage and Spring is supposed to be here now, I decided that I would start getting Oliver (1978 MGB) convertible-top-down, road-ready. Today's post starts us off with replacing the ignition. Like so many things with these old cars, nothing is as simple as it seems on it's surface, so this post ends without a running car. Maybe I'll get it next time. 

Most folks think of "Pertronix" when they think of an electronic ignition upgrade. PerTronix makes great electronic ignition systems, for sure. I ran one on the old bus engine for a little while. I decided I would try the AccuSpark system because it was less expensive and appeared to be UK-based. The UK-based nature had me thinking that maybe it was better suited to an English motor. So, back in the Fall 2019 when I was thinking that I would be spending the winter finishing Oliver, I bought one of their kits, and put it on a shelf. Things happened, CoViD-19 blew up, the wiring in Hapy got complicated and I forgot about it... until I was cleaning my garage and found the box with the international shipping labels still attached. Jeez.

For around 70 pounds plus postage, you can order a drop-in ignition. When I did it, it ran me about $150US including shipping to the States. So, what's in the box?
- a pre-fitted electronic ignition in a 45D distributor (correct dizzy for a rubber-bumper / later model MGB)
- a distributor cap
- 4 Champion AC9C 3-prong spark plugs
- a blue ballast coil (though you can upgrade to a "Viper Dry" coil)
- a set of spark plug leads ("wires" to US folks) with a di-electric grease pouch
- you can optionally include a timing light

It is recommended that you set your engine to top dead center (TDC) or firing position for the #1 cylinder. For most cars, the distributor could go in multiple ways, so making sure it is oriented correctly requires that you know which cylinder is ready to fire. The MGB engine is unusual in that it has a one-way-only key in the distributor drive and corresponding gear in the engine. If you try to put it in any other way it will not seat and not turn. Still, I wanted to make sure I had clear timing marks so I got a 1-15/16" socket for my 1/2" ratchet, removed the air-directing tin between the front of the engine and the rear of the radiator and turned the engine (clock-wise) until I could see the little nick on the rear of the main pulley. I aligned it with the top-most saw-tooth on the timing meter and then marked both with a dot of yellow paint. I also marked the tooth representing 15* advanced, since that is the stock amount of advance desired at idle (per the Bentley manual). This mark later turned out to be incorrect.
To find true TDC, I had to pull the valve cover and watch the valves move while I rotated the engine with the 1/2" ratchet. The mark that I believed was TDC was about 90* off. I found a different mark on the front (front is front) of the main pulley that was around midway between the points where the intake/exhaust valves opened. Accepting that this was the true TDC, I marked the nick with a dot of yellow.

With the replacement kit and tools on a nearby table, and the radio playing the Saturday morning KBOO shows, I was now ready for some ignition fun.

Spark Plug Remove and Inspect
My old ignition basically had not been looked at since I bought this car. Since Oliver has not had more than a couple hundred miles on him, clearly I did not drive his old system into it's current state. Frankly, it sat for a few years before I bought him, and without any servicing since, my expectation was that it needed, well... a lot. His performance had not been all that great when I bought him, and the new exhaust did not create the pick-me-up I had hoped for. So, away we go.

I started with removing the old spark plugs with a 13/16" socket. Spark plugs are a great gauge for how well your combustion has been performing: too rich, too lean, oily, too hot, etc. My findings were mixed, and odd. From firewall to fan, Plugs 4 and 3 were black, but dry, indicating that they had been burning inefficiently. Plug number 2 was black as well, but had some soot on it. Plug number 1 had a little bit of soot around the edge, but the prong was grey. So, I look at my plugs and compared them to the AutoLite Plug diagram (image posted here) to try to figure out what's going on.

Based on the mapping document here, numbers 3 and 4 are "carbon fouled", with lots of possible causes, ranging from "wrong plug" to weak ignition to mis-tuned carb, too-gentle driving or a bad choke. Neat, but at least they look like twins. I am more concerned about plugs from 1 and 2. These are fed by the same intake runner, yet they look very different. Plug 2 looks like a worse version of plugs 3 & 4, and plug 1 has a clean or normal looking ground electrode with some carbon around the edge.

L-to-R: plugs 4 to 1
These plugs may have been wrong for the engine. Spark plugs are a religion, and I have followed a relatively simple rule (that will probably frustrate / anger / enflame some people): I align the plug manufacturer country with the engine/car manufacturer country. So, I'll find NGK's for our Subaru, Bosch for Audi/VW's and Champions for US/English cars. The plugs I pulled out are single-prong NGK's; to be fair, NGK may make a perfect plug for the MG. I don't know. I can say that the AccuSpark folks (from the UK) sent me 4 Champion AC9C 3-prong plugs. MAybe they subscribe to the same flavor of spark plug religion that I do. I don't expect the plugs were the problem, but they may have been a contributor. Regardless, I gapped (0.035" per the Bentley manual) the three prongs on each plug and then installed the delivered-with-the-kit spark plugs.

Dizzy Swap
With the plugs done, I pulled the distributor cap and the lead from the coil. This created some access to the distributor mounting bolts. The prior owner installed a remote inverted oil filter unit, so the access to the front mount bolt was still a little challenging, but not impossible. Still, a little fun with an old-skool 7/16" spanner got the bolts out. Once the bolts were removed, the dizzy popped right out. I immediately settled it back in, and made a mark in the dust on the starter cover which direction the rotor was pointing. As I said above, the dizzy can only go in one way, so whichever way the rotor is pointing when you pull the cap is the only way it could be pointing. In the images I have seen on the internet, the number 1 plug is usually in the 1 o'clock position, but that is not always the case. Sometimes the dizzy gear is put in upside down so your #1 plug position is actually on the bottom. When I rotated the engine, the cap was on, and the leads had not been touched. I noted at this point, though, that the distributor rotor was pointing down and slightly to the left. When I removed the cap, the #1 lead was in that area, so I figured I would put it back together the same way. If things were 180* off, it will run, but poorly. Re-setting the leads 180* the other way would be easy to test later. I took the picture below so I could run the replacement leads the same way as they were.

old leads
With the old dizzy off, I loosened the clamp on the new one so it would move a little bit. The timing of the old set up might have been off, based on the "carboned" color of the plugs. Regardless, I want to be able to change it from the shipped-to-me state. I will clamp it tight (5/16" wrench) once the timing is verified. Satisfied that the dizzy would turn in the clamp, but only with some effort, I popped the new dizzy into place. The new mounting plate only had a single bolt hole for mounting, above and to the left, but otherwise, install was trivial. The lower bolt was the hard one to get to, so I was not going to complain. I put the now-extra bolt into the lower right hole anyway and snugged them both down tight.

The kit delivered with a new coil that corresponds to whether there is a ballast in your system. The electronics require the more powerful coil. Swapping a coil is easier than swapping a distributor, usually. In this case, because so many things leverage the mounting bolts for a ground, it is more of a juggling act than the coil swap in the old bus. I unmounted the original, and set it upright on the starter, leaving the wires attached to the various tabs. This gave me room to install the new coil and remain focused on the grounds and such. Once in place, I moved the female wire-disconnect-tipped wires from old coil to new coil.

Cap and Leads
The hard part, if you want to call it that, was over. Now, it was a simple matter of installing the leads. The prior set-up had the #1 lead around 7 o'clock. From there, they run the same way on all MG's: 1-3-4-2 anti-clockwise. The set of leads delivered with a clear dielectric grease, which is recommended on the electric connection to prevent corrosion and encourage a better current delivery. I smeared some on each end of all connections and plugged things in nice and snug. The electric connector floated around in the spark plug boot end a little bit so I made extra-sure the encased metal spark-plug end was definitely snapped on.

At this point, I would have started the engine and set the timing. Unfortunately, the turn of the key resulted in nothing. No noises, no clicks and definitely not starting. A quick glance at the voltmeter told me all I needed to know: flat battery. Awesome. So, this post got kinda long, and the resulting diagnostic did as well. So, I'll get back to Oliver next time. Thanks, as always, for following along--