Running engines without catalytic converter

Didn’t pick up on that earlier. I’m guessing you don’t mean a straight-through pipe (i.e. with no silencer) but a non-cat exhaust?

IOW the car originally had a cat but passed emissions without one. If so that’s very interesting.

I’ve never had a car with a catalytic exhaust system, and I’d like to keep it that way. Do you know what adjustments were made? My understanding is that a catcar has to run rich to keep the cat cooking, so presumably your mechanic had some way of leaning the mixture, since you imply he wasn’t using the old (cross my palm with) silver catalyst with the test centre.

Anyway, assuming it isn’t a cat, how about upper cylinder lubricant if you want to extend engine life?

Rather than buy snake oil, I’d speculate that a very little 2-stroke oil should be the right stuff, though I’ve never heard that suggested so there might be something wrong with it I havn’t thought of, apart from cost and slightly higher HC.

[quote=“Ducked”][quote=“sandman”]
Do you know what adjustments were made?
[/quote][/quote]

Its possible that the testing centre simply calibrated their machine to let a car pass. There are at least one or two in my area that do that. Otherwise the sales would simply find another testing centre that passed them and they would loose their business.

The mech didn’t have anything to do with the test centre. I always jsut go there myself. Its close to home and only takes a few minutes.

[quote=“sulavaca”][quote=“Ducked”][quote=“sandman”]
Do you know what adjustments were made?
[/quote][/quote]

Its possible that the testing centre simply calibrated their machine to let a car pass. There are at least one or two in my area that do that. Otherwise the sales would simply find another testing centre that passed them and they would loose their business.[/quote]

So is it fairly common practice to retrofit non-cat exhausts, and if so, are there leaning procedures that’ll actually meet the standards for real (at least for the duration of the test)?

I’d thought the control system was fairly black box/non tamper without special gear, but I don’t really know that much about it. Maybe it can be fooled.

Separate question: How about the 2-stroke UCL idea?

[quote=“Ducked”][quote=“sulavaca”][quote=“Ducked”][quote=“sandman”]
Do you know what adjustments were made?
[/quote][/quote]

Its possible that the testing centre simply calibrated their machine to let a car pass. There are at least one or two in my area that do that. Otherwise the sales would simply find another testing centre that passed them and they would loose their business.[/quote]

So is it fairly common practice to retrofit non-cat exhausts, and if so, are there leaning procedures that’ll actually meet the standards for real (at least for the duration of the test)?

I’d thought the control system was fairly black box/non tamper without special gear, but I don’t really know that much about it. Maybe it can be fooled.

Separate question: How about the 2-stroke UCL idea?[/quote]
Ed, Sandy is talking about the testing center adjusting their own equipment to read the car’s exhaust gas as cleaner than it really is. One simple way of doing this is just not putting the sensor all the way into the tailpipe, so that it’s sample is diluted with outside air. There are other ways.
A carburetted car can be tweaked via the idle mixture screw, though on an older vehicle it may be that the car would be basically undriveable set up this way. A lot of agents help in this way. Car fails the sniffer, they take it to their mechanic buddy who leans the mixture via the idle mixture setting. Car passes the sniffer, buddy puts it back the way it was. Everybody happy.

Note that vehicles fitted with cats are usually set up to run richer than those not catted. The extra hydrocarbons make the cats more efficient. If you then remove the cats you have to completely retune the car to run clean without them. A well-tuned car can pass most emissions standards without a cat, but ‘well-tuned’ means individual custom tune and no manufacturer can afford to do that for every car off the line. DIY requires specialist tuning software and a wideband O2 sensor at the least. Very few carburetted cars are catted, so you have to modify the maps in the fuel injection computer.
AFAIK there is no requirement in Taiwan to leave all the emissions equipment as-is. If you personally take a car in for test that has a loud exhaust and emits black smoke you will probably fail. If the right mechanic takes it in, it could easily pass. A lot of people drive catless 364 days a year and refit the stock exhaust for testing. The testing centers are under instructions these days to make life hard for modified cars, so going in with a huge aftermarket muffler is not going to go well.
BTW, although current standards require the cars to be certified with emissions tested under load, the inspections of used cars only check emissions at idle, since none of the testing stations, and probably not even the DMV stations, have a rolling road. With a healthy car, in theory you could remove the cats and still pass inspection as long as the muffler looks and sounds stock. The idle mixture on modern injected cars is set to stoichiometric and the cats usually don’t light at idle as there’s not enough energy in the exhaust gas, so the cats are doing nothing at idle.

redwagon does not approve of running catless on the street. :no-no: Turbo cars need to run quite rich to make their best power and that’s very dirty if catless. N/A cars need very accurate fueling to run clean without cats and that’s beyond the abilities of the average tuner.

Maybe there’s another thread for 2-stroke discussions? :wink:

I know that, except I think it was Sulavaca.

I said that above

[quote=“redwagon”]If you then remove the cats you have to completely retune the car to run clean without them. A well-tuned car can pass most emissions standards without a cat, but ‘well-tuned’ means individual custom tune and no manufacturer can afford to do that for every car off the line. DIY requires specialist tuning software and a wideband O2 sensor at the least. Very few carburetted cars are catted, so you have to modify the maps in the fuel injection computer.
AFAIK there is no requirement in Taiwan to leave all the emissions equipment as-is. If you personally take a car in for test that has a loud exhaust and emits black smoke you will probably fail. If the right mechanic takes it in, it could easily pass. A lot of people drive catless 364 days a year and refit the stock exhaust for testing. The testing centers are under instructions these days to make life hard for modified cars, so going in with a huge aftermarket muffler is not going to go well.
BTW, although current standards require the cars to be certified with emissions tested under load, the inspections of used cars only check emissions at idle, since none of the testing stations, and probably not even the DMV stations, have a rolling road. With a healthy car, in theory you could remove the cats and still pass inspection as long as the muffler looks and sounds stock. The idle mixture on modern injected cars is set to stoichiometric and the cats usually don’t light at idle as there’s not enough energy in the exhaust gas, so the cats are doing nothing at idle.[/quote]

The crux of the matter, and thanks for that full explanation. IOW you can probably get away with just about anything, but doing it properly isn’t likely to be a practical DIY job (and probably still less a practical “black hand” job).

Of course if you say that I believe you, but IF the system is using feedback from an O2 sensor I don’t quite see why it can’t keep itself clean whether there’s a cat there or not. I suppose you lose too much power running that close to stochiometric?

I think I’ve read something somewhere about making a DIY “EGA” using the oxygen sensor from a scrap car (doesn’t actually analyse the gases of course, just gives you a null indicator of stochiometry). Wouldn’t get you round the computer re-mapping requirement, but it might be a very useful aid to tuning a non-cat car or motorcycle. I’ll look it up. Might be a vacation project, though probably not.

If there is this wouldn’t have been an appropriate question for it, since I’m inviting comment on whether 2-stroke oil would be useful as an UCL for an uncatalysed 4-stroke engine, as apparently found in Sandman’s Sentra.

A priori it seems to me it would, but I’ve never heard of it being done and there might be a good reason that it isn’t.

Edit: I originally considered suggesting it on a forum for those Chinese copies of prewar sidevalve BMW’s, which apparently break a lot, but since I’m never likely to own one of them, I couldn’t be bothered to register. They’d probably just have ignored it anyway. Goddam pseudo-Nazi’s.

The sensor is to keep the mixture to its set opperating range, not to keep the exhaust free from CO.

The typical O2 sensor found in most production cars is a cheap, narrow-band unit. That means it can keep track of the oxygen present in the exhaust over a narrow range around stoich (roughly 15.0:1 to 14.2:1). This is fine at idle and under light load and most cars run in ‘closed loop’ mode in these conditions. IOW, the ECU checks the accuracy of the fuel mapping against the feedback from the primary sensor* (fitted before any cats) and adjusts the maps accordingly to remain at stoich. The problem with this is that under heavy load the engine will detonate if you try to run at stoich… more fuel is needed (up to 10.5:1 in turbo cars) to prevent that and the narrow-band O2 cannot read that far. So, the ECU stops reading the sensor output under heavy load and goes into ‘open loop’ mode, whereby it simply reads from airflow (or manifold absolute pressure) sensor, throttle position and rpm, cross-checks those inputs to a 3D map table, and the injectors are pulsed accordingly. No further checks or feedback is involved.

  • Many modern cars are fitted with a secondary O2 sensor, downstream of the cats, which checks the cats are working efficiently. Obviously, uninstalling them will trigger a CEL.

Yes you probably could kludge one up but as most cars have a narrow-band sensor, one from the scrapyard wouldn’t get you very far. A few cars do have wideband O2 sensors from the factory as they have a wider operating range in closed-loop mode, but even if you find one of those you still have to create a power supply, controller and some method of outputting the data is a useful form, unless you are good at converting current to AFR in your head when it’s a logarithmic scale between the two. A wide-band kit is available from many vendors at a price that’s reasonable if you need it, like this. As Sulavaca points out, oxygen content is only one indicator and a low amount of unburnt hydrocarbon doesn’t guarantee health or passing inspection as it doesn’t have any bearing on oxides of nitrogen in the gas.

Ah. I didn’t understand where that came from or what it was about. Obviously you don’t want to use an UCL in a car with cats as it can easily foul them. I’d be dubious about using one at all in a modern engine as it obviously reduces the effective octane of the fuel and causes detonation. Without some way of modifying the fuel/air ratio you are effectively making the engine run leaner, as some of the liquid being ingested is now oil instead of fuel, with a far lower calorific value. The engine will run hotter, make less power and be more prone to detonation. An alternative would be to rig up a separate pump and injection system like an autolube 2-stroke motorcycle. Just adding 2T to the fuel tank will not get you any good results I think.