I've been playing around with my Picoscope and wanted to share. I checked the latency on the Bosch 24lb injectors at normal idle. At 14.3v on my scope, the pintle opening event is clearly visible at 1.386ms on the injector amperage ramp.

Let me say that again..... the 14.3v latency of the Bosch III 24lb injector is 1386 microseconds.

Now - given that $6E, $8D, and other stock BIN profiles have the latency at this voltage level at about 500 microseconds..... it is pretty obvious exactly WHY they run lean and they are ABSOLUTELY NOT direct replacement injectors. Regardless of what any of the vendors of these injectors say. Since the ECU isn't waiting the additional nearly 900 microseconds required to open the injector, the BLM and INT MUST climb to raise the pulse width of the injector and compensate for the incorrect voltage offsets.

I'm going to test some other voltages by both disconnecting the alternator as well as adding some resistance to drop the voltage. With my LINK ecu I can actually put the whole bank of injectors into test mode and it will pulse them at any rate I specify. Really useful for on-car testing. I'll post my results at further voltage levels:



GD

 

And here's the latency at 12.09v (alternator unplugged). We're up to 1709 microseconds



Looking at these injectors compared to a stock injector - seems reasonable to conclude they are so slow because of how skinny they are. The stock injectors are much larger, and made of steel so they can accommodate a larger, more powerful coil, and can provide cooling for it. The Bosch III injectors are plastic and thin. Also an old design but not quite as old as the RP Multec's,etc - which suffer from heat induced coil insulation failure. Too strong of a coil - generates too much heat I'm guessing.

GD

 

Nice info and work, what does it mean to the average layman??

 

After further review and testing it looks like the data-point I was looking at is likely the pintle stopping at the top of it's travel, so that's not the dead time. On the second scope trace you can see there is a smaller dip in the amperage right before the one I marked which seems to correspond well with the offsets that work in practice. I find that the 14.3v dead time is more like 1050 microseconds, while the 12.1v dead time is about 1250 microseconds.

Lower values but still quite a bit longer than the original injectors and still explains the higher BLM's with the Bosch III's.

GD

 

Average layman? As in people that don't tune themselves? So not likely to be looking in this sub-forum? LOL.

Well. Here's what it means:

1. If you are going to buy injectors and you don't want to have to reprogram (tune) the computer, buy EXACT stock replacements.

2. If you are going to buy non-stock injectors (DIY or pro tuning), MAKE SURE you buy from a seller that has the correct published voltage latency information for use at your planned operating pressure. If they don't - find another seller/product/etc.

GD

 

Playing around with the numbers..... disconnecting and reconnecting the alternator, observing AFR's, trying different areas of the scope trace..... seems that about 1/3 of the way up the "pintle hump" yields a pretty close result. I landed on 1260 microseconds at 14v, and 1480 microseconds at 12v.

I don't quite understand the large discrepancy with "published' data on similar injectors, but the scope trace doesn't lie. You can clearly see the opening "pintle hump" in the amperage and the closing pintle hump in the voltage trace. I can't reconcile this discrepancy. I can only report what I see.

GD

 

According to what I've seen in modern ECU,
It's not a set number at some voltage,
as Fuel pressure and manifold pressure play a role in latency


If you aren't sure what latency to use a typical starting point is 1ms delay
If you over-do the latency this will reflect in the VE table as an unnatural decreasing slope near idle regions,
thus you don't NEED injector data to figure out the ballpark latency, you can determine by inspection after logging with a wideband

 

So far the best I've been able to do is estimate the latency with a scope to get into the ballpark, then do fine adjustments by varying the system voltage and watching the AFR while keeping other variables constant.

And yeah - I've seen plenty of cars tuned without the latency, but I would rather do some testing and see how the injectors are performing under real world conditions - especially if they are unknown but also even if they are known - as you correctly point out fuel pressure and manifold pressure have an effect on latency as well.

More information in this regard isn't a bad thing and can only lead to better/faster tuning. I understand what you're saying and yeah - if all I have to work with is stone knives and bear skins I'll make that garbage run and drive but that's not all I have. I have a lab scope and am a curious person by nature.

GD

 

Are the 22 lb injectors going to exhibit the same latency? I have seen different tables for voltage offsets and the one I currently use (610.36@14.4 and 762.95@12.8v) is quite a bit different. Another I found is much closer (946.06@14.4v and 1098.65@12.8v) to your numbers.

 

For injectors below 60lb/hr the delay being 100-400us off isn't going to hurt or change much

so little fuel comes out of a small injector that it barely matters to anything but a stock engine at 550rpm idle, and even then there is so much variance event to event, and due to temperature, that any 'exact number' is really just a range of possibilities, a 'ballpark'

I like what you are doing and I think its a good idea, but I don't think anyone with 20-30lb injectors is going to do much harm if the latency is off by 100 to 400us range

 

This is just a trial run to get a handle on the procedure necessary for quantifying the offsets on any given injector that comes in already installed on something we need to tune. The size of this particular injector just happens to be what I have on hand for testing that I didn't know the offsets for and they are in a good running daily driven car so I know that fuel pressure is correct, and the engine, electrical system, etc is healthy. I have an injector flow bench as well so I can test flow rate, clean, and quantify some of the data needed to be able to use large injectors. We routinely use 200 lb/hr injectors where this data is very important for getting any kind of decent idle quality.

Discussing this on the PicoScope forum I see that I really need to use a pressure transducer to see when the injectors begin to open and correlate that to the pintle hump in the amperage waveform while triggering the event with the voltage drop from the injector driver. So I'm going to do that next to get the exact time at which the injector starts to squirt.

GD

 

I was curious, so I pulled up a psd file I had saved for my Holley injectors. These are Bosch/ev1 style 36# injectors. Holley lists their off time at 14 volts at 450us. First capture is at the pintle hump, about 1300us, not unlike what your capture shows. I then moved the rulers to show 450us, and there is a defined change in the shape of the current ramp, which I would imagine is the pintle beginning to move. I can't see anything like that in your first capture, but I can see it in your second, at around the 1300us mark. Interesting differences in injectors for sure. I'm currently swapping to 66lb injectors, I'll try to remember to hook a pressure transducer up to the rail and see what it shows with the old and new injectors, if it can get detailed enough to see a drop when the injector first begins to fire.

 

There's a fellow nerd over on the Pico forums that claims to have a reliable way using voltage to the injector, control signal, amperage, and fuel pressure. So I'm going to give a try at adding in some additional data streams and see if I can't figure this out. It does seem clear that the pintle hump occurs at the end of the pintle travel, not the beginning. I'll report back my findings. I have a couple ideas how to get the most accurate information.

What I noticed about all my traces is there's a consistent undulating pattern across the entire waveform that sometimes shows up right where you think there's a change in the slope.... in many of my traces I saw that also. But in others there's really nothing visible.

GD

 

So talking with some people that really know and understand this stuff better than I do - the valley right after the pintle hump is indeed the opening event - as I originally suspected. There is more to these traces than just what I was looking at though and I will have some updated lap scope traces with fuel pressure, injector input voltage, injector signal, and current ramp to show soon. But suffice to say that *for my configuration* (that takes into account my injector supply voltage, injector temperature, idle manifold pressure, fuel pressure @ idle, and the electronic characteristics of my ECU driver), the 14 volt dead time of these 24lb injectors is in the range of 1.4 to 1.5 ms.

Another method of "seeing" the dead time (mostly to corroborate the lab scope data), is to do an XY plot of the injector actual PW (not effective PW), and 1/Lambda while the engine is at a steady idle and the voltage offset is slowly run up and down to vary the Lambda. Then doing a linear fit of the low slope area of this plot. The linear fit came up with 1.57 ms dead time which is quite close to the lab scope numbers at the valley (the "check mark") of the pintle hump:




GD

 

Some really cool info here, GD. Your test data seems to contradict what 84Elky has posted, a few months ago, with his theory on the B3s. You should remember that post. I would like to see what some stock Multec injectors test at. I would also like to hear 84Elky opinion on your test results. It would seem like the B3s should be more efficient than the original Multecs. If they are smaller, should be able to move the pintle faster than a larger injector. Thus dont need a big driver coil to move it. The test results at FIC certainly prove that they flow very well.

Also, cannot ignore what King says above. Off by a few hundred uSec, and it really wont be noticeable. My stock values for IPW correction are no where near 1.0ms, at 12.8v and 14.4vdc. I will be replacing my injectors in a few months, and will likely go with the B3s. And yes, it would be nice if there were published correction values for those injectors........i will be surely asking about them at time of purchase.

 

I replaced the stock injectors in my 85 Corvette with 24lb Bosch IIIs. The engine is now difficult to start, and the MPG display reads high. But, actual MPG seems to be slightly better. Would these issues be related?

 

Your vette should be running 22lb injectors if it's a stock ECU. Not 24lb. And yes the Bosch III's are not direct replacements. They require tuning.

GD

 

From what I have read, 85s came with 24 lb injectors and ran slightly lower fuel pressure.

 

That could be but the Bosch III "24" lb injectors usually test out more like a 22.5 - 23 lb/hr, and the voltage offsets are all wrong for the stock ECU programming so it needs to be tuned either way. The hard starting is probably an effect of improper offsets. The better fuel economy may be because the injectors you replaced were shot or because the fuel trims are routinely on the lean side (high BLM, etc) and that's manifesting as "fuel economy". It could, however, be bad for engine longevity if it's left that way.

GD

 

I understand. Maybe I'll put the old ones back in since they seemed to be ok anyway. I have no way of tuning this 1985 ECM. Thanks for the info!

 

Would you have the offsets for these? 24lb Bosch injectors? (blue)
that would be super to help me out with tuning and I can’t find them anywhere.
the battery correction
thx

 

The Bosch III "24lb" injectors I was running for this test are black. You should post the part number on yours and I can look at mine but they are likely different. I can probably look up the data in my old files. Since this post I have upgraded the engine to a 400 HP 350 and am running 42lb LS injectors (with published data) for the ability to support a supercharger in the near future.

Second rule of tuning is - buy injectors that come with data. Learned that the hard way......

GD

 

Go here; and scroll down a ways, you will find exactly what you need.
Mike

https://www.thirdgen.org/forums/tpi/...injectors.html