Page 35 of 37

Re: Widebands That You Would or Wouldn't Buy

Posted: Sat Apr 01, 2017 10:34 am
by Fred
https://www.bmotorsports.com/shop/produ ... gr9hd34806

I was going to ask you to link it and how much you charged, but ^ & : 89 bucks! :-)

For that price, have you addressed everyone's number one complaint of not being able to correctly index it without purchasing extra parts?

Complaints: https://www.amazon.com/Innovate-Motorsp ... 1+innovate

Re: Widebands That You Would or Wouldn't Buy

Posted: Mon Apr 03, 2017 4:08 pm
by bmotorsports
Fred wrote:https://www.bmotorsports.com/shop/produ ... gr9hd34806

I was going to ask you to link it and how much you charged, but ^ & : 89 bucks! :-)

For that price, have you addressed everyone's number one complaint of not being able to correctly index it without purchasing extra parts?

Complaints: https://www.amazon.com/Innovate-Motorsp ... 1+innovate
Presently the HBX-1 Style bung we offer is made from high quality materials, made in the USA, and the washer is captive which are the only distinctions. I ordered some copper washers that should work for spacing / orientation. I'll either add that as an option or make it a kit that comes with some. $89 is the retail pricing which is the same as Innovate.

Re: Widebands That You Would or Wouldn't Buy

Posted: Wed Apr 19, 2017 2:50 am
by HelmutVonAutobahn
Finally got hold of the new Innovate MTX-L PLUS. A bit strange.

The hardware is ALMOST identical to the regular MTX-L The only visible difference is that they added a proper DAC, to replace the old PWM DAC in the previous design. It's a bit weird, as they used a 2-channel DAC ( LTC1661C ). But, the new gauge has only a single DAC output. Externally, the second analog output wire has been re-purposed as a headlight/dimmer sense line. They have an OPAMP buffering the output. But, for whatever reason, they did not make the output differential. In spite of ground/voltage offset issues.

As for response speed, it's a bit of an oddball. If you catch it at just the right time, it can have a quick response, in the 25ms range. But, more often, it has a strange pattern, where, it make an initial move towards the right value. Then, it kinda tapers in the right direction for a few ms. Then, it "snaps" to the final value. Usually over a period of about 40ms or so. I'll try to get some traces. It does seem to update the DAC at about a 3ms interval. But, there appears to be some kind of buffering of the data. So, you are not getting "new data" every 3 ms.

It is definitely faster than the regular MTX-L. But, it doesn't, consistently, give the faster responses. Maybe one out of 4 is fast. But, even the "strange" responses are 2X faster than the older units. So, it's a real improvement.

Re: Widebands That You Would or Wouldn't Buy

Posted: Wed Apr 19, 2017 3:29 am
by HelmutVonAutobahn
Here's what I mean about the output. 1st, it's pretty glitchy/noisy. So, its hard to get a clean trigger.

These pics are several traces from the same repeated lean-to-stoich events.

The best trace that I got looks pretty good. It was at about 25ms, overall. i.e. the big change happens about 25ms after the actual solenoid firing. I got traces like this about 1/4 or 1/5 of the time.

The other traces show combinations of tapering/ snapping to a fixed value. Usually, over about 40-60ms There seems to be some kind of filtering algorithm that looks at the last few internal samples, and, decides what to output on the DAC. It seems to miss the kind of detail that the LC-1 "instant mode" had. But, it doesn't have the "jitters" like the LC-1.

I re-verified the setup using the AEM 30-0300 and got consistent results in he 20ms range, as before. So, all was good, there.

Conclusions:

Strange. But, in the end, about twice as fast as the regular MTX-L ( 82ms + ) . Maybe even 3X as fast, depending on how you count. So, definitely, a step in the right direction. Would recommend, over the regular MTX-L, unless you need both analog outputs. The higher pump frequency SHOULD give some advantage in heater control. So, MAYBE the sensors will last longer.

Re: Widebands That You Would or Wouldn't Buy

Posted: Wed Apr 19, 2017 12:29 pm
by Fred
How weird! Gotta be dodgy software behind that behaviour... as usual, thanks for posting up! :-)

Re: Widebands That You Would or Wouldn't Buy

Posted: Wed Apr 19, 2017 2:50 pm
by bmotorsports
Interesting. What are you sweeping to/from on your gasses / AFR values? Sorry if this was covered before, will be nice to list test constraints & conditions near the post though on such a long thread.

Re: Widebands That You Would or Wouldn't Buy

Posted: Thu Apr 20, 2017 6:13 am
by HelmutVonAutobahn
It probably is a firmware thing. It's hard to imagine what physical thing would have those attributes. It LOOKS MAYBE like they are trying to filter out noise, creatively. When they get a clean transition, it looks good, and they just output it to the DAC. But, if they get noisy samples, they, maybe wait until they get enough samples to know where they are, then, snap to the average value. Or, interpolate to it, depending on something?

I figure, that's why it's inconsistent. Or, not.

This way, they can post a new value to the DAC, before waiting for X samples to be averaged together. So, in a test environment, they may show better than they actually are. Which would be sneaky, but, smart. Imagine a test protocol, where, you take 100 traces and average out the best 10. They would look pretty good, then :)

Like I said; even with the goofy looking traces, this is a large improvement over their standard MTX-L / LC-2. Because of the different responses to the same stimulus, it's hard to put a number on it. But, eyeballing it, I'd call it about 35ms, if you throw out the worst traces. This, compared to their old value of 82ms+ for the MTX-L and LC-2.

35ms would put them up near the top of the AEM speed chart; along with MOTEC, PLX, and the old LC-1.
http://aemelectronics.com/files/info/AE ... _Chart.jpg

Still, about, 15ms behind the X-series. BUT, THIS IS NOT THE EXACT SAME TEST. it's close, though :)

The gas sweep, here, is from free-air ( 20.8% O2 ) to lambda 1.0 ( CO2 ). It simulates coming out of fuel cutoff. Which is, usually, the nastiest real-world event.

Re: Widebands That You Would or Wouldn't Buy

Posted: Thu Apr 20, 2017 9:48 am
by Fred
Interesting table, but Alan's products don't make an appearance, and nor do Hentai's favourites the ECM1000 and AFR500(v1 or v2). Given his presence in this thread, one has to wonder if the Alan omission is intentional?

Re: Widebands That You Would or Wouldn't Buy

Posted: Thu Apr 20, 2017 11:30 am
by HelmutVonAutobahn
I think I remember, during the alan-vs-delsolid wars, that they chose the most popular / most units in the field, when selecting the models for the test. So, it may have been intentional. But, for a different reason.

I think the AFR500 V1 is the same as the NGK Powerdex ? Generally, any wideband that uses an NTK sensor is going to be limited to about 80ms, at best, in response speed. So, the AFR500 V2 may be significantly faster, with the Bosch sensor. The NTK sensors are durable but, sluggish.

Maybe I'll buy one of Alan's units to test here. Which of his models is the fastest ? Then, I'll have, at least, a ballpark number to compare :)

Re: Widebands That You Would or Wouldn't Buy

Posted: Thu Apr 20, 2017 11:39 am
by Fred
I imagine they're all comparable, and newer is probably better. Let me have a look at the current range... https://www.14point7.com/products/spart ... ntroller-2 125 with a 4.9 sensor (if you don't have a fresh one). I'd appreciate that :-)

I confess to being tempted by the AEM stuff, but I'd still rather support the little guy over the big corporate, especially if the product is as good or better. Or even if it's not much worse. I've put more miles on one of Alan's units than my old LC1 and although neither has let me down, and both are fast, I know what I'll be buying every time unless you can show me not to! :-)