Ted75zcar's questionable logic - the '91 compound turbo-super
#21
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
here are some pics of the data acquisition boxes. I used to design sensor technologies and sensor systems, so I am an instrumentation nut.
Each of the output connectors shown in the pic above contain 2 ADC channels, a sensor ground, and an ADC referenced 5V source for ratiometric correction. The sense resistors for resistive sensor technologies are built into the sensor side pigtail of the sensors harness. This allows me to swap/move channels to any of the connectors at will. The larger connector on the box is some ebay military knock-off.
Each of the output connectors shown in the pic above contain 2 ADC channels, a sensor ground, and an ADC referenced 5V source for ratiometric correction. The sense resistors for resistive sensor technologies are built into the sensor side pigtail of the sensors harness. This allows me to swap/move channels to any of the connectors at will. The larger connector on the box is some ebay military knock-off.
#23
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
two things. The TinyIOx requires a regulated 5V source, and the box not only measures the sensor output, it also powers the sensor. There are actually 2 versions of the box. A 9 channel, and an 18 channel. Once we start powering 18 arbitrary sensors, the power requirements go up.
#24
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
Time for a weekend update, left off with DAQ stuff, so might as well start there ...
Cobbled a manifold together that would allow me to characterize cheap ebay pressure traducers and Chinese MAP sensors against a high quality calibrated reference sensor. The Arduino samples the voltage from any of the connected sensors along with the ref sensor and provides the data via serial stream. I import the data into a spreadsheet program that gives me the slope and offset. I had originally coded it to generate inc files, but the sensors are actually really linear, so a simple slope/offset conversion should be good enough. The overall objective is to be able to accurately characterize pressure drops across all things (heat exchanges, charge piping, air filter, throttle body(s), ...). As mentioned, these sensors are super linear, I suspect they incorporate some sort of 2 element Wheatstone bridge or something. The sensors do have significant offset error, so this was a good time investment IMO.
here are some pics ...
Cobbled a manifold together that would allow me to characterize cheap ebay pressure traducers and Chinese MAP sensors against a high quality calibrated reference sensor. The Arduino samples the voltage from any of the connected sensors along with the ref sensor and provides the data via serial stream. I import the data into a spreadsheet program that gives me the slope and offset. I had originally coded it to generate inc files, but the sensors are actually really linear, so a simple slope/offset conversion should be good enough. The overall objective is to be able to accurately characterize pressure drops across all things (heat exchanges, charge piping, air filter, throttle body(s), ...). As mentioned, these sensors are super linear, I suspect they incorporate some sort of 2 element Wheatstone bridge or something. The sensors do have significant offset error, so this was a good time investment IMO.
here are some pics ...
#25
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
Also made some progress on the valve seats. All of the exhaust valve seats are cut. I will say this ...
This is the last effin time I will be cutting valve seats by hand, it effin sux monkey n\/ts. It takes me something like 1 to 1.5 hours to cut each seat. My builder will charge $25 per seat on his Sunnen. Granted, it is a standard 3-angle, but the extra angles get me what, 0.5% or worse improvement? F that. He will do a whole head for $320, which includes strip, seat cut, valve replace/clean, head clean, buzz, and reassemble. I am considering having him do the intake side, but some recent rod-related complications may force me to be a little more budge aware. If you missed it ... F-this seat cutting BS.
Now for some pics
This is the last effin time I will be cutting valve seats by hand, it effin sux monkey n\/ts. It takes me something like 1 to 1.5 hours to cut each seat. My builder will charge $25 per seat on his Sunnen. Granted, it is a standard 3-angle, but the extra angles get me what, 0.5% or worse improvement? F that. He will do a whole head for $320, which includes strip, seat cut, valve replace/clean, head clean, buzz, and reassemble. I am considering having him do the intake side, but some recent rod-related complications may force me to be a little more budge aware. If you missed it ... F-this seat cutting BS.
Now for some pics
#26
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
Picked up an OEM NB2 alternator, came in with bunk bearings, so decided to rebuild. The Teflon coated bearing in the rear is a biznatch to source, so here:
Welcome to HodgePodgeia!
2 x B10-50D 2RS with teflon rings (B10-50D w/rings) = $9.52
2 x 6303 2RS (6303) = $6.76
These guys have Chinese versions, ship super fast, and at a killer price. They even refunded a couple bucks for the shipping.
Anyway, spent too much time tearing the thing apart, sourcing the bearings, masking and painting and reassembling. Then went to install the pulley, and my mind must have been in a bad spot because I lost patience with a cross thread and fubar-d the shaft threads. Love it when life provides me repeated opportunities to pause. Replacement scuzzy OEM from ebay ordered, yeah!
Couple notes on the selection of the NB2 alternator.
The NB2 is an externally regulated 80A, as we all know. I know the trend around here is to swap in the internally regulated versions, 'cause, easy button. On the 1.6 setup, I ran both a stock internally regulated, and an aftermarket high-amp internally regulated. I was always seeing a falling voltage when the car reaches temperature. I am speculating that what-ever reference these alternators is using to regulate against has a pretty crappy temperature coefficient. Since my alternator is located on the hot-side, I am much more prone to issues. If I go to the NB style external regulation, I will be able to regulate from in the cabin, where the temperature will be less of an issue. I will be using the Westfield board with my MS3PRO module, so the the HW based alternator control circuitry is already there. In an effort to reduce the temperature dependence, I heat shielded the alternator, ran the radiator fan constantly (flowing air past the alt), and installed big MFing wire from the alt back to the junction box. These all provided some improvement, but didn't achieve my goals.
until next time ...
Welcome to HodgePodgeia!
2 x B10-50D 2RS with teflon rings (B10-50D w/rings) = $9.52
2 x 6303 2RS (6303) = $6.76
These guys have Chinese versions, ship super fast, and at a killer price. They even refunded a couple bucks for the shipping.
Anyway, spent too much time tearing the thing apart, sourcing the bearings, masking and painting and reassembling. Then went to install the pulley, and my mind must have been in a bad spot because I lost patience with a cross thread and fubar-d the shaft threads. Love it when life provides me repeated opportunities to pause. Replacement scuzzy OEM from ebay ordered, yeah!
Couple notes on the selection of the NB2 alternator.
The NB2 is an externally regulated 80A, as we all know. I know the trend around here is to swap in the internally regulated versions, 'cause, easy button. On the 1.6 setup, I ran both a stock internally regulated, and an aftermarket high-amp internally regulated. I was always seeing a falling voltage when the car reaches temperature. I am speculating that what-ever reference these alternators is using to regulate against has a pretty crappy temperature coefficient. Since my alternator is located on the hot-side, I am much more prone to issues. If I go to the NB style external regulation, I will be able to regulate from in the cabin, where the temperature will be less of an issue. I will be using the Westfield board with my MS3PRO module, so the the HW based alternator control circuitry is already there. In an effort to reduce the temperature dependence, I heat shielded the alternator, ran the radiator fan constantly (flowing air past the alt), and installed big MFing wire from the alt back to the junction box. These all provided some improvement, but didn't achieve my goals.
until next time ...
#27
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
been a while, progress has been made. Might as well get some of it documented.
SC bracket and the beginnings of the outlet manifold are in progress. This bracket aligns the SC pulley face to be parallel with the crank pulley face, it is adjustable front to back. One of the issues with the last system that I wanted to improve was the install/removal process. I am trying to make this as easy to get in and out as possible. I am using a Honda knock-off with a BP4W flange. It may not be the best performance option, but I need the space. The turbo is back from getting balanced at G-pop. This is essentially a new GTX2867R (Gen 1) with a 0.86 A/R turbine housing. Got the 3.307 R&P back from MFactory, they performed a surface treatment on it. Hopefully they work! Picked up another diff to install them. This way if they bomb out I can throw the 4.10 back in there. Nabbed an NB1 6-speed.
Block is back from the engine builder:
Wiseco, Manley H, ARB mains King bearings, BE street strip with 2 shims
Had to go with the Manley instead of the Carrillo because everybody's favorite vendor around here did a FABulous job of not ordering the correct replacement rod that I specified. Of course I failed to check what was sent and didn't realize it until 2 years later when I went to get the engine put together, so probably my fault. As expected, email is launched into the order issue black hole of nothingness over there.
SC bracket and the beginnings of the outlet manifold are in progress. This bracket aligns the SC pulley face to be parallel with the crank pulley face, it is adjustable front to back. One of the issues with the last system that I wanted to improve was the install/removal process. I am trying to make this as easy to get in and out as possible. I am using a Honda knock-off with a BP4W flange. It may not be the best performance option, but I need the space. The turbo is back from getting balanced at G-pop. This is essentially a new GTX2867R (Gen 1) with a 0.86 A/R turbine housing. Got the 3.307 R&P back from MFactory, they performed a surface treatment on it. Hopefully they work! Picked up another diff to install them. This way if they bomb out I can throw the 4.10 back in there. Nabbed an NB1 6-speed.
Block is back from the engine builder:
Wiseco, Manley H, ARB mains King bearings, BE street strip with 2 shims
Had to go with the Manley instead of the Carrillo because everybody's favorite vendor around here did a FABulous job of not ordering the correct replacement rod that I specified. Of course I failed to check what was sent and didn't realize it until 2 years later when I went to get the engine put together, so probably my fault. As expected, email is launched into the order issue black hole of nothingness over there.
#30
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
I am actually moving to the MP62. It is more appropriately sized for the 1.8, where the OA2076 is a good size for the 1.6. I have an unused OA3150, but it is more difficult from a packaging perspective. Additionally, my configuration acts as both a compound and a twin, and I believe the roots is a better option while in twin mode. The twinscrew is probably better in compound. In twin, the twinscrew continues to add heat due to internal compression, where the roots should add less heat when in this mode. The pressure ratio across the SC when in twin-charge will be at or near 1. After I am on the road, I may fabricate a solution using the 3150, just to see how it compares. The internal compression on the 3150 is only 1.4, so the incremental temperature add may not be significant.
#31
Elite Member
iTrader: (8)
Join Date: Dec 2008
Location: Kingston, Ontario
Posts: 2,910
Total Cats: 51
Time for a weekend update, left off with DAQ stuff, so might as well start there ...
Cobbled a manifold together that would allow me to characterize cheap ebay pressure traducers and Chinese MAP sensors against a high quality calibrated reference sensor. The Arduino samples the voltage from any of the connected sensors along with the ref sensor and provides the data via serial stream. I import the data into a spreadsheet program that gives me the slope and offset. I had originally coded it to generate inc files, but the sensors are actually really linear, so a simple slope/offset conversion should be good enough. The overall objective is to be able to accurately characterize pressure drops across all things (heat exchanges, charge piping, air filter, throttle body(s), ...). As mentioned, these sensors are super linear, I suspect they incorporate some sort of 2 element Wheatstone bridge or something. The sensors do have significant offset error, so this was a good time investment IMO.
here are some pics ...
Cobbled a manifold together that would allow me to characterize cheap ebay pressure traducers and Chinese MAP sensors against a high quality calibrated reference sensor. The Arduino samples the voltage from any of the connected sensors along with the ref sensor and provides the data via serial stream. I import the data into a spreadsheet program that gives me the slope and offset. I had originally coded it to generate inc files, but the sensors are actually really linear, so a simple slope/offset conversion should be good enough. The overall objective is to be able to accurately characterize pressure drops across all things (heat exchanges, charge piping, air filter, throttle body(s), ...). As mentioned, these sensors are super linear, I suspect they incorporate some sort of 2 element Wheatstone bridge or something. The sensors do have significant offset error, so this was a good time investment IMO.
here are some pics ...
I was considering these sensors for things like fuel and coolant pressure (I think that they are a liquid sensor, yeah?)
#32
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
I will be using the sensors to measure fractions (hopefully) of a psi drop across various flow channels. The sensors are thankfully linear, which means I don't need LUTs, but the scaling between them are not exactly the same. I will be able to use the control system and peripheral interfaces to apply a linear correction based on slope and offset. The full scale response of these sensors is not controlled enough to simply leverage the published sensitivities and measure pressure differentials at less than say 10kpa with a 100kpa bias.
#33
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
yes, they are good for air/gas/oil/water. If you are interested in a resolution/accuracy of say 1psi, you should be fine. You can probably even get away with a hard-coded offset correction, as opposed to a re-cal for offset at turn-on if you want.
#34
Elite Member
iTrader: (8)
Join Date: Dec 2008
Location: Kingston, Ontario
Posts: 2,910
Total Cats: 51
you mind sharing your source? im sure there are some cheap sensors out there that aren't as good as what you have.
For my needs, a hard offset with linear scaling is fine. Fuel correction based on pressure (if your pump starts to fail or cant keep up for some reason) and general monitoring of water/airbox pressures.
For my needs, a hard offset with linear scaling is fine. Fuel correction based on pressure (if your pump starts to fail or cant keep up for some reason) and general monitoring of water/airbox pressures.
#35
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
I didn't qualify these at all, really just picked these guys because they have several different sensors.
http://stores.ebay.com/Industrial-st...sub=2911276017
http://stores.ebay.com/Industrial-st...sub=2911276017
#36
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
Mostly finished with the Inlet and outlet manifolds. This is a major milestone, so figured I would post some pics.
Little details left:
Need to plumb the "Twin Valve" (name I am calling it) into the outlet manifold.
Pressure sensor bung on the inlet manifold
Weld a couple of lock-down mounts to the steel brackets
Mount for the Twin Valve control solenoid
fabricate a belt drive solution for the MP62 internal bypass
Now some pics ...
Edit: Tiny-pick bomb
Little details left:
Need to plumb the "Twin Valve" (name I am calling it) into the outlet manifold.
Pressure sensor bung on the inlet manifold
Weld a couple of lock-down mounts to the steel brackets
Mount for the Twin Valve control solenoid
fabricate a belt drive solution for the MP62 internal bypass
Now some pics ...
Edit: Tiny-pick bomb
Last edited by Ted75zcar; 07-22-2020 at 10:33 PM.
#37
Tweaking Enginerd
Thread Starter
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,816
Total Cats: 371
So it seems Mazda might be interested in this as well. Maybe not such a terrible idea.
Mazda Patent Application Shows Wild Engine With Twin Turbos and Electric Supercharger
Mazda Patent Application Shows Wild Engine With Twin Turbos and Electric Supercharger