When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
In This Thread, We Discuss Gauges, Digital Dashes, HUDs, etc.
SEARCH NEWB!!!
Now that we have that our of the way...
I am looking into a typical gauge setup for one of my cars. I had no idea the subject of gauges was so deep and wide. I also had no idea decent gauges were so expensive. The overarching question is:
What is the best value in decent gauges?
I am thinking I need AFR, boost, water temp, and oil temp gauges. At $175 each for something like AEM, that adds up to ~$700, which is well into digital dashboard territory.
What is out there that looks decent (preferably somewhat understated, but easy to read), is accurate, is durable, and is cost effective?
If you had it to do over again, what would you buy?
Sure, I'm researching this on my own, but I value the opinions of the community greatly, so let the ideas fly!
Personally I'm making a project out of this, and not buying any real gauges at all.
For oil and fuel pressure I am using a couple of Honeywell PX3 transducers. (PX3AN2BS150PAAAX) They run about $25 each.
For oil temperature I picked up a Boost Monkey "Replacement temp sender" for $7 on amazon
For boost pressure I grabbed an MPX4250 for about $10 (I'm a sucker for NXP)
I considered buying a couple of k-type thermocouples ($10-20 a piece) and some MAX31855 breakouts (themocouple amplifiers) to use as EGT sensors. Though I don't think i'll end up getting much value out of this.
Finally the entire thing is going to be wired into the CAN bus using a Nucleo F446RE ($15) and two $7 can bus transceiver breakout boards. Gauge data will be published over the CAN bus and read by a completely different system to display the values. This greatly simplifies the microcontroller code as I can simply read analog values and push them over the CAN when they're requested. The entire program should be less than 100 lines of code.
Minus the cost of the wiring, and my time to program the system. It cost me less than $100 to order those parts.
This does not provide you digital gauges in itself though.
To accomplish this, I purchased a fully working nav screen for the dash ($150), a liliput to throw in there ($150) and a minnowboard MAX ($150 3 years ago) to use as both a carPC and a gauge display. You'll also need a USB to CAN adapter, hard drive, and the ability to hack everything into your nav hood. You could accomplish close to the same thing using a Nexus7 tablet for half the price though.
In the end, not that far cheaper than a set of gauges. I'd also much rather have everything on a single display integrated into the vehicle rather than gauges all over the cabin though. That's the real value to me.
None of this is actually installed though. I ordered the parts early this week and hope to have it running by the end of the month.
I personally got myself DEFI BF gauges a year ago and I love them. The gauges are very reliable and the looks is very similar to the RX8, but they do require a control unit, which adds 150$ to the total cost. Because of that I hesitated with cheaper Prosport units, but after reading about the accuracy being hit and miss, I preferred paying more but be 100% confident with the readings.
One thing they do, which you might want to consider adding to your requirements, is they have an LED and audible warning when they gauge reaches a determined limit (temperature, pressure, etc...) This can be useful at the track when you're not constantly looking at your gauges.
The only gauge DEFI don't do however is the AFR, which you would have to source from AEM or others.
I personally got myself DEFI BF gauges a year ago and I love them. The gauges are very reliable and the looks is very similar to the RX8, but they do require a control unit, which adds 150$ to the total cost. Because of that I hesitated with cheaper Prosport units, but after reading about the accuracy being hit and miss, I preferred paying more but be 100% confident with the readings.
One thing they do, which you might want to consider adding to your requirements, is they have an LED and audible warning when they gauge reaches a determined limit (temperature, pressure, etc...) This can be useful at the track when you're not constantly looking at your gauges.
The only gauge DEFI don't do however is the AFR, which you would have to source from AEM or others.
Here's what they look like:
Man. Those DO look nice with the 8's stock cluster.
The gauges I need are going in my 2003 Miata. Its dash has gauges with white backgrounds and orange (now pink) analog needles. I think I want to match that. Programmable warnings would be a definite plus, since it is my primary track car.
I think I am going to replace the stock idiot water temp gauge in the dash with a real one. Beyond that, I may just take precautions to cool the oil and forget the gauge. I don't really NEED a boost gauge, so I may leave that out. That leaves the AFR gauge for the wideband as the only other gauge I need, which is good, considering my very limited mounting options.
Personally I'm making a project out of this, and not buying any real gauges at all.
For oil and fuel pressure I am using a couple of Honeywell PX3 transducers. (PX3AN2BS150PAAAX) They run about $25 each.
For oil temperature I picked up a Boost Monkey "Replacement temp sender" for $7 on amazon
For boost pressure I grabbed an MPX4250 for about $10 (I'm a sucker for NXP)
I considered buying a couple of k-type thermocouples ($10-20 a piece) and some MAX31855 breakouts (themocouple amplifiers) to use as EGT sensors. Though I don't think i'll end up getting much value out of this.
Finally the entire thing is going to be wired into the CAN bus using a Nucleo F446RE ($15) and two $7 can bus transceiver breakout boards. Gauge data will be published over the CAN bus and read by a completely different system to display the values. This greatly simplifies the microcontroller code as I can simply read analog values and push them over the CAN when they're requested. The entire program should be less than 100 lines of code.
Minus the cost of the wiring, and my time to program the system. It cost me less than $100 to order those parts.
This does not provide you digital gauges in itself though.
To accomplish this, I purchased a fully working nav screen for the dash ($150), a liliput to throw in there ($150) and a minnowboard MAX ($150 3 years ago) to use as both a carPC and a gauge display. You'll also need a USB to CAN adapter, hard drive, and the ability to hack everything into your nav hood. You could accomplish close to the same thing using a Nexus7 tablet for half the price though.
In the end, not that far cheaper than a set of gauges. I'd also much rather have everything on a single display integrated into the vehicle rather than gauges all over the cabin though. That's the real value to me.
None of this is actually installed though. I ordered the parts early this week and hope to have it running by the end of the month.
I thought about going that route, too. I am a software engineer by trade, so I have the skills to do it. Sometimes, I just don't want another project, though. Sometimes I want to push the Easy Button, plug some stuff in, and drive the car.
Some of the less expensive temperature senders look interesting to me, but not all gauges can be calibrated with other sensors. It seems they like to hold you hostage to their offerings. I guess there is really no way to guarantee accuracy otherwise, however.
.
Last edited by Steve Dallas; 04-17-2017 at 02:01 PM.
I thought about going that route, too. I am a software engineer by trade, so I have the skills to do it. Sometimes, I just don't want another project, though. Sometimes I want to push the Easy Button, plug some stuff in, and drive the car.
Some of the less expensive temperature senders look interesting to me, but not all gauges can be calibrated with other sensors. It seems they like to hold you hostage to their offerings. I guess there is really no way to guarantee accuracy otherwise, however.
.
Same here, and I know that story all too well. If it's done right though, the code should be pretty minimal.
I thought about running everything off of one board, not interfacing with the CAN, and draw the gauges using the microcontroller itself. But that would've easily turned into a 2-6 month project. With all kinds of EMF concerns, timing issues, the whole lot.
As long as I don't run into any horrendous issues with the CAN interface, the entire thing should pretty much be a small switch statement that reads an analog value and a few lines to convert each voltage to a digital value to send over the CAN. Everything else I'm just going to use off the shelf software/hardware for.
I considered dropping the minnowboard for a nexus tablet but having a linux box with a sizeable SSD in the car leaves a lot of room open for future additions without being constrained to Android.
Regarding the senders, that's another bonus. If a sensor breaks I can buy quite literally whatever I can find that will fit, as long as it has a datasheet.
I have, but they don't make them for old Miatas. What I may do for the AFR gauge, is mount it in a single A pillar pod. I generally don't like those pods, but I guess they don't look too stupid with a single gauge installed.
Originally Posted by Reoze
Same here, and I know that story all too well. If it's done right though, the code should be pretty minimal.
I thought about running everything off of one board, not interfacing with the CAN, and draw the gauges using the microcontroller itself. But that would've easily turned into a 2-6 month project. With all kinds of EMF concerns, timing issues, the whole lot.
As long as I don't run into any horrendous issues with the CAN interface, the entire thing should pretty much be a small switch statement that reads an analog value and a few lines to convert each voltage to a digital value to send over the CAN. Everything else I'm just going to use off the shelf software/hardware for.
I considered dropping the minnowboard for a nexus tablet but having a linux box with a sizeable SSD in the car leaves a lot of room open for future additions without being constrained to Android.
Regarding the senders, that's another bonus. If a sensor breaks I can buy quite literally whatever I can find that will fit, as long as it has a datasheet.
Yeah, there are a lot of upsides to be sure. I still just want to hit the easy button on this one, and it being a track car, I don't want to add any unnecessary weight.
What companies like AEM want for a sender is crazy at 5x the lower cost solutions.
I like the idea of a single simple gauge, pulling data from multiple sensors; programmed to flash/buzz a big DANGER DANGER WILL ROBINSON if something goes out of a programmed range.
Eliminate the need to continuously watch a fistful of gauges. Simplify the installation. Let the technology do it. Ignore the inside of the car; focus on the track. If the warning sounds/lights, back off immediately, then take your time to figure out what is going on.
I've seen a few custom solutions, similar to what Reoze is building. There are some off-the-shelf solutions such as the PLX multigauge https://www.plxdevices.com/ProductDe...e=897346002436. Although it looks like this device can only monitor 2 channels to trigger warnings.
I personally went a bit gauge-crazy in my old '90 RX7 - and it was simple n/a! Coolant, oil, AFR, etc etc. And yet I managed to miss the coolant temp skyrocketing, and melted an engine. A piece of the cooling system failed (plastic t-stat neck cracked, slow leak) so the gauge read good - good - good - VERY BAD all in 5-10sec, too fast for me to catch while manually scanning.