I already use a bath filter, but am using just plain water. It is a down draft marble filled bed with a 5 gpm pump blasting the wash over the bed a flooding it. Its works quite well and is self cleaning for the pump re circulation.
I do plan to experiment with re-purposing the used motor oil here Ive looked the static filter but can not justify the cost and added labor time. Need to go the other way its a hard sale already cant afford to add anymore cost
Not that I know of. However I think what we need to do is use A/C and chill the gas down, then pressurize the gas above atmosphere to jar that stuff loose. It drops out very readily at the engine, I think the gas being in a vacuum is working against us.
But If I can get this sensor to work, then none of those extra steps will be needed. The machine should not make tar when its working and tuned right. But if something breaks, specifically my hopper agitator it will be prone to making tar as the fuel flows will osculate. going cold to hot then back to cold and repeat. I had an agitator motor go bad and it caused all kinds of problems for me. lol So had to redesign that and now have some fixing to do 
This sensor one way or another I will get it to work. It will be very easy to make and implement at hardly any cost. Once the code is wrote, its wrote and the probe is just a 8/32 threaded rod mounted in a pvc pipe cap. that will get threaded into the sheave that will be welded into the engine adapter. So fairly simple to make.
Electrical resistance sensing could work but perhaps not directly into the Arduino. You may need a simple circuit to give the Arduino pins something it can work with.
Get two probes, one tarred up in a way youâd like to avoid and one thatâs only been in clean gas. Separate from the Arduino, get the resistance measurements you want with a multi-meter. Then figure out how to replicate the multi-meter reading with a simple circuit that can feed an Arduino pin as a second step.
Different idea⌠what about spectroscopy? I see research papers characterizing coal and petroleum tars by their color spectrum. You wonât need fancy scientific instrument. LEDs and/or simple optics can read a color âfingerprintâ of tars. It would be non-contact which is a huge plus.
Sparkfun even has a spectroscopy demo board that connects to Arduino with a software library
It does a wide range of light frequencies - wider than you probably need. The scientific literature I found on tar spectrums focused on only on infrared. Actually you might get away with a few carefully chosen IR LEDs and some phototransistor sensors for ~$5 tops. Message me if this is a direction youâd like to go. I have some background here that could be useful.
I think you guys maybe not understanding? The tar has high resistance and the condensate has very low resistance. In a working flow (both mixed together) there will be a varying resistance. The built in pull up resistor can already detect the full range with room to spare on both sides of the spectrum. Or am I missing something?
Ive so far tested.:
Dry tar (no moisture) just gooey tar. = full insulation
Tar with moisture = moderate resistance
Pure condensate = very high resistance.
I can detect all at varying voltage drops as analog signal.
Hi Matt, not sure if this will be a factor on your unit, but also temperature changes will change the resistance. So that might have to be factored in to your calculations.
Bob
Yes indeed, I didnt even think of that.
Uploading a video now so all can smell what Im steppin in. 
Very cool ideas here, I will begin research on this and will need to pursue this with more time. I need to get this working before the end of the week as this machine needs to ship asap and it must have something in place before I can let it go. But that stuff is very interesting
Well Alfred sent me an email asking for more info. I sent back what we would want to achieve and have not heard back yet.
However, just for kicks and grins, I looked them up on Amazon. Ten years ago, bought these and I think at the time I paid at 50 bucks or more for one of them. Now you can ten of them for around 23 bucks.
Supposedly they produce 60 watts at 12volts? I think this is the input for the cooling effect not sure if they would produce the 60 watts going the other way. But for 23 bucks plus cost of some heat sinks this is definitely something worth looking back into. With the engine gasifiers we maybe able to direct focus heat to these. So like a small torch directed at each one. Im going to see if I can justify my very limited budget to purchase a set with heat sinks for an experiment. Alfreds I know will be much more expensive, so for proof of concept these may do the trick. I already have plans to build the micro gasifier,
So all ten would give us a potential of 600 watts for 23 bucks. Not bad four sets of these would then have a potential of 2.4 kW. 
Ok video has uploaded. I have it unlisted so let me know if you are able to view it or not.
The video is working good.
Bob
You may be able to use these, but there is a difference in the construction of the Peltier TEC (thermoelectric cooler) and the Seebec TEG (thermoelectric generator).
Yeah thats what Alfred has said in the past. The actual TEGâs are more expensive. A Peltier is a cooler a TEG is a generator.
Matt, you seem to be biting off a big chunk on this one, but it would be of immense value if you could perfect this ability to detect tar. As always, I applaud your inventive pioneering spirit.
What youâre describing is very interesting. Purely playing devilâs advocate rather than shooting down the approach, my first thought is that other side products carried by the moisture (e.g. acids) will also have a substantial impact upon conductivity due to the ions they place into solution. Iâd expect those compounds to have a greater impact upon resistance than tars etc. BUT if production of those compounds follows a similar trend to the production of tar, then you may still have a valid proxy for inferring the relative amount of tar.
Exactly, It needs to be built and see how it acts in practice. Cheers!!
One point is its the sensor itself that is going to change as tar is produced. Once its on that sensor it can not self clean. So it will be locked in the range of detection of what ever is stuck onto it. By this time the controls will have faulted out and the user will need to clean the sensor and figure out whats wrong. Like in the video once I fouled the probe tips with the gooey tar it could no longer read the condensate.
My client in Scandinavia has his own pellet fuel plant. Indeed you can make your fuel and this unit is far cheaper than a commercial chipper with out need for sorting and drying process. This thing is the whole kit and caboodle. That has a hammer mill integrated so raw materials in, fuel grade pellets out ready to run. A small wood chipper with mulcher could be employed to prep larger fuels. But you could not use a cheap chipper like that for wood fuel gasifier at least it would not be very practical as 75% would go to waste.
I will update as I learn more. He just received it, he is going to let us know its continuous energy consumption rate is and possibly do some video.
Spec wise its 15kW start up and can produce 60 kgs pr hour. The continuous power consumption should be much lower.
