No announcement yet.

HOCOC Open (Gravity) Class Control and Driveability

This is a sticky topic.
  • Filter
  • Time
  • Show
Clear All
new posts

  • HOCOC Open (Gravity) Class Control and Driveability

    This is going to be a long read, so grab your favorite beverage and a snack, sit back, and enjoy!

    Gravity Class cars, similar to the HOCOC Open Competition Class, are an interesting drive. While there has been considerable experimentation in the design and construction of these cars, I've yet to see anything regarding the controlling of these cars.

    This is an interesting class in that we are taking small permanent magnet DC motors originally rated for 3 to 6 VDC and running them at 12 VDC. Some groups are going as high as 18 VDC.

    Al Thurman is one of the originators and early experimenters in this class. His series of LandShark cars in both inline and anglewinder configuration are widely distributed, well respected, and provide a great baseline for comparative performance data. An interesting feature of his ready-to-run builds is the inclusion of a 10 ohm resistor. At 12.0 VDC input , this results in a measured 2 volt drop across the resistor. This means that the car is actually running on 10 VDC. The car is fast, but even more important, quite driveable. This resistor is an important feature that will become clearer as this discussion progresses.

    I started building brass chassis back in the early 1990s. Ed Bianchi and I were experimenting with direct drive cars using 18 VDC motors but running them at 12 VDC. Ed ended up developing and selling the "Rattler". This was a great handling chassis that used the concept of a floating weight so as to absorb vibrations and help keep the chassis "nailed" to the track. I developed and sold the Micro-Cuc II. This was an iso-fulcrum chassis based on technology that had worked so well for me in 1/24th and 1/32nd scale racing back in the late 1960s/early 1970s. All good fun, but the concept of both cars never caught on amongst the various racing groups.

    Jump forward in time to "now" and we have this current crop of interesting miniature permanent magnet DC motors available. This allows the building of a wide range of fabricated chassis in inline, anglewinder, and true sidewinder configurations utilizing flex, iso-fulcrum, and center hinged configurations so as to maximize handling.

    I started building my current crop of Open Class Gravity Cars sort of in a vacuum, having not really seen what others might have been doing. Straight line speed was insane, but the cars were so fast that corners were an issue. Turns out the issue was voltage control. Ed and I did some experimenting during a joint practice/test session in which we ran the track voltage as low as 9 VDC. Definitely helped with driveability, but this was not a solution as HOCOC rules specify 12 VDC for this class.

    So let's go back to the 2 volt drop as exhibited by the Thurman cars. The voltage drop doesn't really slow the car down all that much, but the softening of the power delivery makes the car extremely driveable. Short of installing 10 ohm resistor, how could I get that voltage drop? It turns out that there are a couple of ways, some better in regards to driveability than others.

    I'm going to base the remainder of this discussion around "electronic" controllers. And by this I mean controllers that either use diodes, transistors, FETs, or MOSFETs for voltage control. Some electronic controllers include a feature called "Choke". This feature typically reduces voltage by approximately 2 VDC. Hey, here's that 2 volt drop again! How convenient. But there is a downside to this feature in that with most controllers, pulling the controller trigger all the way to the high throttle stop will bypass the choke circuit. This can result in a burst of full track power going to the car at a most inopportune moment. This characteristic is exhibited by both my Difalco and OS3 controllers. Another facet of this control scheme is that the sensitivity adjustment of the controller can change considerably when choke is used.

    So is there a way to get the benefit of Choke without changing or messing up your sensitivity adjustment? There sure is! And diodes are the answer.

    Diodes are an interesting electronic component when used in a DC circuit. In one direction of voltage (reverse) they have an almost infinite resistance. This effectively prevents any voltage/current flow. In the other direction (forward) , they have an extremely low resistance. An interesting and consistent characteristic of diodes is that they result in a forward voltage drop of 0.7 VDC. Yes, I know some might be a little more or less, but on average 0.7 VDC us a good number. So going back to our desired 2 VDC voltage drop, three diodes in series results in a 2.1 VDC drop. Certainly close enough for government work.

    This leads us into the controller discussion using the following controllers.

    Professor Motor diode based controller. This is a no-frills controller void of any adjustments.

    IMG_8324 by gcullan, on Flickr

    JetStream controller as made by Jim Segreto. Uses MOSFETS. There is no Choke. There is a changeable bias resistor that changes the "feel" of the sensitivity control.

    IMG_8325 by gcullan, on Flickr
    IMG_8326 by gcullan, on Flickr
    IMG_8327 by gcullan, on Flickr
    IMG_8328 by gcullan, on Flickr

    Difalco, an earlier model, transistor based, and features Choke

    IMG_8329 by gcullan, on Flickr
    IMG_8330 by gcullan, on Flickr

    OS3 3T Pro, not sure, but guessing it is FET based, possibly MOSFET, features Choke

    IMG_8331 by gcullan, on Flickr

    OS3 HO All Pro, again possibly FET or MOSFET based, features Choke

    IMG_8332 by gcullan, on Flickr

    Diode tree using 3 amp diodes. Note the code bar on the diodes. This configuration is correct for a track that is wired "Positive Gate". Reverse the position for tracks wired "Negative Gate". The connections between the diodes were deliberately twisted such as to provide a robust connection point for the controllers "White" coded alligator clip. The diode tree then clips to the "White" post of the track's controller station.

    IMG_8333 by gcullan, on Flickr

    Track testing was conducted on the White lane (2nd lane in from the outside) on my Slider (tm) oval with the track power set 12.0 VDC. The most recent version of TrakMate was used for timing. Testing was conducted in 25 lap blocks. Two cars, both of my own construction, were used for this test sequence. The Micro-Cuc IV is an inline configuration iso-fulcrum chassis that features a conventional guide pine and wiper pickups. The GCLS is my take on an inline configuration chassis but features a center hinged construction that allow flex and motion in both the pitch and roll axis. Data is presented in the following format: Controller, Micro-Cuc IV Median/Best Lap pair and GCLS Median/Best Lap pair. When testing with the diode tree, the third connection was consistently used resulting in a 2.1 volt drop. When testing Choke, full Choke was always used resulting in an approximate 2 volt drop.

    Professor Motor: 2.08/1.848; 2.16/1.831
    Professor Motor w/diode tree: 1.83/1.728; 1.87/1.769

    JetStream: 1.92/1.797; 1.90/1.772
    JetStream w/diode tree: 1.86/1.750; 1.81/1.732

    Difalco, no Choke: 1.92/1.770; 1.86/1.720
    Difalco, full Choke: 1.93/1.789; 1.90/1.744
    Difalco, no Choke w/diode tree: 1.87/1.737; 1.91/1.762
    Difalco, full Choke w/diode tree: 1.95/1.834; 1.94/1.789

    OS3 3T Pro, no Choke: 1.83/1.722; 1.98/1.739
    OS3 3T Pro, full Choke: 1.92/1.773; 1.94/1.801
    OS3 3T Pro, no Choke w/diode tree: 1.86/1.783; 1.88/1.782
    OS3 3T Pro, full Choke w/diode tree: 2.13/1.927; 2.10/1.859

    OS3 HO All Pro, no Choke: 1.87/1.782; 1.87/1.742
    OS3 HO All Pro, full Choke: 1.90/1.805; 1.93/1.781
    OS3 HO All Pro, no Choke w/ diode tree: 1.75/1.622; 1.72/1.602
    OS3 HO All Pro, full Choke w/diode tree: 1.84/1.792; 1.87/1.803

    1. All controllers performed better with the diode tree
    2. Stand alone Choke was not as good as stand alone diode tree
    3. Choke in conjunction with the diode tree resulted in consistently poor performance and driveability
    4. The best overall combination was the OS3 HO All Pro, no Choke w/diode tree in regards to both lap times as well as overall feel and driveability.

    This complete test sequence will be duplicated on my Banzai BuckTrax. Data will be posted as soon as it becomes available.

  • #2
    Thank you for posting this test. I recently purchased the OSE HO All Pro from Harden Creek, for use when I run my "hobby/competition" cars (BSRT, Viper, Wizzard & M-Tech) away from home (Gary Fast's custom Bowman track, for instance). I will be trying it for the first time in February, and I am really looking forward to using this controller's adjustability to determine each car's "sweet spot" on this challenging track.........


    • #3
      I don't think that you will be displeased with the OS3 HO All Pro controller, but in all fairness there are several really great controllers currently available.

      Two that I would have liked to test but don't own are the newer Difalco multi-band units as well as the Lucky Bob's wiperless controllers.Hmmm, I wonder if they would like to submit them for product testing. I do, however, report the good/bad/ugly aspects of all products that pass through my hands.


      • #4
        For Open class cars I use a Difalco Genesis controller, that has 30 bands and a replaceable resistor board in the control circuit. For 12 volt racing I use a 377 ohm board. The controller has the usual brake and sensitivity adjustments. There is a blast relay in the full power circuit and a defeat able full brake relay as well. In addition there is a choke control that is helpful if it can be used in moderation. With my controller the choke is bypassed when the trigger is pulled all the way. Difalco does offer a switch to defeat that feature, but my controller does not have it.
        I also use diodes to reduce the voltage when I race open cars. Rather than use a diode chain I built a box that has diodes connected to a rotary switch. The box goes between the controller white wire and the white track connection.


        • #5
          Notice the code bands on Richard's diode box. His unit is also set up for a "Positive Gate" wired track.

          I had built a box with a rotary switch but have either lost it or more likely accidently disposed of it when doing some clean-up/reorganization of my slot car work area. The diode trees are a quick, easy, and inexpensive solution.


          • #6
            Like Richard, I have used a separate choke box, wired in series with the white wire of my controller.

            I use a very basic Professor Motor controller, which does not have a built-in choke. My add-on choke box has a 100 ohm ten-turn potentiometer (variable wire-wound resistor) instead of diodes. The resistance of my choke box increases smoothly from zero to 100 ohms as I turn the knob, through 10 full rotations.

            Using a variable resistor instead of diodes gives me (at least in theory) finer adjustment than the 0.7 volt steps of diodes.

            (Given the choice, I would prefer a 20-ohm ten-turn potentiometer, but I've not been able to find one. Wire-wound pots are becoming very hard to find.)

            So why use resistors or diodes to reduce the power to your car in the first place? It comes down to how well your car couples its power to the ground, and your own ability to control how fast you feed power to your car.

            A gravity car can develop wheelspin if you feed power to it too fast. Wheelspin under acceleration is very undesirable. It makes your car handle poorly and reduces your car's ability to achieve maximum acceleration. Worse, if your car develops wheelspin coming out of a corner, it can go sideways, spin out, or deslot.

            It is also possible for a gravity car to just pop out of the slot if it accelerates too powerfully.

            Now a skillful driver can compensate for these issues by smoothly applying power rather than simply jerking the controller trigger all the way to the max.

            But even skillful drivers find that hard to manage consistently. Using a choke box to reduce the maximum power just makes it easier to avoid wheelspin and pop-up deslots.

            You don't really give away much speed if you adjust the choke box properly. You dial back the power just enough to keep the car from misbehaving. The car will accelerate just as fast as it is capable of without losing control. Your race will benefit because you will be able to consistently accelerate your car at the practical max, without having to rely so much on your own skill. You will give away some top speed, but few gravity cars achieve top speed on most tracks.

            A choke box is especially useful in oval track racing. My own race performance improved noticeably when I dialed back the power on my Rattler car. When I used the choke box in competition for the first time I was able to achieve an easy win against the very tough HOCOC field simply because I had better control of my car. (My very fast car!)

            You can make the argument that a truly skillful driver can run a faster race with access to the full power of the track. I won't dispute that. But it is the rare driver who can maintain that level of performance for a full race. Some humility and a little less power will often be the winning combination.

            Ed Bianchi
            Last edited by HO RacePro; 02-01-2018, 12:24 AM.


            • #7
              A problem for me with gravity gravity cars is the motors, which are a total crapshoot. Those motors do not usually come with any specs, you would lucky if you even knew if they had carbon/copper brushes. Just putting a resistor in the car is one solution if the motor is too powerful, providing that it is the right resistor. I don't have a braided track, so I can't test my cars at home. It is easier for me to use a choke box. One advantage to using a choke box is that it is easy to make on the fly adjustments. On a large banked oval it is sometimes nice if you can run a higher voltage in the outside lanes. I don't find much of a disadvantage to the stepwise adjustment that you get with diodes.


              • #8
                Viper will be doing a motor for the Mega-G+, and the crap-shoot will then be over


                • #9
                  Originally posted by NicoRosberg. View Post
                  Viper will be doing a motor for the Mega-G+, and the crap-shoot will then be over
                  Maybe so, but meanwhile Gerry's excellent Thread will be allowed to remain on topic. Rumour has it that there will be further developments revealed, later.


                  • #10
                    As promised, what follows is data resulting from testing on my Banzai BuckTrax road course. As with the prior testing, room temperature is at 69 degrees F. Track power is set at 12.0VDC. Timing and counting is by means of TrakMate Ver. 8.5. The diode tree was consistently used at the 2.1 volt tap. Choke, when tested, was on full. The tested cars are once again my Micro-Cuc IV and my GCLS. As will be discussed later, I also tested a somewhat conventional LandShark that is set up with 8/21 gearing versus the 8/15 gearing as installed on the other two chassis. Data will be presented in Mean Lap/Best Lap pairs in the order of Micro-Cuc IV, GCLS, and LandShark.

                    Without further ado:

                    Professor Motor transistor : 7.30/6.922; 6.28/6.023
                    Professor Motor transistor w/diode tree: 6.50/6.193; 6.27/5.887

                    JetStream: 6.60/6.304; 6.50/6.118
                    JetStream w/diode tree: 6.45/6.112; 6.10/5.825

                    Difalco no Choke: 7.70/6.749; 6.91/6.251
                    Difalco full Choke: 7.35/6.535; 6.34/5.995
                    Difalco no Choke w/diode tree: 7.73/6.658; 6.92/6.022
                    Difalco full Choke w/diode tree: 7.03/6.552; 6.71/5.989

                    OS3 3T Pro no Choke: 7.01/6.059; 6.45/6.066
                    OS3 3T Pro full Choke: 6.53/6.081; 6.25/5.971
                    OS3 3T Pro no Choke w/diode tree: 6.01/5.841; 6.36/5.958
                    OS3 3T Pro full Choke w/diode tree: 6.50/6.085; 6.23/5.774

                    OS3 HO All Pro no Choke: 6.44/5.984; 5.96/5.597; 6.12/5.637
                    OS3 HO All Pro full Choke: 6.10/5.818; 5.89/5.589; 5.83/5.564
                    OS3 HO All Pro no Choke w/diode tree: 6.14/5.784; 5.86/5.450; 5.64/5.424
                    OS3 HO All Pro full Choke w/diode tree: 6.32/6.006; 6.07/5.672; 5.70/5.452

                    We can draw the following conclusions from the presented data:

                    1. In general, the best performance was achieved from the controller/no Choke/diode tree combination. The OS3 HO All Pro, at least for this driver, presented the best overall performance and drive "feel".
                    2 Full Choke presented drivability issues, especially if the full speed bypass band was inadvertently activated
                    3. The 8/15 gear ratio that works so well on the oval was not optimal for the road course. Roll out was excessive and braking affects were minimal. This led to the testing of the LandShark chassis that featured 8/21 gearing. Roll out was considerably reduced and braking was quite notably stronger.
                    4. The GCLS center hinged chassis ran very smoothly and performed well. I will build a second chassis with 8/21 gearing so as to test this format on the road course.

                    A comment about the flex elements of the GCLS chassis. The inner two piano wires connect the bottom of the motor box to the rear hinge plate. This arrangement allows little to no flex in the side to side (roll) axis but does allow vertical flex in the pitch axis. The outer two piano wires attach at the front of the hinge and to the top of the motor box by means of a 90 degree bend. This feature, in conjunction with the chassis hinge, allows a degree of rotational movement of the chassis. The 90 degree bend in conjunction with the mounting to the top of the motor box allows for the stretch/shrink of these piano wires as the chassis twists. They also act as a anti-roll or anti-sway bar so as to help allow but also control the chassis rolling motion.

                    Many thanks to a BB member whom has volunteered the use of two new high end Difalco controllers as well as the testing of a Luck Bob's wiperless electronic controller. This data will be presented upon completion of testing.
                    Last edited by gmcullan; 02-05-2018, 10:21 AM.


                    • #11
                      I use a 377 ohm resistor board with my Difalco Genesis controller. The brake pot is 250 ohms. With the Difalco controller you can use a resistor board that takes plug in resistors. You can tailor that for a non-linear response.


                      • #12
                        Regarding future controller testing, I'm in the process of ordering the components so as to update my 10 band Difalco controller to the 30 band Genesis configuration. Difalco also has a new 2 transistor model coming out. I hope to test that one too. It sounds interesting. I'm still waiting to get my hands on a Luck Bob's wiperless controller for testing.


                        • #13
                          The Difalco components arrived today. This will allow updating my 10 band controller to the 30 band Genesis configuration.

                          I will document this update in a separate thread and then repeat the testing with the updated controller.

                          Stay tuned for film at 11 !


                          • #14
                            I am so far behind in posting controller/performance data. Between slot cars, working part time at the local hobby shop and repairing guitars, I'm about as busy as can be.

                            As earlier discussed, I did receive the components from Jim Difalco to convert my original 10 band controller with choke to the 30 band Genesis configuration with choke. I also purchased the 377 ohm plug in board to test against the 290 ohm standard plug in board.

                            Long story shortened, the Genesis configuration is heads and tails ahead of the older 10 band configuration in both feel and overall control.

                            Testing was conducted on my HORacePro Slider (tm) banked oval, 12.0 VDC, on the Red (2nd lane in from the outside) lane. Timing was by means of TrakMate 8.5.2 and run in 25 lap blocks. The room and track temperature was 68 F.

                            Chassis tested in order were LandShark, Micro-Cuc III, Micro-Cuc IV, and GCLS. Data is presented in Mean/Best Lap pairs in the order that the chassis are presented. When a diode tree (DT) was used, the connection was consistently made three diodes out from the track lane post.

                            Genesis w/290 ohm board: 2.09/1.878; 2.09/1.943; 1.85/1.699; 1.87/1.802
                            Genesis w/290 ohm board & DT: 1.90/1.730; 1.94/1.867; 1.79/1.612; 1.81/1.746

                            Genesis w/377 ohm board: 1.88/1.747; 1.93/1.840; 1.79/1.610; 1.88/1.813
                            Genesis w/377 ohm board & DT: 1.84/1.719; 1.93/1.838; 1.68/1.605; 1.79/1.656


                            1. As with the earlier testing, the use of a diode tree had a strongly positive affect on reducing lap times.
                            2. The 377 ohm board allowed better control, thus also contributing to lower lap times. I would have expected this on the road course, but maybe not so much on the oval. But the results do speak for themselves.

                            In comparison with testing against the older 10 band Difalco controller with the diode tree, the Micro-Cuc IV turned in a best of 1.87/1.737. With no changes to the car, track, or power supply but using the Difalco Genesis with the diode tree, lap times dropped to 1.68/1.605. This is a significant reduction in lap times and would make quite the difference in racing conditions.

                            Please note that all the aforementioned data is relevant only to my cars and driving style. Your results may very well vary. And it is entirely possible that your best results might be with a different controller than works for me. But clearly, controllers can make a significant difference in your on-track performance.


                            • #15
                              Well, Gerry has proved that a different controller and reducing the track voltage with diodes can have a major impact on lap times. In recent unlimited gravity competition on his oval an average 1.800 second lap time was a winning pace. Being able to knock that down by more than a tenth of a second is a huge advantage.

                              It is not something you can assign to statistical scatter. It means you would lap the exact equivalent car and driver, with a different controller, 3 times a minute! That is a runaway advantage!

                              Gerry keeps advancing the state-of-the-art with his methodical studies. And he is good enough to share them with us. I find it freaking amazing that, 50 years on, there is still unrealized performance available in gravity cars. But, there you are...

                              Can I, or anyone, answer the challenge that Gerry has laid down? Um, maybe. I have my own ideas about chassis design. I know that a Rattler miracle lap can beat the times Gerry has reported. I just need to figure out how to take the 'miracle' out of that lap time.

                              Ed Bianchi