Snowman 8 Transistor SSG Build - Trigger Problems?

[Yaro1776]

Snowman SSG Build Baseline Data

Hello to the group,

Continuing to run the Snowman SSG in a basic stable state to gather data for establishing a performance baseline that can be used as a comparison point for other improvements to the system.

Becoming more familiarized with SSG2 (new board) and have started to get a sense of how it all works - more tuning required, however most of the data points are relatively consistent with the existing set up using my mower batteries and a car battery. Info below is based on about 10 run cycles.

Using the car battery (Only for initial tests) as a Primary source it usually takes about 5.25 to 5.35 hrs to charge the mower battery to15.4v at about a 1.55 amp rate - 8.3 amp hrs or slightly less. The Charge battery has been receiving about 0.75 amp rate or total ~4.01 amp hrs from the SSG. Upon discharging at a 0.48 amp rate the Charge battery drops to the shutdown voltage of 12.2v after 8.5 to 9 hrs yielding ~4.1 amp hrs.

So far the overall charging system efficiency (Output/Input) is close to 50 percent, but improving. The Charge battery input and output are about equal with some reads above 1:1 on the output side.

Using a Sears clamp-on to do the amp reads - there is some reading bounce that is interpolated to the mean. Bounce is ~ +/-4 percent. Discharge of Charge battery is via a 6watt automotive bulb 0.48 amp rate using DMM to measure the amp draw. The clamp-on meter is used during the actual runs and agrees with the DMM amp read.

Conditioning of the batteries to be used for the actual tests continues.

So the above information creates a great starting point for the next modifications - first up is getting the freewheel time to increase - the magnets and tape definitely slow the wheel down due to air resistance.

More to come in this adventure!

Yaro

[Yaro1776]

New Bicycle Wheel Upgrade

Greetings,

Have the new bicycle wheel with primitive cup and cone ball bearing setup vs. the original wheel with hybrid cage ball bearings. The original wheel gave 75 secs. of free spin time in unaltered state - removing seals and degreasing the bearings upped this time to about 120 secs. This bicycle wheel (559mm x 17mm) is outfitted with 21 magnets from the Teslagenx kit. Thought that this would be the hot setup -wrong! The free spin time has degraded over multiple test charging runs to ~ 100 secs. Aargh!

So the new wheel (same diameter of 559mm) when first spun (no magnets) yielded about 150 secs - degreased the bearings and put on a drop or two of Triflow equivalent. OOPs lost a ball during assembly. Delayed a bit until replacement of the lost component.

Spun the unbalanced wheel after getting all the pieces together (still no magnets on wheel) and it rotated (to my delight) for 5 min and 15 secs. Way better! So does this translate into improved performance? Expect needed extra magnets from TeslaG in the next few days. They will be slapped onto the new wheel, with meticulous care of course, and then the new results will tell the story! Stay with us on this...

Have a Rheostat (potentiometer) 100 ohm 12.5 watt from Digikey on the way for fine tuning of the trigger circuit... Is this needed? Don't know, so we burn the bucks...

This remains a "Tremendous" adventure,
Yaro

[aln]

Yaro,
looks like you are having fun. Don't make it too efficient or you'll have to put a fan on it to slow it down. J/k Keep up the good work. It is an enlightening/frustrating and sometimes expensive process. That pot will more than suffice as long as you are w/i 100 ohms on the base of where your machine likes to run. Aln

[Yaro1776]

Potentiometer Test on 8 Trans SSG

Received the potentiometer from Digikey and had some time to put it through its paces. Hooked up the unit to the trigger terminal bypassing the regular 12 ohm resistor. Varied the pot resistance as the SSG wheel was spinning to get a feel for the operation of the pot. The bike wheel speedo was very handy for this operation giving a quick and responsive readout to all the changes.

Ran a series of tests varying the pot from 0 ohms to 20 ohms - anything greater than 20 ohms was deemed not useful for tuning purposes since the wheel would not accelerate past 140 rpm at the given gap of 3/32". The data is attached if you wish to look it over.

Results show that the SSG board kit, as setup in the Snowman build, is tuned very closely to its potential maximum rpm, give or take a few ohms, with the standard 12 ohm resistor. It is interesting to note that within the ohm range used the Primary output and Charge input values decreased with increased resistance, while the ratio of Input/Output remained constant. The maximum rpm is in a narrow range of 12 to 18.6 ohms.

Kudo's to John B. for designing a great board for us Beginners - excellent.

Next up is the new wheel freespin test and its initial testing. Can't wait, but work always interferes here!

Best to All!
Yaro

Rheostat Test 1 Board 1.pdf

[Yaro1776]

New Wheel Spin Time and Variable Gap Test

The new wheel was prepped, balanced and run for free spin time - result is a free spin time of 10 min and 10 secs. The old wheel (both with identical diameters and number of magnets) was able to run a measly 120 secs and had been degrading slowly in free spin time down to 105 secs.

Next step was to fire this puppy up and see what gains are realized by improving the free spin time! For these runs the gap was varied to find the best performance points with and without the 12 ohm fixed trigger resistor. The test data is presented in the .pdf attachment.

The results show nearly a gain of 10% in operating RPM along with a rise in maximum output and input amps. This wheel performs much better, however it does suck up the Primary amps but also produces a higher input amp to the Charge battery. You will note in the data that the Input/Output ratio shows an overall increase, this may be due to the reduced friction from the wheel and yet to be verified. One would speculate that the Primary output amps would go down with reduced friction...

The Test Runs 3 to 6 show a marked increase in the I/O ratio over the other runs, not sure why at this point, but will repeat.

Another strange run was Run 1 where the RPM with the resistor was a paltry 157, yet it was pushing a high amp rate to the Charge battery with a high I/O ratio. Another item to repeat.

Plastic shim stock (0.025") was used to vary the height of the coil along with a 0.120" thick, smooth paint stirrer.

Overall results: Better RPM, Wider Gap range, Higher Primary and Charge amps.

Life is good!
Yaro

[Yaro1776]

Start of Snowman 8 Trans COP Runs

Today marked the start of the formal testing of the Snowman 8 trans SSG - mind you this project was started in the middle of November with many interruptions and other hiccups. Good to go, and up and running as this is being written.

Free wheel spin time is at 10 min, 55 sec.

The Primary output is at about 1.85 amps and Charge Input 0.94 amps with fully charged Primary battery - basic ratio at 0.51- 0.52. The Charge battery was conditioned and then load tested with just under 10 amp hrs available. Running wheel RPM is 295 with a coil to magnet gap of just under 0.200". This appears to be the best configuration of the many tried.

Cup and cone bearings on this wheel make a lot of extraneous noise, not happy about this since it offends my mechanical sensibilities. The other wheel sang a smooth song, though with poor free wheel times. The test wheel runs very smoothly though.

Coil temp is 3 degrees above Lab temp of 71 F, transistor heat sinks are at room temperature.

Mentioned in previous post that runs 3 - 6 had a higher Charge Input amps/Primary Output amps ratio of about 0.60, not sure how this happened since most other runs have shown a ratio of 0.51 to 0.53, however suspect high initial battery voltage or operator error. The amp readings are tedious with the clamp on DC meter since there is a bounce in the readings of about +/- 4 percent. Assume that taking the median read from high and low reads is about as good as can be done for now - mind that the overall readings from run to run are very consistent with this method. A straight analog type meter may be better, but it does have to be wired in each time for a reading.

The data from today's run will be presented in a simplified format in the next post, and yes, all the data will be recorded in the approved spreadsheet format for others to review once there are a sufficient number runs completed.

Enough for today - best to all,
Yaro

[Yaro1776]

Snowman First COP Run Results

Greetings to all,

The first test run yesterday operated flawlessly and yielded predictable results based on earlier learning tests. Overall the following basic info was recorded over the 10 hour charging run on a NAPA 8224 N 230 cca battery.

1) Charge battery start voltage 12.45 v with an end voltage of 15.36 v.
2) Primary batteries just about eked out the Charging time for the two cycles required for the Charge battery.
3) The Primary output amperage was averaged at 1.78 with the Input amperage to the Charge battery at 0.91 - this yields an I/O ratio of 0.51. This works out fairly closely to the individual batteries' ratings, albeit the Primary amperage being rather high for the battery spec.
4) The Charge battery has been on its discharge cycle for about 14 hours at a nominal discharge rate of 0.5 amps/hour - still have about 6 hours or so before it is depleted to the 12.2 volt level. The total discharge time will tell the story here.
5) The wheel rpm started at 292 and gradually climbed to 297. The pitch of the noise and rumble from the wheel was becoming annoying after a few hours, never mind being exposed to it for 10 hours. Earplugs are next!

So the positive aspects of this test run were:
a) Nothing broke or blew up over the 10 hour run!
b) Data was consistent.
c) Survived the marathon effort.

The major negative aspect of this test run was:
a) The Primary amperage draw is too high for the battery rating and may lead to an early failure. The C20 rate for this battery is approximately 0.50 amps and it is being forced to produce 1.78 amps for the way the SSG is configured for maximum speed and gap. The Primary battery is being discharged close to a C5 rate based on an approximate capacity of 10-12 amp hrs. The next step is to improve this to a reasonable discharge rate.

According to the "Bible" this can be corrected by either adjusting the Coil to Magnet gap, modifying the base resistance or changing the number of magnets.

With the current SSG configuration the gap can be increased to 0.390" or so and the resistance tweaked a bit higher to about 20 ohms total to yield 236 rpm with an Output of 0.98 and Input of 0.46, however this would probably increase the overall Charge time to about 20 hours. The Output rate would be about C10, a livable value.

The second method is to change the number of magnets, well this has not been researched as yet - perhaps someone can provide a bit of direction here.

The third way is modifying the base resistance of the SSG - this is the last choice... It can be seen in the prior post data that as the base resistance is lowered the amperage rate increases - wrong direction. On the flip side increasing the resistance over a certain value causes the wheel rpm to drop dramatically.

A more reasonable option was suggested by BobZ on another thread to parallel two or three identical batteries on the Primary side to eliminate the battery switching struggle - this is the simplest approach and will lower the Primary discharge rate for each parallel battery to a reasonable level below the C10 discharge rate.

So turning to the SSG experts - what is the best option here?

Thanks to all for your attention in this great experiment!
Yaro

[romeo-kilo]

Hi Yaro,
for me it looks like adding (parallel) batteries to the primary is the easiest solution. The other solutions would involve a lot of effort and time!

Other Info - reading just the Amps in and Amps out is (for me) not a valid approach for measuring CoP.
I am measuring the Amp hours charged into the Primary battery vs. the Amp hours getting out of the secondary after SG charged it. To be generous I ignore the fact, that the Voltages are different, and the disadvantage in my calculation would probably be between 15 and 20 percent. So the best CoP on my machine using my calculation is 0.6 (my machine is running since End of November on an almost daily basis)

Ralf

[Yaro1776]

Hi Yaro,
for me it looks like adding (parallel) batteries to the primary is the easiest solution. The other solutions would involve a lot of effort and time!

Other Info - reading just the Amps in and Amps out is (for me) not a valid approach for measuring CoP.
I am measuring the Amp hours charged into the Primary battery vs. the Amp hours getting out of the secondary after SG charged it. To be generous I ignore the fact, that the Voltages are different, and the disadvantage in my calculation would probably be between 15 and 20 percent. So the best CoP on my machine using my calculation is 0.6 (my machine is running since End of November on an almost daily basis)

Ralf

Hello Ralf,

Appreciate your feedback on the battery question - I am leaning towards doing another battery in parallel which appears to be the simplest solution until the capacity is increased and charge time is lowered on the Charge battery. My Charge battery discharge test yesterday was incomplete - fell asleep at the wheel and allowed the voltage to drop to 12.1v before I awoke to stop it. I will correct this to get a very close calculation of actual discharge time...

I am considering raising the Charge battery maximum voltage to a higher level from the present 15.3v to 15.4v limits - after many test runs I noted a diminishing capacity on the first Charge battery. I decided to up the maximum level to gauge the impact on capacity. Last run done was to 15.86 volts and subsequent discharge capacity increased from 4.5 hours to about 7 hours, major jump and almost back to original 8 hours. Note that Branch Gordon has been pushing the max voltage on the Charge battery until the charge curve plateaus at about 16 volts.

I am still considering the best solution for the amperage measurement. For now I am measuring the Primary output amps, Charge Input amps and Charge Battery discharge time at a constant 6 watt load down to 12.2 volts. Primary battery is being charged w/ an EnergenX 2A12 or 10A12.

By the way, read about your Aha! moment on John K's Thread - very nice.
Keep spinning,
Yaro

[aln]

JB says lead acid are fully charged at 16v, old ones are a struggle to get there at times. depending on the size of your charger. Beef up the front end (bigger batt) or increase the resistance on the base as you will kill your primary discharging it at that rate. Aln

[Yaro1776]

The Struggle Continues

JB says lead acid are fully charged at 16v, old ones are a struggle to get there at times. depending on the size of your charger. Beef up the front end (bigger batt) or increase the resistance on the base as you will kill your primary discharging it at that rate. Aln

Aln,

Thanks for your suggestions - they are right on. The current set-up is running on the hairy edge with respect to gap and resistance (no scope for fine tuning). The maximum charge voltage for the Charge battery will be upped to the 16 volt plateau to ensure a full desulfation and radiant charge. The primary battery set-up will be upgraded today with additional similar batteries to eliminate the switching BS.

So welcome to the world of radiant charging, yet again another diversion,
Yaro

[romeo-kilo]

Yaro,
did you experience any advantages in charging yet? I mean that the time to charge will decrease. Unfortunately I didn't, yet - even after 60+ runs.

Thank you for providing the pdf files with your rpm and in vs. out amps results, I really love seeing results from others so I can compare my setup.
Here is my stuff (just ignore the Km/h, it is the German mph)


btw - I also calculated the rpm vs mA to find the sweet spots with best speed vs current.

Ralf

[Tom C]

you will not see 1 to 1 with a 16 volt Termination point for a very long time. the last volt and a half above 14.5 is harder to achieve. that is another reason battery companies use 14.5 as termination voltage. if you have good batteries 15.3 is good enough. it takes a long time and lots of cycles, with no resting to achieve higher capacity, and you will lose it with letting the batteries sit. this is from Lindemans battery books.

1 to 1 is very easy at 14.5

Tom C

[Yaro1776]

Ralf,

I did check out your graph and according to your method the sweet spots are self evident. I am unfamiliar with your build so I have no reference point to what I am looking at. The data that you collected, is that from varying the resistance or gap - perhaps it is real time data from the wheel as it is ramping up?

My first run Charge discharge time needs to be corrected and I am in the process of completing the correction. My last run ended in a failure due to my battery charger having issues. So, in reality I have not accomplished much this week beyond spotting a weird set of readings on my Charge input voltage during the test run battery switch. This intrigues me and I am trying to duplicate it again.

Yaro

[Yaro1776]

Tom,

Thanks for your input on 15.3v test termination level. Looking at my data it shows that an additional 35-40 minutes were required from 14.5v to reach 15.3v. I will slide that into the calcs for grins since we know the input amp rate.

I can see that this will be a marathon effort to measure the gains since they are additive according your reference to Peter L. Uncertain if the 20 continuous runs can be pulled off by a single person when one considers a 20 hr discharge time and ~10 hr charge time with the present SSG set up. Looks like it is time to bite the bullet again and buy the CBA IV to help with the discharge cycle program. I need more toys...

BTW - All the batteries are new NAPA 230cca.

Yaro