Gary's "Complete Advanced" SSG Build

[pearldragon]

Hi Gary, RS,

For me as an engineer to live with a 25% difference in performance without knowing what caused it, is hard… not to say unacceptable hahaha. I even would ask myself if it would make sense for me to continue with this research if I can’t find satisfactory answer for this difference in performance.

But not all hope is lost yet : although I did buy my core from TeslaGenX, they forgot to supply the core material/welding rods… So I bought those myself, what at that point I thought would work. I’ve bought R45 welding rods https://certilas.com/en/product/g1.

Now in the beginner’s handbook they speak about R45 (page 60) and in DVD24 @+/-51.3 about R60. TeslaGenX sells R60 on their website too. With a little research I learned that this R values says something about the tensile strength of the welding rod material, not necessarily about the composition. In other words, different types welding rod materials can all have a R60 classification. So, to pinpoint exactly which composition (material nr) you have in your core, my first question to you is: Are the welding rods in your coil-core supplied by TeslaGenX? http://www.teslagenx.com/parts/tx-r6...category=parts
I just contacted them to chase down the material nr of their rods.
In the case that instead of matching your welding rods, I want to try to go beyond the specs of those welding rods, I assume I need to look for welding rods with better specs in regards to magnetic remanence and magnetic coercivity?

Best regards,
Rodolphe

[Gary Hammond]

Hi Rodolphe,

I've pretty well ran out of ideas to suggest to you. I doubt that the difference in our base resistor arrangements would result in a 25% difference in performance. I've pointed out all the obvious differences I see between the two builds, but nothing you've changed has resulted in a big change in performance. So most of the differences appear to be somewhat insignificant.

Don't get too hung up on the electrical COP of your first build. These machines are basically a learning tool where you can observe how several different variables affect the overall performance; how they interact with each other and produce different results. They also demonstrate how energy from the inductive collapse can be recycled to charge a battery while still being able to produce some mechanical work.

Gary Hammond,

[pearldragon]

Hi Gary, RS,

PCB connections
Response to you post# 85 Gary. Thanks for the picture.
I ran two tests with the washers and nuts removed, so that the soldered rings would mate directly on the PCB (no serrated washers). First test with 2Amps as a starting input (cycle C911), the second test with 2.4Amps as starting input (cycle C912), I was so sure that this was the issue I was looking for, I also connected my 8^th power core again… but as you can see in the attachment: performance is again in the same ballpark…
In the attachment I have some pictures of the new connections. As you can see, my wires to the batteries are a bit long. I do not expect that I could gain 25% improvement by shortening them, especially after seeing no improvement after removing the washers/nuts on the PCB connections.

Batteries
I just charged up my AGM output battery again this time using the CBA to monitor the charging voltage. I want to do the same with my flooded lead acid battery. So I then can compare the graphs and determine which voltage of the current charging curve of the flooded L.A. battery would correspond +/- to the 15.3V of my AGM battery. (Flooded L.A. still is not charging till 15.3V). And then maybe run a cycle with the Flooded L.A. at the output… to cross the AGM batteries off the list as a possible issue. I don’t think they’re causing the problem, based in previous cycles and tests, but will be one more to cross of the list…

Potmeter
What I haven’t done yet is trying to replace my pot meter (2W) with a/some fixed resistor(s) (with a higher wattage)… But not sure if it would change anything. If I remember correctly you (Gary) had a variable pot meter as well, but a version with higher wattage. If 25% of the energy would be used up in my 2W pot meter, I guess it would have burned out a long time ago…

Resistors
Just re-read the response from Erik from TeslaGenX, he mentioned as prime candidates the amount of magnets on my rim. That was at that point Rotor1, with 24 magnets (see the signals attachment of my post #78). The second reason for the low performance he thought would be the low base resistances, but I think he didn’t realize that I have a pot meter installed too (I might have not told him). Is there any reason why 470-Ohms base resistors (+pot meter) could account for 25% better performance over 100-Ohms resistors (+pot meter)? We talked about it since #55, but after the testing I’ve done (or apart from the testing I’ve done), is there another reason to swamp to 470 ohm?
While writing this I’m re-reading your post #29 and calculating total base resistance values:
7 coils, gap 0.31” = 7.8mm gap: (7x 470 in parallel) + 12 in series = 79.1 Ohm
In my cycle C903/C904 (7 coils, 12mm gap): (7x 100 in parallel) + 50.5 in series = 64.8 Ohm
In my cycle C905/C907 (7 coils, 9mm gap): (7x 100 in parallel) + 57.8 in series = 72.1 Ohm

@RS, will try to respond to your post tomorrow, it’s getting bedtime for me on this side of the “lake”.

Best regards,
Rodolphe
2020-09-21 - Attachment measurements.pdf
2020-09-21 - Attachment hardware.pdf

[RS_]

[QUOTE=pearldragon;n73153]Hi RS, Gary,



-Does your SG perform at COP 1 or better too (Ah wise), like Garry's?

Please understand, I have build some where around 50 SG or SSG variants over the last 20 years, so the answers to your questions are not all clear cut....

I have had some models that were COP 1 + Mostly ones that had a genny coil / coils to help out. The vast majority were Not COP 1. I learned a LOT of little things along the way to help this issue, and having access to JB later on was a big help. His graciousness in answering my questions over the phone, and talk about all kinds of topics too, for hours at a time, helped a lot, and is why I am here helping others like he helped me.

-Do you charge your input batteries with a radiant charger?

SSG, SG's cap pulsers, you name it, I have built and tested it....... some i charged with a standard battery charger some i charged with a SSG, and some i charged with a cap pulser. A SSG should not be used to charge primary input battery's. A cap pulser is ok to use on primary input battery's


-Do you have 100-Ohm or 470-Ohm base resistors?

Just depends on the setup on a given circuit. I have been all over the place on base resistors.
My big 48 transistor SG that charged L16 size battery's at 24V had matched 22ohm branch R's on each base with matched 47ohm branch R's feeding the 22ohm R's for every 4 transistors, and 3 matched 10W 10ohm branch R's feeding the 4 47ohm R's on every 16 transistors, with a 25W 250ohm rheostat and a auto tail light feeding the 3 10W 10hohm R's.

It's too bad that most of the models were destroyed in the shop fire a few years back......

[Gary Hammond]

Hi Rodolphe,

All wire connections to my PCB are soldered on ring end terminals that are fastened to the board with a bolt and two nuts with serrated washers to bite into the board traces and ring end terminals. And the bolt heads are soldered to the board traces on the under side.

Here's a photo of my PCB. You can click on it to enlarge.



Gary Hammond,

[pearldragon]

Hi Gary, Rs,

I made a new “CG-cable” yesterday, bypassing the Radian/CG-witch, and left out my amp meter as well. See first picture in attachment.
Then did a discharge/charge cycle today. Although I couldn’t read the input amps, I left the pot meter at the same setting as the previous test, so could judge by just charging time whether the performance would be better or worse. Result; the same.

Now during the day I thought a lot about what would be the next likely thing for the low performance. I was about to go to bed, and thought about the bolted connections to my PCB. Then I had a closer look at specifically the connections between the coil wires and my PCB… and then saw That these coil wires are NOT directly bolted onto the metal of the PCB, but have a washer and a nut in between there… see second picture in attachment. I try changing these connections hopefully this weekend, getting the washer and nut out of there. If performance improves, I might consider soldering them. (while writing this, I realize that I have more connections like this: that all my connections to the PCB do not touch directly, but that there is first a nut and a washer….If these connections are critical, than this might be the problem I was looking for…)

Any chance you could send me a picture how your connections from the coil wires to the PCB look like?

Best regards,
Rodolphe
2020-09-19 - Attachment.pdf

[Gary Hammond]

Hi Rodolphe,

Gary, in your test from post #29, you charged up to 15.3V too, correct?

Yes, that is correct.

Gary Hammond,

[pearldragon]

Hi RS, Gary,

I also use conductive grease on most connections. Helps lower the resistance of each connection that would other wise add up with the more connections you have in the system. I also try to use high quality, high strand count audio power cable on all connections from the circuit to the battery's

So far I've used 10 AWG cable (see attachment). Conductive grease I've not used so far. I would reckon that that would be a good one the get the max out of the machine, but at the moment I'm still about 25% performance wise away from Gary... so my guess is that there is still a big bottle neck somewhere... although I've crossed quite some candidates of the list already as you can see in the posts here...

I've got a couple of questions regarding your SG:
-Does your SG perform at COP 1 or better too (Ah wise), like Garry's?
-Do you charge your input batteries with a radiant charger?
-Do you have 100-Ohm or 470-Ohm base resistors?

Gary, in your test from post #29, you charged up to 15.3V too, correct?

Best regards,
Rodolphe
2020-09-17 - 10 AWG stranded wire.pdf

[pearldragon]

Hi Gary,

Started today with a test run with an AGM at the input and one of my Flooded L.A. at the output. Only drained 0.5Ah out of the F.L.A., just wanted to see if I could get it up to 15.3V already, but stuck at 14.96V.

After that I followed up on your comments/recommendations and made new connection cables to the batteries/SG:
-Connections to the SG and Amp meter are now bolt on
-Connection to the batteries is now with cables-shoes
See attachment for photos of before and after.

Just finished a discharge/charge cycle (with AGMs): performance still in the ballpark as before… Tomorrow probably try to do a run without the amp meter (so that eliminates a switch), and will consider temporarily bypassing the Radiant/CG switch too. But would be highly surprised if this would increase performance with 25%. I had hopes that eliminating those jacks would help, but if so, only very slightly. That something in these connections was the main issue sounded plausible; since with some Radiant tests I did in the past, I would start with 1.4A as input current, but then after hours it would still not get the output battery to 15.3V, only when I increased the amperage (by turning the pot meter) I would get it to 15.3V. So that there is a spot with too much resistance sounded very plausible to me.

Thanks for your compliment on my SG; high praise from an expert like yourself

Best regards,
Rodolphe
2020-09-17 - Attachment.pdf

[RS_]

I also use conductive grease on most connections. Helps lower the resistance of each connection that would other wise add up with the more connections you have in the system. I also try to use high quality, high strand count audio power cable on all connections from the circuit to the battery's.

[Gary Hammond]

Hi Rodolphe,

1-I’ve taken some pictures of my hardware (connection). Could you have a look over it and see if you see anything out of the ordinary? Something that could be the culprit for the bottle-neck. See attachment.

The only thing I see is that you have jacks on each end of both your input and output leads, as well as on your ammeter. Each one of these connections will have a small resistance, which added together may cause slower charging. I found with my machine that the fewer and more positive connections I could use gave me the best charging. Especially when running in CG mode. Switches can also cause unwanted resistance.

I use all screw terminals and/or soldered connections, even on the alligator clips. And I try to keep the number of connections to a minimum. Bolting the leads directly to the battery also helps.

Every connection in any circuit is a potential trouble spot.

You have a very nice looking build. I admire your craftsmanship, but the excessive use of jacks may be the culprit.

Gary Hammond,

[pearldragon]

Hi Gary,

I’ve run the tests with the Rotor3 (21 magnets, 22mm wide), with a 12mm gap and a 9mm gap. Both tests are still within range of all the previous tests… I also analyzed the signal again to determine the ON time with the 9mm gap. What I noticed is that the voltage of the input battery while running at the end of the charge cycle is lower than in previous cycles.
For all the results, see attachment.

I’m scratching my head a bit that after changing and checking so many parameters one by one, nothing seems to have a significant effect on the performance of my machine. This gives me the impression that there still is a bottle neck somewhere that I cannot seem to find. I’ve been thinking how to proceed from here and came up with the following plan:

1-I’ve taken some pictures of my hardware (connection). Could you have a look over it and see if you see anything out of the ordinary? Something that could be the culprit for the bottle-neck. See attachment.
2-If you don’t see anything strange in my hardware, I send my results to TeslaGenX too, see if they have any more input.

For 3.x I’m still considering which one to choose/choose first.
3.1-Considering replacing my 100-Ohms resistors with 470-Ohm. Although my earlier scope measurements didn’t show any reason to doubt my 100-Ohms and they are specified in the intermediate manual, I have to try something else, besides it would bring my machine even more in accordance with yours.
I think this would be a good one to start with

3.2-Considering soldering a new PCB board with 8 matched transistors from TeslaGenX incl 470-Ohms resistors instead of 100-Ohms.
If 3.1 doesn’t work, this would be my second choice
3.3-Make a 4^th rotor with 24mm wide magnets as per spec of the Beginners manual.
A new rotor seems the least likely candidate to yield results as my current rotors/gap sizes didn’t seem to change the performance either..

3.4-Test with flooded lead acid batteries at input and output.
Since I’ve switched/tested around with batteries quite a bit already, including using a flooded 24Ah at the input without any results, I’m inclined to leave this option off the list for now too.
3.5-Test again with current rotor with another gap and/or input current value
Since I’ve tested around with different gaps in the past and different input current values, I’m inclined to leave this option off the list for now too.

What are your thought about this approach?

Best regards,
Rodolphe
2020-09-15 - Attachment hardware.pdf
2020-09-15 - Attachment measurements.pdf
2020-09-15 - Attachment signals.pdf

[pearldragon]

Hi Gary,

Gauss rating of new 22 wide magnets lies between 3800-4000Gauss (Y30BH)
Gauss rating of original 20 wide magnets lies between 4300-4500Gauss (Y35)
Rated pounds of pull spec I do not have, but can inform with the suppliers if you need it to assess the differences better?

Good point regarding the gap and these new magnets/difference in Gauss rating. I’ll first run some tests with the 12mm gap and after that I’ll reduce it to 9mm.
One of these days my machine has to yield and give me some improved performance!

Best regards,
Rodolphe

[Gary Hammond]

Hi Rodolphe,

(although the scope percentage ON time of Rotor3 (new one) is the same as Rotor2, I think the max rpm was higher. But this would not be what I expected since Rotor3 has 1 magnet more, although the shape is different… I’m not a 100% sure I remembered the max rpm correctly, but this will be confirmed during the new test cycles).

It could be that differences in the magnet field strength at any given coil gap could also affect the percentage of on time. What is the gauss and rated pounds of pull of the new 22mm magnets compared to the original 20mm ones? A 12mm coil gap is a little too wide for the 22mm magnets I have on my machine which are rated at 3850 gauss with 4.5 pounds of pull force. My machine seems to work best with an 8.9mm gap with the magnets I have.

I realize you wanted to keep the gap the same to reduce the number of variables for each test, but magnet strength may be a variable as well that you didn't account for?

Gary Hammond,

[pearldragon]

Hi Gary,

The 25% on time is the theoretical calculated maximum on time based on magnet size and wheel circumference. The 20% time is the actual observed on time, viewed on the o-scope, while the machine is running.

Ok this cleared things up for me. I finished my new rotor but before installing it, I checked the ON time too with the scope for but by rotors I already had, then I installed my new rotor and did the same. In the attachment you can see a summary of the scope readings, compared to the calculated values. Hopefully tomorrow I’ll be running test with the new rotor.
(although the scope percentage ON time of Rotor3 (new one) is the same as Rotor2, I think the max rpm was higher. But this would not be what I expected since Rotor3 has 1 magnet more, although the shape is different… I’m not a 100% sure I remembered the max rpm correctly, but this will be confirmed during the new test cycles).

I agree. The differences seem to be just from the probes and not from any circuit imbalance.

Instead of the probes, it might also be that one of the channels is a bit off / that I need recalibration of my scope. For now it doesn’t matter to much for me which of the two it is. What is important is to determine whether the resistors and the transistors are sufficiently matched and so far I understood that they are and we can cross that one off the list as possible bottle neck.

Best regards,
Rodolphe
2020-09-13 - Attachment.pdf