Beta matching for multi-coil/filer builds

[min2oly]

Yes, that is some nice ringing... begs the question - what are you charging there :-)
That's pretty much it. here's an old vid, look about half way through.. I show the scope quite a bit.

Bedini SS Charger

[Howard Wetsman]

Thanks Patrick,
The primary is 7AH SLA and the charging battery is a 20 AH FLA. I haven't done any runs yet as I went on vacation and then got sick. Also I was waiting for my 100 MJL 21194s and my Atlas DCA. Very disappointing results. Of the 100 here are the results for Hfe
28 - 1
30 - 7
31 - 2
32 - 2
33 - 1
34 - 2
35 - 4
36 - 6
37 - 9
38 - 7
39-7
40-8
41-15
42-4
43-9
44-3
45-5
46-2
49-3
50-3
So only 15% are at 41. I'll have to buy another 150 to get 40 matched.
H

[Howard Wetsman]

Should have mentioned I bought a used 10-coiler but having just had the little test I did for a few minutes with the ss ssg, maybe I don't need 40 anymore.

[Howard Wetsman]

Patrick,
I watched the video (and I've seen a bunch of your, very nice). You seem to have moved to a bigger coil with bigger wire. Is that the principle, rather than adding coils to a bus, to increase the size of your charger by increasing the size of the winding? Is there a point of diminishing returns? A point at which you'd add a second coil rather than make your coil bigger in this scenario?
thanks,
h

[min2oly]

Hi Howard,
that's about the same results I'm getting - very expensive to match these things. I'm always astounded when I hear others having better luck. It's no wonder so few of us in the past have had any luck getting good results with these energizers, and the transistor is just one component. This new forum is inspirational - Thanks John B.

I found this web page very helpful
Curve Tracer Measurements - Microwave Encyclopedia - Microwaves101.com


I'm still very curious and am hoping some one with a curve tracer will be able to answer
Does this simple ckt provide enough or even better information on Beta Matching, "for our purposes here", than a curve tracer does?

kind regards,
Patrick

I see while I'm slow typing, you posted a couple more...
yes, there is a point of "diminishing returns" I have not completed enough experiments to find what that point is. At this point, I would build to JB's spec. He is here for help - you have 15 matched, that's enough for 3 x 18awg 4filer coils, place them at 120 degrees around the wheel get it all tuned and place the 23awg you have on the other slots in series for genny pickups ... just my from the hip thoughts :-)

[Howard Wetsman]

Thanks Patrick,
But I was referring to the solid state. I'm not going to move anymore on my 10 coiler until the replacement wheel becomes available. I think someone is working on sourcing some that will be superpole ceramics. I don't see the point of using the wheel I have with NEOs when John has said no NEOs.

So in the mean time I'm going to build an SS SSG with the CPD mod after your video. So the question is, without a rotor, is there a point at which you would say, for instance, a max of this many wires, at this size. Your aluminoso seemed really big compared to the regular 8 filar #23 coil in the other video. Would you ever consider more than one coil on an SS build if the components were matched and we used a joint buss like we would in a multicoil machine with a rotor.

I've got to echo your thanks to JB and all the mods. I was about ready to give up and was very frustrated with not getting good results but buying and building what I thought were good enough kits. Tom C was especially helpful in hanging in there until things could change. I've learned so much with the new forum, and especially from John's more frequent presence. Thanks.

So if I don't need full curve tracing and my matching has been sufficient I have 15 circuits matched and that's good. I imagine I can do something with them in a solid state build while I'm awaiting to get a new rotor.

[min2oly]

Hi Howard,
Someone had a video of a multicoiler w/ the CPD mod working nicely. can't find it....
I'm still trying (not that hard) to come up with the ultimate build. I have not been able to beat or even match the one in the vid you watched. The coil is a 9 filer 23awg 3.7 ohms on each wire. the trigger is also 23 awg. the base resistors are 218ohms1/4watt and the transistors are mje13009g ($1.00each) the cap is the 10joule yellow's that Rick sells. So far, I have not found a cap that works better for that size coil. the core is paramagnetic, however, welding rods work fine - not as good, but fine. I still have the resistors on the EB and BC, this transistor runs about 15 degrees hotter than the rest, however, I will not touch this build until I can replicate it's efficiency. hope this helps :-)
my 18awg coil seems to give a longer lasting charge to the bank, however, it is not as efficient. I may have over heated a couple of the transistors - too lazy to hunt them down - too many new projects in the pipeline...
RS posted on the other Beta matching thread today about branch resistors... http://www.energyscienceforum.com/be...sistors-2.html what he posted leads me to wonder if we only need to match the transistors w/in each branch????

Patrick

[Howard Wetsman]

Thanks Patrick,
matching in each branch would certainly be cheaper!
I'll do a 7 circuit build with a #23 8 filar. I'll get the yellow cap. I've got so many 21194's. Can I use those? Not sure what you mean by EB and BC resistors, but I assume that's specific to the MJE13009. I want to get an efficient charger because my real research interest in in doing the runs and observing battery behavior.
thanks,
h

[Robert Darrah]

Hi All,

This thread got me thinking a little. Seeing Howard's results when he tested his transistors, the gains were anywhere between 28 and 50. That's a huge difference. So let's assume your transistors are all matched, but what gain value do you need for any particular machine?

If precision is so important in other aspects of the machine, it seems a gain of anywhere from 30 to 50 in the transistors is anything but precise, and having the right gain for any particular machine might make a huge difference in performance. Maybe like tuning the base, if you have just the right gain it will bring the machine more into resonance and increase it's efficiency. So maybe the proper gain needs to be matched to each individual machine.

Has anybody played with this or have suggestions on how to test this without just using brute force and swapping different sets of transistors to see what happens? I haven't seen anybody talking about this, but it's such an obvious question that I'm sure others must have experimented with it.

thanks,

Robert

[John_Koorn]

Robert, perhaps this link will help: Transistor - Wikipedia, the free encyclopedia

The gain is not as important in the circuit because we tune the machine with the base resistor.

John K.

[totoalas]

Solar SSG less battery going to ALUM BATT/ SJR LOAD 260512 - YouTube

This is a replication of the 1000 ah charger SS SSG by mod by Minoly
Using a toroid from a CRT and additional ceramic capacitor increased the input current reange and fine tuning
totoalas
http://www.youtube.com/watch?v=cRELOGnr7Qw

http://www.youtube.com/watch?v=D98w_2ckjnA

http://www.youtube.com/watch?v=NKy7Bs_gNlw

[Nofear]

What's wrong with this picture? My build is on hold because I can't pass the transistor matching step
It's a simple enough test, and I can't believe I am staling
Basically all my MJL21194 give me the same reading. I practically have 58 perfectly matching transistors !!!??
I've tried two similar circuits (with lower resistance values then the ones suggested by john koorn/JB and in the SG handbook) With large resistance I am always working on the lower end of the ammeter.
I will give further details on the circuit and test method is someone is willing to comment or simply share if you had similar experience.
Thank you.
AA

[min2oly]

What's wrong with this picture? My build is on hold because I can't pass the transistor matching step
It's a simple enough test, and I can't believe I am staling
Basically all my MJL21194 give me the same reading. I practically have 58 perfectly matching transistors !!!??
I've tried two similar circuits (with lower resistance values then the ones suggested by john koorn/JB and in the SG handbook) With large resistance I am always working on the lower end of the ammeter.
I will give further details on the circuit and test method is someone is willing to comment or simply share if you had similar experience.
Thank you.
AA

[ATTACH=CONFIG]1778[/ATTACH][ATTACH=CONFIG]1779[/ATTACH]

Hi AA,
Someone else posted a similar experience a while back.
post or PM the details.
I have ordered supposedly matched transistors only to find that using my method, they were really not matched under the conditions we place them with JB's SSG.
Thanks for sharing,
Patrick

[brodonh]

The hardest part was matching the voltage drop of the diodes. Out of about 50 1N5408's on an ammo roll, the voltage drop varied between 530 and 580. I picked 4 where the voltage drop was + or - 1.


John K.[/QUOTE]

Hi JOhn K,
I spent a few hours at matching diodes.
First tried the diode checker on the meter and each diode would rise in voltage reading indefinitely.
Probobly increasing heat.
The I tried 9 volts from a PSU and a 300ohm resistor and the readings were much more stable accross the diode.
However when I went back to recheck my group of 7, they mostly gave different, but close, readings.
They seem to be very heat sensitive and I had to test and retest to get 7 with consistent readings. Ambient temp and heat from hands seem to affect readings plus temp change from amp draw even when it is very small amount.

Thanks,

bro d

[Nofear]

Hi All,
The last week or so, I have been trying to get a better handle on this transistor beta matching.
For that I have built all the beta matching circuits proposed in this forum with the expectation that the final transistor "sorting" would be independent of the chosen circuit.

I have measured my 58 transistors using three circuits ( SG intermediate handbook, circuit proposed by John Kroon and the circuit used by Patrick). But now I have serious doubts about my work and if it is even relevant to how these transistor will be used in the final SS SG circuit.
Here are few observations I noted:
-The readings using SG handbook circuit were hard to stabilize, The Ice was lower then indicated in the book.
-The circuit proposed by john kroon was first used as is, but later modified to protect the base and to have adequate readings.
-Patrick's circuit was the one that gave me most satisfaction.
-After trying all circuit as first indicated I moved the ammeter closer to the transistor so to measure only the collector current and not the total current.
-After experimenting with different circuits, I have found that it helps if the "collector-emitter" resistance is much smaller in value then the base resistance.

And here's the million dollar question:
- If we are using the transistors for switching why is it important to match them based on their Hfe? In my mind that is a static snapshot.
- Shouldn't we be more concern about their dynamic behavior?. i.e how fast do they open, how fast do they close. How abruptly are they closing? (leakage current etc)

I have attached the circuits aforementioned (for reference) Plus a pic of my transistors sorted using Patrick's circuit.
Although I am going thru the motions I can't see how it is relevant.

Disclaimer: These results and observation are for my own understanding. Although I have performed my tests with diligence, I will discourage anyone to draw over reaching blanket conclusions as these test are relevant to my own particular setup and equipment.

Having said That I welcome all form of feedback.