JM-SST feedback

[JimmyM]

Hi folks,
Since I've got 2 batches built and sold, I'm thinking about batch 3.
Before I do, I'd like to hear back from those who've used them.

This will not only go into the next batch of JM-SSTs, but also go into the design of my regulator that is currently showing real promise. Breadboarding is going very well.

So let's have it!
Too fragile? Too big? Not adjustable enough? Starts too fast? Starts too slow? Sh*tty build quality?
And the good stuff too.
Post your successes and failures.

Thanks, all.

 

[jimjones3630]

Jimmy congrats on two sold out batches of sst's. This thread is a great idea as often comments in a sales thread are buried. I bought some from both batches and wish had some real feedback to give but haven't got them wired in yet.

I didn't want to be in the position like with Winny's PWM's which where and are desirable but not available. I have seen the quality of your work on another project and so anyone thinking about getting one better get while the getting is good so to speak.
jim

 

[SafetyBob]

Jimmy, I am ready for batch 3 and will put that puppy in a light within 15 minutes of it coming in the mail. I am especially ready for your regulated one too.

Please someone who has one from batch one or two put it in a light and let us know how you like it. If there is anything that could be fine tuned, we need to let Jimmy know.

Bob E.

 

[JimmyM]

I know and plasmaman put one into a 600W Larry14K and it works.
But he and petrev tried it in a 1000W version but cooked the MOSFET.
So at 600W it was way over what I originally intended it for but it kept working. The 1000W was WAY more.
It's on its way back for repair.

 

[JimmyM]

Anything, guys?
Batch 3 will have to wait for a couple other projects in line.

 

[petrev]

Anything, guys?
. . . .

Hi Jimmy

As Jim knows I had some teething troubles with my first V1/B1 but you fixed it Quick-Smart (trim pot needed a bit of extra soldering) and have now fitted :-

2 in Maglites (1 Std. and 1 with an external pot for Volume control JM-SST(V) ?)

Plus 1 in a (PlasmaMan-Version)LarryK14++++1000W Sleeper PK1000 (whatever you call it) and it's working great - had to fit an extra FET and mount them on a heatsink to enable the 1008W - 36A flow - but working great.

You have to set up the trim for each pack/cell count in a host as the changes in Voltage seem to move the ideal trim point a bit but that's fine.

Bottom line - GREAT ! ! !
Can't wait for the full regulated kit too . . .

Thanks Jimmy

 

[cnjl3]

MY JM-SST is powered by 18 A size CBP 2500mah cells in a MAG 5D tri-bored host. I initially tried to light up a 64458 but my Delghi 3" rig grabbed the bulb and I twisted one of the bi-pins off - what a bummer! I had to settle for a 65W IRC. It sure shines nice. I like the audible twang when you fire it up! The only knock/negative (for me at least) is that since it is PWM my cheap DMM will not give me a true reading. I still dont want to spring for a "Fluke". I am looking forward to your regulated version. Keep up the good work.

 

[JimmyM]

Your meter will have the same problem with the regulated version since it will be PWM as well.
There are other meters that measure RMS without having to do for the expense of the Fluke. I've got a Fluke 189 and don't need the newer 287/289, but am just itching for a good reason to buy it.

 

[cat]

Well, you can sell me the Fluke 189 and then you''ll have buy the new one. :devil:

Because either my cheap DMM is horribly inaccurate or my BC-6 is charging my batteries to 4.6V or so.

 

[rizky_p]

ETA on the third run?

 

[cnjl3]

I did a quick check on Google for a Fluke 189 and it is still $380 or higher. Yikes! I probably will get one later. Hey I musta been half asleep when i posted the cell count on my homemade "A" size CBP 2500mah battery pack -it is not 15 cells but 18 cells.
PM sent.

Your meter will have the same problem with the regulated version since it will be PWM as well.
There are other meters that measure RMS without having to do for the expense of the Fluke. I've got a Fluke 189 and don't need the newer 287/289, but am just itching for a good reason to buy it.

 

[JimmyM]

ETA on the third run?

Not soon. That's the best I can do. When the 3rd run is made. I think I'll have the boards made professionally. Funds have been sapped a bit by regulator development.

 

[JimmyM]

Well, you can sell me the Fluke 189 and then you''ll have buy the new one. :devil:

Because either my cheap DMM is horribly inaccurate or my BC-6 is charging my batteries to 4.6V or so.

Don't tempt me! I've got the PC interface, software, and the extended run battery pack.
That's bizarre. The output of those chargers is usually filtered and NOT straight PWM. I'd like to get my Fluke or 'scope on there to see what's going on. What kind of meter do you have?

 

[Mr Happy]

If you disconnect the batteries after charging and measure the open circuit voltage, isn't that what counts? That's what the charger reads and charges to.

 

[cat]

Open circuit. - If I remember correctly. But now I'm confused, thinking about it now. I'd used it before checking new CR123's and they were around 3.0, 3.02, etc., so that seemed ok. ...Right now, I cant remember what batteries it happened with, and a couple days ago I charged my WE Storm batteries and iirc, they came off at 4.2-something.

 

[petrev]

If you disconnect the batteries after charging and measure the open circuit voltage, isn't that what counts? That's what the charger reads and charges to.

Hi

The open circuit voltage is not to be relied on as the charging voltage as this generally falls when current flows at the set rate - The initial applied charging voltage must be read accross the cell to gain any relevence

Good chargers for lithiums generally follow a constant current then constant voltage method (CC/CV)

Initially the applied voltage may be higher than the 4.2V ending voltage that is considered fairly standard for LiIons but this should become ~4.2 when the charger gets into CV phase.

when the battery is fully charged it will come off the charger at 4.2 and then fall back as the cell rests and cools.

Good chargers switch off momentarily many times a second and read the actual voltage of the cell as they charge (FMA)

Pete

 

[cat]

hi Pete

When the battery is fully charged it will come off the charger at 4.2 and then fall back as the cell rests and cools.

ok, That's why I'm measuring immediately after charging, to see what it charged to - according to my multimeter, which I doubt is more accurate than the BC-6.

Initially the applied voltage may be higher than the 4.2V ending voltage that is considered fairly standard for LiIons but this should become ~4.2 when the charger gets into CV phase.

ok, that explains why...I put an AW C cell on this morning, the charger showed 3.4 (iirc) but in a few minutes it showed 4.2 and was still there..about 40-50 minutes later, and then I took it off and went to work.

Good chargers switch off momentarily many times a second and read the actual voltage of the cell as they charge (FMA)

I think the BC-6 works like that. After reading Silverfox about his WE Storm cells becoming unbalanced on the WE charger, and balancing them "passively", I checked mine. They were a bit out of balance, so (because I haven't set up anything like his magnet wires yet) I decided to discharge them and then charge them on the BC-6. The voltage shown was changing up and down quickly by .1 before settling and then going on to the next point.

 

[petrev]

Hi Cat

BC6 is another noted excellent charger.

Sounds like your stuff is working correctly

Pete

 

[JimmyM]

Hi Jimmy
You have to set up the trim for each pack/cell count in a host as the changes in Voltage seem to move the ideal trim point a bit but that's fine.

I'm such a dolt! I've been fighting this problem the whole time with the regulator. RMS versus Average voltage. At duty cycles less than 100% RMS voltage is higher than average voltage by a certain percentage that varies depending on the duty cycle. So Petrev is correct. You need to set the JM-SST for the specific pack voltage you're using.
Set the JM-SST to full clockwise (Minimum duty cycle). turn it on. Then turn the pot counter-clockwise until you reach the correct measured RMS voltage at the pins of the bulb. You can't set it without a bulb either. It needs SOME load. You can use a different bulb for setting. A pair of 12 bilbs in series would work well since it would be tough to blow that with NORMAL setups. But using the actual bulb you intend to use later would be best since you will be measuring the RMS voltage under load which will allow you to set the duty cycle while taking into account all other resistances.

 

[JimmyM]

Some changes I'll be making to future versions.
1) I'll be decreasing the resistor values used in the soft start / Duty cycle setting circuit.
2) Increase the soft-start capacitor size (to compensate for #1 above.
3) Add additional internal filtering. This will make it less sensitive to switching noise.
4) Increase the size of the FET gate pullup resistor. I was using 3.3K which worked fine, but changed to 1K for faster switching. I'll be moving back to 3.3K (maybe 2.2K). This coupled with the larger internal filter capacitor should make it more immune to switching noise.

I believe this has been the cause of some issues where the duty cycle jumps from ~80% right to ~100% with almost no movement of the pot. It works fine in testing when I sweep the pot from low to high duty cycle, but under load this problem has arisen in a couple of units.