Maybe explanation for not-so-great current output on some EAST-092 drivers?

27 posts / 0 new
Last post
ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US
Maybe explanation for not-so-great current output on some EAST-092 drivers?

Hi,

I’ve been testing several different drivers and types of drivers in preparation for building a new triple.

I also happen to have a number of the “original” EAST-092 drivers, I think from HD-2010s, that I bought from another BLF member awhile ago, so I was testing (actually, re-testing) one of those as part of this “exercise”. I had tried these awhile ago, and one of the things that I wanted to re-check was that I recalled that on high mode, even with good batteries, emitter current was only in the 3.6+ amp range.

Now, the EAST-092 is supposed to be a DD type driver, so it seemed like it should be able to provide higher current, since it should be, I think, similar to the BLF17DD and the “NANJG-092”?

So, I hooked it up, and tested both with several different 18650s, including a couple of -20Rs that I had, and also with my bench supply. These were hooked up to a triple XP-G2 on Noctigon, and sure enough, emitter current never went higher than about 3.6 – 3.7 amps…

Puzzled, I dug out my old scope, and hooked it up to the gate and source pins on the FET (a 70N03), and compared the signal in low, medium, and high mode.

What I found was that in medium and low modes, the signal would toggle between “0” and 3V, whereas in high mode, the signal was steady, but only about 2.5V.

So, it seems like, at least on these EAST-092s that I have, the firmware doesn’t raise the gate signal to 100% in high mode, and that is probably why I am not seeing much higher emitter currents, even with what should be pretty good, high discharge batteries.

I don’t have any of them yet, but, on the other hand, probably the firmware in the BLF17DD and NANJG-092 drivers does raise the gate signal to 100% in high mode?

Does this make sense?

Thanks,
Jim

EDIT: If this is true, then it’s probably time for me to try making some NANJG-092s :)!!

Edited by: ohaya on 10/18/2014 - 15:03
Major
Major's picture
Offline
Last seen: 6 years 7 months ago
Joined: 11/10/2011 - 13:59
Posts: 232
Location: Peak district, UK

Interesting thanks

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 2 years 3 weeks ago
Joined: 07/22/2013 - 13:10
Posts: 4005
Location: Ohio

The voltage on the gate means very little, what did the wave forms look like and what duty cycle %? Can you post pic’s / screen grab’s?

As long as the voltage is within range it is controlled by the duty cycle, lower voltage on high make sence since it would drop under load, that’s ok that it does that, again big thing is overall duty cycle as well as nice clean square sine waves with no spikes.

 RIP  SPC Joey Riley, KIA 11/24/14. Now I am become death, the destroyer of worlds.

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

Cereal_killer wrote:
The voltage on the gate means very little, what did the wave forms look like and what duty cycle %? Can you post pic’s / screen grab’s?

As long as the voltage is within range it is controlled by the duty cycle, lower voltage on high make sence since it would drop under load, that’s ok that it does that, again big thing is overall duty cycle as well as nice clean square sine waves with no spikes.

Per my post in OP:

Quote:
whereas in high mode, the signal was steady

Also, doesn’t the gate-source voltage govern the drain-source resistance?

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

CK,

I don’t know if I’ll be able to get pics. I think I did something to that one driver I was working with earlier when I did the OP, and it won’t change modes any more. I’ve tried another driver, but while I can get that to go to fast blink and SOS, I can’t seem to get it to go to any other mode.

comfychair
comfychair's picture
Offline
Last seen: 5 years 8 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

It depends on the FET's gate characteristics. The SUD70N02-03P & AOD510 have been tested and proven to work with the Nanjg-compatible firmware & hardware. Other FETs have not. Using them hacker-style like this you can't always predict how they will behave just from what's published in the datasheets.

comfychair
comfychair's picture
Offline
Last seen: 5 years 8 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

Jumper B+ direct to the FET's gate while the driver is running on high mode, see if/how much the output increases.

Muto
Offline
Last seen: 4 hours 7 min ago
Joined: 09/04/2012 - 16:42
Posts: 2699
Location: Southeast, PA

About 2 years ago I got my first HD2010, it was a Fandyfire from LightsCastle.
I did a review of it then and it measured 3.8 amps on a 26650, however I had not hardwired my VOM at that time or braided the driver spring.
After doing both of those things, the light pulls just over 5 amps on a good battery. Still does it to this day.

Now 2 years later, I have 1 Ultrafire HD2010 and 2 Tangsfire HD2010’s and none of them pull the amps of the Fandy.
All the newer East-092 drivers seem to have less balls than the old ones.

The only exception to this is the Tangsfire C8 XM-L2 from Tmart from the beginning of this year which will pull over 5 amps With the XM-L2 which is said to be harder to drive than an old XM-L T6.

So that begs the question, where did they source that driver and can it be sourced for individual sale?
Later,
Keith

The difference between Hoarding and Collecting is the illusion of Organization
.
.“I will get one of flashlight from patrol car”

“History doesn’t repeat itself, but it sometimes rhymes,” Mark Twain

After the Apocalypse there will be only 2 things left alive, Cockroaches and Keith Richards

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US
comfychair wrote:

Jumper B+ direct to the FET’s gate while the driver is running on high mode, see if/how much the output increases.

Hi,

I just tried that & there was no change in output/emitter current, measured by clamp meter.

Jim

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US
comfychair wrote:

It depends on the FET’s gate characteristics. The SUD70N02-03P & AOD510 have been tested and proven to work with the Nanjg-compatible firmware & hardware. Other FETs have not. Using them hacker-style like this you can’t always predict how they will behave just from what’s published in the datasheets.

I think that it may be time for me to give in/up, and order one of the BLF17DDs, and see how that works. Maybe there is some special combination.

Also, FYI, the EAST-092s I have came with the 70N03s on them, i.e., I didn’t swap those FETs in.

Thanks,
Jim

TheGloriousTachikoma
Offline
Last seen: 5 years 11 months ago
Joined: 08/22/2013 - 09:23
Posts: 946
Location: Ann Arbor, MI

I makes me wonder why the EAST-092s from Gearbest are different. A properly spec’ed batch or a batch of old stock?

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

Muto wrote:
About 2 years ago I got my first HD2010, it was a Fandyfire from LightsCastle.
I did a review of it then and it measured 3.8 amps on a 26650, however I had not hardwired my VOM at that time or braided the driver spring.
After doing both of those things, the light pulls just over 5 amps on a good battery. Still does it to this day.

Now 2 years later, I have 1 Ultrafire HD2010 and 2 Tangsfire HD2010’s and none of them pull the amps of the Fandy.
All the newer East-092 drivers seem to have less balls than the old ones.

The only exception to this is the Tangsfire C8 XM-L2 from Tmart from the beginning of this year which will pull over 5 amps With the XM-L2 which is said to be harder to drive than an old XM-L T6.

So that begs the question, where did they source that driver and can it be sourced for individual sale?
Later,
Keith

When I got mine, I had to try several times to get an HD2010 from Tmart with a “good” EAST-092. I was fortunate that a couple of folks from Tmart used to post here at the time, and I sent the HD2010 back to them at least once, and maybe twice, before I got one that was good.

If I recall, a similar sort of thing happened when FT start selling the “new” EAST-092s (there’s a thread somewhere) where some people got good ones, and some got not so good ones.

I was told that the EAST-092s that I got (the ones I’ve been testing in this thread) were all the “original” “good” ones, and I guess that 3.6+ amps could be considered “good”, but the ones I tried out of the batch I have just don’t seem to go “true” DD on high mode… so I’ve gone ahead and ordered a BLF17DD and some of the FETs from RMM, and will see how that does.
I just went ahead and ord

comfychair
comfychair's picture
Offline
Last seen: 5 years 8 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

There is no single cell driver that will match the 17DD/SUD70N02-03P combo.

The 'good' east-092s use a DTU06N03 FET, but still have the weird MCU & firmware issues that limit max output.

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

Hi,

I finally was able to get one of the drivers to get to the modes, and took some pics of the gate signal on my scope (sorry the scope is so filthy. I had to dig it out of my basement to do these, and haven’t had a chance to clean it up :():

Time scale (20 usecs/div):

Voltage scale (2 volts/div):

Low mode:

Medium mode:

High mode:

I mis-read the high mode earlier. I think that it’s actually more like 2.2V?

comfychair
comfychair's picture
Offline
Last seen: 5 years 8 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

If jumping B+ to the gate doesn't increase the output, then the gate voltage isn't important. At least, not with that FET on that driver.

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US
comfychair wrote:

If jumping B+ to the gate doesn’t increase the output, then the gate voltage isn’t important. At least, not with that FET on that driver.

Understood.

texaspyro
Offline
Last seen: 2 years 6 months ago
Joined: 04/29/2011 - 12:43
Posts: 4593

Measure the cell voltage while on high. I bet it is being dragged down by the load. You need a cell with better dilithium crystals in it.

wight
Offline
Last seen: 1 year 10 months ago
Joined: 11/27/2013 - 16:40
Posts: 4969
Location: Virginia, USA

Cereal_killer wrote:
The voltage on the gate means very little, what did the wave forms look like and what duty cycle %? Can you post pic’s / screen grab’s?

As long as the voltage is within range it is controlled by the duty cycle, lower voltage on high make sence since it would drop under load, that’s ok that it does that, again big thing is overall duty cycle as well as nice clean square sine waves with no spikes.

I’m not so sure about that… I haven’t done any experiments to prove it, but all the datasheets make it very clear that gate voltage is really critical.

texaspyro wrote:
Measure the cell voltage while on high. I bet it is being dragged down by the load. You need a cell with better dilithium crystals in it.
Big Smile
ohaya wrote:
comfychair wrote:

If jumping B+ to the gate doesn’t increase the output, then the gate voltage isn’t important. At least, not with that FET on that driver.

Understood.

texaspyro makes a good point. What’s the cell voltage under load? If this is a freshly charged 20R I’d expect it to be >2.5v under only a 3A load. Since you’ve got 100% duty cycle you can also just use your DMM to measure the gate voltage in case your scope is crazy…

Still fine, still on a break. One day I’ll catch up with you folks! previous wight catchup Wink
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

wight wrote:
Cereal_killer wrote:
The voltage on the gate means very little, what did the wave forms look like and what duty cycle %? Can you post pic’s / screen grab’s?

As long as the voltage is within range it is controlled by the duty cycle, lower voltage on high make sence since it would drop under load, that’s ok that it does that, again big thing is overall duty cycle as well as nice clean square sine waves with no spikes.

I’m not so sure about that… I haven’t done any experiments to prove it, but all the datasheets make it very clear that gate voltage is really critical.

texaspyro wrote:
Measure the cell voltage while on high. I bet it is being dragged down by the load. You need a cell with better dilithium crystals in it.
Big Smile
ohaya wrote:
comfychair wrote:

If jumping B+ to the gate doesn’t increase the output, then the gate voltage isn’t important. At least, not with that FET on that driver.

Understood.

texaspyro makes a good point. What’s the cell voltage under load? If this is a freshly charged 20R I’d expect it to be >2.5v under only a 3A load. Since you’ve got 100% duty cycle you can also just use your DMM to measure the gate voltage in case your scope is crazy…

With 20R battery:

- Vbat voltage under load: ~4.02V
- Gate voltage using meter: ~3V

wight
Offline
Last seen: 1 year 10 months ago
Joined: 11/27/2013 - 16:40
Posts: 4969
Location: Virginia, USA

This is with the Vishay 70N03? I’ve just glanced at the datasheet and sure enough 3v+ looks like it should be fine.

What kind of current do you see when you bypass the driver (eg true DD)?

Still fine, still on a break. One day I’ll catch up with you folks! previous wight catchup Wink
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

wight wrote:
This is with the Vishay 70N03? I’ve just glanced at the datasheet and sure enough 3v+ looks like it should be fine.

What kind of current do you see when you bypass the driver (eg true DD)?

Hi,

I can’t remember if it’s a Vishay, but it was the one some of us bought from Ebay during one of the older threads about when FT brought back “an” EAST-092/-092B:

http://budgetlightforum.com/node/21657

Re. your last question, good timing :). I was just curious, about that, and had measured it a few minutes ago, and got 6.68 amps (triple XP-G2 on Noctigon). This was with only limited optimizing of the cables (it’s a mess, sorry).

Edit: That was with a -20R battery.

wight
Offline
Last seen: 1 year 10 months ago
Joined: 11/27/2013 - 16:40
Posts: 4969
Location: Virginia, USA

Hmm, that’s a lot of thread to fish through. Instead I glanced at the current pics on FT: they show the ON Semi logo. Maybe this FET? I’m not really confident, just a guess.

Anyway doing a ton of additional work here may be a moot point: once you have some AOD510 or Vishay 70N02 or IRFR3711ZTR (or whatever RMM is sending) in hand it’ll be easy to see whether the FET is the hangup. (Which it certainly is Wink :-p)

Still fine, still on a break. One day I’ll catch up with you folks! previous wight catchup Wink
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

wight wrote:
Hmm, that’s a lot of thread to fish through. Instead I glanced at the current pics on FT: they show the ON Semi logo. Maybe this FET? I’m not really confident, just a guess.

Anyway doing a ton of additional work here may be a moot point: once you have some AOD510 or Vishay 70N02 or IRFR3711ZTR (or whatever RMM is sending) in hand it’ll be easy to see whether the FET is the hangup. (Which it certainly is Wink :-p)

LOL Smile you don’t speed read?

Yes, we shall see…

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US

ohaya wrote:
wight wrote:
Hmm, that’s a lot of thread to fish through. Instead I glanced at the current pics on FT: they show the ON Semi logo. Maybe this FET? I’m not really confident, just a guess.

Anyway doing a ton of additional work here may be a moot point: once you have some AOD510 or Vishay 70N02 or IRFR3711ZTR (or whatever RMM is sending) in hand it’ll be easy to see whether the FET is the hangup. (Which it certainly is Wink :-p)

LOL Smile you don’t speed read?

Yes, we shall see…

Ahh. Hopefully the link’ll still work, but I found the item in my Ebay watch list:

http://www.ebay.com/itm/251144695278?_trksid=p2055119.m1438.l2649&ssPage...

5 PCS NTD70N03 TO-252 T70N03 70N03 ON N-Channel Enhancement Mode MOSFE

wight
Offline
Last seen: 1 year 10 months ago
Joined: 11/27/2013 - 16:40
Posts: 4969
Location: Virginia, USA

Yes, I can see the page. The FET does look similar to what FT shows in their current pic. It’s definitely ON Semi. (It’s still quite possible that I’ve linked to the wrong datasheet though.)

Still fine, still on a break. One day I’ll catch up with you folks! previous wight catchup Wink
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

ohaya
Offline
Last seen: 1 month 5 hours ago
Joined: 03/16/2013 - 19:01
Posts: 5332
Location: US
wight wrote:
Yes, I can see the page. The FET does look similar to what FT shows in their current pic. It’s definitely ON Semi. (It’s still quite possible that I’ve linked to the wrong datasheet though.)

I think that this is the part of that older thread where that Ebay item was referenced (by moderator007):

http://budgetlightforum.com/node/21657?page=2#comment-473532

wight
Offline
Last seen: 1 year 10 months ago
Joined: 11/27/2013 - 16:40
Posts: 4969
Location: Virginia, USA

My glance at the datasheet didn’t make them look that bad, but as comfychair pointed out we aren’t exactly using them in their intended role either. (EDIT: ohaya is referencing post #108 in that thread)

Still fine, still on a break. One day I’ll catch up with you folks! previous wight catchup Wink
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)