Is that a synchronous boost converter controlled by just an attiny? Pretty slick if you can pull it off. That gives me hope for a high power buck boost driver.
Thanks for the schematic. I was wondering what F2 was being used for but still don’t see the need for it.
I believe the original driver switched at 300khz, it would be nice if that could be increased. Maybe replacing F1 with a Schottky diode might be easier than trying to figure out the band gap between F1 and the pfet. The band gap may need to change with load and temperature, and there could be some shoot through if the attiny gets too busy.
The attiny 1626 and 1627 are almost available. 20mhz clock speed and 12 bit adc would be nice for this application.
I was looking at a micro-driven switching controller a couple weeks ago but, with the chips I have available, couldn’t get a sufficiently high switching speed and decent resolution at the same time due to hardware PWM limitations. 6.8uH inductance probably helps with that; I wanted to stay under about 2uH to get high saturation and low DCR in a compact package for a 17mm driver. Slick use of the analog switch as a gate driver too.
The 1616 has a 10 bit adc but it may not matter if you are using an op amp. The little bit I have read about the two series claims very fast adc conversions and a more powerful cpu. Either one would be an improvement on the old 85.
My biggest concern is the attiny could be spending too much time on other tasks and not spending enough time on the boost converter. If this thing works and the switching speeds can be increased it opens the door to higher power versions and maybe buck boost as well.
Its impressive that you have got it down to 18mm. Is there a reason you dont want to go double side?
This design could easily fit 17mm if you stuck some of the smaller components on the oppisite side. Just leave the center for positive to solder on a copper stud or small spring.
In time, I think Aundril will be adapted to work with boost. There’s a few members working on it.
17 mm is considered the most common, across the board. But I've been seeing 20, 21, 22 mm's pretty often lately. Going smaller, AA size, 15 mm would be the next smaller size, I would consider AA size. But of course there are exceptions to every rule or assumption .
17mm is probably the most common as Tom E said. The easiet way is to make it as small as you feel comfortable working with then just change the pcb size to accomodate different size lights. That way the driver can fit many lights just by changing the pcb size in the design without having to go thru the process of designing a new layout.
I do not know any popular light with e-switch and 17mm driver. Why go small? Better change host with more space ;)). Tomorrow last hope. I will get FET with same specs only more powerful as is in original Tamagoshi schematics.
On BLF any light can become a host. There is plenty of lights out there that use 14500/AA with e-switch and a lot of 16340 lights also. Its your design, you design it as you see fit.
I have a Ultratac A1 using this driver with a XHP50.2 making 2000 lumens with a 14500 very floody. Its not a e-switch but I could have used a Sofirn SP10S, Thorfire TK05, or a Utorch UT01 for a host if the driver had been eswitch capable.. A XHP35.2 might be interesting in one of these host.
More info please.
It’s boost driver for 1 cell and 6V led I guess?
Is it programmable with blf firmware?
Awesome…..I guess?
We shall see.;))
If it’s as fun as the other drivers I got from you then count me in!
Quadrupel your pleasure
Quadrupel your fun
It's the right one
The Quadrupelmint gum
How to Post Images on BLF // Many knives for sale (USA only)
How to move a thread
Is that a synchronous boost converter controlled by just an attiny? Pretty slick if you can pull it off. That gives me hope for a high power buck boost driver.
What kind of switching frequency do you expect?
Thanks for the schematic. I was wondering what F2 was being used for but still don’t see the need for it.
I believe the original driver switched at 300khz, it would be nice if that could be increased. Maybe replacing F1 with a Schottky diode might be easier than trying to figure out the band gap between F1 and the pfet. The band gap may need to change with load and temperature, and there could be some shoot through if the attiny gets too busy.
The attiny 1626 and 1627 are almost available. 20mhz clock speed and 12 bit adc would be nice for this application.
I was looking at a micro-driven switching controller a couple weeks ago but, with the chips I have available, couldn’t get a sufficiently high switching speed and decent resolution at the same time due to hardware PWM limitations. 6.8uH inductance probably helps with that; I wanted to stay under about 2uH to get high saturation and low DCR in a compact package for a 17mm driver. Slick use of the analog switch as a gate driver too.
So, ordered PCB and components. Coils same size 7×6.5mm, 3 values: 5.6, 6.8, 8.2uF and caps 22, 47 and 100uF for experimenting and tuning.
How are they better than 1616?
The 1616 has a 10 bit adc but it may not matter if you are using an op amp. The little bit I have read about the two series claims very fast adc conversions and a more powerful cpu. Either one would be an improvement on the old 85.
My biggest concern is the attiny could be spending too much time on other tasks and not spending enough time on the boost converter. If this thing works and the switching speeds can be increased it opens the door to higher power versions and maybe buck boost as well.
Just took a closer look…I really like it.
I wonder what kind of power can it deliver and at what efficiency.
It would be better in 20mm size and 2.5A @12V
WTB Titanium 4sevens 2xAA tube
I guess efficiency should be about 80-85% and power.. maybe 20W?. Ordered few 20mm too.
To be honest I find the efficiency underwhelming…what makes it so much less efficient than say GXB172?
50mOhm series resistor and the rectifier drop, to start. Without a synchronous rectifier, conduction losses really pile up.
Hahahaha, what was that all about
Yes, yes the best is synchronous rectification , but Ill be very happy with “only” 85% eficiency
It's a (modified) commercial jingle for Wrigley's Doublemint gum.
How to Post Images on BLF // Many knives for sale (USA only)
How to move a thread
Great job! Looks like getting it down to 17mm is possible.
Awaiting testing data to see how these perform. Very interesting design.
V11R and V10R Current Mod
Now components is in 18mm circle all in 1 side, so 20mm possible, but 17mm … 1 side … hmmm ;)) Boost driver without big coil wont work ;))
Am I the only one who gets the Big Ass Spider reference, or am I just late to the game?
Its impressive that you have got it down to 18mm. Is there a reason you dont want to go double side?
This design could easily fit 17mm if you stuck some of the smaller components on the oppisite side. Just leave the center for positive to solder on a copper stud or small spring.
In time, I think Aundril will be adapted to work with boost. There’s a few members working on it.
V11R and V10R Current Mod
Which e-switch flashlight using 17mm driver? Another question. Which e-switch drivers diameter most common and popular and desirable?
Most desirable are currently for 14500/AA size if you ask me.
WTB Titanium 4sevens 2xAA tube
14500/AA…. Its not a driver size
17 mm is considered the most common, across the board. But I've been seeing 20, 21, 22 mm's pretty often lately. Going smaller, AA size, 15 mm would be the next smaller size, I would consider AA size. But of course there are exceptions to every rule or assumption
.
17mm is probably the most common as Tom E said. The easiet way is to make it as small as you feel comfortable working with then just change the pcb size to accomodate different size lights. That way the driver can fit many lights just by changing the pcb size in the design without having to go thru the process of designing a new layout.
V11R and V10R Current Mod
I do not know any popular light with e-switch and 17mm driver. Why go small? Better change host with more space ;)). Tomorrow last hope. I will get FET with same specs only more powerful as is in original Tamagoshi schematics.
On BLF any light can become a host. There is plenty of lights out there that use 14500/AA with e-switch and a lot of 16340 lights also. Its your design, you design it as you see fit.
I have a Ultratac A1 using this driver with a XHP50.2 making 2000 lumens with a 14500 very floody. Its not a e-switch but I could have used a Sofirn SP10S, Thorfire TK05, or a Utorch UT01 for a host if the driver had been eswitch capable.. A XHP35.2 might be interesting in one of these host.
V11R and V10R Current Mod
Pages