The Texas Buck driver series, Q8 / Skyray King 2S/4S buck driver RELEASED!

467 posts / 0 new
Last post
Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

So I'll work on the last changes to the shopping cart soon, but what about about heat sinking?  There is no exposed copper on the edge.  I don't have an SRK.  If the copper there was exposed could it be made to contact the case?  Maybe with some thermal grease at least?

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas
Flintrock wrote:

I’m a bit surprised you left the Ruv’s in.  It won’t hurt though.  I guess you’re worried we won’t ever figure out R1?  Anyway, nice job.  I will try to dig through source and code and attiny manual code and if I can get to the bottom of the R1 thing.  

I left then in because taking them out didn’t free up any edge space and I am always a fan of cramming every possible option into something if there are negatives.

Look it over carefully, I will give it a day or 3 to let us all figure out any possible issues and then I will put together an official release.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas
Flintrock wrote:

So I’ll work on the last changes to the shopping cart soon, but what about about heat sinking?  There is no exposed copper on the edge.  I don’t have an SRK.  If the copper there was exposed could it be made to contact the case?  Maybe with some thermal grease at least?

Yeah, I was debating this myself. Obiously we don’t want it grounded to the case and shorting out but as long as the tailcap side is not shorted to the case then it should be fine. It is easy enough to expose the copper on the edge like normal.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Can't the outer edge of the tailcap be ground?  connected directly on tail-cap side to ground pad, and by vias to outer ring on component side?  Ground touching the case should be fine, right?

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

Ground touching the case is fine, the issue is anything from the tailcap also touching the case will cause a short.

As long as the tailcap is isolated by using the matching multi-cell tailcap properly setup for this driver, it should be fine.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

of course I meant tailcap side of this board.  I think you got that.  I think we want all the heat sinking we can get.  

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas
Flintrock wrote:

of course I meant tailcap side of this board.  I think you got that.  I think we want all the heat sinking we can get.  

I agree, I will take care of that before releasing it.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

I found this in bistro:

    // 1.1v reference, left-adjust, ADC3/PB3
    ADMUX  = (1 << V_REF) | (1 << ADLAR) | CAP_CHANNEL;

 

And 17.13.1 of the manual explains the bits clearly, so:

 

ADMUX =(1 << V_REF) | (1 << ADLAR) | CAP_CHANNEL | 1<< 4;

 

Would set the 2.56V reference for attiny 25 and attiny 8, but really should add a define, and might be different in attiny13, and needs changing in a couple of places and might interfere with temperature readings (but I don't think so, it's got separate code), etc... So doable, but not quite just flipping a switch.

 

Anyway that line and the details behind it confirms that the way things are setup now is for 1.1v, so I agree, that's what we should use for now at least.

 

The only question then is what total resistance to aim for for acceptable drain current.  is 95k enough or should we double R2 and bring it all up to the 200k range?  I don't like 900k.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

And the reason it's setup for 1.1V is 2.56V cannot be used if Vcc drops below 3V.  So 1.1V is more universal.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

Ah, that would explain why the 2.56v is used and I think we should keep using it for that reason. We don’t need more accuracy then that honestly.

Far as resistance the parasitic drain is much lower the higher the resistance goes. I think the 200k+ range would be a good goal.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Texas_Ace wrote:
Ah, that would explain why the 2.56v is used and I think we should keep using it for that reason. We don't need more accuracy then that honestly. Far as resistance the parasitic drain is much lower the higher the resistance goes. I think the 200k+ range would be a good goal.

 

Well sure it is, exactly inverse proportional.  Ok, I think 11k and 220k was good when I checked.  No time to run the numbers now, but something very close to that is good.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

We missed one.  According to your LDO TA series docs, C2 should be 10uF when using an LDO, which makes sense.    This one at least is low voltage, so really doesn't need to be physically larger.    This is why we waited a few days.  Updates coming soon.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas
Flintrock wrote:

We missed one.  According to your LDO TA series docs, C2 should be 10uF when using an LDO, which makes sense.    This one at least is low voltage, so really doesn’t need to be physically larger.    This is why we waited a few days.  Updates coming soon.

Good catch, I have dealt with too many variations in the designing of these drivers to keep track of my own notes lol.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Yeah, I've just had enough of parts lists for a couple of days.  I'll find a new cap and edit it all in a day or two.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas
Flintrock wrote:

Yeah, I’ve just had enough of parts lists for a couple of days.  I’ll find a new cap and edit it all in a day or two.

No problem, I am swamped with things myself so it will most likely by late this week or this weekend before I get around to releasing it anyways.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

 Well rats, so yeah, C2 is only available in 6.3V X5R in a 10uF 0603.  Price isn't bad at 0.13 per one.  Your usual comment TA is that an 0805 will fit there.  Is that still true in this one?  It looks possible to slide C1 north about a mm, if it goes back to not jumping the trace (trace turns between c1 and c2).  Is there room to squeeze an 0805 on as it is?

 

On the other hand, X5R is probably ok.  There's an X7t also.  Not sure which is better, probably makes little difference.  In the long run, it looks like C1's are headed toward 0805 with this new OTC trick.  sigh.  There is a way, to even get the footprint, moving the trace between them, and then sliding C1 left a bunch.  I know you don't want to make that a long trace.  If we can just have enough clearance to fit an 0805 on the 0603 pad, that's good enough too though.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

An 0805 fits on the 0603 pads, overhanging just slightly.

Although in this case I should be able to change it to an 0805 pad.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Texas_Ace wrote:
An 0805 fits on the 0603 pads, overhanging just slightly. Although in this case I should be able to change it to an 0805 pad.

 

Ok, I knew it would fit the pads, but was worried there wasn't room here for the overhang.   I'm thinking of dropping the clickie R1/R2 reccomendations for now since a) there's no clickly application for this in its current form I guess, b) the new OTC is going to make changing the resistors irrelevant anyway if it pans out.  I wrote it up just for it to be there for later, but now it's not clear what later will look like anyway.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

R1/R2 should be setup for the LVP first and everything else second. I agree the new off-time setup should really improve things.

Come to think of it, it should also fix/vastly improve the issues with lighted tailcaps as well.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

There's nothing wrong with the clicky R1/R2 10x lower values. They just aren't going to be needed anymore if OTC gets offloaded.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

Ah, I see.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

Ok, 0805 C2 and ground ring:

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Awesome, new parts table coming soon.  I was starting on it when I ran into that.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Is the battery side ground ringed too? Thermal vias between the two sides might even help.  

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas
Flintrock wrote:

Is the battery side ground ringed too? Thermal vias between the two sides might even help.  

I thought about that but just seemed like it was too much risk for short circuiting if there was any mis-alignment.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

I don't understand.  If the case hits a battery pad that's a short with or without the ground ring there.  The case is going to be at battery ground now either way.  Could just make the battery pads a little smaller.

Texas_Ace
Texas_Ace's picture
Offline
Last seen: 19 hours 48 min ago
Joined: 03/24/2016 - 07:44
Posts: 8639
Location: Everything is brighter in Texas

I should but that requires doing a fair amount more work I didn’t want to do lol.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Texas_Ace wrote:
I should but that requires doing a fair amount more work I didn't want to do lol.
 That's fair, but I don't think leaving off the ground ring really helps the problem.  It's the case that will short.  It's already at ground now, or hopefully is anyway.  Of course even that, I don't know if it's as simple as hitting the fill bucket icon and selecting copper and clicking.   I've never used the software. 
DB Custom
DB Custom's picture
Offline
Last seen: 6 hours 15 min ago
Joined: 01/13/2013 - 22:28
Posts: 20562
Location: Heart of Texas

I see no less than 8 places where traces have 90º turns. I’m no electronics engineer, but I’ve seen the guys who are argue this many times, no 90º turns.

Edit: Oops! My bad, 10 places.

Flintrock
Offline
Last seen: 2 years 3 weeks ago
Joined: 09/10/2016 - 20:29
Posts: 1544

Doesn't seem very open and shut to me at worst:

http://electronics.stackexchange.com/questions/226582/pcb-90-degree-angles

 

Some other discussions suggest the most practical reason for this may be that some older hardware and thus software didn't deal with creating these well.

 

Pages