Easy OKR-T VV mod
- Thread starter bstedh
- Start date
- th_trl_thread_readers 0
- Status
Why did you choose 200Ω which means it's running out of spec? From the datasheet.
With a stacked pair of Li-Ions there isn't enough voltage headroom for 6+V anyway, even 5.5V is pushing it.
A 240Ω would be a more appropriate minimum.
As far as input caps, it's the cells they are helping, not the IC. These PWM driven devices basically pound away at them. We tend to overwork the cells anyway, a little buffer/cushion will lessen the beating and hopefully not add to the premature aging we get.
That's the same route I'm going, single cell boost converter.
I've had a lot of trouble getting the efficiency I want out of a buck/boost configuration. I tried it with a SEPIC topology using this controller and couldn't get it over 80% at 10 Watts.
The simplest option is to use two cells with a buck converter, you get the efficiency, simplicity, and full voltage range. The only downside is you have two cells which is a biggie for me. I want to carry on-board charging and it gets real tricky with multiple cells.
Another option is to use a buck/boost converter than uses 4 FETs to switch between buck and boost mode. That's probably the most efficient method, especially since it's synchronous, but those kind of controllers have limited availablility and require a lot of circuit board real estate because of all the high current FETs.
At this point, I'm just going to stick with the boost topology and be happy with the 4 to 6 volt range using a single cell. I wanted a 3.3 to 6V range, but like I said, I've had too much trouble with the buck/boost topology.
If someone comes out with a buck/boost module that has high efficiency (+90%), input voltage range suitable for a single Li-Ion batt, and suitable output current capacity, that would be quite a find. That makes it easy to carry on-board charging. There's much higher availability for low current stuff, much easier to design too, but with e-cigs we need those high currents which can really make things difficult.
To get catch up with this thread, that CNC method for making PCBs looks really cool, but pricey. How much does it cost?
I've been making my own PCBs for years now. In the old days, before laser printers, I used to lay out my etching paterns in tape and sticky pads designed for that purpose. The laser printer transfer method was a revolution for me, a world of difference. You can use any drawing program to make your patterns, but one that prints true size is going to work best. I use an inexpensive CAD program myself.
I use paper for my transfers. It's Office Depot Photo Basic Gloss, item #471865. The toner goes on the dull side, not the shiny side. The paper is good because it's easy to iron and and seperates easily from the toner when soaked it in water. Once it's saturated, you just peel it off. A little rubbing with your thumb under running water and a soft brush removes any that sticks behind.
When I etch my boards, I use nothing other than a shallow level of Feric Acid in a plastic container. Sometimes a little light brushing with a soft brush to speed it up, but it's not necessary. Though Feric Acid is like ink if you spill it on anything. The alternative chemicals you can use are less messy, but the Feric Acid works really fast and easy. I'm just careful not to spill it on anything.
I get the Feric Acid in powdered form. You just add water to make the etching solution. It gets hot when you add water to it. Definitely don't want to mix it up all at once in a plastic bottle, almost melted one that way. You have to add a little at a time and let it cool off.
I also tend to drill my holes before etching. Though it's usually not many because I use SMD components as much as possible. Usually, it's just a few holes for wire connections or maybe a through hole switch or something else like that.
Last edited:
The Buck converter is the winner in efficiency.
The sepic design I've been vaping with for the past 7 months is running about 85% efficient in the 8-10W range across the full input range (2.7-18V). Ideal inductors just took around 4 months to arrive and I haven't had a chance to test them yet since I've been building the CNC machine to make more boards. I am looking to gain a couple of % with this double sided board and a couple more % when I go to a multilayer.
The total cost I have invested in the CNC was around $2,000 including pcb routing tooling. They can be made for cheaper, and there are kits that can be done for under $1,000. I made sure I can mill aluminum housings out of it too.
I have probably spent twice that on components and other modding supplies, but it's still cheaper than smoking.
The sepic design I've been vaping with for the past 7 months is running about 85% efficient in the 8-10W range across the full input range (2.7-18V). Ideal inductors just took around 4 months to arrive and I haven't had a chance to test them yet since I've been building the CNC machine to make more boards. I am looking to gain a couple of % with this double sided board and a couple more % when I go to a multilayer.
The total cost I have invested in the CNC was around $2,000 including pcb routing tooling. They can be made for cheaper, and there are kits that can be done for under $1,000. I made sure I can mill aluminum housings out of it too.
I have probably spent twice that on components and other modding supplies, but it's still cheaper than smoking.
Last edited:
I pretty much would agree with you on this one. I never get above about 4.7V, sometimes I may hit 5V if I'm trying to clean off a coil.
Also the more I look into it the more I'm thinking of installing a cap on the input leg. I am just keeping it as simple and strait forward as possible and the components to a minimum. I can definitely come up with way more efficiency but the more efficient you get the more it's going to cost in money and real estate. At least from my research so far.
The capacitance included on this OKR converter is adequate, but I've been adding a 10uF on the input when I have them available, it can't hurt especially since you are not using the shutdown pin and doing a full power-up with each hit.
Specs where written to be broken =] Haven't you ever over clocked your cpu? Seriously though you do have a point. Even though I do run most everything beyond spec. Sometimes it's fun just to see how far you can push something before it cracks.... Also "most" specs have a certain amount of head room built into them to cover the manufacturers ..... Like Scotty says - under promise and over deliver and you will always be a hero.
Ok, I have a question about the placement of the input capacitor.
In some of the schematics I have looked at, the input capacitor is being placed after the switch. This is on models that are not using the control pin.
Shouldn't it be prior to the switch to buffer the initial charge up of the regulator? If it's placed after the switch it would actually buffer the off cycle not helping the battery at all. In fact it would be increasing the initial hammer on the battery.
If you scroll to the buck/boost circuit referenced in this thread you will see what I mean.
Also I am assuming that if you use a big enough capacitor you would need a resistor in series to keep from zapping the regulator. I think the ones that are being discussed however are not big enough to do this.
Along the same line of though - What is the output capacitor doing? Isn't it actually increasing the initial output load of the regulator as the capacitor has to charge when you first power up and then it discharges when shutting down?
I am assuming the heaviest load times are initial power on. They will help with transient spikes in demand during the on period but I am thinking this would be minimal compared to the initial start up. If we where dealing with a sensitive load I could see this being helpful but the Atty doesn't care if there are ripples in the power it receives.
I guess what I am getting at is that if we are mostly concerned about protecting the battery then we should have a high capacity capacitor in parallel with the battery to help it out by reducing the sharp on/off power surge. I really could care less about ripples in the supplied current.
In some of the schematics I have looked at, the input capacitor is being placed after the switch. This is on models that are not using the control pin.
Shouldn't it be prior to the switch to buffer the initial charge up of the regulator? If it's placed after the switch it would actually buffer the off cycle not helping the battery at all. In fact it would be increasing the initial hammer on the battery.
If you scroll to the buck/boost circuit referenced in this thread you will see what I mean.
Also I am assuming that if you use a big enough capacitor you would need a resistor in series to keep from zapping the regulator. I think the ones that are being discussed however are not big enough to do this.
Along the same line of though - What is the output capacitor doing? Isn't it actually increasing the initial output load of the regulator as the capacitor has to charge when you first power up and then it discharges when shutting down?
I am assuming the heaviest load times are initial power on. They will help with transient spikes in demand during the on period but I am thinking this would be minimal compared to the initial start up. If we where dealing with a sensitive load I could see this being helpful but the Atty doesn't care if there are ripples in the power it receives.
I guess what I am getting at is that if we are mostly concerned about protecting the battery then we should have a high capacity capacitor in parallel with the battery to help it out by reducing the sharp on/off power surge. I really could care less about ripples in the supplied current.
Last edited:
Never mind most of my post.... Pulse modulated... We are evening out the draw on the battery...I need to quit asking questions in the middle of the night! I wind up way over thinking things and missing the basic point. =]
However, I do think putting a capacitor parallel to the battery prior to the switch is valid just that one big enough to do any real good on start up would be size prohibitive.
Last edited:
Talking about unwieldy caps, one thing about boost versus buck, you need high capacitance and low ESR on the output with boost. When you get into the higher frequencies we use with a switching converter (~500kHz), the capacitor's ESR makes just as much of a difference as anything else. You need a certain amount of capacitance to start with, but after that, ESR has the biggest effect on ripple voltage. Unfortunately, a side effect of reducing ESR is reduced stabilty. The output capacitor's ESR dampens the ringing you get from the LC filter formed by the output capacitor and the power inductor. Get's tricky to stabilize. I've been using these expensive Tantalum Polymer SMD caps. The output capacitors on these boost converters have always been a headache, ridiculously expensive when you need low ESR and high capacitance in a small package.
I think I got away with cheaping out on the input caps. MLCC caps are small, inexpensive, and have very low ESR by nature so I'm running a couple 10µF 1206 packages in parallel as input capacitors. Its low on capacitance, but ultra low on ESR. It seems to do the job. I'm only getting about 20mV input ripple which is good enough and I save a whole bunch of space and cost that way.
I think I got away with cheaping out on the input caps. MLCC caps are small, inexpensive, and have very low ESR by nature so I'm running a couple 10µF 1206 packages in parallel as input capacitors. Its low on capacitance, but ultra low on ESR. It seems to do the job. I'm only getting about 20mV input ripple which is good enough and I save a whole bunch of space and cost that way.
Nice layout bstedh!
I had looked at using this regulator a while back but nixed it since I vape at 5.4 - 5.5v.
So I'm curious - are you having any problems with the Vin needing to be 2v or higher than Vout? I'm guessing if you vape under 5v you wouldn't have a problem with the regulator cutting out before the batts need a charge. With vaping at 5.5v, I'd have to maintain a total batt voltage above 7.5v (or about 3.8v individually).
I had looked at using this regulator a while back but nixed it since I vape at 5.4 - 5.5v.
So I'm curious - are you having any problems with the Vin needing to be 2v or higher than Vout? I'm guessing if you vape under 5v you wouldn't have a problem with the regulator cutting out before the batts need a charge. With vaping at 5.5v, I'd have to maintain a total batt voltage above 7.5v (or about 3.8v individually).
It basically just cuts off sooner so your vape time is decreased. I use 2.0 ohm or lower cartomizers so I never really need the higher voltages.
This isn't a regulator, so LDO is not applicable. 2 V higher than Vout is a margin for maximum efficiency.
This was discussed on the evercool thread
Last edited:
Thanks. I knew there would be a reduced vaping time with the 2v difference. Was just wondering if it affected the regulator in any other way with that particular spec relative to running it at higher output voltages.
Not a regulator?? LOL. I understand what you're attempting to say, but I would think even for switching regulators lower input–output differential voltage is certainly applicable. Who says max efficiency must be a margin of 2v??
If I compare the 2 switching regulators - the PTR08100W that blue uses in the evercool mod with this OKR-T regulator, and based on the fact that I like to vape around 5.5v, it appears the PTR08100W regulator is better equipped to give me a longer vaping time since it has a lower input–output differential (minimum input voltage = Vout/0.83) than the OKR-T (minimum input voltage = Vout + 2v).
PTR08100W regulator: output 5.5v, min input 6.6v
OKR-T regulator: output 5.5v, min input 7.5v or higher
Or am I missing something here?
The only downside to the PTR08100W regulator is it maxes at 5.5v. Although breaktru broke that barrier and maxed it at 5.66v.
Last edited:
You are missing something here.
This converter does not require 2v higher voltage to operate. As I tried to point out in that post, this converter requires ~.5V overhead (6VIn for 5.5Vout) but efficiency will drop from ~97% to ~90%. The same thing happens with the TI converter.
I've done a side by side comparison, I recommend you do the same and come to your own conclusions.
To do a side by side comparison:
TI minimum input voltage = Vout/0.83
OKR minimum input voltage = Vout/0.86
OKR will give you longer vape times
Last edited:
Yes, I've come to my own conclusions. The datasheet does say 2v. So obviously I missed the fine print.
LOL - just yankin your chain... err "converter".
My teachings have revealed that “Wisdom” derived from believing all that you read is true - is not wisdom at all. True Wisdom is, and can only be derived from, firsthand experience and the knowledge gained from that experience.
Prove it for your self – and only then, will you Truly Know………
Prove it for your self – and only then, will you Truly Know………
You're right - I will do just that. I've been trying to decide which switching regulator I want to put in my next mod, so instead of trying to choose between the two I've ordered both the TI and Murata regulators and will give em a side-by-side comparison from a "user" perspective.
If what Cape is saying is spot on, I'm thinking there won't be a BIG difference between the two, so will be interesting to actually find out.
I have found that this regulator setup is running one of the batteries to its undervoltage limit and the protection circuit built into the battery is shutting it down. At first I thought the regulator itself was shutting down due to lack of sufficient input voltage. I am going to look into a voltage cut out circuit so the device itself has a protection circuit. For now I will have to remember to switch out batteries before they die.
- Status
Similar threads
- Locked
