Tinkering with the OKL2-T20 20A 110W dc/dc converter...

[mamu]

Got the OKL2-T/20-W12 in yesterday and got it breadboarded this morning.

So far, so good with testing (i.e., it's doing what it's suppose to ).

Size comparison with the Naos Raptor 20A 120W board...

Min output voltage (I'm using a 1K ohm POT + 1.43K ohm resister - for a lower min voltage (~3.0v), switch to a 1.5K ohm POT)...

Max output voltage...

Testing the P-FET for reverse polarity... good to go with protection...

Wiring guide...

External parts that I have breadboarded with the OKL (all bought at mouser.com unless otherwise indicated):
20K ohm resistor (CMF5520K000FKEK)

1.43K ohm resistor (RN60D1431FB14)

2x 22uF input caps in parallel (C3225X5R1C226K250AA)

1K ohm POT (PT10MH01-102A2020) (use 1.5K ohm POT for min voltage at ~3v)

2x 10A fuses in parallel (RGEF1000)

P-FET (SUP75P03-07-E3)

1S-6S voltage reader (ebay)

3-pos slide switch (can't recall the part no - it's a mini C&K 3-pos slide switch)

 

[david4500]

Congratulations on successfully getting a new convertor working.

I believe I have a link for your on-off-on slide switch:

1103M2S3CQE2 C&K Components | Mouser

Here is a very similar alternative that I recently found with a nice lower price:

48ASSP3S1M2RBT Grayhill | Mouser

 

[mamu]

Note: I just realized I had pins 1 and 2 mislabeled in my wiring guide. That's been corrected.

Also, I'm using the OKL2-T/20-W12N2-C version.

Congratulations on successfully getting a new convertor working.

I believe I have a link for your on-off-on slide switch:

1103M2S3CQE2 C&K Components | Mouser

Here is a very similar alternative that I recently found with a nice lower price:

48ASSP3S1M2RBT Grayhill | Mouser

Thanks, David! Good find for the switch.

 

[mamu]

ok... just had a duh-oh moment while tinkering with the breadboarded setup some more - with this board I tied the switch to GND, with a zener diode as a low voltage cutoff inline with GND like this that's a no-go. Removed the zener and corrected the wiring guide.

 

[Phone Guy]

Cool.... I can smell the solder of new box mods all around me

 

[cturboaddict]

I ordered a few. Going to follow along here and build my own. I prefer things that way anyhow.

 

[mamu]

After a few days of tinkering and testing with a breadboarded setup off and on, I've run into a few glitches.

The loaded output voltage is now higher than the unloaded output voltage. At max output, loaded voltage is 5.9v - no load voltage is 5.6v. The load/no load voltages at lower output voltages aren't as extreme as this, but are still reading higher than no load voltage.

Load voltage higher than no load voltage just doesn't make sense.

Although I was getting no output voltage drop under load during all this testing, I wired up a second OKL and this time connected the +sense and -sense pins. Dialed the pot to max at 5.6v no load, connected an atty and the loaded voltage shot up to 6.1v with the 1st firing, then fluctuated between 5.8v - 6.0v with repeated firing of the fire switch. It finally settled in at 5.9v.

Other than the strange initial voltage fluctuations, no difference there with the Sense pins connected vs not connected. So the Sense is failing to compensate for the unwanted voltage increase - was curious if it would.

So... yeah I'm shaking my head by now and thinking well crap - nothing has changed here.... oh wait - the input voltage has gone from 8.4v no load to now 7.3v no load over these last few days since I initially breadboarded the OKL.

Replaced those batts with a freshly charged set and now no load voltages match load voltage. So then connected the 12v power supply and yup - no load match load voltage.

Put the 7.3v battery set back in and got the weird higher loaded voltage of 5.9v vs no load voltage of 5.6v. Same results with and without the Sense pins.

Am letting this set of batts run down to 7v, then 6.5v, then 6.0v, to see what this converter does with load vs no load output voltages with lower input voltages.

This is a summary of what I'm experiencing at max output voltage:
OKL #1 - without Sense pins connected:
Input voltage-----------Output no load----------Output load
7.3v-----------------------5.6v---------------------5.9v
8.5v-----------------------5.6v---------------------5.6v
12v---------- -------------5.6v---------------------5.6v

OKL #2 - with Sense pins connected:
Input voltage-----------Output no load----------Output load
7.3v-----------------------5.6v---------------------5.9v
8.5v-----------------------5.6v---------------------5.7v
12v------------------------5.6v---------------------5.6v

What *appears* to be happening is that the OKL is not stable with regulating output when input voltage is < 7.5v. This is not good.

 

[mamu]

ok... been tinkering again and *think* I've now got a stable OKL.

Although the datasheet says input and output caps are optional and not necessarily needed, I feel they are necessary for a stable input and output for our application. The datasheet says if using caps, use 44uF for input and 188uF for output. You're not going to find 44uF MLCC caps or 188uF MLCC caps, so the best option is to parallel a lower value to add the capacitance for the total we need.

Also, with my testing without an output cap it makes no difference whether the Sense pins are connected or not - I'm not getting a stable output with or without the Sense when input voltage drops to <7.5v. It is important then to use both an output cap and connect the Sense pins to get a stable output voltage else you'll be back to an unstable output under load.

Added 4x 47uF 16v caps (soldered together in parallel) to the output, connected +Sense to Vout and connected -Sense to GND and am now getting stable voltage out under load no matter the input voltage level (min voltage should still be ~6v).

I'm still going to be doing a few more days of testing, but hopefully this wiring is now good to go. I've updated the OP with the latest wiring guide, but can't edit the parts list to add the output caps.

Here's the parts I've now got breadboarded in addition to the OKL:
20K ohm resistor (CMF5520K000FKEK)

1.43K ohm resistor (RN60D1431FB14)

2x 22uF 16v input caps in parallel (C3225X5R1C226K250AA)

4x 47uF 16v output caps in parallel (GRM32ER61C476ME15L)

1K ohm POT (PT10MH01-102A2020) (use 1.5K ohm POT for min voltage at ~3v)

2x 10A fuses in parallel (RGEF1000)

3-pos slide switch (1103M2S3CQE2)

P-FET (SUP75P03-07-E3)

1S-6S voltage reader (ebay)

This has been a rather *fun* project and a great learning experience for me (I say that with a bit of sarcasm) but I think we're good to go. I'm checking in to adding a voltage detector for low voltage protection for the batts per Craig's recommended parts (thanks Craig!) so will update that when I get the parts in.

 

[CapeCAD]

Also, the function for the Sense pins is not described in the datasheet - just a blurb regarding pin configuration. With my testing without an output cap it makes no difference whether the Sense pins are connected or not - I'm not getting a stable output with and without the Sense when input voltage drops to <7.5v. With an output cap, it is very important to connect the Sense pins to get a stable output voltage else you'll be back to an unstable output under load.

Hi Mamu,

The plus and minus sense pins are used to sense the output voltage at a specific place. You can tie them directly at the converter, or you can sense the output at the atty and the converter will use that voltage feedback to regulate the voltage. This takes the losses from the wiring out of the feedback loop.

 

[mamu]

Thanks Cape! Good to see you!!

When I was researching this earlier, I found a good tutorial for Sense and overcoming issues and it explains it just as you indicated: Applying the Remote Sense and Trim Functions on DC DC Converters

It's still strange though that Sense for it to work correctly with this converter requires an output cap when the datasheet says external caps are not necessary. With no output cap, makes no difference if Sense is connected or not. Weird stuff - or friggin magic, right?

I probably will tie Sense directly at the converter since I'm only working with a few inches of line, but if I run into issues will tie Sense directly at the atty.

My first project with this converter is a passthrough mod.

 

[bapgood]

How does it compare to the Raptor?

 

[branflan]

Guy on reddit posted these breakout boards OSH Park ~ OKL2-T/20 Breakout

 

[mamu]

I did some high amp testing with the OKL. I didn't max it out - I ran 5.5v with 0.4 ohms (76W and 14A) and the converter handled the load aok.

But I feel like I'm in the twilight zone with this converter. With the high amp testing I just did, there is the expected Vdrop under load, but with lighter loads the Vout loaded is once again higher (as much as 0.3v higher) than Vout no load at Vout settings >5v. And this is no matter the input voltage level.

I thought the output cap stabilized it, but apparently not. If you dial in 5.5v unloaded you expect to vape at 5.5v, not 5.8v or whatever it fluctuates up to. It's like the converter is overcompensating and maybe the feedback loop or whatever is off, especially at higher set output voltages.

I give up - it is what it is.

How does it compare to the Raptor?

I think this OKL is going to be an issue with those who vape >5 volts and especially if they max it out with Vout. Other than that, it's comparable to the Raptor.

 

[mamu]

A Murata technical engineer finally got back to me this morning. I sent another email yesterday evening lol - that makes 4 now. He probably got tired of the repeated emails and thought he better respond to get me off his back.

There were a few questions I had that didn't get answered. But he said there are 10 ohm resistors internal connecting the Sense pins to Vout and GND, so the DC-DC will work ok if left open circuit but you sacrifice a small amount of output voltage regulation if left open circuit.

He also said with or without an output cap, should not affect the Sense. He also said "We guarantee our output voltage regulation at the point where the Remote Sense lines connect into the Vout and GND busses."

I took everything off the breadboard and wired all components directly to the converter and also tied +Sense directly to +atty connector and -Sense directly to -atty connector.

and... success... no more funky Vout... max unloaded = 5.5v, loaded = 5.5v (this was with a load of ~4A) and with higher loads the expected Vdrop.

I'm going to still monitor input voltage level and Vout load/no load just to make sure I don't re-visit the twilight zone with this, but I think I'm finally ready to put it in a mod. It took me a while to get there and to get it right, but all is good in the end.

I updated the wiring guide in the OP (you may need to clear your cache or hit refresh).

 

[bapgood]

Where did you have the sense lines connected before?

*Edit - So the variation was due to the breadboard?

 

[mamu]

yeah - most likely the breadboard. I didn't realize the feedback loop would be sensitive to it. First time I've worked with Sense and a feedback loop so was learning as I go.

aah well.. it was a journey and a good learning experience for me. If I get hit with something like this again though, somebody just shoot me.

 

[cturboaddict]

Thanks for sharing all your experience and knowledge. Just got my OKLs in today. Scavenging up the rest of the parts and excited to give it a go myself soon.

 

[mamu]

Welcome, cturboaddict.

OKL is doing great now and has been stable with the output with the direct connections. So it was def the breadboard.

Thinking back on this, it was just strange that for over a day the OKL worked correctly until the input volts <7.5v - so I associated that with the OKL. Then put an output cap and tied in Sense and everything aok for a while. Then got funky again - so associated that with the OKL too. Strange stuff.

Am curious now if I tied Sense directly at the converter instead of directly at the load, but aint gonna mess with it anymore. Is good to go as is.

 

[T1lVl]

The board a lot smaller then the raptor?

What would be your opinion on each chip? Which one you like more?

I want to build one of these devices. I vape at higher watts.

Sent from my SGH-T999 using Tapatalk

 

[mamu]

The board a lot smaller then the raptor?

What would be your opinion on each chip? Which one you like more?

I want to build one of these devices. I vape at higher watts.

Sent from my SGH-T999 using Tapatalk

I posted a side-by-side comparison pic of the OKL with the Raptor in the OP.

Wiring wise, I prefer the Raptor - no need for extra wiring for Sense since the Raptor doesn't have it. Plus I can easily have under voltage protection with the Raptor for the batts with a zener diode. With the on/off control open collector/drain circuit the OKL has, no can do a zener. Plus the Raptor maxes at 6v while the OKL maxes at 5.5v.

I did run into one issue with the Raptor when I was breadboarding for a passthru mod. If using a 12v power supply the pull-down resistor needs to be lowered from the recommended 7.5K. I used a 4.7K and having good results with that at all voltages so am switching to using a 4.7K resistor for the Raptor instead of the 7.5K resistor.

The 20K pull-up resister recommended for the OKL does fine at all voltages, so no worries there.

I don't know how the OKL will handle heat if running at max load, but the Raptor is handling heat aok. May need some kind of vent or airflow with the OKL. Not sure with that.

I think I'll know better how to answer which I like more though after I have the OKL in a mod and vaped with it for a while. I absolutely love the Raptor and the vape I get with it - it's a total vaping beast for sure.

arrow.com had 160 Raptors yesterday and by the time I went to order, friggin oos already.