Page 1 of 15

Li-ion power mod

Posted: Sun Apr 02, 2017 12:10 am
by qu1ck
I like this little scope for it can be a great tool for low frequency work. Biggest selling point to use it over the big proper scope is that it is small and portable.
Except it's not quite portable if you also need to carry a power source and look for an outlet, now is it?

Well I decided to fix that:

Image

I used a separate charger board with protection and a step-up module built on fp6291 high frequency pwm DC converter IC. I had to swap the resistor in the feedback divider to 130k (circled on the pic) to bump output voltage from 5v to 8.4v:

Image

Also I modified the main board so that I could use the built in power switch. It was quite easy, just cut two tracks to the switch and add a jumper bypassing the protection diode (we don't want to waste 0.7v there, also bringing the voltage on par with what 9v with a diode would be):

Image

Now to add a bit of thermal glue to stick the modules on the enclosure and add a cut in the bottom panel for the charging port:

Image

Stick the battery pack on the back of the case with some double sided tape and tidy up the wiring to complete the whole thing. I now have an awesome portable scope:

Image

If you want to buy the same things I used here are the links: This one is 4000mah but it's a bit too big (6x60x90mm), barely fits in between the screw holes on the back. I recommend buying one a bit smaller. Power consumption is around 250ma so this pack should last me quite a few hours before I need to recharge.
I also recommend to recharge early because even though the charge controller has over discharge protection, it's quite low at 2.5v. You shouldn't let your li-ion battery fall below 3.0v, unless it's specifically rated for low discharge.
If you use another step up module, make sure it's switching frequency is high. With the one I used VCC ripple (a few mV at 1mhz) has no noticeable effect on the waveform acquisition because it's much higher than the bandwidth of the scope. With a lower frequency switching module you most likely will have to fight the ripple with some caps and filters.

Wishlist:
  • Bare li-ion pack hanging out is no bueno, it can get punctured and then it's galaxy s7 all over again. I would like to add some cover for it, ideally a 3d printed one with a backstand.
  • Maybe when the firmware sources are released it can be hacked to use one of the unused controller pins to monitor battery voltage, that would be awesome.

Re: Li-ion power mod

Posted: Fri May 19, 2017 10:40 am
by MorganFlint
Hello, qu1ck, very good thread! I found it after I'd started a similar mod on my own with very similar ideas but different approach.

I've been thinking on how to make room for the battery inside the case, and found the main problem is that the encoder mini-pcb reduces the clearance for the battery in the spare space bellow the analog board and under the main board to only 4mm or less, which shortens a lot the choices for a suitable battery (fo example, "classical" Nokia batteries are 5mm thick).

So I've been working to increase this clearance and found that it was possible to cut 2-3 mm off the two pods for the mini-pcb screws, just enough so the encoder's front face sits on the main board. To avoid the knob to protrude from the case, I also cut 2 mm from it's knurled spindle. This way I have 6 mm clearance between the main board and the case, so I can use nokia or game boy batteries I have liying arround, or even a bigger 604060 size (e.g.this 1200mAh one or this 1650mAh one).

Another (easier) idea to modify the encoder installation I had after doing that was to put aside the mini-pcb and bend all the encoder pins towards its front face, so it could be soldered directly to the main board (also would require cutting the spindle and both screw pods flush with the mainboard). This would have been easier (and leave more free space) if done from the beginning, but more complex once the encoder was soldered to the mini-pcb.

After that the plan was to fix the microUSB charging module to the main board (hot melt glue) so the microUSB connector sits in place of the white power connector, and the booster module (I bought this one, which is adjustable, but also found another one that outputs 9V without mods) behind the encoder min-PCB (both modules are very low profile, so they wont reduce battery clearance). BTW, microUSB charging modules come in 2 flavours: with battery protection (as the one you used), and without it (like this one), so if the cell has it's own protection (as is the case in the ones I pointed), maybe it's better to use the unprotected one).

But after I ordered the components, I found that there are also modules that combine battery charging and boosting in one board, like this one (with battery protection) and this one (without it), so I also ordered them.

I'm also planning (if there's enough room for it) to insert a 12V to 5V buck module in the place of D2, so I could use also 12V wallwarts with 5.5x2.1mm jack to charge the battery and power the scope, alternatively to the microUSB.

So, once I receive the modules I'll finish the mod and post the results.

As one picture is worth a thousand words, here you have some pictures of the mods:
Space for the battery:
Image

Encoder mod front view:
Image

Encoder mod side view:
Image

Encoder mod top view:
Image

Shortened spindle:
Image

Battery clearance:
Image

Alternative encoder mod proposal (bent pins):
Image

Re: Li-ion power mod

Posted: Sun May 21, 2017 8:34 pm
by qu1ck
Awesome ideas, thanks for posting!
I like the encoder mod, putting the battery inside the scope crossed my mind but I didn't explore possibilities to increase the space available inside.

Be careful with choosing the boost converter though. Like I mentioned the switching frequency matters. Eevblog forum user here reported increased noise when using a MT3608 based converter. In theory that one is also >1Mhz but in practice I've had a few converters that were producing serious output noise at 2.1 khz. That frequency will bleed right through your analog stage into your waveform.

Also how are you wiring the on/off switch if you use combined charging-boost module? With my implementation I can charge with the unit on and off and I can use it while charging as well. I don't see how you can do that if charging and boost are combined. If you put the switch between the module and scope you will likely have at least a few milliamps of quiescent current that will drain your battery in a just a few days.

Re: Li-ion power mod

Posted: Thu May 25, 2017 2:25 am
by spdrfx
Hello,

This http://www.ebay.com/itm/112324437388?_t ... EBIDX%3AIT one is working fine, I asked the seller, and he told me this module sw frequency are 1.2MHz. I also used a 800mAh Li cell, http://www.ebay.com/itm/111832711745?_t ... EBIDX%3AIT this one is perfectly fit inside.

Unfortunately, the DC-DC module is wider 1mm, so i haveto cut 1mm from the potentiometer's PCB edge, but there are many other ways...

Re: Li-ion power mod

Posted: Sun Jun 04, 2017 12:58 pm
by MorganFlint
qu1ck wrote:...Also how are you wiring the on/off switch if you use combined charging-boost module? With my implementation I can charge with the unit on and off and I can use it while charging as well. I don't see how you can do that if charging and boost are combined. If you put the switch between the module and scope you will likely have at least a few milliamps of quiescent current that will drain your battery in a just a few days.
Sorry for the late response, but i've been waiting to complete the mod before answering.

You're right, if the module is combined, probably it's not possible to avoid some quiescent current from tthe battery to the upconverter. Anyway, this also should happen in many USB power banks and, according to the ebay ad, it's only 0.2 mA, so 5000 hours on a 1000 mAh battery. I didn't measure it, and it's not elegant anyway, so I'll see if I can intercalate the switch between the charging and boosting (for this, I'd try to separate completely one of the sections of the DPDT switch, so the other section still works in case of using the 9V input, see next paragraph).

Another thing: Instead of connecting the charger-booster board GND directly to the scope's GND, I connected it to the normally closed switch inside the original 9V jack (pin 2 on the original schematic). This way, if I plug a 9V adapter, the booster gets automatically disconnected from the scope (but battery can still be charged from a 5V adapter).

Some pictures of intermediate stages and the final result (click on them for bigger versions):

Prior to fixing it to the mainboard with hot melt glue, I isolated with "kapton" tape the areas of the bottom part of the converter-booster which could make unwanted shorts:
Image

Converter-Booster already fixed and charging the battery:
Image

General view with battery fixed with two small spots of hot melt glue:
Image

More detailed view of the connections. See some hot melt glue spots on battery cables, both battery and charger-converter sides, as strain relief. See also the module protrudes about 2 mm from the mainboard. This allowed the microUSB male to plug in correctly without the need of making a very big hole in the faceplate (the two small "ears" ob the module's PCB at both sides of the microUSB female had to be filed, and also the corner at the right to match the corner of the mainboard. All this required some adjustment by heating the hot melt glue with a rework tool (100ÂșC only) and slightly sliding the module until it's final position. Note also the connection of the black wire to the 9V jack:
Image

It would have been better to solder the yellow wire to the other side of the diode (D2). This way, the forward voltage drop is avoided, and it also would block current flowing back from the booster to something plugged in the 9V jack, in case of problems in then 9V jack internal switch (or in the transient while it opens). I'll do it next time I open the case...

Closed shell without plate:
Image

Idem after installing the previously drilled plate:
Image

Look ma, no wires!
Image

While charging, the module's red led is visible through the screw's hole:
Image

And also the blue led when fully charged:
Image

That's all by now! Maybe it would be better to start a new thread to describe my version of the li-ion mod in one post...

Re: Li-ion power mod

Posted: Sun Jun 04, 2017 5:34 pm
by qu1ck
Anyway, this also should happen in many USB power banks and, according to the ebay ad, it's only 0.2 mA, so 5000 hours on a 1000 mAh battery.
Power banks only have to output 5v fixed. They usually use specialized ICs with very low quiescent current that don't engage the switching mosfet unnecessarily. Jelly-bean boost converters don't do this.
You should measure current anyway. Also use battery with protection.

Otherwise, very well done mod! Thanks for posting.

Re: Li-ion power mod

Posted: Mon Jun 05, 2017 8:38 am
by MorganFlint
I've seen power banks with "all in one" IC (charge+boosting, e.g. FM6316FU), but also others with a configuration like this module (charging and boosting with different ICs), the latter is good as they can be used as mini-UPS, because they can charge the battery and feed a load at the same time (I've used one of them with a WiFi IP camera). In some of both cases, Vout can be modified by changing a resistor in the feedback, but in other it's fixed as there is no external feedback pin.

According to what I've found in the internet, these Nokia batteries and clones are supposed to be protected... but with clones there's always a chance to be scammed.

Re: Li-ion power mod

Posted: Tue Aug 01, 2017 5:11 pm
by MorganFlint
Hello, I modded a second DSO150 with some lessons learned from the first one, so I'll post some photos of the differences with the first one.

First: separation of SW5 sections, so one of them can be used afterwards to switch the battery:
Image

Second: Encoder mod. Instead of using the small PCB, I bent the puns upwards and soldered it to the mainboard. This leaves more space for the battery, but you need to cut a bit the two posts where the small PCB is screwed to, and also the encoder's shaft to avoid the knob to protrude in excess from the case:
ImageImageImage
Image

I cut the shaft more than I should, and I tried to solder the case's flanges to the PCB, but this is not recommended because it's easy to damage a track (see the reconstructed track in the last picture between the encoder and the top post). To add mechanical strength, if the battery is not too thick, you can screw the mini-pcb in its place (as it doesn't have solders, it still leaves more clearance that if installed in the conventional manner). The following photos show the "battery bay" height (without mini-pcb) that would allow for a battery thickness of up to 9 mm except in the DC jack area, where it's 1.5 mm less:

Image
Image

without the encoder mod, the maximum battery thickness would be not more than 4-4.5 mm (maybe 5mm would fit, but the solders in the encoder's mini PCB could damage the battery).

Third: Charger+Booster module modification to avoid quiescent current to drain the battery when not in use. I had to unsolder C2, cut the tracks as shown and resolder C2 to the pads for battery connection:
ImageImage
Image
Image

Black goes to 9V jack, brown is battery negative, orange is booster input (from SW5, afterwards I marked it with black, see next picture) and yellow booster output. Charger output, still not connected, is the top of C2 in its new location.

Two more cables are needed from the top side of C2, one goes to the battery (orange) and the other one to SW5 (orange-black):
Image

So the two orange-black cables go to SW5, the threaded orange-brown to battery + and -, respectively, and the black and yellow ones feed the scope, connected the black one to the appropriate pad of J7 (so if a 9V jack is inserted the boost module is disconnected from GND), and the yellow one to the anode of D2. Click on the picture for a bigger image.

The following picture shows connection of the isolated section of SW5:
Image

... And that's all for now! I have the idea that it could be possible to add a miniature USB to serial module connected to the same microUSB of the charger and to the serial port of the mainboard (two arrows at the left of the module), so it could be possible to update the firmware, and also make use of the possibility to export captured data trhrough the serial port with toshi's firmware, but that will be another day...

Re: Li-ion power mod

Posted: Wed Sep 27, 2017 6:21 pm
by upsss
Here is my implementation of of the internal Li-Ion battery power. I studied carefully all my options and I believe it accomplishes ALL my requirements.
1. Complete battery protection. Over/short current protection, over/under charge/discharge voltage protection.
2. Lowest possible battery drain in the Off position.
3. Longest battery run time.
4. Quick battery charge time.
5. Low PWM switching noise.

Most cell phone type batteries have some sort of battery protection but NOT all the needed protections. Some of the battery protections like under voltage protection is implemented by the phone software so it is NOT typically implemented by the battery itself. In any case, I would NOT trust any eBay, AliExpress vendor to tell me what kind of protection a battery has.

To be able to fit the largest possible (1800 mAh) battery I followed MorganFlint idea by eliminating completely the encoder daughter board. BTW, I didn't have to reduce the length of encoder shaft at all, I am not sure why MorganFlint had to do that. To reduce the drain on the battery in the Off position <20uA, I split the charging/protection circuit from the step up converter so the On/Off switch can turn off the power to it. This step up converter is very tiny and it fits in the original external power jack location which I didn't need anymore. What so special about this converter is that it includes an extra LC filter on the output which provides a very low noise output. The only minor problem with this converter is that it is configured for 5V output so you will have to change one of the super tiny SMD feedback resistors for a 9V output.

As far as any Chinese no brand name battery capacity ratings, ALWAYS be very skeptical of their ratings. They NEVER specify the current load at the Ah specified rating. Any reputable company will typically specified Ah at least at 1/10C capacity. Without specifying the current load you can come up with very high and ridicules Ah capacity. My particular battery capacity is specified at 1800 mAh and the actual measured capacity at 200mA was only 1340 mAh! Most people have seen on eBay and Aliexpress those 18650 batteries specified at 10,000 mAh capacity which is ridiculously high. If such a battery existed, the electric car companies would be able to get more than 3 times the currently advertise driving range, wouldn't that be great!

Here are some pictures:
Image
Image
Image
Image

Re: Li-ion power mod

Posted: Sun Oct 01, 2017 11:54 am
by MorganFlint
upsss wrote:... BTW, I didn't have to reduce the length of encoder shaft at all, I am not sure why MorganFlint had to do that...
Very good implementation!
I had to do it because I wasn't able to push the knob as close to the box as I wanted, but it might be because the fit was too tight and/or I didn't push hard enough...