Teknetics T2(chinese knock off)AA to LIIon 18650 conversion

[nonkapo]

Good morning again,

As i write somewhere else, the plan is to convert the fake Teknetics T2 from four AA batteries to one or two 18650 Lithium ion accumulators. The target is mainly cutting the batteries cost, slightly reducing the weight and possibly increasing the battery life.

A bit of an introduction for those who are not so familiar with this particular machine:

T2 have a separated from the control box battery container which includes small board DC-DC converter(at least as per me;-). The first part of the project was to measure the weight of the 4xAA vs 1x18650 or 2x18650.
Here the results:
1x Overrated Chinese LiIon claiming 3800mAh - 36grams;
1x Samsung Lion unknown capacity(probably 2600mAh) - 45grams;
1x HP laptop battery element Lion rated 2600mAh - 45grams;
Let's average the weight to 40 grams.
1xAA Sony Ultra - 17G;
1xAA Ikea Alkaline - 23grams;
1xAA Camelion super heavy duty - 14 grams;
For now those are the only types i have and let's average the weight to 20grams per piece. Exchanging the 4x2AA with one 18650 reduces the weight by half(40 grams - not a big deal). If replaced by two 18650 we have the same or a bit more weight but increased capacity.

The conclusion is not very bright - in fact the only benefit from the MOD is the cost cutting of the batteries price, especially if you use elements from failed laptop battery as i do.They are free:-).
The weight and capacity benefits are either neglect able or must be checked/measured to prove and depend on the type of LIion you are using. That will come at later stage.

To be continued...

[Skippy]

I wouldn't recommend it.
I have the F75, the power supply / regulator is identical to the T2.
Two cells in series is too high a voltage for the DC-DC converter, it's a boost-mode converter , set for 6 Volts output. If you put more than 6.5V into it, it shuts down and simply passes the DC input through to the output unregulated.
A single cell may be possible ( or two in parallel ), but the regulator will need modifying, which is a lot of effort. I have the circuit, but haven't studied it to see if it will work OK on lower input voltages. I have myself wondered if a single AA (14500) Li cell would work it.
The weight of the 4 x AA battery pack is in a useful place, counter-balancing the coil weight slightly. I don't think a lighter battery will make the machine feel any lighter to swing.
It consumes very little power ( about 50mA ), so even cheap basic NiMH cells, like 1800 mAh capacity, will run it for 30 hours. This is my recommended solution.

[nonkapo]

Partly agree skippy,however let me try. Do not worry, I'm planning to detach the detector from the psu when connecting serial for 7.2. So far the test with single cell is successful. Don't know when it is discharged, but when 4.2volts the regulator brings 6Volts out and the detector is working. Will test later. Might think to leave it in external box to have option going back to AA.

[Tibuck19]

This might help you https://www.youtube.com/watch?v=aND0j2Y2IkM

[nonkapo]

Thank you, that's nice three in one solution. Well i'm having concerns about the interference from it. Hope someone more advanced than me in the electroncs can prove me wrong? Meantime i've notice big ferrite choke fixed on the coil cable inside the control box of some garret model. Just thinking if i decide to put such, which particular type ferrite i must use?

[Skippy]

Here's the circuit diagram of the switching voltage regulator:
https://web.archive.org/web/20170607...-layout-01.jpg

And the regulator datasheet:
https://www.analog.com/en/products/lt1613.html

So the device will certainly work on one Li cell, it's just the question what needs changing ... the inductor maybe. PSU decoupling ?

[nonkapo]

From the datasheet - For boost converters, inductance should be 4.7μH for input voltage less than 3.3V
and 10μH for inputs above 3.3V. So even the induction should not be changed as discharged cell should stop working below 3.3V. This way you don't need discharge indicator either ;-).

[Skippy]

"So even the induction should not be changed"
I think you've misread the circuit diagram, which shows a 27 microhenry inductor in the regulator. BUT .... we don't know if that circuit diagram is correct. I don't have my machine in pieces to see if the inductor is marked with a value, I don't have any detailed photos of the board, the ones on MD-hunter don't show anything.
Which is why I said it may not be simple to convert to a single Lithium cell. You need to understand the design of the circuit, in order to make the right decision when modifying it. I was intending to run the regulator on its own, with dummy loads, and try and observe how it behaves. I would like to probe a 'scope on it, but I know from experience that simply putting a probe on a circuit like this can mess up its operation.

[homefire]

Those ( Any ) Boost er Buck converters generate piles of unwanted signals and noise on the power output . I don't see why a simple Choke/Capacitor Filter would not do the job cleaning it up to a usable power source. Encasing the thing in a Shielded enclosure may be required.

[nonkapo]

"So even the induction should not be changed"
I think you've misread the circuit diagram, which shows a 27 microhenry inductor in the regulator. BUT .... we don't know if that circuit diagram is correct. I don't have my machine in pieces to see if the inductor is marked with a value, I don't have any detailed photos of the board, the ones on MD-hunter don't show anything.
Which is why I said it may not be simple to convert to a single Lithium cell. You need to understand the design of the circuit, in order to make the right decision when modifying it. I was intending to run the regulator on its own, with dummy loads, and try and observe how it behaves. I would like to probe a 'scope on it, but I know from experience that simply putting a probe on a circuit like this can mess up its operation.

For now I'm waiting for the original coil, as with the counterfeit one the behavior of the detector is a bit erratic, despite it works. So once i get hand on the original coil and test it i will continue the topic.
Otherwise if you interested i can try measuring the inductor value as is on the board or even unsoldering it. If you have a doubt i can even check all the values and reverse engineer that board - is not too complicated. As far as it may help someone i do not bother. Just a question of free time. I guess no one have access to free detector which he don't care to sacrifice ;-))) The Russians say - Всичко за фронта, всичко за победата!(Everything for the frontline, everything for the win!).

[Skippy]

The 27 uH inductor value is looking plausible, despite what the datasheet recommends. The energy stored in an inductor is 0.5 x L x I². If you have a larger L value, you need to pass less current through it to store the same amount of energy. Less current means less dramatic turn-off, less interference is generated, and efficiency is higher. The output current requirement is lower than the datasheet example circuits ( eg. 5V at 200mA ) , it's more like 50mA at 6 Volts most of the time, except when driving the internal loudspeaker. This also allows you to change the L value ( and physical size) to a different value.

One potential problem with running at a lower battery voltage is the battery monitoring circuitry/microprocessor. I think the machine shuts down because it measures the battery voltage as too low , not because the regulator itself stops working. So you need to modify the monitoring circuit to ensure the machine keeps on working.
There are two obvious options:
Modify the resistors in the circuit so that a 4.2 V Li cell reads as a 6.2V battery, for example. Then a flat 3.3V Li will read as a 4.9V battery.
Rewire the monitoring circuit so it measures the regulated 6V supply. Then it will always see a full battery, and it's up to the user to determine if it's gone flat or not.
( remember some Li cells are only 3.7V fully charged ,not all are 4.2V , so some re-calculation may be needed ).

There's only a simple divider network for the battery monitor, 3 resistors - R74 (10k), R53 (20k) , R52 (20k), it will be easy to change this , the 10K is part of a 4-in-1 package, that's one thing to remember. It goes straight into a micro input pin.

[Skippy]

Nonkapo: no need to fully reverse-engineer the regulator/audio amp board, it's already been done. I've done it on my own machine, many years ago, and there are two other independant rev-eng's probably done by "the Russians", that agree with mine. I just haven't worked out the inductors, they are unmarked, and small. I'm reluctant to unsolder my own inductor, they are tricky to get off, usually needing two soldering irons and two people , unless you heat the heck out of it. And they are fragile.
I should be able to put that regulator on the test bench, and work out what the L value is from the waveform timings. Feed it with a 5 volt power supply, apply varying loads to the 6 volt output, and observe the switching times. You can then calculate L reasonably accurately. Inductors don't come in many values, it shouldn't be hard to work out.

[nonkapo]

That's very kind of you, but think also if one like still to keep the AA option?!? I just come with screw on power socket where the Lions can be connected externally. i know free lunch and perfect happiness does not exist ,but still do you think its possible?

[Skippy]

I don't like the idea of using an additional switching regulator, it's just going to make electrical noise. And if you look at the Tek regulator, you will see they have added additional inductor-capacitor filtering on it, to clean up the output. So I think it's unwise to spoil their good design.

However .... if you modified it to run on a single Li cell , it should also work on three x AA batteries. So if you fitted a dummy 'blank' battery in one of the slots in the battery compartment, and 3 cells in the others, it should work as normal.

Personally, I think an external battery is an ugly solution, and only useful if you need really long run-times, like 100+ hours without access to a charger.

[nonkapo]

I try measuring the inductance,but my Chinese meter cant get it. Seems below the range.