Notice To Unregistered Users: By visiting this forum, you are automatically agreeing to the terms and conditions set forth in our Safety Notice & User Agreement. All electrical work shall be completed by a qualified and licensed electrician. Admins, moderators, and members do not accept any liability for the information posted here. Please make sure to read the Safety Notice & User Agreement for the full disclosure. To remove this message, please register for an account.
Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
TP4056: The Clone Wars
#1
Okay, so I got a few TP clones as I mentioned before. I'll include some cheap "western" equivalent or similar in the test later.

Quick list:
TP4056, BL4056, TC4056A, CL4056, ME4057, BQ24090, BQ21040, MCP73830
(click for bigger picture)
[Image: fercsa_tp_clones.jpg]
Finally I made my testing jig on a breadboard(detailed post later) and here is the first result from an genuine TP4056 IC.
[Image: fercsa_cgr18650e_tp4056.png]
Here is the CSV file  too, if somebody want to play with it or make a nicer graph.

But you can see the important stages, I label it this time.
[Image: fercsa_cgr18650e_tp4056_explained.png]
-No battery means charger is idle, no battery present.
-CC stage, constant current stage. Looks like the voltage rising slowly and not delivering the maximum set of current, just limiting it. It's a little bit weird so I'll repeat my test one more time. Because, when resting the battery the voltage is around 3.6-3.7V, then this is stage is mostly skipped.. or just too quick.
-CV stage, constant voltage stage
-On the end the voltage and current drops, so charging is over and we measure the battery voltage and ~zero current.

Of course EN pin working. Spoiler alert: on all of them.

I'll try to test one chip per day, but it's take time to discharge and charge the same battery, which is a purple guinea pig aka CGR18650E.


(If you have a different TP clone which is not on my list and willing to post it to me in a simple envelope for comparison, please drop me a msg)
BlueSwordM, Rad, Korishan like this post
Reply
#2
This very nice. I presume that spreadsheet and graph is of the genuine TP4056?
I like how nicely it keeps the cell at 4.25v, my TC4056A's spike up to 4.29v.

How long between each of your readings?
I count 117 readings where it was charging @ 1.2A, Id like to figure out how long in minutes that was because mine charge @ 900mA (+/- 100mA) all the way to 4.25v before CV makes it taper off

Heres my hacked up log of the charging portion of my tester for my TC4056A (readings approximately every 1.2 seconds)

https://docs.google.com/spreadsheets/d/e...mI/pubhtml
Reply
#3
What qualifies as a "genuine" TP4056? Is there a specific brand name?
Reply
#4
(06-13-2018, 04:46 PM)Stephen Ryan Wrote: This very nice. I presume that spreadsheet and graph is of the genuine TP4056?
I like how nicely it keeps the cell at 4.25v, my TC4056A's spike up to 4.29v.

How long between each of your readings?
I count 117 readings where it was charging @ 1.2A, Id like to figure out how long in minutes that was because mine charge @ 900mA (+/- 100mA) all the way to 4.25v before CV makes it taper off

Heres my hacked up log of the charging portion of my tester for my TC4056A (readings approximately every 1.2 seconds)

https://docs.google.com/spreadsheets/d/e...mI/pubhtml

Yes that's right. The two graph is the same and for the genuine TP4056.
Your readings are way too bounce, what do you use for voltage and current messurment?

My sample rate is 3SPS (sample per seconds). 
Looks like I should log the elapsed time too. OOpsy.

(06-13-2018, 07:39 PM)mike Wrote: What qualifies as a "genuine" TP4056? Is there a specific brand name?

Genuine which developed and manufactured by NanJing Top Power, it has a TP logo. Probably a lot of clones made by "licensing" the ASIC from Top Power, but in china licensing can mean anything..


Also this is the second run on the same battery and genuine TP4056 IC:
[Image: fercsa_cgr18650e_tp4056_2.png]
CSV

The same weird behavior on the CC stage's current chart. I'm thinking I should put a few 18650 in parallel to rise the battery capacity, maybe it's gonna make a difference.
Reply
#5
(06-13-2018, 08:10 PM)FERCSA Wrote: Your readings are way too bounce, what do you use for voltage and current messurment?
I cant understand why your TP IC isnt doing a proper CC phase.

I use an Arduino Nano (10-bit ADC i believe) set to internal vref and then have a set of voltage dividers for each cell. Gives a minimum stepping of 0.004v
The current readings are hugely bouncy. Maybe heat is a factor? They still get super toasty even with three small heatsinks on each charger module. One on the IC, two underneath the board.

Im using the current sensing directly off the PROG pin of the TC4056A IC. Except that its going to an external 10-bit ADC with a vref of 5v (yet still a stepping of 0.004v somehow).
The formula is given on the datasheet as:          Current = (PROG pin volt / resistor value in ohms) * 1200 
Since my chargers resistor value is 1.2k ohms, whatever voltage is at the PROG pin is the current (in Amps) that the TC4056A itself is measuring.
Maybe some software smoothing in the form of averaging should be done but they are the raw readings

Edit: I also have 4056E IC's lying around. Im going to replace the TC-A one i just tested with the E onto the same charger board and redo the whole test. I can post you a couple of the E's if you dont already have them Wink

Update: Just finished testing the 4056E. Identical charge curve as the TC-A, charge time was within margin of error
https://docs.google.com/spreadsheets/d/e...Yw/pubhtml
Rad likes this post
Reply
#6
(06-13-2018, 10:58 PM)Stephen Ryan Wrote: Maybe some software smoothing in the form of averaging should be done but they are the raw readings

Yeah, try oversampling your readings.

Here is my TC4056A result:
[Image: fercsa_cgr18650e_tc4056a.png]
CSV
First time I didn't make a unified graph, sry.
I ran the test one more time for sure, my second result:
[Image: fercsa_cgr18650e_tc4056a_2.png]
CSV
If you make your own graph from the CSV, then zoom in, the weird CC stage is still there..

I'm using the same method and measuring the PROG pin for current, except I'm using a way better ADC, but I'm sure I lost a few bits because of the breadboard. I'll write a detailed post from my setup later.
Reply


Who read this thread?
48 User(s) read this thread:
Katarn_89 (06-13-2018, 06:32 PM), Bubba (06-18-2018, 07:19 PM), happygroover (06-13-2018, 07:20 PM), oldserg (06-14-2018, 11:46 AM), solar_18650 (06-16-2018, 10:57 AM), Geeksboro (06-16-2018, 02:05 PM), Stephen Ryan (06-14-2018, 08:00 PM), Ibiza (06-14-2018, 08:56 PM), Redpacket (06-14-2018, 01:43 AM), DonaldBitterman (06-13-2018, 08:22 PM), drchips (06-15-2018, 03:58 AM), JDish (06-14-2018, 10:04 AM), Cherry67 (06-15-2018, 08:36 AM), Bart (06-14-2018, 08:35 PM), polargeek (06-16-2018, 05:18 AM), Rasmus Godske (06-14-2018, 12:52 AM), 100fuegos (06-13-2018, 10:08 PM), Riplash (06-14-2018, 09:11 PM), Unleashed (06-15-2018, 03:50 AM), Sholphin02 (06-15-2018, 02:36 AM), jakonthestak (06-16-2018, 03:56 PM), BrianMonks (06-18-2018, 09:33 AM), CBlack (06-13-2018, 01:35 PM), brwainer (06-15-2018, 04:04 AM), Jason Morris (06-13-2018, 05:07 PM), IronSky (06-13-2018, 02:57 PM), jesusangel (06-13-2018, 08:03 PM), watts-on (06-14-2018, 08:17 PM), Kaspars Krumins (06-14-2018, 07:47 AM), wim (06-13-2018, 07:00 PM), jdeadman (06-15-2018, 12:09 PM), FERCSA (06-14-2018, 07:51 PM), PAF (06-15-2018, 10:15 AM), A + A (06-14-2018, 11:12 PM), Korishan (06-14-2018, 07:52 PM), Stefaan De Ridder (06-14-2018, 02:28 PM), BlueSwordM (06-14-2018, 11:46 PM), acidice333 (06-14-2018, 08:53 PM), Charly144 (06-14-2018, 06:06 AM), rev0 (06-20-2018, 11:38 PM), 12ax7 (06-21-2018, 04:28 AM), Grumplestiltskin (06-14-2018, 01:05 AM), mr_hypno (06-15-2018, 03:19 PM), Jon (06-14-2018, 07:56 PM), Rad (06-15-2018, 09:42 AM), Geek (06-15-2018, 12:09 AM), ivnz (06-13-2018, 09:52 PM), mike (06-14-2018, 12:38 PM)

Forum Jump:


Users browsing this thread: 1 Guest(s)