Home
JAQForum Ver 20.06
Log In or Join  
Active Topics
Local Time 16:37 29 Mar 2024 Privacy Policy
Jump to

Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.

Forum Index : Electronics : Reverse engineering a TL494 based Chinese

     Page 2 of 2    
Author Message
disco4now

Guru

Joined: 18/12/2014
Location: Australia
Posts: 839
Posted: 06:33am 30 Oct 2020
Copy link to clipboard 
Print this post

Thanks for the circuit.
  Techie007 said  
 Thanks for the AP1509 part suggestion, that does look like the chip they're using—or at least a plausible replacement!  


On my 1800W board looks like XL7005A is used.Orientation is opposite to the LM358
Latest F4 Latest H7
 
Techie007

Newbie

Joined: 25/10/2020
Location: United States
Posts: 3
Posted: 02:42pm 30 Oct 2020
Copy link to clipboard 
Print this post

  disco4now said  Thanks for the circuit.On my 1800W board looks like XL7005A is used.Orientation is opposite to the LM358

You're welcome, and thank you for the buck converter information—that one looks correct, so I have updated the schematic accordingly!

  wiseguy said  Does the output resistor/shunt exist as a part or is it a PCB trace also ?

The output resistor is a traditional shunt, being a bar of metal just like the ones in a multimeter, with a resistance of 0.005 ohms.  I drew the schematic to be an accurate representation of the original circuit design, and I intend to keep it that way.  Wires/circuit traces cannot be assumed "equipotential"—they all have an inherent resistance.  Layout really matters in analog design!  Someone looking at the schematic will have to ask themselves, "Why did he draw it that way?" just like I had to ask myself that question as I studied the circuit board.  To make the enigma easier to realize, I will add a negative voltage reading under load where those resistors connect once I measure it.
 
Edited 2020-11-03 06:04 by Techie007
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 10:42pm 30 Oct 2020
Copy link to clipboard 
Print this post

Sorry but IMO it is probably not a totally accurate representation of the original circuit design which you did not get to see.

The pcb track on which the 52mV is generated from the input current did not happen by accident, it was part of the design.

99% of connected tracks for this design can be considered equipotential and their associated inductance, capacitance, resistance and that every soldered joint is a thermocouple is of monumental inconsequence & can be safely ignored, if were talking RF that is a different discussion.

I have designed schematics & PCB tracks to work as shunts for less accurate current measurement, component placement around those areas and kelvin connections are part of that process.

I will not comment further with regard to the schematic. I have been looking at schematics for over 54 years and designing my own for over 45 years - I know a well laid out easy top follow schematic when I see one.
Edited 2020-10-31 16:52 by wiseguy
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 04:02pm 31 Oct 2020
Copy link to clipboard 
Print this post

I retract the suggestion of the AP1609, it has the right pin out but its input voltage is too low for this application.  The XL7005A is the right part and has an 80V input capability.

I intend to design a PCB layout for my version of a high power boost converter so I decided to re-enter this schematic into Altium before I start to modify it.

I found it a bit easier to read & follow after laying it out a bit differently, linked below if interested.

1800W SCH.pdf
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 405
Posted: 08:06pm 31 Oct 2020
Copy link to clipboard 
Print this post

  wiseguy said  I retract the suggestion of the AP1609, it has the right pin out but its input voltage is too low for this application.  The XL7005A is the right part and has an 80V input capability.

I intend to design a PCB layout for my version of a high power boost converter so I decided to re-enter this schematic into Altium before I start to modify it.

I found it a bit easier to read & follow after laying it out a bit differently, linked below if interested.

1800W SCH.pdf


Interesting, will you do a pcb design too?  
By looking @ your schematic you gave me a few tips to improve my next schematic layout, thank you.

One question, -5R and +5R is the same potential?

Was playing with the idea to design a powerfull boost converter to capacity test my batteries without just wasting the energy.
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 11:08pm 31 Oct 2020
Copy link to clipboard 
Print this post

Hi Nicks, I maybe going cross eyed but I cant see a -5R, I'm wondering if it is a print/display issue ? Please point it out if it exists. Thanks for the compliment, I'm still not sure if it will improve or detract readability if I move the main switching FET into the upper capacitor diode inductor block, will have a play later.

The resistors showing res2 are all 10K, the output LED resistor should be 75K not 100K, the perils of copy paste......Pin8 of the LM358 is missing the net Label 13V5. Check values against techies original schematic if in doubt.

I wont update the schematic for a day or two in case anyone else discovers any other issues/errors.

Re the PCB re-design it is about 4 in line currently - & we have a new puppy which needs a lot of watching ATM ( its ferally crazy !! ) so its sapping a lot of my (what seems to be ever decreasing) available time.
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
disco4now

Guru

Joined: 18/12/2014
Location: Australia
Posts: 839
Posted: 12:42am 01 Nov 2020
Copy link to clipboard 
Print this post

33K resistor from FET Gate to GND is missing.
Designation of driver transistors not carried across.
Latest F4 Latest H7
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 02:57pm 01 Nov 2020
Copy link to clipboard 
Print this post

Thanks Disco, I missed the 33k omission, the rest was on a list to do.

Updated Schematic with main FET shifted to Power block for improved layout.
I hope to get good at this analog design stuff yet.....

Just noticed that the Altium PDF outputs omitted the pin numbering on the LHS of the parts ? It also omitted to print net names, these were added manually - painful !
Just fixed and re-uploaded

1800W SCH.pdf
Edited 2020-11-02 11:38 by wiseguy
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 405
Posted: 07:03pm 16 Dec 2020
Copy link to clipboard 
Print this post

I'm thinking of doing some experiments with those tl494 pwm drivers, is there an easy way to calculate/determine  or choose the RC components for the feedback pin3? Sometimes it's left open, but could give unstable close loop control?
 
InPhase

Senior Member

Joined: 15/12/2020
Location: United States
Posts: 178
Posted: 02:04am 17 Dec 2020
Copy link to clipboard 
Print this post

  nickskethisniks said  I'm thinking of doing some experiments with those tl494 pwm drivers, is there an easy way to calculate/determine  or choose the RC components for the feedback pin3? Sometimes it's left open, but could give unstable close loop control?


The feedback pin on the TL494 is a bit mysterious unless you dig around the application notes. It is just a direct connection to the internal PWM comparator. Driving it low will result in max PWM. You don't necessarily need to do anything with the FB pin if you are using the error amplifiers for feedback.

You could wire the error amps so that they output low and then use a pot between Vref and ground with the wiper to FB and manually control the PWM.
 
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 405
Posted: 03:09pm 17 Dec 2020
Copy link to clipboard 
Print this post

Would a hall sensor with opamp behind it be fast enough for current measurement? I would like to switch at 50khz and regulate the input current, to use as load.
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 03:43pm 17 Dec 2020
Copy link to clipboard 
Print this post

  nickskethisniks said  Would a hall sensor with opamp behind it be fast enough for current measurement? I would like to switch at 50khz and regulate the input current, to use as load.


I have used a hall effect (ACS715) type to control current directly without an opamp just by juggling reference level with divider resistors and comparing it to ACS715 output.

I agree with Inphase that sometimes no compensation is required but I find that doing step changes to the load and watching the overshoot/undershoot, empirical choice of components is often faster than doing the maths required.

  InPhase said  You could wire the error amps so that they output low and then use a pot between Vref and ground with the wiper to FB and manually control the PWM..


Almost correct, if we assume inputs used are pins 1+ & 2- link pin3 fb to pin 2- and connect pin 1+ to pot wiper with pot between 5Vref and gnd. This sets up internal opamp as a non inverting buffer and fb follows it and adjusts pwm proportionally.  Note: also tie pin 16 to gnd and pin 15 to 5ref if opamp 2 unused.

Or for feedback regulation, ground pin2 and set pin1 at 5Vref, then the fb output (pin3) which is a current source can be pulled low with an opto or transistor to reduce PWM for regulation
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
InPhase

Senior Member

Joined: 15/12/2020
Location: United States
Posts: 178
Posted: 03:54pm 17 Dec 2020
Copy link to clipboard 
Print this post

  nickskethisniks said  Would a hall sensor with opamp behind it be fast enough for current measurement? I would like to switch at 50khz and regulate the input current, to use as load.


I don't know exactly what you're trying to do, but for the most part 50 kHz is a piece of cake for any op amp and, I would guess, many hall sensors. I have a hall sensor that is accurate up to 400 kHz.
 
InPhase

Senior Member

Joined: 15/12/2020
Location: United States
Posts: 178
Posted: 04:17pm 17 Dec 2020
Copy link to clipboard 
Print this post

  wiseguy said  Almost correct, if we assume inputs used are pins 1+ & 2- link pin3 fb to pin 2- and connect pin 1+ to pot wiper with pot between 5Vref and gnd. This sets up internal opamp as a non inverting buffer and fb follows it and adjusts pwm proportionally.  Note: also tie pin 16 to gnd and pin 15 to 5ref if opamp 2 unused.


The circuit I use for adjustable low frequency PWM  experimentation is a TL494 wired so: Tie pins 1 and 16 to ground, and 2 and 15 to Vref. This pulls the error amplifiers low and basically off. Then swinging FB between ground and Vref with a pot offers a clean sweep between 0 and max PWM. No feedback at all.
 
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 405
Posted: 07:45pm 17 Dec 2020
Copy link to clipboard 
Print this post

I know I'm going a bit off topic here, so I probably will start a new thread soon about an electronic load with buck topology controlled by an Arduino nano.

So I made a pcb with all power parts for a buck regulator, I want to test batteries/power supplies with a fixed current. A next step would be a push pull stage for upping the voltage and isolation. So I could theoretically charge a 48V battery with a 12V battery.

I know It could be done in 1 stage. But I prefer not at this moment. Even the ultimate would be to test a dcdc supply with energy recuperation, sending the output energy back to the input again. That’s why I need current regulation.

My board has A mosfet switch/resistor for pre charge the capacitors, a mosfet for for bridging that, no problem with that. Then a normal buck circuit. I have 3 voltage measurements and 2 current measurements, 1 on the input before the input capacitors and a current sensor between input capacitors and mosfet, to measure switching current.

I tried a MC33025, controlled with an Arduino nano with DAC for voltage reference and acs 712 current measurement. The thing was oscillating badly under current mode. Voltage mode was no problem with a resistive load. Then during testing I noticed that the ic was not really suitable  for single ended purpose, only push pull. The ic could only regulate between 0-45% 10% deadtime and then 0-45% again, so I think that’s the problem.

Edit:

I just had some good results with the tl494, but I had to add a capacitor between the negative opamp input and the comparator pin (pin4 and 15). It will need to be tweeked further, I tried different capacitors but 220nf gave good results.
The input was 15-30VDC max 3.2A, the output +-13.8V 0-6,1A in to a small lifepo4 battery. The different input current settings were stable under input voltage variation, it was also fast acting.
I tested inductors of 100-400µH with the same results.

I used an input resistor of 0,1R to measure the current. I wil try later with the hall sensors. Good tips thanks, I hope to do a write up this weekend, to show my progress. Allready spent more then 2 weeks on the project... wife is not so happy...
Edited 2020-12-18 07:12 by nickskethisniks
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 10:16pm 17 Dec 2020
Copy link to clipboard 
Print this post

  InPhase said  
The circuit I use for adjustable low frequency PWM  experimentation is a TL494 wired so: Tie pins 1 and 16 to ground, and 2 and 15 to Vref. This pulls the error amplifiers low and basically off. Then swinging FB between ground and Vref with a pot offers a clean sweep between 0 and max PWM. No feedback at all.


Ok I can agree, provided you use a relatively low value pot it will alter the PWM but you will be fighting the internal current source on fb so it will not be very linear and for a higher value pot will be terrible and cramped towards the 5V end.  

If you connect as I suggested it wouldn't matter if you use a 2K or 500K pot it will perform smoothly and linearly - try it.
Edited 2020-12-18 08:16 by wiseguy
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
prinx49
Newbie

Joined: 15/03/2021
Location: Italy
Posts: 2
Posted: 12:12pm 21 Mar 2021
Copy link to clipboard 
Print this post

Hello, I'm new to this forum and first of all I immediately send you my like for the competence of how the subject is treated.
I am Italian and I apologize why not knowing English perfectly I get help from Google.
I also bought the 1800W step-up that you described extensively to recharge the service batteries of my motorhome faster. I am not at your level of knowledge so I ask for some advice. My doubt is about the operation of the device in my specific case, as the difference between the input voltage (about 13 V) and the output voltage (about 14.5V) is not that described in the characteristics of the device (about 2 V ), so I asked your advice if it is appropriate to put 2-4 large diodes in or out to obtain a voltage drop and therefore increase the voltage difference between input and output even if the method does not excite me very much.
Having to recharge 2 lead-acid batteries of 100Ah each, I thought I would consider about 10-15A.
I also agree that the direct intervention of the fan at 60 is a bit late, an early and gradual intervention would be better.
Thanks in advance for your advice.
 
wiseguy

Guru

Joined: 21/06/2018
Location: Australia
Posts: 988
Posted: 01:35pm 22 Mar 2021
Copy link to clipboard 
Print this post

Hello Marco & welcome to the forum, it doesn't feel right for me to be the one to welcome you when I have been rather absent here for a few months.

I am not sure that you (maybe I) have interpreted the poor specifications of the converter correctly.

Depending on exactly which unit you are considering one specification says input 10 - 60V and output 12 - 90V. Your input voltage of 13V is within the input spec and your output requirement of 14.5V is within the output range adjustment (ok very close to one end of the spec). As far as I can see it should just work fine, without the need for diodes at the input or output to increase the difference between the input & output levels.

That would just mean the converter working a bit harder to overcome the diode losses we just introduced. If I have missed your interpretation or you want to post the specific specification that concerns you, I will try to help further. I also apologize in return for not being able to speak Italian (at all, let alone perfectly)  !

PS I agree the Fan starting to run at say 45 degrees and at full speed above 60 would be nice but not really feasible to implement for these boost supplies - a separate little dedicated circuit & sensor would be one answer.
Edited 2021-03-23 08:31 by wiseguy
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike
 
prinx49
Newbie

Joined: 15/03/2021
Location: Italy
Posts: 2
Posted: 09:02pm 23 Mar 2021
Copy link to clipboard 
Print this post

Thank you.
I think that as far as the fan is concerned, the simplest thing is to connect it directly to the step-up entrance, given the working range and since it will still work only during the travel period.
To verify the operation I give that the best thing is to try it in the "laboratory" with a power supply.
The model should be: QS-4884CCCV-1800W, the one with single terminals both in and out.
Hi Marco.
_________________
P.S.  ... but if one does not jump, he will never experience the thrill of flying
 
     Page 2 of 2    
Print this page


To reply to this topic, you need to log in.

© JAQ Software 2024