U10: Intersil ISL6255HRZ Integrated battery charger with automatic power source selector for laptops.
U2, U1 : Fairchild FDS4435BZ P-channel MOSFET (connector control on circuit)
U42: Fairchild FDS6298 N-Channel MOSFET (Vcore-Hi CPU)
U39: Fairchild FDS6676AS MOSFET N-Channel (CPU Vcore-Low)
C437: SMD ceramic capacitor, not less than 25v
The circuit of U2 and U1 is not exactly similar to what appeared in the test circuit of the ISL6255 data sheet. This is more similar to the circuit in the case of TI bq24751 except that the exhaustion and source are reversed.
Connector - > S-U2-D - > D-U1-S - > SYSTEM
Frame - > D-U42-S - > Vcore - > D-U39-S - > weight
U42-D - > C437 - > U39-S (Ground SYSTEM for U42+U39)
The frame is 19v from the connector or 11v from the battery.
I have this PC for profit. It powers any power source in the outage. A short film should be easy enough for me to get to work. Since my first workstation motherboard required more than a power outlet, I didn't recognize what to look for. I assumed that if I pulled enough coins, I would find out the problem.
The power connectors are simply short and close, so I add power from the seat to the power outlet and increase the voltage. At around 1.1v 1 ampere, the primary U2 in-line MOSFET warms up. She tried not to be short-circuited, but rather to warm herself up, so she went out. At that time, I put the control power supply from the seat to the failed circuit point by U2 and U1 warms up. It goes on and on. Each MOSFET on lines U2, U1, U39 and U42 warms up and exits. All DCA55 tests except U42 which was devastated during the evacuation. The shorts are still there and this time C437 is warming up. The C437 comes out and the shorts are gone.
At the moment, everything bodes well. U42 and U39 probably did not warm up since the current for C437 does not know them. C437 is directly among them and I expected the MOSFETs to get hot, not the top. U2 and U1 heated up because they were legitimately working and trying to provide a short circuit.
I got an FDS6298 swap for U42 and pulled a top shot from a comparable area on a terrible board and the workstation works. Another electronic device was spared by the residue receptacle.
It's a terrible summit, but finding it isn't quite the same as finding terrible electrolytes.
Thursday, August 23, 2018
Tuesday, August 14, 2018
How does SPW47N60C3 POWER MOSFET work?
could someone help me solve this problem?
I have to drive a Power mosfet SPW47N60C3 for 600V, 47 An and it has an input capacity of about 12 nF.
The recurrence exchange should be 200 kHz and with a short rise and fall time (about 50 ns). Everything is on the top driver's side and must be fast. I will use the appropriate optocoupler if it is as well for the pilot on the low side.
Would you be able to give me a correct sign for the driver?
To evaluate the exchange speed, Qgd emits more than Ciss. Qgd is frequently called the Miller door todeplete load since it is the input load required when the no-load voltage of the MOSFET oscillates from completely on to off (or back). This refers to figure 11 of the data sheet, and is shown as an average of 121nC, which is not terrible for a high voltage FET of 415W. Adept's adaptation of a similar piece includes "Low Miller Capacitance" on the first page of the fact sheet.
The load equation tells us that you will need I = Q/t = 121/50 = 2.4A door drive to switch this part in 50ns. I keep a combination of FET driver chips to process FETs of different sizes. For example, the TC4427 is a double conductor rated at 1.5A, the TC4424 is a double 3A and the TC4420 is a single conductor 6A. Yet I would suggest section 6A for your huge FET. I would use a small input resistance of 1.5 ohms and keep the FET conductors to the short door wires. Keep the path from the source to the driver's GND handle short and instantly parallel to the inductance of the door prompt limit. Similar jacks for the 0.1uF shot at by bypassing the IC driver.
I would start my tests at low voltage and without load. I would evaluate the input and source voltages by changing to check the nature of my FET-druve low inductance wiring. For example, I would run the high side FET with its opto-coupler, but without the 400V on its exhaustion, and without the low side switch. Indeed, I would ground the source of the FET (to maintain the gate and source voltages within the range for my degree) and I would use a short resistive load to HV. I would have the FET beat for short heartbeats (to avoid overheating the resistance) and deliberately measure the lead inductance peaks, which will increase as the heap current develops. Finally, I would use an inductive load to evaluate high current destruction peaks.
I have to drive a Power mosfet SPW47N60C3 for 600V, 47 An and it has an input capacity of about 12 nF.
The recurrence exchange should be 200 kHz and with a short rise and fall time (about 50 ns). Everything is on the top driver's side and must be fast. I will use the appropriate optocoupler if it is as well for the pilot on the low side.
Would you be able to give me a correct sign for the driver?
To evaluate the exchange speed, Qgd emits more than Ciss. Qgd is frequently called the Miller door todeplete load since it is the input load required when the no-load voltage of the MOSFET oscillates from completely on to off (or back). This refers to figure 11 of the data sheet, and is shown as an average of 121nC, which is not terrible for a high voltage FET of 415W. Adept's adaptation of a similar piece includes "Low Miller Capacitance" on the first page of the fact sheet.
The load equation tells us that you will need I = Q/t = 121/50 = 2.4A door drive to switch this part in 50ns. I keep a combination of FET driver chips to process FETs of different sizes. For example, the TC4427 is a double conductor rated at 1.5A, the TC4424 is a double 3A and the TC4420 is a single conductor 6A. Yet I would suggest section 6A for your huge FET. I would use a small input resistance of 1.5 ohms and keep the FET conductors to the short door wires. Keep the path from the source to the driver's GND handle short and instantly parallel to the inductance of the door prompt limit. Similar jacks for the 0.1uF shot at by bypassing the IC driver.
I would start my tests at low voltage and without load. I would evaluate the input and source voltages by changing to check the nature of my FET-druve low inductance wiring. For example, I would run the high side FET with its opto-coupler, but without the 400V on its exhaustion, and without the low side switch. Indeed, I would ground the source of the FET (to maintain the gate and source voltages within the range for my degree) and I would use a short resistive load to HV. I would have the FET beat for short heartbeats (to avoid overheating the resistance) and deliberately measure the lead inductance peaks, which will increase as the heap current develops. Finally, I would use an inductive load to evaluate high current destruction peaks.
Tuesday, August 7, 2018
Issue for my designed board with W5300
I designed my board with W5300 which worked with FPGA Spartan 6. I draw the diagram as the reference diagram of W5300, I just used the connector HR911105A RJ45 with an internal transformer from the company HANRUN, instead of the reference connector RJ45 "RD1-125BAG1A". I draw my circuit board with all the details refer to the W5300 documents.
My problems are:
1. when I have attached the W5300 it consumes 220mA that I think it is not normal, also the W300 will be hot.
2.as you see in the diagram when I remove 3.3VA current drop to 70mA and W5300 will not be hot.
3.when I receive PING from a PC, it does not recognize any Ethernet hardware.
When I connect a cable to the RJ45 connector, the RJ45 LEDs start flashing.
5. I checked that the schematic and PCB values are compatible, especially ON DNP parts like pin 1.8vo.
6. I checked all the power grids again. Everyone's OK.
7. I checked the Rsensens value which was 12.3K.
8. I checked the clock for a 25MHz oscillation. The test mode is'0000'. When the cable is connected to the card, the crystal does not work! But when I disconnect the cable, the Crystal works with a frequency of 25MHz with 100mV Peak-to-Peak and 0.9 volts DC offset.
9. I checked the Link LED when I connect the cable, it will light up, and the ACT LED starts flashing erratically.
in the schematic, I used crystal and oscillator, but I just welded crystal.
I'm really confused, any help would be greatly appreciated.
I do not use an oscillator, it is optional but I used 1.8 volts for its supply voltage.
if all my patterns are true, why isn't it working?
The dissipated power of the W5300 is 180(Typ)~250mA(Max).
And here's my comment on your diagram.
When using an oscillator in internal PHY mode, be sure to use a 1.8V level oscillator and connect it only to XTLP(Pin 96).
Otherwise, I don't see any problem.
Please check the reset time if at least 2uSec after power up.
My problems are:
1. when I have attached the W5300 it consumes 220mA that I think it is not normal, also the W300 will be hot.
2.as you see in the diagram when I remove 3.3VA current drop to 70mA and W5300 will not be hot.
3.when I receive PING from a PC, it does not recognize any Ethernet hardware.
When I connect a cable to the RJ45 connector, the RJ45 LEDs start flashing.
5. I checked that the schematic and PCB values are compatible, especially ON DNP parts like pin 1.8vo.
6. I checked all the power grids again. Everyone's OK.
7. I checked the Rsensens value which was 12.3K.
8. I checked the clock for a 25MHz oscillation. The test mode is'0000'. When the cable is connected to the card, the crystal does not work! But when I disconnect the cable, the Crystal works with a frequency of 25MHz with 100mV Peak-to-Peak and 0.9 volts DC offset.
9. I checked the Link LED when I connect the cable, it will light up, and the ACT LED starts flashing erratically.
in the schematic, I used crystal and oscillator, but I just welded crystal.
I'm really confused, any help would be greatly appreciated.
I do not use an oscillator, it is optional but I used 1.8 volts for its supply voltage.
if all my patterns are true, why isn't it working?
The dissipated power of the W5300 is 180(Typ)~250mA(Max).
And here's my comment on your diagram.
When using an oscillator in internal PHY mode, be sure to use a 1.8V level oscillator and connect it only to XTLP(Pin 96).
Otherwise, I don't see any problem.
Please check the reset time if at least 2uSec after power up.
Tuesday, July 31, 2018
Top246YN in three terminal modes
I have a Top246YN in three terminal modes trying to determine if it works.
My setup:
48v power supply with 300mA capacity
LED and 10k resistor to test drain current.
47uF cap attached to control line and negative side to neutral.
10k and 100k series power supply pot for control.
Everything seems to work perfectly, the led lights up and when I increase the current to the control pin, the led goes out as if the work cycle was reduced. I look at the voltage of the control line and the shunt goes into action and keeps the line at ~5.8V. The problem is that even if I don't give the current of the control line, the thing still works. I even tried to add a tensile strength down. What do you think's going on? The reason why the IC was removed in the first place was that it did not appear to have a stable reversal. Any ideas?
The reason you see a 5.8V on the control pin is that there is an internal shunt regulator in the TOP246YN. As long as you put a voltage higher than 7V or more in the drainage pin, you will get 5.8V on the control pin. Details can be found in the Top with data sheet.
On the right, when the unit starts, the internal power supply loads the RC circuit at 5.8v to start the soft start. However, if no external feedback is applied before the RC circuit reaches 4.8v, an automatic restart will begin. The automatic start must then put the device into low power mode and attempt to restart the device at a frequency eight times higher than the RC time constant. My problem is that by testing the device as described above when no external feedback is applied, the device continues to work. I wonder if this could be caused by damage to the control spindle.
Yes, if there is no external feedback to the control pin, it means that it is an open loop. At 4.8 V, automatic restart is activated, disabling the output MOSFET and putting the control circuit into low-current sleep mode. The high voltage current source lights up and recharges the external capacity. An internal hysteresis undervoltage comparator keeps the VC within a 4.8V to a 5.8V window by turning the high voltage current source on and off.
My setup:
48v power supply with 300mA capacity
LED and 10k resistor to test drain current.
47uF cap attached to control line and negative side to neutral.
10k and 100k series power supply pot for control.
Everything seems to work perfectly, the led lights up and when I increase the current to the control pin, the led goes out as if the work cycle was reduced. I look at the voltage of the control line and the shunt goes into action and keeps the line at ~5.8V. The problem is that even if I don't give the current of the control line, the thing still works. I even tried to add a tensile strength down. What do you think's going on? The reason why the IC was removed in the first place was that it did not appear to have a stable reversal. Any ideas?
The reason you see a 5.8V on the control pin is that there is an internal shunt regulator in the TOP246YN. As long as you put a voltage higher than 7V or more in the drainage pin, you will get 5.8V on the control pin. Details can be found in the Top with data sheet.
On the right, when the unit starts, the internal power supply loads the RC circuit at 5.8v to start the soft start. However, if no external feedback is applied before the RC circuit reaches 4.8v, an automatic restart will begin. The automatic start must then put the device into low power mode and attempt to restart the device at a frequency eight times higher than the RC time constant. My problem is that by testing the device as described above when no external feedback is applied, the device continues to work. I wonder if this could be caused by damage to the control spindle.
Yes, if there is no external feedback to the control pin, it means that it is an open loop. At 4.8 V, automatic restart is activated, disabling the output MOSFET and putting the control circuit into low-current sleep mode. The high voltage current source lights up and recharges the external capacity. An internal hysteresis undervoltage comparator keeps the VC within a 4.8V to a 5.8V window by turning the high voltage current source on and off.
Thursday, July 26, 2018
STM8S003F3U6TR, a cheap and fast EEPROMs alternative
I just went to check my shopping cart at DK and they didn't have much stock of the common 24LC256 EEPROMs I use so I looked around. It turns out that there are piles of one piece faster stock from 1MHz to 75% of the cost. So I changed my order, of course. The part is an FT24C256A-USR-T and costs 50cents in quantities of 250 and even better in quantities of 1K. They have about 15,000 in stock at present and although it is described as an "SOP" part, this is the standard width of 0.150" SOIC. All specifications are same or better than regular parts.
This also revitalized my interest in the fast EEPROM and the higher capabilities that bring me to the next device, the M24M01-RMN6TP, a fast 128KB EEPROM with an additional 256 bytes, a lockable ID page and a minimum endurance of 4 million cycles. In terms of bytes per dollar, it is tied with the cheap part of the FT and costs $2.10 to $250 off.
So I ordered 25 to play because I thought my Tachyon Forth could make good use of that extra capacity and speed without introducing extra chips. The source code can be edited, saved and loaded in this small space as well as the additional memory for all help functions where the comment and description of each fourth word in the stack can be referenced.
Now, that leads me to mention something: What happened with Holly's wonderful finds that we used to love?
With EEPROM, there is a price premium for endurance and fine writing. If you just need storage, try dropping W25Q into DK engine. Now you will find that 8M Bit of serial memory, is just 36c/250+ - even less than your 256kb device.
The very low cost of W25Q parts, raises the question of when it makes sense to use a small uC bridge, to let the Prop 1 boot from W25Q parts? - Such a bridge could either look like a serial host PC or an i2c slave chip. Which would be faster? There is also a C8051T606-GT in MSOP10, at 73c/100+, which could be used when the code is stable. Also new, without price indication, this new MSOP10 is this new device -[Vcc wider, Flash, and adds ADC on the T606].
STM8S003F3U6TR QFN20 ( ~44c/100+), it will be cheaper or similar in price to similar ST part.
This also revitalized my interest in the fast EEPROM and the higher capabilities that bring me to the next device, the M24M01-RMN6TP, a fast 128KB EEPROM with an additional 256 bytes, a lockable ID page and a minimum endurance of 4 million cycles. In terms of bytes per dollar, it is tied with the cheap part of the FT and costs $2.10 to $250 off.
So I ordered 25 to play because I thought my Tachyon Forth could make good use of that extra capacity and speed without introducing extra chips. The source code can be edited, saved and loaded in this small space as well as the additional memory for all help functions where the comment and description of each fourth word in the stack can be referenced.
Now, that leads me to mention something: What happened with Holly's wonderful finds that we used to love?
With EEPROM, there is a price premium for endurance and fine writing. If you just need storage, try dropping W25Q into DK engine. Now you will find that 8M Bit of serial memory, is just 36c/250+ - even less than your 256kb device.
The very low cost of W25Q parts, raises the question of when it makes sense to use a small uC bridge, to let the Prop 1 boot from W25Q parts? - Such a bridge could either look like a serial host PC or an i2c slave chip. Which would be faster? There is also a C8051T606-GT in MSOP10, at 73c/100+, which could be used when the code is stable. Also new, without price indication, this new MSOP10 is this new device -[Vcc wider, Flash, and adds ADC on the T606].
STM8S003F3U6TR QFN20 ( ~44c/100+), it will be cheaper or similar in price to similar ST part.
Wednesday, July 18, 2018
Alternative Mosfets working for brushed ESC
I'm constructing a brushed ESC that can be effectively arranged to have any present rating you want just by picking an alternate mosfet!
I at present make them work with the front just, I might want to make it fit for turning around too. I have a few thoughts on the best way to do this, however, I invite proposals.
I will show the source code document (written in C) and any updates and utilitarian HEX records.
Hobbyking offers a 1S Lipo 3amp ESC brushed mic that I have and it functions admirably. This ESC utilizes a lowercase13 and a mosfet. I drew a graph for the HK ESC and I will go along with him to this position. It is on this premise I base my underlying endeavors.
From the chart, I comprehended that the PWM engine presumably utilizes the main equipment PWM unit in lowercase13 (stick 5, OC0A) and that the servo PWM input utilizes either the PCINT on stick 2 or perhaps the ADC yet it's most likely a standard IO port and everything is done in code.
I am at present utilizing PORTB3(pin2) for the info and OC0A(pin5) for the PWM yield to the mosfet.
Legitimate level mosfets must be utilized on the off chance that you drive them specifically with the microcontroller, else they won't illuminate totally!
Clarification: Regular N-ch mosfets pivot the distance when the door achieves 10 volts, legitimate level mosfets turn the distance within the vicinity of 2 and 5 volts on the entryway relying upon the mosfet. ATtiny and ATmega microcontrollers can't transmit up to 5 volts on their pins and less if their supply voltage is under 5 volts as though you are utilizing a 1 cell lipo, so, utilize a rationale level mosfet.
(cheap) Mosfets that should work correctly: (can be found on HQEW by looking up the part number)
SO-8 :
FDS6570A 15A, 20V (would be very good for a single-cell ESC as if the power-on resistance were lower than the single-cell voltage).
FDS6680A 12.5 A, 30 V
IRF7821 13.6A, 30V
FDS8813 18.5A, 30V
TO-252 (DPAK) :
FDD8447L 50A, 40V
IRL3502S 110A, 20V
I at present make them work with the front just, I might want to make it fit for turning around too. I have a few thoughts on the best way to do this, however, I invite proposals.
I will show the source code document (written in C) and any updates and utilitarian HEX records.
Hobbyking offers a 1S Lipo 3amp ESC brushed mic that I have and it functions admirably. This ESC utilizes a lowercase13 and a mosfet. I drew a graph for the HK ESC and I will go along with him to this position. It is on this premise I base my underlying endeavors.
From the chart, I comprehended that the PWM engine presumably utilizes the main equipment PWM unit in lowercase13 (stick 5, OC0A) and that the servo PWM input utilizes either the PCINT on stick 2 or perhaps the ADC yet it's most likely a standard IO port and everything is done in code.
I am at present utilizing PORTB3(pin2) for the info and OC0A(pin5) for the PWM yield to the mosfet.
Legitimate level mosfets must be utilized on the off chance that you drive them specifically with the microcontroller, else they won't illuminate totally!
Clarification: Regular N-ch mosfets pivot the distance when the door achieves 10 volts, legitimate level mosfets turn the distance within the vicinity of 2 and 5 volts on the entryway relying upon the mosfet. ATtiny and ATmega microcontrollers can't transmit up to 5 volts on their pins and less if their supply voltage is under 5 volts as though you are utilizing a 1 cell lipo, so, utilize a rationale level mosfet.
(cheap) Mosfets that should work correctly: (can be found on HQEW by looking up the part number)
SO-8 :
FDS6570A 15A, 20V (would be very good for a single-cell ESC as if the power-on resistance were lower than the single-cell voltage).
FDS6680A 12.5 A, 30 V
IRF7821 13.6A, 30V
FDS8813 18.5A, 30V
TO-252 (DPAK) :
FDD8447L 50A, 40V
IRL3502S 110A, 20V
Subscribe to:
Comments (Atom)