Film resistor vs Carbon resistor.

[gary350]

Removed for mods

I figured I would need a 1 ohm resistor in series with pin 3 I already have coffee can of them, 50 watts 1 ohm with heat sink mounts.

What is the capacitor across pin 2 and pin 3? What is that for?

The datasheets I looked at, one said 28 amps continuous, another said 30 amps, another said 39. My DC amp meter shows they work at 37 amps but they are about 130 degrees after 1 min 45 seconds that is when I turn it off. I'm not sure what continuous current will be just have to test them and see how fast they warm up.

 

[chemelec]

That is NOT a Practical way to Drive a 5 HP Motor.

Any Power Between Full ON and Full OFF will Destroy the Mosfets with Excessive HEAT.

As Jpanhalt Said, PWM would be MUCH BETTER.

 

[gary350]

That is NOT a Practical way to Drive a 5 HP Motor.

Any Power Between Full ON and Full OFF will Destroy the Mosfets with Excessive HEAT.

As Jpanhalt Said, PWM would be MUCH BETTER.

LOL. That circuit works for a toy motor I just made the circuit bigger for a bigger motor. I am no expert at DC motor controls. What does PWM mean?.

 

[chemelec]

PWM is Pulse Width Modulation. The Output Continuously Switches ON and OFF at a Specific Frequency.

Simple Method is just using a 555 to drive the Mosfets.

This is Just an Example:

 

[spec]

Not according to the previously cited IR document:
View attachment 96075

Can you support your spec_ulation with similar data?

John

Hi jp,

Appologies for the blunt response- I was in a hurry.

That is an unsubstantiated report which does not gel with me. But it is only for the TO- 220 case and not the MOSFET in question in a TO-220 case. This MOSFET is designed for switching applications with short bursts of high current not continuous high current like in gary's application. In fact some manufacturer's data sheets don't even include DC in the SOA graphs. Also, I am going by experience with the TO- 220 case which is totally unsuited to this application: high thermal resistances and poor contact with the heat sink due to the single hole fixing and the insubstantial metal base. This application is crying out for a TO-3 case or next best TO-247. Have you done the thermal budget for the MOSFET at 55A? I havent yet but I'm expecting it to be bad.

 

[jpanhalt]

The pictures I posted were for a speed control for a heavy duty Ford 6V starter (without a limiting coil, model 3110) run at 12 volts and used as a high-speed winch for launching model sailplanes. Exact current draws are not available. Most users report at least 225A. Some users estimate it as high as 5 HP (about 300A). I used 5 IR power mosfets in TO-220 packages in parallel, and we could stall the motor without popping a mosfet. I used a mosfet driver and PWM. For launch, the PWM was as close to 100% as we could get. A more reasonable drive used a 12V treadmill motor and only 4 mosfets in parallel. I do not have the specifications for that motor. I believe it was rated at least 2 HP. The tabs were soldered to copper buss (0.025 thick) in the former and to 4 oz copper pcb in the latter, which were attached to an aluminum heat sinks. I have used it for the past 15 years, but have not built any controllers of similar power since.

Here is a picture of the latter control:

As I pointed out above, soldering the tabs is not so much for electrical reasons as it is for heat dissipation. Of course, both machines were intermittent duty only. The winch was considerably less on time than the second device.

I have not done any calculations for what Gary350 is planning. For one, I have given and can give additional links for him to do that, should he desire. Second, I don't think any of us know what he is planning to do. The schematic he provided allows a max of 12 mA. That is a whole different story than the 35 A being discussed in the other thread.

John

 

[spec]

Not according to the previously cited IR document:
View attachment 96075

Can you support your spec_ulation with similar data?

John

The pictures I posted were for a speed control for a heavy duty Ford 6V starter (without a limiting coil, model 3110) run at 12 volts and used as a high-speed winch for launching model sailplanes. Exact current draws are not available. Most users report at least 225A. Some users estimate it as high as 5 HP (about 300A). I used 5 IR power mosfets in TO-220 packages in parallel, and we could stall the motor without popping a mosfet. I used a mosfet driver and PWM. For launch, the PWM was as close to 100% as we could get. A more reasonable drive used a 12V treadmill motor and only 4 mosfets in parallel. I do not have the specifications for that motor. I believe it was rated at least 2 HP. The tabs were soldered to copper buss (0.025 thick) in the former and to 4 oz copper pcb in the latter, which were attached to an aluminum heat sinks. I have used it for the past 15 years, but have not built any controllers of similar power since.

Here is a picture of the latter control:
View attachment 96081

As I pointed out above, soldering the tabs is not so much for electrical reasons as it is for heat dissipation. Of course, both machines were intermittent duty only. The winch was considerably less on time than the second device.

I have not done any calculations for what Gary350 is planning. For one, I have given and can give additional links for him to do that, should he desire. Second, I don't think any of us know what he is planning to do. The schematic he provided allows a max of 12 mA. That is a whole different story than the 35 A being discussed in the other thread.

John

Hi John,

Intresting project. What MOSFETS were used.

gary said that the schematic he posted was for a model motor. The 1K resistor is his own thing and we can all see that in the final application it is meaningless and does not need further discussion. As for none of us knowing. We know a lot from what gary has told us, even the supply volts, but not the answers to the two questions that I posted previously. In a nut-shell; what is the maximum current that the motor takes under stall/start. That is the defining parameter for determining the MOSFET bank current capability.

 

[gary350]

PWM is Pulse Width Modulation. The Output Continuously Switches ON and OFF at a Specific Frequency.

Simple Method is just using a 555 to drive the Mosfets.

This is Just an Example:

There is a problem with that. I have worked with 555 times for 45 years. You can change the pulse width to anything you like from about 10% to 90%. Assume you have it set for 50% it is always 50% no matter what frequency you adjust it for. If the mosfet is on 50% at 60 Hz or 50% at 10KHz that is the same ON time 50%. Only thing that is changing is the Hz. If you have adjustable 555 pulse width that is totally different, I have never made 555 adjustable pulse width I need to check my book to see if that can be adjustable.

 

[gary350]

The pictures I posted were for a speed control for a heavy duty Ford 6V starter (without a limiting coil, model 3110) run at 12 volts and used as a high-speed winch for launching model sailplanes. Exact current draws are not available. Most users report at least 225A. Some users estimate it as high as 5 HP (about 300A). I used 5 IR power mosfets in TO-220 packages in parallel, and we could stall the motor without popping a mosfet. I used a mosfet driver and PWM. For launch, the PWM was as close to 100% as we could get. A more reasonable drive used a 12V treadmill motor and only 4 mosfets in parallel. I do not have the specifications for that motor. I believe it was rated at least 2 HP. The tabs were soldered to copper buss (0.025 thick) in the former and to 4 oz copper pcb in the latter, which were attached to an aluminum heat sinks. I have used it for the past 15 years, but have not built any controllers of similar power since.

Here is a picture of the latter control:
View attachment 96081

As I pointed out above, soldering the tabs is not so much for electrical reasons as it is for heat dissipation. Of course, both machines were intermittent duty only. The winch was considerably less on time than the second device.

I have not done any calculations for what Gary350 is planning. For one, I have given and can give additional links for him to do that, should he desire. Second, I don't think any of us know what he is planning to do. The schematic he provided allows a max of 12 mA. That is a whole different story than the 35 A being discussed in the other thread.

John

This is very interesting that the mosfets have 25 ohm resistors. I have not experimented with resistance that low. It is also interesting the tabs are soldered to a PC board how can that be a heat sink. How can a PC board handle 225 amps, seems like the copper would burn off of the epoxy board. I have seen PC boards burn with less amps than that.

 

[spec]

I figured I would need a 1 ohm resistor in series with pin 3 I already have coffee can of them, 50 watts 1 ohm with heat sink mounts.

Hi gary,

I presume pin3 is the source pin. Afraid 1 Ohm would be too high. I expect 0R1 or so woulod be about right.

Or did you mean the gate resistors. If so, they would be 10R or 22R and are there to stop the MOSFET from oscillating. They must be mounted right on the gate pin to be effective.

What is the capacitor across pin 2 and pin 3? What is that for?

The capacitors are there for frequency stability: to keep the MOSFETs happy. The capacitors would be about 100nF ceramic type. The MOSFET has around 1nF there anyway, but the value is modulated, mainly by the drain voltage. In practice the capacitors may not be needed, but for a one-off design it is simpler to just fit them. A motor is an inductive load and will reduce the amout of current the the MOSFET can reliably supply. The capacitors may help in this area.

The capacoitors may have to be removed though when the switching characteristics of the MOSFET are analysed when the pulse width modulation functionj is designed. A snubber circuit will probably have to be designed too to cope with the inductance of the motor and protect the MOSFET.

The datasheets I looked at, one said 28 amps continuous, another said 30 amps, another said 39. My DC amp meter shows they work at 37 amps but they are about 130 degrees after 1 min 45 seconds that is when I turn it off. I'm not sure what continuous current will be just have to test them and see how fast they warm up.

The current rating on the datasheet is only a rough guide and is not normally the limiting factor. Other parameters affect what the MOSFET can reliably handle and be reliable in a real application. You coul probably get 70A out of that MOSFET but it wouldn't last long. When you tell what the maximum current is that your motor takes, then a circuit can be designed with all the values and you will have a reliable piece of equipment which will go on for years. This is achieved by a normal desing procedure.

 

[chemelec]

There is a problem with that. I have worked with 555 times for 45 years. You can change the pulse width to anything you like from about 10% to 90%. Assume you have it set for 50% it is always 50% no matter what frequency you adjust it for. If the mosfet is on 50% at 60 Hz or 50% at 10KHz that is the same ON time 50%. Only thing that is changing is the Hz. If you have adjustable 555 pulse width that is totally different, I have never made 555 adjustable pulse width I need to check my book to see if that can be adjustable.

This PWM is Adjustable PULSE WIDTH.
So NO the Motor is NOT on 50% of the time.
It may only be on 1%, 99% or anything in between these.

THIS VERY EASLY CONTROLS MOTOR SPEEDS, POWER and CURRENT DRAWS

 

[gary350]

I have an old 1970 TTL Cookbook. I recently bought the newest version of the TTL Cookbook and it has new information the old book does not have like adjustable Duty Cycle. The new book only shows 50% to 99.9% duty cycle but I can see it can be 5% to 50% or 5% to 99% duty cycle with a duel variable resistor. 555 is very unstable with temperature changes. I can see a new experiment very soon.

 

[KeepItSimpleStupid]

Take a look at this https://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_120539_-1 kit. Take a look at the datasheet/schematic. I used it with a motor once and it sang. Then I used it on a solenoid to make it close slowly, It was near a 1 mm thick $1000 4" diameter piece of glass. The glass was used for spectral shaping. The solenoid was part of the shutter mechanism.

I made other shutters (fiber high intensity light source) that used a rotary solenoid. No PWM ing.

 

[spec]

I have an old 1970 TTL Cookbook. I recently bought the newest version of the TTL Cookbook and it has new information the old book does not have like adjustable Duty Cycle. The new book only shows 50% to 99.9% duty cycle but I can see it can be 5% to 50% or 5% to 99% duty cycle with a duel variable resistor. 555 is very unstable with temperature changes. I can see a new experiment very soon.

Hi gary,

Just to put answer your concerns, the 555 chip, or perhaps a more suitable chip, can be incorporated in a circuit to do 0% duty cycle to 100% duty cycle (effectively) without too much bother. The actual frequency of the switching waveform is unimportant, within reason: it is the mark to space ratio of the waveform that counts.

 

[spec]

Hi again garry,

can you say which case your MOSFETs have, from the attached image (TO-220 or TO-220F).
can you give the full part number of your MOSFET.

 

[jpanhalt]

Regarding the 1K resistor, here is what Gary350 posted in response to a request for HIS schematic:

I found this circuit online I think it is for a toy motor. Like I said already I changed this circuit this was the basic circuit I started with. This [sic, that?] was the original idea for the fork lift motor speed control.

He "thinks" it is for a toy motor and he has "changed" it. Neither comment tells me that he recognizes the problem with the 1K resistor. He has yet to post his circuit. Why not?

On the matter of a gate resistor, there used to be an excellent exprimental piece by Fairchild that is no longer on its site. I have uploaded it here (AB-9). This current application note from Fairchild incorporates much of that information: https://www.fairchildsemi.com/application-notes/AN/AN-9005.pdf If one compares Fairchild's advice with IR's numerous notes on gate drives, you will find that both recommend a gate resistor to address ringing when using a good, low impedance driver for power mosfets. IR tends to recommend a slightly higher value resistor than Fairchild recommends. When using paralleled mosfets, remember that the gate resistors are effectively in parallel when viewed from the driver. The resistors shown in the most recent image are 15 Ω -- the first band is brown. I have a problem with brown and violet, which makes a bigger difference.

Regarding soldering the tab, that was common practice when I was designing those drives (last 1990's). While most us assume the die is attached to each pin with internal wires, that may not always be the case. In this note from IXYS (https://www.ixys.com/Documents/AppNotes/IXAN0061.pdf ) discussing temperature gradients across the die suggests that the electrical connection to the drain may be different than to the source:

In the first design, I used copper bus bars as mentioned and those bars were attached directly to the heat sink with a thermally conductive, electrically insulative, hardening material from Loctite. I took advantage of the fact that the intact surface of anodized aluminum is an insulator. Heat was not a problem, so in the second design, I used heavy copper traces. Heat has not been a problem either. I understand your concern, Gary, about heat transfer:

How can a PC board handle 225 amps, seems like the copper would burn off of the epoxy board. I have seen PC boards burn with less amps than that.

. Please show your calculations. BTW, the second drive was for considerably less power than the first -- only about 2 HP.

In any event, it seems most of the banter in this thread is about opinions. I have documented my opinions with reference citations and working models. If we are to move ahead, I hope others will meet similar standards of scientific discourse. Most important, Gary350 needs to post his schematic and describe what he is trying to do. There is a huge difference in controlling a tiny motor drawing less than 12 mA and one drawing 35 A or more. Do we know whether Gary's motor is brushless, PM, shunt or series wound? If he is interested in fork-lift motors, I highly recommend that he review 4QD's site (https://www.4qdtec.com/pwm-01.html ).

John

 

[spec]

 

[gary350]

Regarding the 1K resistor, here is what Gary350 posted in response to a request for HIS schematic:

He "thinks" it is for a toy motor and he has "changed" it. Neither comment tells me that he recognizes the problem with the 1K resistor. He has yet to post his circuit. Why not?

On the matter of a gate resistor, there used to be an excellent exprimental piece by Fairchild that is no longer on its site. I have uploaded it here (AB-9). This current application note from Fairchild incorporates much of that information: https://www.fairchildsemi.com/application-notes/AN/AN-9005.pdf If one compares Fairchild's advice with IR's numerous notes on gate drives, you will find that both recommend a gate resistor to address ringing when using a good, low impedance driver for power mosfets. IR tends to recommend a slightly higher value resistor than Fairchild recommends. When using paralleled mosfets, remember that the gate resistors are effectively in parallel when viewed from the driver. The resistors shown in the most recent image are 15 Ω -- the first band is brown. I have a problem with brown and violet, which makes a bigger difference.

Regarding soldering the tab, that was common practice when I was designing those drives (last 1990's). While most us assume the die is attached to each pin with internal wires, that may not always be the case. In this note from IXYS (https://www.ixys.com/Documents/AppNotes/IXAN0061.pdf ) discussing temperature gradients across the die suggests that the electrical connection to the drain may be different than to the source:

View attachment 96095

In the first design, I used copper bus bars as mentioned and those bars were attached directly to the heat sink with a thermally conductive, electrically insulative, hardening material from Loctite. I took advantage of the fact that the intact surface of anodized aluminum is an insulator. Heat was not a problem, so in the second design, I used heavy copper traces. Heat has not been a problem either. I understand your concern, Gary, about heat transfer: . Please show your calculations. BTW, the second drive was for considerably less power than the first -- only about 2 HP.

In any event, it seems most of the banter in this thread is about opinions. I have documented my opinions with reference citations and working models. If we are to move ahead, I hope others will meet similar standards of scientific discourse. Most important, Gary350 needs to post his schematic and describe what he is trying to do. There is a huge difference in controlling a tiny motor drawing less than 12 mA and one drawing 35 A or more. Do we know whether Gary's motor is brushless, PM, shunt or series wound? If he is interested in fork-lift motors, I highly recommend that he review 4QD's site (https://www.4qdtec.com/pwm-01.html ).

John

My plan is to put about 20 TO-220 mosfets in parallel. I had to remove the 1k resistor in the above circuit to get my little motor to run. I was hoping 20 mosfets will be enough. I don't have a

Regarding the 1K resistor, here is what Gary350 posted in response to a request for HIS schematic:

He "thinks" it is for a toy motor and he has "changed" it. Neither comment tells me that he recognizes the problem with the 1K resistor. He has yet to post his circuit. Why not?

On the matter of a gate resistor, there used to be an excellent exprimental piece by Fairchild that is no longer on its site. I have uploaded it here (AB-9). This current application note from Fairchild incorporates much of that information: https://www.fairchildsemi.com/application-notes/AN/AN-9005.pdf If one compares Fairchild's advice with IR's numerous notes on gate drives, you will find that both recommend a gate resistor to address ringing when using a good, low impedance driver for power mosfets. IR tends to recommend a slightly higher value resistor than Fairchild recommends. When using paralleled mosfets, remember that the gate resistors are effectively in parallel when viewed from the driver. The resistors shown in the most recent image are 15 Ω -- the first band is brown. I have a problem with brown and violet, which makes a bigger difference.

Regarding soldering the tab, that was common practice when I was designing those drives (last 1990's). While most us assume the die is attached to each pin with internal wires, that may not always be the case. In this note from IXYS (https://www.ixys.com/Documents/AppNotes/IXAN0061.pdf ) discussing temperature gradients across the die suggests that the electrical connection to the drain may be different than to the source:

View attachment 96095

In the first design, I used copper bus bars as mentioned and those bars were attached directly to the heat sink with a thermally conductive, electrically insulative, hardening material from Loctite. I took advantage of the fact that the intact surface of anodized aluminum is an insulator. Heat was not a problem, so in the second design, I used heavy copper traces. Heat has not been a problem either. I understand your concern, Gary, about heat transfer: . Please show your calculations. BTW, the second drive was for considerably less power than the first -- only about 2 HP.

In any event, it seems most of the banter in this thread is about opinions. I have documented my opinions with reference citations and working models. If we are to move ahead, I hope others will meet similar standards of scientific discourse. Most important, Gary350 needs to post his schematic and describe what he is trying to do. There is a huge difference in controlling a tiny motor drawing less than 12 mA and one drawing 35 A or more. Do we know whether Gary's motor is brushless, PM, shunt or series wound? If he is interested in fork-lift motors, I highly recommend that he review 4QD's site (https://www.4qdtec.com/pwm-01.html ).

John

This in my circuit with 20 mosfets in parallel.

**broken link removed**

But this circuit will be much better.

**broken link removed**

 

[jpanhalt]

Your first circuit has the problem that the mosfets will dissipate a lot of heat, depending on the motor being driven, as already discussed. What are the specifications for the motor or whatever are you driving?

The second circuit is much better but is limited by the max sourcing current of the 555 (200 mA). Hence, the rather large gate resistors. It will be problematic with 20 power mosfets at a current level that would justify using that number of devices. A purpose-designed driver can source several amperes. So, moving forward, what circuit do you plan to build?

John

 

[chemelec]

GARY350........
The Circuit I Posted Can go from a 1% to a 99% Duty Cycle, Depending on the Total Resistance of the Frequency Potentiometer.
Yes 555's are Not very Frequency Stable with Temperature or Supply Voltage Variations, But Frequency STABILITY is NOT a Problem with Pulse Width Modulation.

You Don't say Which Mosfets you have?

But Mosfets have Two Different Current Ratings.

As an Example:
An RFP50N05 is rated at 50 Amps RMS Continuous Current.
And Also Rated at 120 Amps When PULSED, Depending on the Pulse Width.