DC-DC boost converter design process, build, and testing
6905
4
I have already made a DC-DC converter which takes 12V or 24V from batteries and boosts it to about 300 VDC for use as the DC link voltage for a standard 240 Volt VF drive and three phase motor. This was actually the result of a long process of various trial and error attempts, simulations, and tests, and now I am trying to complete the process by making observations about my results and implementing changes which will result in a reliable unit that will do what I need. And then I may try a new design which should be more efficient, less expensive, smaller, and more practical. So I'd like to use this thread to explain the design process from my perspective, and perhaps be of use to others who may want to attempt something similar. And as I proceed (since there is yet more work to be done), I'd appreciate any suggestions on better ways to do this. 
To begin, I'll provide the basic specifications. I want to be able to power and control a 2HP three phase AC motor, on a small utility vehicle or lawn tractor, using batteries of 12 volts to 48 volts total. The vehicle weighs roughly 100 pounds and I would expect a fully loaded weight of about 500 pounds including batteries, motor, controller, and rider. For this project it will be used only as a utility vehicle, maybe pulling a small cart, and it should have an operating time of 1/2 to 1 hour. Speed will be no more than 6 MPH.
So here's a rough calculation of what is needed. This should actually be done before selection of the motor, but I'll just verify that it's adequate. I'll start with the maximum grade, which I'll say might be 20%. So for a 500 pound vehicle that requires a force of 100 pounds exerted by the tires. They are roughly 16" diameter, for a radius of 8". Thus I will need a torque of 100*8/12 = 67 Lb-Ft. I want a top speed of 6 MPH or 528 ft/min, and each rotation of the wheel is 3.14*16 = 50.2 inches = 4.18 feet. So I need a shaft speed of 126 RPM.
I'll use a 3450 RPM motor since that is similar to the ICE speed, (and its what I already have), so I need a reduction of 27:1. The rated torque of a the motor is 2HP*5252/3450 = 3.04 lb-ft. So with the reduction drive I will have about 82 lb-ft of torque. I needed 67 so it seems more than adequate. Of course there are losses due to friction, but 81% efficiency seems reasonable. But this is for the rated motor torque, and it should have short term overload capacity of 2x to 3x. So I'm good!
Now I'll calculate the amount of energy I will need. For a tractor, rolling resistance and grade are the overwhelming factors. I can estimate the rolling resistance on the expected rough surface by simply pushing or pulling it with a spring scale. I need to do this, but for now I'll estimate that it takes 20 pounds to push it. For a round trip, the average grade will be zero, but since regeneration is only good for about 20% and I might not even implement it, I'll assume a 5% grade over half the distance, or effectively 2.5%. This requires a force of 12.5 pounds plus the 20 for rolling resistance, or 32.5 pounds. This is a torque of 21.6 lb-ft and at top speed of 126 RPM that is about 1/2 HP or 389 watts. So if I want to run at this speed for 1 hour I need 389 W-Hr. If I use a single 12V 105 A-Hr deep cycle battery, that is 1260 W-Hr and 389 watts will draw 32 amps. This is 1/3C so I can probably get at least 50 A-Hr and maybe 1.5 hours of use. So far, so good. Even allowing for 67% overall efficiency I meet my 1 hour goal.
Enough for now. I want to break this up into reasonable chunks, so I'll get into the design considerations for the DC-DC converter in another post in this thread. BTW, as a reality check, I have already taken this vehicle on a couple of test runs and it drew about 15 amps at 24 volts, or 360 watts.
To begin, I'll provide the basic specifications. I want to be able to power and control a 2HP three phase AC motor, on a small utility vehicle or lawn tractor, using batteries of 12 volts to 48 volts total. The vehicle weighs roughly 100 pounds and I would expect a fully loaded weight of about 500 pounds including batteries, motor, controller, and rider. For this project it will be used only as a utility vehicle, maybe pulling a small cart, and it should have an operating time of 1/2 to 1 hour. Speed will be no more than 6 MPH.
So here's a rough calculation of what is needed. This should actually be done before selection of the motor, but I'll just verify that it's adequate. I'll start with the maximum grade, which I'll say might be 20%. So for a 500 pound vehicle that requires a force of 100 pounds exerted by the tires. They are roughly 16" diameter, for a radius of 8". Thus I will need a torque of 100*8/12 = 67 Lb-Ft. I want a top speed of 6 MPH or 528 ft/min, and each rotation of the wheel is 3.14*16 = 50.2 inches = 4.18 feet. So I need a shaft speed of 126 RPM.
I'll use a 3450 RPM motor since that is similar to the ICE speed, (and its what I already have), so I need a reduction of 27:1. The rated torque of a the motor is 2HP*5252/3450 = 3.04 lb-ft. So with the reduction drive I will have about 82 lb-ft of torque. I needed 67 so it seems more than adequate. Of course there are losses due to friction, but 81% efficiency seems reasonable. But this is for the rated motor torque, and it should have short term overload capacity of 2x to 3x. So I'm good!
Now I'll calculate the amount of energy I will need. For a tractor, rolling resistance and grade are the overwhelming factors. I can estimate the rolling resistance on the expected rough surface by simply pushing or pulling it with a spring scale. I need to do this, but for now I'll estimate that it takes 20 pounds to push it. For a round trip, the average grade will be zero, but since regeneration is only good for about 20% and I might not even implement it, I'll assume a 5% grade over half the distance, or effectively 2.5%. This requires a force of 12.5 pounds plus the 20 for rolling resistance, or 32.5 pounds. This is a torque of 21.6 lb-ft and at top speed of 126 RPM that is about 1/2 HP or 389 watts. So if I want to run at this speed for 1 hour I need 389 W-Hr. If I use a single 12V 105 A-Hr deep cycle battery, that is 1260 W-Hr and 389 watts will draw 32 amps. This is 1/3C so I can probably get at least 50 A-Hr and maybe 1.5 hours of use. So far, so good. Even allowing for 67% overall efficiency I meet my 1 hour goal.
Enough for now. I want to break this up into reasonable chunks, so I'll get into the design considerations for the DC-DC converter in another post in this thread. BTW, as a reality check, I have already taken this vehicle on a couple of test runs and it drew about 15 amps at 24 volts, or 360 watts.
1 - 5 of 5 Posts
Joined
·
120 Posts
its interesting that you should start this thread. i've been running simulations for the last three weeks, for a hv dc converter. im designing a converter with input of 250v dc, to output 650v dc.
im simulating with proteus and everything looks good on simulation. i noticed you use ltspice for your simulations. i have the full version of orcad pspice version 16.5, and its just too complicated ffor me. i just got done figuring out proteus.
for those questioning this direction, (converter design) we need this higher voltages to power ac industrial controllers, which feed on 320v dc and 640vdc (240v or 480v ac depending on drive).
getting ready to buy control parts. glad you have a working version, and keep us updated with the developments.
im simulating with proteus and everything looks good on simulation. i noticed you use ltspice for your simulations. i have the full version of orcad pspice version 16.5, and its just too complicated ffor me. i just got done figuring out proteus.
for those questioning this direction, (converter design) we need this higher voltages to power ac industrial controllers, which feed on 320v dc and 640vdc (240v or 480v ac depending on drive).
getting ready to buy control parts. glad you have a working version, and keep us updated with the developments.
dc-dc I built is 390v to 765v at 330 amps based on 765v.
the 765V was derived from the output of the 150KW fuel cells from Ballard. The 390v was my orginal Battery bank when working with 96 kw 3 phase.
I did not want to invest more into batteries at the time, waiting for better price and specs.
the 765V was derived from the output of the 150KW fuel cells from Ballard. The 390v was my orginal Battery bank when working with 96 kw 3 phase.
I did not want to invest more into batteries at the time, waiting for better price and specs.
Joined
·
3,143 Posts
I won't go into the mechanical design very much as it is posted elsewhere. I just mounted the motor (which happened to be a C-face) on a plate where the ICE had been, and I attached a small pulley which connected with a V-belt to the large horizontal pulley on the existing 5 speed tranny. It used a sprocket and chain drive to the differential.
I tested it using an extension cord and 240 VAC. Fortunately the VFD could be powered with single pahse, and most of the default motor settings wotked so I could run it from the front panel pushbuttons.
So, on to the design of the DC-DC. I started with a small 175 watt inverter I had, and I was able to make about 320 VDC by using a FWB doubler circuit with two large capacitors I had from surplus. The inverter provided a +/-160 V peak at 30% duty cycle as a stepped sine wave approximation. Here is the circuit:
![]()
In the simulation it provides about 720 watts with 760 watts input, and this was for a larger inverter I planned to buy. I was able to get the motor to spin, but any load tripped the VFD on undervoltage. With a 300 watt inverter I was just barely able to move. Obviously I needed more power. My first two videos show the progress up to this point, although they do not show my more successful ride on 220VAC, and I did not film the first attempts with the inverters:
http://youtu.be/SGd8i6dp4SY
http://youtu.be/DdvscTp3thw
I determined that the automotive inverters have an internal DC bus of about 160 VDC but I need twice that, and it is inefficient to use the 30% switched outputs. I also thought about using two in series with two batteries, but these inverters are not isolated, and so there would be high voltage on the batteries. I was going to purchase a 220 VAC unit but the 1000 to 3000 watt units were a couple hundred bucks, and still might not be what I really want.
So, I decided that I needed to build my own DC-DC converter. My next post will get into that design...
I tested it using an extension cord and 240 VAC. Fortunately the VFD could be powered with single pahse, and most of the default motor settings wotked so I could run it from the front panel pushbuttons.
So, on to the design of the DC-DC. I started with a small 175 watt inverter I had, and I was able to make about 320 VDC by using a FWB doubler circuit with two large capacitors I had from surplus. The inverter provided a +/-160 V peak at 30% duty cycle as a stepped sine wave approximation. Here is the circuit:
In the simulation it provides about 720 watts with 760 watts input, and this was for a larger inverter I planned to buy. I was able to get the motor to spin, but any load tripped the VFD on undervoltage. With a 300 watt inverter I was just barely able to move. Obviously I needed more power. My first two videos show the progress up to this point, although they do not show my more successful ride on 220VAC, and I did not film the first attempts with the inverters:
http://youtu.be/SGd8i6dp4SY
http://youtu.be/DdvscTp3thw
I determined that the automotive inverters have an internal DC bus of about 160 VDC but I need twice that, and it is inefficient to use the 30% switched outputs. I also thought about using two in series with two batteries, but these inverters are not isolated, and so there would be high voltage on the batteries. I was going to purchase a 220 VAC unit but the 1000 to 3000 watt units were a couple hundred bucks, and still might not be what I really want.
So, I decided that I needed to build my own DC-DC converter. My next post will get into that design...
1 - 5 of 5 Posts
