Buck Boost transformers on VFDs

[Jraef]

there you go, sounds like a solution

ot question
is it true vfds don't like a really stiff supply?
3-5% vdrop is prefered?

Stiffness? What Golddigger said.
VD? No.

Hmmm...taken out of context, that doesn't look so good. :roll:

Here's how I explain it to laypersons, you of course will likely understand more of the nuances behind this. When a rectifier is rectificating, the diodes don't really conduct in one direction continuously, they conduct only at the peaks of each sine wave; hence the "non-linear" nature of current draw by a rectifier. When feeding a DC bus with capacitors (as in a VFD), the caps charge and discharge instantly. So there is a ringing transient on the line side and a motor is running, a diode (or two) may cease to conduct for a cycle (or two) when the ringing of that transient takes the voltage below the forward conduction threshold of the diode(s). Because the caps were depleted by the transistors firing into the load, those caps will attempt to re-charge themselves instantly by pulling current through the NEXT valid diode circuit, AT THE AVAILABLE FAULT CURRENT. So the stiffer the system, the more "violent" that current spike through the diodes will be. You would have the same scenario every time you energize a VFD if it were not for the pre-charge circuit, usually a current limiting resistor. But once the VFD is already running, that pre-charge resistor is bypassed out of the circuit so it no longer protects the diodes. Adding impedance slows the rise time of that current surge, spreading the capacitor re-charging energy across more cycles and more diodes, lowering the peaks seen by any one diode. In an experiment we did a few years ago, we showed how a grid transient into a 25HP VFD caused a peak of 805A for one cycle through one diode. Adding a 3% reactor dropped that to 55A over 3-4 cycles, so multiple diodes as well.

 

[petersonra]

Can't you also put in a bus reactor though? I've often wondered why vfd manufacturers don't do this more often. Or at least offered as an option.

Incidentally those of you suggesting that Allen Bradley drives are probably not getting that this is apparently a low-budget project. I like Allen Bradley drives but they can be insanely expensive compared to the competition. And all the accoutrements are more expensive like circuit Breakers, line reactors, break resistors, Etc. You can easily be double the price with an Allen-Bradley solution compared to Schneider or Siemens. Unless the marketing manager really wants the project and beats on the circuit breaker product manager to get the price of the circuit breakers to a reasonable level. I Remember When Alan Bradley first came out with the 140g circuit breakers. The product manager assured us they would be competitive with other similar IEC style Breakers adapted to UL standards. The first time we tried to use them they were almost three times the price even with our alleged good discount. They have since adjusted their pricing somewhat but they are still the most expensive option to us. And they've given up claiming they are competitive. They just aren't. Most of the time the drives aren't particularly competitive either. Although sometimes the really large ones are. Which I find really strange since they don't make any of those themselves they just buy them and Rebrand them.

 

[Besoeker]

Stiffness? What Golddigger said.
VD? No.

Hmmm...taken out of context, that doesn't look so good. :roll:

Here's how I explain it to laypersons, you of course will likely understand more of the nuances behind this. When a rectifier is rectificating, the diodes don't really conduct in one direction continuously, they conduct only at the peaks of each sine wave;

That's not correct.

 

[petersonra]

That's not correct.

Sure it is. They only pass current during that part of the cycle when the voltage at the input is higher than the bus voltage.

 

[Ingenieur]

Sure it is. They only pass current during that part of the cycle when the voltage at the input is higher than the bus voltage.

plus the diodes conduction voltage 0.5-0.7 v


j/k

 

[kwired]

Can't you also put in a bus reactor though? I've often wondered why vfd manufacturers don't do this more often. Or at least offered as an option.

Incidentally those of you suggesting that Allen Bradley drives are probably not getting that this is apparently a low-budget project. I like Allen Bradley drives but they can be insanely expensive compared to the competition. And all the accoutrements are more expensive like circuit Breakers, line reactors, break resistors, Etc. You can easily be double the price with an Allen-Bradley solution compared to Schneider or Siemens. Unless the marketing manager really wants the project and beats on the circuit breaker product manager to get the price of the circuit breakers to a reasonable level. I Remember When Alan Bradley first came out with the 140g circuit breakers. The product manager assured us they would be competitive with other similar IEC style Breakers adapted to UL standards. The first time we tried to use them they were almost three times the price even with our alleged good discount. They have since adjusted their pricing somewhat but they are still the most expensive option to us. And they've given up claiming they are competitive. They just aren't. Most of the time the drives aren't particularly competitive either. Although sometimes the really large ones are. Which I find really strange since they don't make any of those themselves they just buy them and Rebrand them.

And for 10 Hp or smaller drives I think often many will just risk not using some of those accoutrements and use relatively cheap drives and just replace them if they fail. During component selection process they may consider how critical the drive is to the process and how much they can afford to be down when something fails. Spending twice as much on a small drive but some assurance it isn't as likely prone to failure can be much preferred over production losses when the cheap drive or one without the added accoutrements is almost assured to have more failures.

 

[Besoeker]

Sure it is. They only pass current during that part of the cycle when the voltage at the input is higher than the bus voltage.

It isn't.
With a well designed six pulse circuit they conduct fat 120 degrees.

 

[Ingenieur]

It isn't.
With a well designed six pulse circuit they conduct fat 120 degrees.

are you saying
For a 1/2 cycle 180 deg sinusoidal voltage
ac voltage (sqrt2 x rms) is greater than bus v (1.35 x rms) for 120 deg of the 180 deg 1/2 cycle?

 

[petersonra]

And for 10 Hp or smaller drives I think often many will just risk not using some of those accoutrements and use relatively cheap drives and just replace them if they fail. During component selection process they may consider how critical the drive is to the process and how much they can afford to be down when something fails. Spending twice as much on a small drive but some assurance it isn't as likely prone to failure can be much preferred over production losses when the cheap drive or one without the added accoutrements is almost assured to have more failures.

Most of the time for the small drives the $50 or $100 difference between an Allen Bradley Drive and a lesser cost one will be eaten up by the amount of extra engineering time it takes to integrated into the system if it is a compact Logix plc. Allen Bradley is done a really good job of making it easy to use their drives with their plc. Schneider has tried to make it easier to make their drives work with Allen Bradley PLC but it never seems to work quite as easily as it should. There are always these mysterious register settings that aren't in the manual that have to be made to get it to work right. And the profiles don't always seem to quite work like they should.

 

[Ingenieur]

Duplicate

 

[kwired]

Most of the time for the small drives the $50 or $100 difference between an Allen Bradley Drive and a lesser cost one will be eaten up by the amount of extra engineering time it takes to integrated into the system if it is a compact Logix plc. Allen Bradley is done a really good job of making it easy to use their drives with their plc. Schneider has tried to make it easier to make their drives work with Allen Bradley PLC but it never seems to work quite as easily as it should. There are always these mysterious register settings that aren't in the manual that have to be made to get it to work right. And the profiles don't always seem to quite work like they should.

I can see that being an issue. Most my drive experiences were either manually entering frequency reference or using analog references, which is roughly the same task with an Allen Bradley drive as it is with an Automation Direct drive. Those that were connected to PLC's were typically already a part of someone else's complete assembly and we just don't even try to mess with something that already works as is, I would opt for the correct transformer like the OP has done in that situation, but for a more simple and somewhat stand alone drive I may have better chance of just risking a cheap drive over a more expensive drive along with purchasing a transformer on top of it - especially if it is temporary in the first place.

 

[Ingenieur]

Did the math
ignore diode cond v
1 cycle = 360 deg
magnitude = 1

a single diode conducts for 68 deg out of 360 deg
about 19% of the time/cycle and when magnitude >83% on the positive swing
sounds like 'peak' is a fair description?

 

[Besoeker]

are you saying
For a 1/2 cycle 180 deg sinusoidal voltage
ac voltage (sqrt2 x rms) is greater than bus v (1.35 x rms) for 120 deg of the 180 deg 1/2 cycle?

I'm saying that, in a well designed six pulse bridge, each diode or SCR conducts for 120 degrees. That's not opinion - it's fact.
Will this fact get deleted too???

 

[Besoeker]

Did the math
ignore diode cond v
1 cycle = 360 deg
magnitude = 1

a single diode conducts for 68 deg out of 360 deg

Not correct for a six pulse bridge.

 

[Ingenieur]

Not correct for a six pulse bridge.

obviously (do you mean a 3 ph full wave rectifier?)
but not germane to the discussion

his comment:
"When a rectifier is rectificating, the diodes don't really conduct in one direction continuously, they conduct only at the peaks of each sine wave;
your response:
"That's not correct." (note no explanation of why it is wrong)


he then responded and clarified by saying that current only flows if input > bus V:
"Sure it is. They only pass current during that part of the cycle when the voltage at the input is higher than the bus voltage."
your reply:
"It isn't.... With a well designed six pulse circuit they conduct fat 120 degrees."
which did not address his comment, he said nothing about duration, only magnitude, but he is still correct for a 3 ph full wave rectifier, current only flows when input V (any phase) > than bus V, which is essentially at 83% of peak

bottom line: current only flows at ~peak voltage, ie, 3 sqrt2 /2 Pi x Vrms < sqrt2/sqrt3 x Vrms or 0.67/0.82 ~ 83%
so he was correct, current flows at ~ peak

 

[Besoeker]

obviously (do you mean a 3 ph full wave rectifier?)
but not germane to the discussion

his comment:
"When a rectifier is rectificating, the diodes don't really conduct in one direction continuously, they conduct only at the peaks of each sine wave;
your response:
"That's not correct." (note no explanation of why it is wrong)


he then responded and clarified by saying that current only flows if input > bus V:
"Sure it is. They only pass current during that part of the cycle when the voltage at the input is higher than the bus voltage."
your reply:
"It isn't.... With a well designed six pulse circuit they conduct fat 120 degrees."
which did not address his comment, he said nothing about duration, only magnitude, but he is still correct for a 3 ph full wave rectifier, current only flows when input V (any phase) > than bus V, which is essentially at 83% of peak

bottom line: current only flows at ~peak voltage, ie, 3 sqrt2 /2 Pi x Vrms < sqrt2/sqrt3 x Vrms or 0.67/0.82 ~ 83%
so he was correct, current flows at ~ peak

A well designed three phase rectifier, controlled or uncontrolled will have a DC link reactor (choke). This results in continuous current. Each device thus conducts for 120 degrees. The DC link current is about 1.225 times the input line current.

 

[petersonra]

A well designed three phase rectifier, controlled or uncontrolled will have a DC link reactor (choke). This results in continuous current. Each device thus conducts for 120 degrees. The DC link current is about 1.225 times the input line current.

This kind of goes back to an unanswered question I posed about why more VFDs don't come with such reactors as standard.

I am not seeing how you get the semiconductors to conduct longer unless you are somehow reducing the bus voltage with the choke. Which I guess it would if the refill to the capacitors is slowed down some by restricting the current.

 

[Ingenieur]

A well designed three phase rectifier, controlled or uncontrolled will have a DC link reactor (choke). This results in continuous current. Each device thus conducts for 120 degrees. The DC link current is about 1.225 times the input line current.

you are skirting the issue
voltage, not current
it conducts only when V line > V bus, and essentially at peak
that was his point

 

[GoldDigger]

you are skirting the issue
voltage, not current
it conducts only when V line > V bus, and essentially at peak
that was his point

And you seem to be deliberately ignoring the voltage drop/rise across an inductor when you state that V line must be greater than V bus for conduction to occur.

Sent from my XT1585 using Tapatalk

 

[Ingenieur]

And you seem to be deliberately ignoring the voltage drop/rise across an inductor when you state that V line must be greater than V bus for conduction to occur.

Sent from my XT1585 using Tapatalk

that is not what was being discussed
V line > V bus only near peak V line
that is correct, not incorrect