Author Topic: AC-DC 340v power supply board design help  (Read 1912 times)

0 Members and 1 Guest are viewing this topic.

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
AC-DC 340v power supply board design help
« on: January 19, 2020, 01:26:07 am »
I use an unconventional 2200w AC servo drive on my CNC milling machine, it's unconventional in the way that it does not have it's own DC power supply. All that is really necessary and what I have done in the past is use a bridge rectifier fed from 120v or 240v mains and push it through a few filter caps. This works fine as far as supplying power but it has the issue of high inrush current that will pop 10amp fuses immediately if the filter caps (I used 4 x 470uf).

I wanted to consolidate this to a single PCB and resolve this inrush issue but I'm having a bit of trouble with it. I worked up a schematic just to put it on paper but I'd like some thoughts and suggestions before I put much more brain power into it. At the moment I'm trying to figure out what value and power rating R2 should be, the shunt value, and I'll likely need a diode and some passives between the output of the INA138 and the IGBT. Wasn't sure whether to go with an IGBT or a mosfet for Q1 but I slapped the IGBT in there for the moment.

The idea there is that the voltage from the shunt should have the output from the INA138 operate the gate of the Q1. The Q1 should bypass the current limiting resistor R2:

Just a schematic to show what I'm thinking. Open to advice or being told I'm doing it completely wrong:
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: AC-DC 340v power supply board design help
« Reply #1 on: January 19, 2020, 01:44:27 am »
Put a NTC in series with the input and bypass it with a contactor. Simple and just works.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #2 on: January 19, 2020, 02:17:31 am »
Put a NTC in series with the input and bypass it with a contactor. Simple and just works.

Thought of this but I have not seen an NTC above 180w and I don't want to use an off board contactor (keeping it all on board is the whole point). I haven't seen a small PCB mount mechanical relay that can handle 30amps and 300v. I need to rate the components at about 30amps since if using a lower input voltage the current will be much higher, also need to maintain the abillity to use 240v on the input side for a 340v output.
 

Offline Prehistoricman

  • Regular Contributor
  • *
  • Posts: 216
  • Country: gb
Re: AC-DC 340v power supply board design help
« Reply #3 on: January 19, 2020, 01:04:36 pm »
Perhaps you could use a triac + diac dimmer circuit to slowly raise the input voltage.

Thought of this but I have not seen an NTC above 180w and I don't want to use an off board contactor (keeping it all on board is the whole point).
You're probably looking in the wrong component category:
digikey link to inrush current limiters

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: AC-DC 340v power supply board design help
« Reply #4 on: January 19, 2020, 01:47:35 pm »
...it's unconventional in the way that it does not have it's own DC power supply...

Incorrect. It is actually pretty conventional, for a servo driver to NOT have its own DC supply. This way multiple servo controllers can share the same DC bus, so if one of them is regeneratively braking (which for a servo driver is pretty common task), other servo driver can consume the energy, instead of dissipating it into a brake resistor.

You have probably meant inconvenient ("pain in the arse to use"), not unconventional ( i.e. "not standard").
 

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #5 on: January 19, 2020, 04:08:44 pm »
Let me backtrack and say that this is common for lower voltage DC servo drives, ~75vDC or so. Similar to stepper motor drives. It's not very typical for 400v class servo drives. This particular drive is suited for a niche market and has a few extremely convenient features so it serves a purpose.

 

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #6 on: January 19, 2020, 04:18:26 pm »
Perhaps you could use a triac + diac dimmer circuit to slowly raise the input voltage.

Thought of this but I have not seen an NTC above 180w and I don't want to use an off board contactor (keeping it all on board is the whole point).
You're probably looking in the wrong component category:
digikey link to inrush current limiters

Well you said "NTC" which is what I knew them as as well. That particular category on Digikey refers to them as ICLs. Look like practically the same thing but the current limits are much higher so they likely may be suitable. I'll have to look at it a bit more, thanks for the heads up.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: AC-DC 340v power supply board design help
« Reply #7 on: January 19, 2020, 05:40:57 pm »
Let me backtrack and say that this is common for lower voltage DC servo drives, ~75vDC or so. Similar to stepper motor drives. It's not very typical for 400v class servo drives. This particular drive is suited for a niche market and has a few extremely convenient features so it serves a purpose.

Then you have probably not seen a lot, nor you know how stuff is done.

Common DC bus is common even in large servo drives, where the DC bus is about 500-700V (400V three phase rectified).

Simply because it does make sense and conserves a bit of power dissipation.
 

Offline WattsThat

  • Frequent Contributor
  • **
  • Posts: 788
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #8 on: January 19, 2020, 06:16:31 pm »
+1 on the common bus arrangement and the standard “pre-charge” circuit is an appropriately sized fusible power resistor and a mechanical contactor to short the resistor when the bus voltage reaches about 66% of the rated voltage. This is bog standard on three phase ac inverter drives. It’s only when you reach 50kw or so do they resort to replacing three of the six the diodes with phase controlled thyristors for the precharge.
 
The following users thanked this post: SeanB

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #9 on: January 19, 2020, 11:15:47 pm »
Let me backtrack and say that this is common for lower voltage DC servo drives, ~75vDC or so. Similar to stepper motor drives. It's not very typical for 400v class servo drives. This particular drive is suited for a niche market and has a few extremely convenient features so it serves a purpose.

Then you have probably not seen a lot, nor you know how stuff is done.

Common DC bus is common even in large servo drives, where the DC bus is about 500-700V (400V three phase rectified).

Simply because it does make sense and conserves a bit of power dissipation.

Arguing semantics about things that do not really even concern the topic of the post is pointless and diversionary, I work with many servo drives with built in DC bus power supplies from old black box Yaskawa's, PacSci's, Bosch Rexroth, Lenze......etc. It would be the type  of machine they are used in that would dictate the use of a common DC bus........something like a machine tool with several large servos in close proximity might make sense, I don't work with those. So maybe it's more common than I thought....OK lets move on and try to mind your manners.
 

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #10 on: January 19, 2020, 11:27:21 pm »
+1 on the common bus arrangement and the standard “pre-charge” circuit is an appropriately sized fusible power resistor and a mechanical contactor to short the resistor when the bus voltage reaches about 66% of the rated voltage. This is bog standard on three phase ac inverter drives. It’s only when you reach 50kw or so do they resort to replacing three of the six the diodes with phase controlled thyristors for the precharge.

Well it's only "common" if there's more than 1 of something using it. If I intended to use a mechanical contractor I really would not have bothered posting because that's rather simple to do. I can read DC bus voltage from my controller and operate a contactor coil to bypass the current limiting device pretty easily but it's not what I'm trying to do. Realistically the solid state components to do this are super cheap, figuring it out is only a bit more difficult which is why I posted on an electronics forum.

For argument's sake we can forget that this has anything to do with servos and just think of it as a self contained AC-DC power supply.
 

Offline WattsThat

  • Frequent Contributor
  • **
  • Posts: 788
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #11 on: January 20, 2020, 04:28:59 am »
Sorry, didn’t see your attachment before commenting. I’m not a power electronics guy but I know enough to know that’s not going to work - but it would be an impressive failure when the power was applied.

The INA138 will not withstand 340 volts common mode.

I’d also expect an IGBT to require 10V on the gate to have it fully turned on, gonna need a gate driver there.

https://www.infineon.com/dgdl/Infineon-Simple_High_Side_Drive_Provides_Fast_Sw-Article-v01_00-EN.pdf?fileId=5546d46254e133b401554de58b225d15

Ignoring the above and to answer your question, selecting R3 is a matter of how fast you want to charge the caps and what the selected resistance will withstand with respect to fusing current. To get to 63% of Vin, t = rc and 100% requires 5 t. It’s hard to find fusible resistors in the size that will be required, I’ve seen regular 5-7w ceramic power resistors encased in fiberglass sleeving, sometimes paralleled parts to get to the required ratings.

At the end of the day, hopefully you’ll figure out why mechanical relays get used in this application. Throwing complexity at a simple problem always fails in some way.
 

Offline blazini36Topic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC-DC 340v power supply board design help
« Reply #12 on: January 20, 2020, 05:51:39 am »
Sorry, didn’t see your attachment before commenting. I’m not a power electronics guy but I know enough to know that’s not going to work - but it would be an impressive failure when the power was applied.

The INA138 will not withstand 340 volts common mode.

I’d also expect an IGBT to require 10V on the gate to have it fully turned on, gonna need a gate driver there.

https://www.infineon.com/dgdl/Infineon-Simple_High_Side_Drive_Provides_Fast_Sw-Article-v01_00-EN.pdf?fileId=5546d46254e133b401554de58b225d15

Ignoring the above and to answer your question, selecting R3 is a matter of how fast you want to charge the caps and what the selected resistance will withstand with respect to fusing current. To get to 63% of Vin, t = rc and 100% requires 5 t. It’s hard to find fusible resistors in the size that will be required, I’ve seen regular 5-7w ceramic power resistors encased in fiberglass sleeving, sometimes paralleled parts to get to the required ratings.

At the end of the day, hopefully you’ll figure out why mechanical relays get used in this application. Throwing complexity at a simple problem always fails in some way.

About the INA138, you're right. I was just trying to get an initial drawing down, like I said just spitballing.....not plugging anything in at the moment. I worked on it a bit more today (attached). The INA138 is pointless anyway because a suitable small shunt does not seem to exist. Relay or not it still has to be controlled, and I don't want to externally control the bypass relay. I need a voltage reference from the DC bus side to control the bypass mechanism that I'm still trying to figure out.

I added the NTC as the current limiting device and moved it and the IGBT to the negative DC bus rail. Added gate driven triacs to the input side. The triacs on both input lines are redundant and I'll likely remove 1. haven't added anything to drive the gate of the IGBT yet

 

Offline station240

  • Supporter
  • ****
  • Posts: 967
  • Country: au
Re: AC-DC 340v power supply board design help
« Reply #13 on: January 20, 2020, 05:45:37 pm »
You'd be better off adding a Power Factor Correction (PFC) circuit on the front.
Easy enough to change the feedback resistors to get 340V DC.
TI and others have dedicated PFC controllers, everything else is just a boost converter.
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: AC-DC 340v power supply board design help
« Reply #14 on: January 20, 2020, 06:39:21 pm »
You'd be better off adding a Power Factor Correction (PFC) circuit on the front.
Easy enough to change the feedback resistors to get 340V DC.
TI and others have dedicated PFC controllers, everything else is just a boost converter.
Only if he's in a commercial space or is running into the current limits of the supply. PFC doesn't really have much benefit in lower powered residential applications.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf