Jump to content


Photo

Soft starting Air Conditioner Compressor


  • Please log in to reply
5 replies to this topic

#1 RCinFL

RCinFL

    Junior Member

  • Members
  • PipPip
  • 2 posts
  • Location:USA
  • Interests:Electrical Engineer

Posted 19 September 2005 - 02:49 AM

I heard you folks down under have or soon to have regulation on starting currents not to exceed twice run current.

I live in S. Florida, USA and have two Xantrex 5548 plus inverters in series connected mode (two 5.5 kw units) along with 15KW Generac Ultrasource generator for hurricane electric backup.

I have a Lennox 3.5 ton central A/C unit with Copeland scroll compressor. It's run current is 16 A (PF about 0.86), when PF corrected to about 95%, run current drops to about 13.5 amps. (single phase 240 vac)

My problem is the starting current for the compressor is about 120 A rms which last for 250 msec (PF about 0.75 during startup period). LRA rating on compressor is 104 A.

I have a closed loop hall effect current sensor that has 200 KHz bandwidth to capture startup profile on digital storage scope.

I bought a IC Electronic STL-3-4025 in attempt to moderate the startup surge. Not much luck, the compressor wants about 90 amps to get rolling and it is very tough to play with softstarter setting with a sealed compressor that you can not see the startup rotation. I managed to get the current down to about 90 amps but it take about 2.5 sec for compressor to come up to speed (vs. 120A for 0.25 sec). Not a very good trade off and rough on the compressor.

The compressor motor is permanent split phase capacitor design. Starter cap is 45 uF and draws about 5 amps during start phase and rises to about 7 amps while running. I connected the softstart only to the run winding, leaving the start winding/capacitor going to full 240 vac.

Most all of the start current is going into the run winding and it is pretty flat at 120 A rms during the 250 msec startup until motor gets near run speed. The softstart ramp from initial torque setting is not really helping me. I need an initial torque setting that, once compressor is rolling at about 85% speed, jumps to full 240 vdc,

I don't believe the A/C is overcharged with Freon as the run current is okay.

I am thinking about doing a transformer step startup. I have a couple of 2KW 120/240 vac isolation transformers that when connected as autotransformer are capable of 4KW each. I don't think I can get the compressor rolling with 120 vac though.

Anybody have some ideas? Anyone with experience with residential A/C compressors and what types have lower startup current.

#2 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,642 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 19 September 2005 - 04:33 AM

Hello RCinFL

Welcome to the forum.
I have not seen the requirment for 2 x FLC start current limit. That will mean that only VSDs can be used with motors.
The problem is that as you reduce the start current, you reduce the start torque by the current reduction squared.
The compressor has a minimum start torque requirement and this is a function of the compressor design and set up. You must generate more than this torque or you will not start the compressor.
I expect that the compressor that you are using is one of the sealed hermetic compressors and there is little that you can do to "off load" during start. If you were able to some how of load during start, you could then reduce the start current.

I see that you are using a hall effect ct and scope to measure the current. Be careful to look at the average current, not the peak current on the scope. The avarage current should be quite a bit lower.

Another thing to try with a capacitor start motor, is to connect the capacitor start winding across the line (full voltage) and reduce the run winding voltage only with the soft starter. You may find that this gives you a better start torque.l

With hermetic compressors, if you slow down the acceleration too much, you increase the starting torque requirement.
Being a positive displacement compressor, there are a number of revolutions required to reach pressure. You need to get to full speed before presure is reached.

Hope this helps,
Best regards,

#3 msdaif

msdaif

    Intermediate Member

  • Full Member
  • PipPipPip
  • 27 posts

Posted 19 September 2005 - 10:05 AM

Dear RCinFL

I think it would be simpler and cheaper to reduce the starting current
with, say, a 3-step autotransformer (step 1 at 80V, step 2 at 160V and step 3 at 240V). This method of starting is commonly applied to huge pumps. You will also need timers and contactors to short out the taps gradually.

However, I urge you to contact the compressor agent and you may need the model number and serial at Copeland Corp website:

http://ccl.copeland-...CCL_Default.jsp

#4 jraef

jraef

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 683 posts
  • Gender:Male
  • Location:USA, California

Posted 19 September 2005 - 03:45 PM

RCinFL,
I urge you to contact the compressor mfgr and tell them what you are planning before doing any further experimentation. Sealed compressors like that have issues with needing to be at-speed within 1 or 2 seconds because the lubrication is dependant on that. Slowing down the acceleration can cause you to damage the compressor beyond repair in short order.

Besides that, there is really nothing you can do with electronics on a split phase cap start motor that will last very long. The capacitors have a tendancy to damage solid state switching devices, the steep rise time of the solid state switches and the harmonics associated with their use tends to damage the capaciotrs. It is usually a race to see which one fails first. I would even check with the mfgr to see if it is appropriate to run them behind an inverter of any sort.

msdaif,
Autotransformer starting is for 3 phase motors, he has 1 phase. Reducing voltage with a transformer on a 1 phase cap start motor may cause the start switch to fail to open in time, leaving the starting cap in the circuit too long and causing it to overheat, thenfail.
"He's not dead, he's just pinin' for the fjords!"

#5 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,642 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 19 September 2005 - 09:42 PM

Hi jraef

Yes I agree, you do need to be careful about extending the start time for to long, however you can usually reduce the start current without causing problems provided that you start quickly. - you need to get to full speed within a certain number of revolutions, or the pressure will be max and you will need full voltage to start. I have seen these units where a slow start has been attempted wher they have happily spun up to half speed and then stopped as the pressure came on. I would expect that you would need about 60% start voltage and about a 1 second ramp or less.

Best regards,

#6 RCinFL

RCinFL

    Junior Member

  • Members
  • PipPip
  • 2 posts
  • Location:USA
  • Interests:Electrical Engineer

Posted 20 September 2005 - 01:18 AM

Thanks guys. I had a scare after about 5 start attempts with different softstarter settings, over 15 minute period, the compressor froze (at 5th start attempt) and not due to heat klixon opening. It still drew 120 A start current. I never let it go more then 2.5 or 3.5 seconds on any attempt.

Letting cool for about an hour (and a few prayers) it started up again. Head pressure looks fine hopefully no permanent damage.

I think it was the heat build up.

Looks like I better let it run a while on each startup profile for lubricant circulation. Needless to say I going to be very careful going forward.

To some of replies, 1) only doing run winding on softstarter, cap-start winding has full 240 v. 2) No starter disconnect switch, it is permanently connected split phase capacitor design. Calling it starter winding-cap is a little mis-labeling. Sort of like a carrot on a fishing pole held in front of a horse.

Problem is rotational startup drag (initial torque needed to startup).

I have modeled the startup on simulation program. Startup looks like a 1.45 ohm with 3.57mH series inductance. Once run is reached looks like 12.7 ohms with 20.5 mH series inductance.

Looks like options are limited. The more revolutions allowed during startup the more head pressure builds up, creating more startup load, requiring longer duration for starting current which does not drop until 85% to 90% operational speed is reached.

Thanks again for comments




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users