What's new
What's new

15 hp vfd for 5 hp motor on compressor

For the following reasons I would not use Basic mode with the Hitachi C1 anymore.

1. The Basic mode rotate tune test made the motor vibrate like hell (the first AC excitation test always does this).
2. I have noticed a retry op after a fault made the motor vibrate excessively.
3. The Extended user's guide is 116 pages longer.
4. The movement from Extended mode to Basic mode transfers the parameters.
The movement from Basic to Extended mode pulls in the default parameters. Then the programming has to be done again.
5. The parameter set for Basic is a subset of the Extended parameters.
6. In Extended mode the rotate tune acts like my Allen Bradley. And the measured motor constants are slightly different on each test.
7. The stuff in post #137 is still the same. After a auto tune in Basic mode the motor constants are always the same numbers.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Tonight:
----------
bA140 (over voltage suppression enable) = 03 (Acceleration)
bA141 (over voltage suppression active level) = 380 V (default)
bA142 (over voltage active time) = 1 Sec (default)
FA-10 (acceleration time) = 7 sec

E007 (over voltage) fault occurs on starting with 69 psi in tank.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Next day: #1
------------
bA140 (over voltage suppression enable) = 03 (Acceleration)
* bA141 (over voltage suppression active level) = 400 V (changed from 380)
* bA142 (over voltage active time) = 3 Sec (changed from 1)
FA-10 (acceleration time) = 7 sec

I got E007 again. Pushed the stop button. I didn't expect the drive to start up again and continue until 120 PSI and then stop.
Using 90 PSI (start) and 120 PSI (stop).
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Next day: #2
------------
bA140 (over voltage suppression enable) = 03 (Acceleration)
bA141 (over voltage suppression active level) = 380
bA142 (over voltage active time) = 1 Sec
FA-10 (acceleration time) = 7 sec
* HC111 (boost value at startup 0 - 50%) = 25%
* AC-03 (S curve) = 01
* AC-05 (S curve bend 01 - 10)= 05
I got E007 again. Pushed the stop button. The drive started up and continued until 120 PSI.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
WEIRD:
That after the first fault everything comes to rest and I press stop 15 sec later and the VFD starts up and continues until 120 psi.
The parameters changed with "*" did nothing to change anything.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------

The air pressure appears to build up twice as fast compared to before and the pump is quieter.
The push of the stop button caused a retry? I don't know yet. When I saw how fast the psi was increasing I knew something got better.

My thoughts are now about that Alfred Hitchcock episode where a man and woman are trying to kill another man in a kitchen.
They tried stuffing his head into the oven, choking him, bashing his head, etc. Went on for a long time. The guy would not die. Like this bug.
(The authorities probably commissioned the movie to show the public it's not so easy to wack someone, to reduce unsolved murder statistics).

Anyway, contact Hitachi support again? And read about different acceleration curves and how to cheat high voltage.
gettyimages-1376252187-612x612.jpg
 
Last edited:
One further idea is that if there is any "springiness" in the system, like belt stretch, some form of rubber clutch, etc, energy might get stored in that and then the motor vs the load might oscillate. It could be a sort of mechanical torque resonance. I think it might take a heavy inertial load, plus a relatively light, low inertia motor, but I am not sure.
I ordered two new Gates belts before all this. One is slightly longer than the other. I can see the longer belt flopping around compared to the shorter one.
 
Last edited:
Assume error code is E07.4, there is not E70 in the manual?
View attachment 438235

Per the manual in sensorless vector mode with automatic torque boost A041=01 adjust A046 and A047. I would also try an S curve acceleration with 2 or 3 second acceleration times. You might also try switching A044 to 00 V/F constant torque and see if that changes the fault mode. You may need to adjust the manual torque boost settings A041=00, A042, A043.
View attachment 438234
I'm going to try this now using Extended mode parameters. I believe that I can do nothing more mechanically to remove the fault.
Would you mind looking at the results?
 
Why such a long acceleration time? I would try something like 3 seconds. Are your motor parameters set correctly? Would an output sine wave filter tame the over voltage error? I can't say I have run into this type of scenario/error before. Maybe post the full set of parameters you are using.
 
Order belts only in matched sets, there are batch set numbers printed on them, refuse belts that are not from a matched set.
 
A list of 20 parameter changes in p.txt. Each test lists a few more changes in addition to those in p.txt.

Test 1:
------------------------------------------------------------------
FA-10 (acceleration time) = 3 sec
HC111 (boost value at startup 0 - 50%) = 50%
AC-05 (S curve bend 01 - 10)= 05

The same E007 fault. Waited a few seconds and pressed stop, fault again. Pressed stop again and same fault.
Immediately set acceleration time = 7 sec and a press on stop. Starts up with no fault. Like in post #161 (WEIRD part).

Test 2:
------------------------------------------------------------------
AA121 (mode) = 00 (V/f constant torque. Manual torque boost is available)

The drive started up and then stopped like it faulted, then automatically started up again and continued with no fault.
Similar to pushing the stop button in sensorless mode after the first fault. Obviously this is a retry with a success.

Test 3:
------------------------------------------------------------------
AA121 (mode) = 00 (V/f constant torque. Manual torque boost is available)
Hb141 (Manual torque boost value 0 - 20%, default = 0) = 20%
Hb142 (Manual torque boost peak speed 0 - 50%, default = 0) = 50%

The drive started up and then stopped like it faulted, then automatically started up again and continued with no fault.

Test 4:
------------------------------------------------------------------
Same result as test 3 with acceleration time changed from 7 sec to 3 sec. Then changed to 2 sec.
FA-10 = 3, 2

Conclusions:
------------------------------------------------------------------
1. Tests 2, 3, & 4 work in V/f mode but with 1 or 2 retries. The drive stops/starts at the end of the acceleration time, whether it's 2, 3 , 7 seconds.
2. Sensorless mode always faults E007. Notice that in Test 1 the HC111 (boost value) was maximum.
This parameter is used in sensorless mode. The torque boost effect like test 3 didn't happen.
3. The pump head rework did not cause the over voltage fault to go away. But the pump was inefficient. Probably 75-100% faster pump time now.
4. If this VFD fault never occurred I might have never been given the heads-up by SAF to look at the valves. Would have never thought anything was wrong.
5. Other parameters to change? I tried changing these before with no success.

bA141 (over voltage suppression active level, default = 380V)
bA142 (over voltage active time, default = 1 sec)
 

Attachments

  • p.txt
    1.6 KB · Views: 0
Last edited:
Sensorless mode
-----------------------
The trip monitor log (10 entries) have just this error with slightly different but close readings.

Fault = E007 (over voltage)
Frequency = 0.52
Output current = 21.58
Volt DC = 400.1
inv status = operation
LAD status = acceleration
inverter control mode = speed control

V/f mode
------------
The retry monitor is empty. This start/stop/start/stop/start/run is not a bunch of retries.
I changed bb-13 (auto reset number default = 3) to zero and there was no difference.
In this V/f mode the motor starts up for 1 to 3 sec then stops and then starts again.

I also tried setting these one at a time (additive). No difference.
bb112 (torque level (0 - 500) def = 200) to 500%. No difference.
bA120 (over current suppression enable def = 1) = 0 (disable)
bA140 (over voltage suppression enable def = 0) = 3 (from my programming) now set back to 0 (disable) again.

 
Last edited:
You might try other acceleration curves A097 either 03 or 04, you would also need to set the respective inflection points (A131/132 or A150/151) for the curve. Check the belt tension that there is no whipping or jarring, the only other thing mentioned is setting the motor slip angle. I would keep the VFD in V/F mode.
 
In Sensorless mode the drive will fault with E007 (over voltage). I press stop on the keypad. The drive starts again and may fault a second time
or keep going. So after 1 to 3 stops on the keypad the drive keeps going.

In V/f mode the drive will stop and within 1-2 sec will restart. This will repeat 1 to 3 times and then the drive keeps going.

The fault information shows that the frequency is 0.52. Even after I run a test with acceleration time changed from 3 sec to 10 sec.
After the long acceleration the drive faults. But the frequency is close to 60 Hz. I don't know what 0.52 means.

Good news:
---------------
When the run input command source is changed from the input terminal to the keypad I get a full start each time. I tried this in sensorless and V/f modes.
Manual torque boost parameters are default at 0%. No special programming for over voltage. The pulley acceleration appears to be faster but maybe because
the drive is not faulting or stopping. When I change back to the input terminal I get the start/stop business in V/f mode and E007 faults in sensorless mode.
AA111 = 02 (keypad)
AA111 = 00 (terminal)

Bad news:
-------------
The terminal command is 24 VDC from a SMC digital pressure that is connect to terminals pins #1 & PLC with jumper from P24 to PLC removed.
I followed the diagram for sink logic with an external 24 V power supply. The diagram is on page 5-4-9 in the Extended mode manual. Using the
output from the SMC (24 V 25mA max) to the VFD is causing the problem. It was fine with the other VFD. I didn't want to modify or make another
control board just to add a miniature relay. Can I use a 24 VDC command to the terminal connector on the VFD?

The SMC switch was used with the Allen Bradley VFD. This is similar to mine:
 
Last edited:
I've been suspecting your solid state pressure control for a while now, but you haven't mentioned it, in a long while now. Your drive tweaking has proven that it's capability is not the issue to me.

It's been telling you that V/H simple mode works fine except for a restart without a fault. The restart is likely being caused by the pressure surge or voltage hiccup on your pressure sensor at drive startup.

I would put your diaphragm actuated pressure switch back in place and change your programming to suit. A jumper could also be used for a temporary manual confirmation test.

Speculation - your believed drive problems started when you installed the solid state pressure sensor.
Your drive control voltage and current inputs, may require a interposing relay with appropriate contacts suitable for low current, since low DC voltage will give oxidation problems from their buildup over time. Line voltage can burn through the oxide tarnish to keep the pressure switch contacts operable.

Order yourself a new set of batch matched belts, so that you can eliminate the flutter in your mechanical drive setup, and you may have good performance in vector mode then.
 
The output of the SMC pressure switch output is PNP open collector (-P).
The input of the VFD terminal block. Using an external power supply with Source logic (lower right).
The diagram shows the equivalent input circuit as a bidirectional switch.
pnp.jpg input.jpg

Wire connections:
Black OUT to inverter input terminal 1
Blue DC(-) to inverter input terminal PLC
 
Last edited:
Top : This circuit with the relay works.
Middle: This circuit does not work with the input terminal. What I first used.
Bottom: This circuit with a optoisolator does not work, I tried another variation with both internal & external 24 V power supplies. Still no good.

DSC_1733.JPG

The Hitachi manual pages describing how to work with a PLC does not work.
I'm going to use the relay.

Leonard-Nimoy-as-Spock.jpg
 
Good to see your moving along, with resistance. Now acknowledging the problem. Old fashioned contact points really aren't that bad, especially on something like a plain Jane compressor. You could have both pressure controls plumbed in, if the digital gauge is that important. I'm partial to the old SqD mechanical ones, just a nut driver required vs opto isolators and thousand page manuals. Reminder to test that tank check valve for leaks, cheap insurance.
 
You must have some small feet to have a shoe box smaller than a mechanical pressure switch enclosure.
: - ) But honestly I took a brief look at the SMC digital models and it has a 35 page selection and programming cut sheet. I can now understand why your having issues getting it to play nice with your Hitachi. No telling what model you have.It's likely neither will support. I looked back in the thread and your picture of it gave no clue either, other than it is tiny.
 

Attachments

  • ZSE20(F)_ISE20.pdf
    7.3 MB · Views: 2
Your optocoupler circuit did not work very likely because of the resistor values, or possibly the connections.

1)
The 4n35 has a 100% transfer ratio with nothing connected to the base lead. You had 24V and 10k ohms, giving approximately 2.2 mA current in the input of the 4n35. So the output was going to be around 2.2 mA also, at 100% transfer ratio (output current same as input).

If the detector in the VFD needs about 5 mA to reliably detect current, which is about what it will get with the 4.7k resistor, you may have been "starving" it.

By changing to a 4.7k or 4.3 k resistor on the input side, it likely would work, or work better..

2)
Also, your direction of current was different, reversed in the VFD input, so it might need the current direction reversed back to be the same as with the relay.
 
No telling what model you have.It's likely neither will support. I looked back in the thread and your picture of it gave no clue either, other than it is tiny.
The model number is on the circuit diagram page in post #172.

Have worked with micro-controller programmable input/outputs before.
The 4N35 values are what I transferred from another design. But will investigate this.
I ordered some dip relays from digi-key and am going the relay design method.
 
Last edited:
Your optocoupler circuit did not work very likely because of the resistor values, or possibly the connections.

1)
The 4n35 has a 100% transfer ratio with nothing connected to the base lead. You had 24V and 10k ohms, giving approximately 2.2 mA current in the input of the 4n35. So the output was going to be around 2.2 mA also, at 100% transfer ratio (output current same as input).

If the detector in the VFD needs about 5 mA to reliably detect current, which is about what it will get with the 4.7k resistor, you may have been "starving" it.

By changing to a 4.7k or 4.3 k resistor on the input side, it likely would work, or work better..

2)
Also, your direction of current was different, reversed in the VFD input, so it might need the current direction reversed back to be the same as with the relay.
1. Works 3 times in a row with 10k input resistor changed to 4.7k.
2. Left it the same because the input is marked as bi-directional. I matched their diagram.

Spec: Voltage between each terminal and [PLC] terminal

ON voltage: min. 18 VDC
OFF voltage: Max. 3 VDC

Maximum allowable voltage: 27 VDC
Load current: 5 mA (at 24 VDC)
Internal resistance:4.7 kΩ

At minimum ON voltage (18 V) the current would be 3.75 mA. My designs attempt to go at least 1/2 under max numbers. But here their 5 mA number is strict.

I did some modifications to the control board for a miniature 24 V relay.
The relay socket (16 pin) has enough unused pins to allow the usage of the 4N35 (6 pin) when plugged into one end of the socket.

Thank you JST.
 
Last edited:
All the above was done with a motor pulley (5.75" dia.) that replaced the original pulley (6.75" dia.). In the belief that had to start problems will go away.
Did not think that the compressor was the problem. Hate to think that the rebuild job was done poorly. At this place that was called Univ Air the co-owner
had a picture of this soldier and he told me his son was killed in Iraq. So I started to feel sorry for the man and bought the compressor anyway, knowing that
it had Ford Taurus green rattle can paint job.

Put the original pulley back and it didn't phase the VFD. Was 575 RPM. Now 675 RPM.
Would a slightly larger pulley change any VFD self-measurements on the rotation tests?
 
No, I do not think it would have a significant effect, and you are running it at a fixed speed so the effect of redoing the autotune is nominal. I would leave it be. Ironically, the original VFD would probably still be working just as well.
 








 
Back
Top