6 replies to this topic

[ianz]

Hi..

I have been reading many topics from this forum and guess this might be a good place to ask some questions. I am hoping to build a softstart using a Mega128 microcontroller but seriously wonder after reading comments in this forum if it is even feasible..given my lack of experience with motors. So I will outline some details below and hope that I will become better informed from the replies.

My machine is an 230Vac, single phase, centrifugal water pump that has a tested' running current of 3.2A. As far as I know it is capacitor start..rather than induction start..but need to confirm that with the manufacturers.

My design has the output pin of the Mega128 micro controller driving a optocoupled triac driver which controls the gate of an Alternistor. The alternistor is on the high side and motor on the low side. From what I can make out an Alternistor is a style of Triac that is designed for improved phase symmetry and therefore lower harmonc output. Please correct me if i am wrong.

The alternistor I have chosen is Treccor Q8025LF rated at 25A 800V so should be up to the job, bearing in mind that the locked rotor current of this pump is in the order of 15-20A and that, in theory, this level of current would not occur with the softstart active. see http://www.teccor.co...nistorTriac.pdf

The optocoupler triac driver is a random phase style driver. Fairchild MOC3052M rated at 600V. see http://www.fairchild...O/MOC3052-M.pdf

I have a current transformer in series with the motor. It is CRMagnetics 8300 type and can be viewed at http://www.crmagneti...ProdName=CR8349

The secondary winding of the CT feeds a Precision rectifier circuit kindly provided by the CT manufacturer and can be seen at http://www.crmagneti...df/ancrct-4.pdf This circuit provides a DC voltage that is in proportion to the motor current and can be monitored using the ADC input of the Mega128.

The purpose of the CT is not only for its potential use with the soft start function, but also to monitor the long term motor current in order to detect filter blockages, overloads etc.

Another hardware feature that I built in is a snubber circuit on the triac to allow it to turn off cleanly, considering the inductive nature of the load.

I also have a seperate mains monitoring transformer that is monitored by the microcontroller and its purpose is to protect against mains brown out, but also allows me to detect the zero cross point of the mains supply.

So thats my hardware. If anyone wants, I can post the circuit diagram.

My initial plan is NOT to use the Current Transformer in a feedback control loop, and to simply to feed the optocoupler with a set of timed pulses, syncronised to a zero crossing reference, as in a simple light dimmer circuit. By reducing the delay of successive pulses, the voltage to the motor will gradually increase, from 50% to 100% and the motor will speed up. My planned ramp time is in the order of 2-5 seconds so is quite quick. Of course I can also watch the motor current for overloads etc, and I guess it makes sense to use it in the ramp determination eventually. But I have never been down this road before so am going for the simplest idea first.

Having said all of the above..and outlined my design, I would like to hear comments from other forums users, in particular with regard to issues that I might not have considered. What other gotcha's should I be aware of before I actually fire this thing up?

Thanks in advance for any comments or replies

Ianz..(yes another kiwi DIY tinkerer)

[marke]

Hello Ianz

You should be able to make this work, but you need to consider that the load is a reactive load that can generate as well as draw current, and you also need to consider that the power factor of the load can vary from very low to quite high.

There are three modes of controlling the triac, single pulse, picket fence and hard fire.
Hard fire is the best, followed by picket fence and lastly by single pulse.

Best regards,

[ianz]

Hello Ianz

You should be able to make this work, but you need to consider that the load is a reactive load that can generate as well as draw current, and you also need to consider that the power factor of the load can vary from very low to quite high.

There are three modes of controlling the triac, single pulse, picket fence and hard fire.
Hard fire is the best, followed by picket fence and lastly by single pulse.

Best regards,

Hi Marke

Firstly..I do understand the nature of reactive..I come from a background in Radio electronics..and there reactive is everything..But motor control is new to me..hence i respect your experience.

My situation is that i have range of pump motors to control..but within a narrow band of difference. Only from 0.5 hp to 1.5 hp and I dont need to go outside of that. All pumps are centrifugal..so that limits it more so.

Obviously all of the variant pumps have different cos/phi factors but only within a specific range..I dont think my problem is that great. My problem is more with my lack of experience..

I have never heard before of the terms 'hard fire' ..'picket fence' or 'single pulse'..its obviously quite specific to motor control. I will google.

**beep**..I have been in electric circles long enough to know that nothing is ever simple..I will research and get back.

[marke]

Hello Ianz

Obviously all of the variant pumps have different cos/phi factors but only within a specific range

The power factor for a given motor changes over w wide range depending on the slip. The power factor can be as low as 0.2 and as high as 0.9 for one motor. That is a pretty wide range.

I have never heard before of the terms 'hard fire' ..'picket fence' or 'single pulse'.

Single pulse firing is where there is a single short pulse applied to the gate to turn the triac on and there is reliance on the triac staying on. Picket fence firing is a series of pulse applied to the gate to ensure that the triac stays on. Hard fire is when DC is applied to the gate for the duration that the triac should be ON. The latter techniques are designed to prevent the Triac from being commutated off early.

Best regards,

[ianz]

Ok..thanks for that Marke..Hard fire looks like the way to go..although I had planned on a single pulse..but only because I never knew the advantages of hard fire.

I also never knew that the power factor changes with slip, and I presume slip changes with load and with supply voltage.

I am also monitoring the supply voltage in order to prevent motor burn out due to supply voltage brown out, and as I understand, the brown out overheating is also because of excessive slip. So these ideas oviously relate together.

The other interesting thing, and a conundrum for me atm, is that the values for the snubber components on the alternistor/triac are calculated from cos/phi, and that is a variable, so I guess it must be comppromise thing in the end.

The unit I am designing has a datalogger function built in, based on an SD card, so I should be able to determine the phase angle between the voltage and current in real time and log that, so that I can see what happens in various load situations.

One of my challenges is to be able to interpret motor current data to assess load characterics that alter according the state of the filters. ie to be able to identify and notify when filters get blocked etc.

Atm I have no idea of what a centrifigal pump, with its input filter blocked will do, in terms of phase angle. All I do know is that this is a lower than normal current situation, as measured with a clamp meter, and as I understand this is due to cavitation. (as always..please correct me if i am wrong)

It is now occuring to me that monitoring the phase angle between the motor current and the supply voltage will provide quite valuable information that can be interpreted in terms of the loading. Just how to interpret this is part of the learning curve, and I welcome the opportuniy to learn new stuff.

What I now have is the idea that the reactance of the motor, as seen by the supply voltage, is changing according to the slip, and that is changing in accordance with load and the supply voltage. So its quite a complex situation.

[AMBA]

I also have a seperate mains monitoring transformer that is monitored by the microcontroller and its purpose is to protect against mains brown out, but also allows me to detect the zero cross point of the mains supply.

So thats my hardware. If anyone wants, I can post the circuit diagram.

Hi Ianz
I'm going to make a soft starter for water pumps too. Would you please send me the hardware drawing?
THANKS

[marke]

Hello Ianz

Just curious as to how you are getting on with this project. Any update??

Best regards,
Mark.