The big TEC driver thread!

Quote:

---

Originally Posted by Solarfire

You defiantly work too much dude! Unfortunately you’re not just around the corner, id come by and pull ya out from behind your PC for beer.. that’ll sweat it out*retardo


Beyond that the weather is f*cked up everywhere. You ever been to Germany? If I add up everything we’ve had this year it may add up to 37°?


Australia was my last hope for that paradise climate change!:cool:

---

I have not been to Germany before, but I am heading to Frankfurt in April, after I do my next tour of China. I here the BEER is good!

After that I am off to Venice, I can't wait. *retardo

Are you in Germany? Maybe we could catch up? I am there from the 6th April for a week.

Quote:

---

Originally Posted by Solarfire

As far as RS232 goes, it’s pretty much getting extinct, maybe do a pole for RS232 or USB?

---

I must disagree about this.. Serial, both RS232 and ttl levels, is still extensively used in industrial process control environments. It is also one of the most easily coded interfaces, and one of the most supported in on chip hardware. Please dont be hatin on RS232 ;)

@dnar: Could you consider using the atmega 88 as the controller? Even if the first version hardware doesnt support serial, at least the controller supports it, so later hardware/code revisions will be easier.

BTW.. Are you in Perth? I am in Melbourne. Perhaps we could catch up sometime over a beer or two :)

Cheers,

Pete

Quote:

---

Originally Posted by ogoun

I must disagree about this.. Serial, both RS232 and ttl levels, is still extensively used in industrial process control environments. It is also one of the most easily coded interfaces, and one of the most supported in on chip hardware. Please dont be hatin on RS232 ;)

@dnar: Could you consider using the atmega 88 as the controller? Even if the first version hardware doesnt support serial, at least the controller supports it, so later hardware/code revisions will be easier.

BTW.. Are you in Perth? I am in Melbourne. Perhaps we could catch up sometime over a beer or two :)

Cheers,

Pete

---

Hi Pete, I agree, RS-232 and RS485 is used heavily in Industry and still in Pro/Commercial Audio. Your correct, it is easy to implement. Yes, I am in Perth and while I get across to Sydney regularly, Melboune (my place of birth) I do not. Sorry.

The choice of MCU for the 1st release is up to Solarfire.

Quote:

---

Originally Posted by dnar

I hope noboddy suggests a TCP/IP interface... O_o

---

Firewire interface? :p

/Thomas

Quote:

---

Originally Posted by Badpip

Firewire interface? :p

/Thomas

---

Now your just pushing my buttons Thomas!!! O_o

We can do 802.11g, n, and Xigbee, but Firewire is for MacWankers!!!! O_o

Quote:

---

Originally Posted by ogoun

I must disagree about this.. Serial, both RS232 and ttl levels, is still extensively used in industrial process control environments. It is also one of the most easily coded interfaces, and one of the most supported in on chip hardware. Please dont be hatin on RS232 ;)

Pete

---

Yes I know these interfaces are still widly in use and also of the rising problems of newer Hard- and Software supporting it. I work at Terex-Demag in Germany R&D for large Telescopic Cranes in we are more and more runing into problems with suport for these interfaces thus moving to USB and Ethernet. We also have a wide palet of PLC's we use which dosn't make it any easier.

I gota go for now, to be continued tomorrow....

I just took a look at the LMD18200 datasheet:

http://www.national.com/ds/LM/LMD18200.pdf

1:
A bypass cap of at least 100uF/A is needed at high currents. A 330uF as close as possible to the power pins would need to be added to comply with this.
(Page 8 in the datasheet)

2:
For pwm frequencies above 1kHz, 10nF bootstrap capacitors are required.
(Page 9 in the datasheet)

/Thomas

Quote:

---

Originally Posted by Badpip

I just took a look at the LMD18200 datasheet:

http://www.national.com/ds/LM/LMD18200.pdf

1:
A bypass cap of at least 100uF/A is needed at high currents. A 330uF as close as possible to the power pins would need to be added to comply with this.
(Page 8 in the datasheet)

2:
For pwm frequencies above 1kHz, 10nF bootstrap capacitors are required.
(Page 9 in the datasheet)

/Thomas

---

I suspect the reason for the big bypass cap has a lot to do with the inductive nature of the load that these chips are usually driving, i.e. a DC motor.

Since the TEC is essentially resistive, I would think the cap could be a bit smaller, though it is good policy to use one, and a low ESR one at that.

Cheers,

Pete

Quote:

---

Originally Posted by ogoun

I suspect the reason for the big bypass cap has a lot to do with the inductive nature of the load that these chips are usually driving, i.e. a DC motor.

Since the TEC is essentially resistive, I would think the cap could be a bit smaller, though it is good policy to use one, and a low ESR one at that.

Cheers,

Pete

---

I agree, this capacitor is were most of the initial drive current is delivered from, hence it's close proximity to the bridge supply pins. With our application running > 100kHz it is less critical but should be low ESR and a high quality brand at that, AVX or Panasonic.

I wonder if it would be best for our west coast surfey admin to split these new TEC project posts into a dedicated thread?

Without wanting to get too ahead of myself, I was bored at 2am and could not sleep, so I coded the guts of a graphical Java application and MCU C code that will interface with ATmega style MCU via RS-232 to provide graphical view of set point and current temperature over time (live) , display of the current process output value (TEC PWM) and permit tuning of the P, I & D gain parameters and also initiate saving these to EEPROM.

Should make observing and tuning the loop easy. For V2 of course. O_o