BLDC and fanuc redcap parameter problems?

andypugh wrote:

gmouer wrote:

After some thought, I decided to run to the local radio shack today and pick up a TTL nand gate which I will use to convert the 3V output of the encoder to a level that will trigger the drives CMOS input. .

I think I would have used a comparator chip so that I could adjust the threshold. But I suppose it depends on what you can get easily.

Yup, I am lucky, Radio Shack locally is open, many all over the country have closed. The 3V encoder is well over the TTL logic one threshold so it should work. I am still lost why the drive does not run away without velocity feedback, that is wierd.

With very optimistic thinking, I reason that "rough" running is actually velocity variation. The motor roughness is to a steady beat. Will a working velocity feedback smooth out the running????? Very curious to see ! Its a fun learning experience.

(have not heard back yet from John Sheerin if he ran his drive in torque or velocity mode, curious about that too ! )

andypugh wrote:

gmouer wrote:

Yup, I am lucky, Radio Shack locally is open, many all over the country have closed.

I could pick this up today until 6pm, or on saturday or sunday
www.maplin.co.uk/p/lm339n-uh31j
But if I wanted three, I would need to visit two separate shops.

Checked, local radio shack has 4 of the LM339 in stock. I will grab those also along with the TTL gates. Always nice to have options and if not used they may come in handy for a future project. All considered, think I will use the LM339 for my attempt, just as easy as a nand gate.

Should have some results to report later tonight.

I have not given up on the redcap project. I got extremely busy with my job lately.

I took a 74ls04 and used it to buffer the encoder A and B channels, thus bringing them up to 5V level for the AMC drive velocity feedback input. This seems to be working nicely now. Didn't make any difference in the rough motor running though.

Just for fun, I hauled out my oscope and took a peek at the Hall1, Hall2, and Hall 3 outputs from the parallel port. I was mostly looking to verify a decent looking square pulse stream BUT..... The hall signals out are very strange indeed and probably causing my rough running. I am presently testing in straight code conversion mode, fanuc grey code to 3 standard hall outputs via BLDC. (cfg=fHT)

Attached is a screenshot of the 3 hall signals with the motor running approx 200 rpm. (guessing)

(I tried IRC hoping to catch Andy but no luck today)

Additional info, I am using pattern 44

Andy, any comments on the strange hall outputs from BLDC?

andypugh wrote:

gmouer wrote:

Attached is a screenshot of the 3 hall signals with the motor running approx 200 rpm.

200 rpm and at a guess, a 4-pole motor is 80 transitions per second, so your 500Hz sample rate ought to be high enough to see what is really happening.

Can you add the 4 Fanuc signals to the plot?

The redcap is a 8 pole motor. I will try and get a screenshot of halscope w the fanuc grey codes together with the halls.

Thanks !

Here ya go. Getting late so I will check back after work tomorrow.

Thanks Andy ! Much appreciated !

andypugh wrote:

If you compare to
www.cnczone.com/forums/attachment.php?at...d=81300&d=1242252956
I think that your signals might be inverted.
Can you try connecting C1 C2 C4 C8 to the ...-not pins instead?

I shut down the shop for the night, early day at work tomorrow. The fanuc sigs are right off the pulsecoder and I verified them with a DVM against "watch pins" in hal monitor. I double checked that just in case I crossed any wires.

I will switch the fanuc pins over to the not inputs tommorrow and report back. I will grab a new screenshot also.

Thanks Andy !