New Centroid Teknic Clearpath officially supported configurations. Updated 5-10-23

[cnckeith]

1.) our results show up to 300,000 works with the 5VDC setup. (there is no wizard setting for 300k step rate, so most only need to set it to 200k, others that need 300k can use CNC12 parameter 968 to set it to 300k)

easy way to increase max rate is to max out the step freq (300k when wired with 5vdc) and drop the steps per rev in Acorn Wizard to 1600 and do the same in the Clearpath software.

2.) yes at the expense of a cleaner signal

keep in mind our supported configurations are a "works in all cases" recommendation so 99% of users have a easy set of clear instructions to quickly get up and running well and don't have to worry or think about any of this stuff, that being said as you can tell there are other configurations that 'work'.

[carboncymbal]

- the max current rating is for all the chip channels combined, the clearpath to Acorn is not using all of them so no risk at exceeding current ratings.

- 1/2 watt 100 ohm at 5 volts runs 'cool', 1/4 watt will run hot (which is ok, just easy to go to 1/2 watt and not have it be hot)

- the 5 volt 100 ohm combo produces the best performance results with the Clearpath. test results show that it is the best way for fast reliable connection to Clearpath.

see detailed test results attached.

Thanks so much for the detail and giving this such a deep investigation. I’ll be using the C86ACCP on my 4 acorn machines I expect.

Thanks,
Clay

[Muzzer]

1. Have you tested at 400khz? I am using that because of a 4th rotary axis setup that has a 15:1 gear reduction, and wanted to maximize steps/rev for that setup.

I don't follow. If you wanted to resolve something like 0.01 degrees (which I doubt you'd ever need) with a 15:1 ratio, you'd only need 360000/15 pulses per rev for the motor. That's 2400 pulses per rev, hardly a high count. With 240kHz, you'd have 100 revs per second at the motor, aka 6000rpm. That would equate to 150rpm at the 4th axis. Do you plan to do mill turn?

Teknic seem to have made a pig's ear of this input circuit business. It's hardly rocket science but I can't help thinking the story has yet to fully play out. Who thought up this "constant current input circuit concept"?

[carboncymbal]

1. Have you tested at 400khz? I am using that because of a 4th rotary axis setup that has a 15:1 gear reduction, and wanted to maximize steps/rev for that setup.

I don't follow. If you wanted to resolve something like 0.01 degrees (which I doubt you'd ever need) with a 15:1 ratio, you'd only need 360000/15 pulses per rev for the motor. That's 2400 pulses per rev, hardly a high count. With 240kHz, you'd have 100 revs per second at the motor, aka 6000rpm. That would equate to 150rpm at the 4th axis. Do you plan to do mill turn?

Teknic seem to have made a pig's ear of this input circuit business. It's hardly rocket science but I can't help thinking the story has yet to fully play out. Who thought up this "constant current input circuit concept"?

I’m not super knowledgeable on this, can you explain what you mean by “constant current input circuit”?

Thanks!

-Clay

[TaylorPilot]

Do you guys have any examples of a proper mount for strain relief of the C86. Also, this thread has convinced me that the C86ACCP might be the best $50 I spend on my build. I am assuming that the board itself provides the proper 5V for the STEP/Direction, as well as the resistors?

[teknic_servo]

TKBOT47
I don't think the change will damage the Acorn from the other threads discussing this issue. The previous Centroid provided Schematic for SDSK's used a +24VDC supply voltage for the Input A+ & Input B+ and the resistor went between these inputs to the motor to the Step (Input B-) and Direction (Input A-)respectively. Many of us have wired our systems up utilizing this schematic. Depending on which version of Teknic manual was referenced, a 1K or 10K resistor was called out.

The latest supported schematic is now recommending a change to +5VDC for the supply voltage to Input A+ & Input B+ for the SDSK's, and a change to the resistor value to 100 ohm 1/2Watt or greater for +5VDC circuit voltage for Step & Direction, and I presume the 1K ohm 2Watt resistor mentioned in the other threads are still valid for the+24VDC supply voltage to inputs A+&B+. To address the timing issue identified subject to the +24VDC supply voltage, and the +5VDC voltage does not experience the timing issue is likely the reason for the supply voltage change.

Tom T. Can verify, but it sounds like this was sorted out using test equipment most of us likely do not have and would be costly and has been validated for the correct values to minimize noise/EMI in the system and completed circuits utilizing an Acorn and provide the best signal to the SDSK's.

viewtopic.php?f=60&t=5389&p=47069#p47069
viewtopic.php?f=60&t=3694&p=47138#p47138
viewtopic.php?f=60&t=5400&p=46583#p46583

RogDC,
Yes, the recent resistor recommendations were chosen based on a series of tests using o-scope analysis with the Acorn controller, ClearPath SD motors, and Teknic controller cables. Centroid is recommending the resistor values (100 watt for 5VDC) that are optimal for the Acorn-ClearPath combination. Teknic is making a more general recommendation for resistor values in the ClearPath manual that should work well with all controllers (though not truly optimized on a controller to controller basis). Either recommendation will work for the Acorn.

In the end, none of the resistor values mentioned will damage the Acorn or the ClearPath circuitry. Simply put, there are different resisitor values that are more optimal for different combinations of controller and drive. Rather than have an extended list of different resistor values based on different controller-ClearPath combinations (that could change over time as controller designs change over time), Teknic is recommending a 1k ohm resistor that should work with all controllers. 1 k ohm 1/2 watt for 5VDC; 1k ohm 2 watts for 24VDC.

If your system is not having any step timing issues, then you do not need to preemptively change the resisitors. That said, you could change the set-up to make the system more optimal if desired.

Additionally, using 5VDC is being recommended over 24VDC because the transitioning voltage being a true logic state and a false logic state (high to low VDC and vice versa) is quicker. Factors such as long cable lengths, high capacitance cabling, controller circuitry, etc. can make the controller signals transition slower (i.e. more sluggish) at 24VDC than at 5VDC (though either voltage can still work well in most situations).
Thanks for your patience and understanding,
Tom T.

[martyscncgarage]

The confusion came when it was suggested to use 5VDC instead of what we'd been understanding up to this time, use 24VDC.
Dave_C your input and help is ALWAYS appreciated as is everyone else's! I can't tell you all how much I appreciate it!!


Marty

[Richards]

When I see a new schematic that seems to change the way things are done, I like to make a very simple schematic showing only the signal that concerns me. I've drawn up this very simple schematic so that I can understand the circuit:

ClearPath.png (6.54 KiB) Viewed 25129 times

Let's look at just the resistor and the Acorn board. If just the resistor were used, the Acorn would pull 50mA through the 100 ohm resistor. The Acorn is rated at 50mA. So that side of the circuit should work. If just the ClearPath driver were used, then the ClearPath driver internal circuitry would determine the amount of current that it allows the Acorn to pull. The ClearPath driver can use 5VDC, so that side of the circuit should work.

When we combine the two parts of the circuit, we add the "pull up" function. The 100 ohm resistor "pulls up" the voltage to 5VDC when the Acorn output is OFF. The low value of the resistance, 100 ohms, assures that the signal will be pulled up very quickly. A quick signal means than the shoulders of the signal are square.

Is there a drawback to using only 100 ohms to pull up the signal? There is not as long as the Acorn can sink 50mA. It is rated to sink 50mA, so there should be no problem.

What do other designers do to pull up a 5VDC circuit? Years ago, I was mentored by George Bauldwin, who had been a chief designer for Kaiser. Kaiser was highly involved with military products. Everything that they designed had to meet military specs - which are very strict. George always pulled every 5VDC signal HIGH with a 1.2K resistor, because that was military spec at the time. When I asked him why 1.2K was the correct value, he took the time to attach one of his "clicker boxes" to a test circuit. A "clicker box" had a rotary dial that allowed him to select a particular resistor to use in a circuit. He turned the value to 1.2K and showed me what the signal looked like on his oscilloscope. Then he picked 2.2K, 4.7K, and 10K. The higher the resistance, the less sharp the shoulder was on the signal. Then he re-selected 1.2K and used another clicker box that used capacitors instead of resistors. He showed me that even 1.2K was not guaranteed to produce a sharp, square signal if the capacitance was too high.

I trust Centroid. If Keith says that 100 ohms works well with a 5VDC circuit, then I'm not going to waste a lot of time to see if 220 ohms, or 1.2K or 2.2K work as well. If I assemble a system that does NOT use the C86ACCP board, and if I use 5VDC as the voltage source for the ClearPath's signals, I'll use a 100 ohm resistor.

[Muzzer]

Without knowing what Teknic have done in their input circuit, it's not a simple matter to analyse the behaviour resulting from these external resistors. It seems that they decided to come up with some form of unusual input conditioning circuit which they have previously described as "constant current". It's not clear why they did this, nor is it something we see in well proven industrial products such as Yaskawa etc. Perhaps somebody might reverse engineer the input circuit to find out what they did, assuming they won't tell us here. I get the feeling they have compromised the speed of the optos and these measures are a sort of workaround, although I may be wrong there.

When I hear phrases such as "recommendations were chosen based on a series of tests using o-scope", I get nervous. These things should work by design, under worst case combinations of temperature, voltage, cable length, component tolerance etc. I wonder what verification and validation testing is done on these products....

[martyscncgarage]

Without knowing what Teknic have done in their input circuit, it's not a simple matter to analyse the behaviour resulting from these external resistors. It seems that they decided to come up with some form of unusual input conditioning circuit which they have previously described as "constant current". It's not clear why they did this, nor is it something we see in well proven industrial products such as Yaskawa etc. Perhaps somebody might reverse engineer the input circuit to find out what they did, assuming they won't tell us here. I get the feeling they have compromised the speed of the optos and these measures are a sort of workaround, although I may be wrong there.

When I hear phrases such as "recommendations were chosen based on a series of tests using o-scope", I get nervous. These things should work by design, under worst case combinations of temperature, voltage, cable length, component tolerance etc. I wonder what verification and validation testing is done on these products....

Centroid Engineering team, did us all the favor of the testing and analyzing the waveforms. We all spent a lot of time trying to address the root cause. Centroid spent two days analyzing long cables and testing. They had conversations with Teknic Engineers. Teknic reviewed the C86ACCP all for the user's benefit.
When people spend that kind of money, they expect it to be solid.
I have to trust the Engineers analysis and recommendations.

Marty