SOLO no longer detected by Windows

General discussion of using Roboclaw motor controllers
evangifford
Posts: 15
Joined: Tue Jun 09, 2020 11:32 am

Re: SOLO no longer detected by Windows

Post by evangifford » Mon Jun 15, 2020 11:41 pm

Ok, makes sense - Thank You!

To Summarize how I managed to kill this SOLO (Presumably applies to all RoboClaws):

When using only a Power Supply instead of a battery to power a RoboClaw, and the battery voltage is set in Motion Studio to match the power supply voltage - the SOLO will send energy generated when the motor is slowing back through the power leads (energy recovery).

Since almost all but the most exotic power supplies will reject this reverse voltage by going high-impedance
- and there is no shunt diode/resistor to dissipate it into a resistor
- and there is no battery to absorb it (which would recharge the battery)

… the voltage will continue to build during slowing and eventually become too high, damaging the electronics on the SOLO.

The damage can present as unexpected behavior, lack of USB connectivity or otherwise malfunction.

-

There are 3 recommended solutions and 1 work-around from BasicMicro.

1) Use a shunt circuit which will send all voltage in excess of the shunt voltage through a resistor to be dissipated as heat - protecting the power supply and RoboClaw.

2) Use a battery as a power supply - excess voltage will be absorbed by the battery (and subsequently, recharge the battery) - protecting the RoboClaw

3) Use a battery in parallel with a power supply - useful during development since you will not need to constantly recharge a battery. The power supply will provide most of the power - but in the case of excess voltage, the battery will absorb it.

One work-around would be to have the SOLO dissipate this excess voltage by braking the motor (power would be dissipated into the resistance of the motor coil as heat) - to do that, you can "trick" the SOLO by setting the main battery voltage to 1 to 2V above the power supply voltage (use the voltage as read by the SOLO since it may be slightly different than a multi-meter). This solution is not ideal, but free.

-

More information about the voltage-clamp / shunt is provided here: https://resources.basicmicro.com/using- ... -roboclaw/


In the RoboClaw SOLO getting started page, this problem is mentioned (but in my opinion, could be emphasized)

"A battery is the better power source if the Solo will be driving a large load or suddenlt starting and stopping. A battery is better able to handle the regenerative voltages created under these circumstances. The battery settings for the Solo should be set according to this https://resources.basicmicro.com/config ... -settings/"
Last edited by evangifford on Tue Jun 16, 2020 12:17 am, edited 1 time in total.

evangifford
Posts: 15
Joined: Tue Jun 09, 2020 11:32 am

Re: SOLO no longer detected by Windows

Post by evangifford » Tue Jun 16, 2020 12:13 am

One point which is still unclear to me (perhaps a subject for another thread?)

In battery-only power source scenario - how can one calculate if the "peak power absorption rate" of the battery is greater than the "peak power generation potential" of the motor?

In other words, how can someone decide Voltage Clamp / Shunt circuit is required to protect the SOLO and battery in battery-only application?

I would assume the best would be to put everything in max generation condition and measure voltage peaks using an oscilloscope - is there another way to do this for those who don't have an oscilloscope?

What is the breakdown reverse voltage of the circuit in the SOLO?

Otherwise, a rule of thumb?

For example, "Lithium Polymer batteries can generally absorb the power generated by a motor with the nominal power output of 5C, or 5 times the capacity of the LiPo, a battery smaller than this will require a voltage shunt circuit to protect both the battery and RoboClaw"
Last edited by evangifford on Tue Jun 16, 2020 12:21 am, edited 1 time in total.

evangifford
Posts: 15
Joined: Tue Jun 09, 2020 11:32 am

Re: SOLO no longer detected by Windows

Post by evangifford » Tue Jun 16, 2020 12:20 am

Also, a question - why does the SOLO not automatically go into braking mode to protect itself in case the voltage is high enough to damage the internal electronics?

If that would interfere with normal energy regeneration, could this be added as an option in the firmware - a "soft shunt protection" option - in a case like mine where motor smoothness is not important?

evangifford
Posts: 15
Joined: Tue Jun 09, 2020 11:32 am

Re: SOLO no longer detected by Windows

Post by evangifford » Tue Jun 16, 2020 12:38 am

Looking at the BOM for the MIT E-Vent (Using a SOLO 30A) - they are using a switching power supply (LRS-150) - but I don't see any shunt/clamp circuit?

How do they ensure reliability of this system without using any of the above strategies to protect the SOLO and power supply from voltage spikes - surely this motor experiences acceleration/deceleration in this application?

User avatar
Basicmicro Support
Posts: 1387
Joined: Thu Feb 26, 2015 9:45 pm

Re: SOLO no longer detected by Windows

Post by Basicmicro Support » Tue Jun 16, 2020 8:21 am

Roboclaw DOES automatically protect from regen. If you read the manual. It specifically talks about using power supplies and setting the maximum main battery voltage to 1 to 2v above the power supply voltage to induce braking. We've written articles on it. I've probably got 100 posts in these forums that repeat the same thing over and over.

The user has to set the maximum main battery voltage. Even your own case you ask if a user could set something. Well they can. The maximum main battery voltage. Set to between 1 and 2v above the power supply voltage reading you see in Motion Studio.

As for the E-Vent. This comes down to mechanics. There IS regen in there system. But the range of motion is so small and the amount of time under decel is so short there is no SIGNIFICANT regen. I specifically worked with them to confirm they had no regen issues.

Regen is an issue when you spin a motor up very fast and slow it down. It is actually worse slowing down slowly than slowing down really fast. There is some calculations behind it if you want to get real fancy. Basically there are two modes of deceleration. Regenerative(meaning energy comes out of the system) and Dynamic(meaning you have to put energy into the system to slow it down). At faster decel rates you actually have to push energy into the motor to slow it down.

Then there is the special case where you short the motor leads together and no energy can go in or come out of the motor. This is braking mode and is what happens when Roboclaw sees and over-voltage.

I had a multiple page message explaining why we cant automatically deal with regen without the user setting the correct voltage limit. We've tried several methods. All have down sides, mostly adding significant power up delays or having false positive problems.

evangifford
Posts: 15
Joined: Tue Jun 09, 2020 11:32 am

Re: SOLO no longer detected by Windows

Post by evangifford » Tue Jun 16, 2020 11:26 am

Ok, I do see that information in the manual now - This is obviously my first time using a RoboClaw product, and I'm in a rush - I mistakenly assumed this value pertained to batteries only - since it's labeled "Main Battery Voltage" in Motion Studio - like it was used to detect the battery percentage remaining and lock-out below a certain voltage - I didn't realize it was critical when using a Power Supply.

I also incorrectly assumed that energy regeneration was not on by default, and that the energy would be dissipated by braking.

Thank you very much for clearing this up for me.

User avatar
Basicmicro Support
Posts: 1387
Joined: Thu Feb 26, 2015 9:45 pm

Re: SOLO no longer detected by Windows

Post by Basicmicro Support » Wed Jun 17, 2020 8:07 am

Roboclaws were originally designed for 4 wheel rovers. The labeling reflects that, eg battery operated.

DC Motors act like generators when they slow down. period. All DC motors do this. Its not optional. Some motor controlers have built-in voltage clamping. Those motor controllers tend to cost an order of magnitude more than other motor controllers.

evangifford
Posts: 15
Joined: Tue Jun 09, 2020 11:32 am

Re: SOLO no longer detected by Windows

Post by evangifford » Wed Jun 17, 2020 9:15 am

Well, it's a great product - I'm very happy with it and have already recommended the SOLO to others.

MIT proved it could be used for a use case you may have never imagined when making Rovers.

Since they also used a Power Supply - I'm sure you explained to them the "Max Battery Voltage" - doesn't apply just when using a battery - but is actually critical when not using an external shunt/clamp with a SMPS.

For someone like me who is unfamiliar with RoboClaw - the term "Max Battery Voltage" in Motion Studio is confusing - and since I don't have a battery at all - I incorrectly assumed I could basically ignore it.

User avatar
Basicmicro Support
Posts: 1387
Joined: Thu Feb 26, 2015 9:45 pm

Re: SOLO no longer detected by Windows

Post by Basicmicro Support » Thu Jun 18, 2020 7:58 am

Thank you.

We've come a long ways from rovers, that is for sure.

Post Reply