YK31A Modifications 60v+ high current

[NStone775]

Im attempting to modify a cheap YK31A 24v 500W DC speed controller for use with 60v (or more) on a larger DC motor. I believe I have a handle on the necessary tasks:

Here is the PCB in question:


Tentative Plan
1) Beef up the traces: Batt->FET->Motor
I will add heavy solder + overlaid copper wire to increase the current carrying capability of the traces.

2) Upgrade Caps
I've soldered in a single 180v 820uF cap in place of the original pair of 470uF 50v caps.

3) Modify resistor values in the resistor divider to adjust the LVC to 3.0v per cell.
I think i've located the resistor divider based on the voltage measurements, see image. I plan on adjusting the resistor value on the VCC side of the divider until I get 2.6v when VCC is at the cutoff voltage.


4) Upgrade FETs
I will replace the 2 stock FET(s) with IRFB7730s & add an additional FET into the unpopulated spot on the board. (the 36v 800W version uses the same board and has a FET there)

5) Decrease the shunt resistance - Add copper wire wrap + solder to reduce the resistance of the shunt.

Outstanding Questions
A) - Will a IRFB7730 be a suitable upgrade for the stock FETs (RU6888R)?]
The IRFB7730 has lower Rds(on) and can handle more watts overall. What other FET characteristics are important in this application?

Datasheet Links:
http://ruichips.com/uploads/file/20200819/20200819171503922.pdf
https://www.infineon.com/cms/en/product/power/mosfet/12v-300v-n-channel-power-mosfet/irfb7730/

B) - I do not see a current shunt anywhere in this circuit. Is it possible there is no current limiting in this circuit? How can i modify the current limiting? or do I need to?

Any input or direction on the two questions / LVC mod plans would be truly appreciated! Thanks!

 

[amberwolf]

Here is the PCB in question:

If you attach the images directly to the post, then anyone that can see your post can see your images. Use the attachments tab below the box you type the post in, or just drag the images into the text box from your computer, and they'll attach automatically.

2) Upgrade Caps
I've soldered in a single 180v 820uF cap in place of the original pair of 470uF 50v caps.

It's usually better to use multiple caps in parallel because their internal resistance (ESR) is lower that way, which means they don't heat up as much and they do a better job of keeping electrical spikes and noise away from the FETs (spikes are one cause for them to blow up).

5) Decrease the shunt resistance - Add copper wire wrap + solder to reduce the resistance of the shunt.

If you do this, the controller will be unable to limit current, and cannot protect itself, the battery, or the motor from overcurrent. One common failure mode of controllers modified this way is unexpected sudden FET failure.

In a brushed motor controller, that kind of FET failure usually results in the motor being stuck ON at FULL POWER.

Outstanding Questions
A) - Will a IRFB7730 be a suitable upgrade for the stock FETs (RU6888R)?]
The IRFB7730 has lower Rds(on) and can handle more watts overall. What other FET characteristics are important in this application?

Possibly the Qg (gate charge energy), if the new FETs require enough more to turn on fully, compared to the old ones, it's possible that the gate drive circuitry in the controller won't be able to turn them on fast enough or fully on enough and they'll still be in their resistive state enough to cause excessive heating. Easy to tell after installing them (before doing any shunt mods etc), and then just using the controller normally, as the controller will get noticeably warmer than it did with the old fets. Or the fets will blow up.

The Vds is important if you want a higher voltage system; use a FET with at least 50% higher Vds than the voltage you will use, cuz there is a heck of a lot of RF noise in a brushed motor, and some of those spikes will be a notably higher than the system voltage.

B) - I do not see a current shunt anywhere in this circuit. Is it possible there is no current limiting in this circuit? How can i modify the current limiting? or do I need to?

There is almost certainly a shunt and current limiting. I've seen some really really cheap brushed cotnrollers that don't hae this, but they tend to blow up frequently even under perfectly normal usage.

If you *really* want to modify the current limit, then if there is just one shunt, you can add a second shunt in parallel with the first, that is the same as the first one. This will double the potential current in the system, because it fools the controller into thinking only half the current is flowing vs what really is. It's not as "dangerous" to the controller as soldering a shunt or wire-wrapping it (which often effectively zeros the shunt's resistance, and thus removes the controller's ability to know what is happening and thus control it)..but there is still the risk of destruction. :/

Or if there's two shunts, add a third, and you increase it's abilty by that amount instead.

If only one shunt, and you don't want to quite double it's current, then if the shunt is marked with a resistnace, you can add one in parallel that is a higher resistance so it doesn't qquite halve the total resistance.


Etc.


Alternatley, you can find the circuit betweent eh shunt and the controller chip, and see if there is a voltage divider in that circuit. If ther eis, you can change that divider to change the way the circuit works. Might be an opamp instead, in which case it depends on how they setup the circuit, but still likely has a resistor that could be changed, instead of messing with the actual shunt.

 

[NStone775]

Thanks for the image tips & all the feedback! (Images should be fixed in the original post)

It looks like Vds(max) is 75v on the IRFB7730, so even 60v is probably too much. That FET came out of a Ryobi 40v(36v) battery, which further supports your suggested 50% safety margin. Looks like I'll need a different FET.

Im still at a loss trying to identify a current shunt though. I've included a comparison shot of the 24v 500W version of this board and the 36v 800W version. I only found 4 things which were different (annotated A,B,C, & D. Not counting the extra FET & caps I removed). Are current shunts functioning as inductors or resistors? Could A) be used for current sensing?

 

[amberwolf]

I've included a comparison shot of the 24v 500W version of this board and the 36v 800W version. I only found 4 things which were different (annotated A,B,C, & D. Not counting the extra FET & caps I removed).

I attached an edited version that points to the shunts.

Are current shunts functioning as inductors or resistors?

They're used resistively. Current flwoing thru them generates a voltage across them, which is then measured by the controller and used to determine how much current is flowing, based on either lookup tables or formulas in the software. Then the software does whatever it is setup to do based on that info.

Change the resistance, and you change the voltage generated. Higher resistance, higher voltage, more response from controller to lower current. Lower resistance, lower voltage, less response from controller to lower current. Low enough resistance and the voltage generated is insufficient to cause any response from the controller to lower current, even when it's on fire.


Both of these controllers use the same shunt resistance, 0.002Ω, so the difference in their capabilities is in the software / control chip or in the electronics (if any) between the shunt and the control chip.

I can't clearly read the numbers on the control chip, but it looks like an LM339, which is a quad op amp chip. There's a number of controller designs based around those, some of which are in this list
https://www.google.com/search?q=lm339+brushed+controller
If so, then if you trace out the circuitry around the chip, you can compare it to the various schematics in that list, and find which stuff is for the shunt response, and tune it that way. That will give you much more control over what happens than the brute force approach of modifying the shunt itself. You can even install an adjustment pot instead of just changing resistors in the circuit. There's some posts several years or more ago by Fechter and Jeremy Harris about doing this with brushless controllers; the same stuff they discuss applies here too.