A4988 smokes with just power supply

Hi,

I’m in the process of testing a board that will use up to four Pololu A4988 Black Edition Carrier boards depending on what machine it is configured for. When I connected +24VDC to VMOT (Pin 16) and GND (Pin 15), the board smoked. I then turned the +5VDC on and the +24V on, and another board smoked. “EN” and “MS3"are connected to GND. “RESET” and “SLEEP” are connected to +_5VDC. There are no motors connected. The “MS1”, “MS2”, “STP” and “DIR” are all low. The motor outputs are not shorted to GND or to each other. What have I missed” Can anyone help me? I have added page two of the schematic for clarification.

Ken
Welch Controller IV Page 2.pdf (87.5 KB)

So to clarify- on the second board, it only smoked once you turned 24v AND 5V on? But the first one smoked immediately when you turned 24v?

Just worth making sure: Have you added the >50uF cap on the Vmot? That is necessary, especially if your power supply has a high starting ripple voltage.

I’ve never actually had an a4988 “smoke.” I’ve only ever applied Vmot to the wrong pins (…twice!!!) and it resulted in faulty performance (I think in both cases, simply wouldn’t run again.)

Hi,

The +24VDC power supply is a switch mode power supply. It is a Lamda 24VDC @6.5A and has 20mV high frequency ripple.

This board measures temperature via type K thermocouples, has two 4-20mA outputs, three pot inputs, 4 slotted thru-beam inputs, an LCD and has 18 1 Watt MOSFET outputs. We wanted to make sure there weren’t any glitches (all outputs are 24VDC) when turned on so I designed a timing circuit that determines when the 24VDC is coming up or down and at this point turns on after 5 seconds so there is no problem as far as glitches because the PIC processor has gone through its program and set all the outputs when it starts up via a relay. I have scoped the +24VDC when it turns on and there is no switch bounce when the relay turns on. The +24VDC snaps on and off due to the timing circuit. The board is in a vise with no inputs or outputs including stepper motors.

Unfortunately, I haven’t added a >50uF capacitor yet. Somehow I didn’t notice it until the board was made, but have added a 100uF 50V capacitor. At this point, I only have one A4988 Carrier left and I’m cautious about connecting it.

I hope I have answered all of your questions and attached page two of the schematic as a PDF file.
Welch Controller IV 2.pdf (112 KB)

Hello.

I am sorry to hear you are having trouble with your stepper motor drivers. It is very likely that the board is experiencing LC voltage spikes above its maximum rated voltage when you apply power. You can find out more about such spikes and how to deal with them here:

pololu.com/docs/0J16

As Tomek and our product page state, the easiest way to prevent LC voltage spikes is to add a high-ESR capacitor (e.g. 50+ uF electrolytic) across power and ground near the driver. It can also help if you shorten your power leads or lower your supply voltage, but you should be able to address it with the capacitor alone if changing these is not an option.

Do you have access to an oscilloscope? If so, you could try applying power to one of your damaged boards with a capacitor across the power leads to verify that LC spikes are suppressed before you try powering your last board.

- Ben

Hi Ben,

I don’t have any stepper motors connected and have scoped out the +24VDC supply when it is switched on and there are no spikes or relay switch bounce. I have added a 100uF 50V low ESR capacitor to one stepper socket but have not connected an A4988 to it. If I insert a smoked A4988 what should I see? Is it shorted out? The board has two mixed signal planes. A +5VDC and +24VDC plane and a Digital GND and a +24V GND connected at one place near the 5VDC regulator. I have also mixed up the bad A4988s with the last good one. I have ohmed pins 16 and 15 and three of them are between 500 and 800 Ohms while one is infinite and runs cool. I have been doing so much testing that I’m misplacing things. Please excuse me as I have no intention of offending anyone because I am asking for help, but I have stated that I do have a scope many times on this thread and a PDF of the schematic. Did you look at the schematic? Please accept my humble apology if I have been rude because as I stated, I am asking for help. I don’t want to sound like a jerk. If I am, please correct me.

Ken

IKen
Welch Controller IV Page 2.pdf (87.5 KB)

Hi Ben,

Did you mean to state a high ESR capacitor of low? I’m thinking low because the power supply has a 20mV high frequency ripple. The page where it describes the A4988 doesn’t specify high or low ESR capacitors, which to me is important.

Ken

Hi Ben,

Did you mean to state a high ESR capacitor or low? I’m thinking low because the power supply is switch mode and has a 20mV high frequency ripple. The page where it describes the A4988 doesn’t specify high or low ESR capacitors, which to me is important.

I’m wrong. I obviously didn’t investigate the A4988 Carrier Board like I should have. I went to the Destructive LC Spikes page and it instructs you to use high ESR capacitors >50uF. Sorry. I sure messed this one up.

Ken

I apologize for not seeing your prior statement about having an oscilloscope (I only see one instance of this being mentioned, though, not “many”). I looked at your schematic briefly and didn’t see any problems with it, but someone else is looking at the one you emailed to us in more detail and will be responding separately to that. I mostly wanted to get in a quick reply that this looks like a text-book case of LC voltage spike problems.

I do not know what will happen if you test one of the damaged drivers; it might have failed as a short, in which case the test won’t reveal much. If it failed open, then you could still potentially use your scope to see what kind of LC spikes occur when you apply power (and how adding a cap affects that). I also suggest you put the scope on your remaining good board and see what power is doing when you apply it (make sure you have the cap on there from the beginning!). If you see any kind of appreciable spike, you might need to take more drastic action to suppress it.

It sounds like you have read the LC spike article now, but to briefly sum it up, the spikes are caused by inductance of the wires coupled with low-ESR (equivalent series resistance) capacitance. When power is first applied, the low-ESR capacitors on the board can draw a lot of current while charging, and the inductance of the long power leads causes the current to continue flowing for a little while after the capacitors are charged, which results in a voltage spike. The way to suppress this is to add a significantly larger high-ESR capacitor. This capacitor chargers over a longer time scale than the low-ESR caps on the board, and it’s higher ESR keeps the charge current from getting too high. The A4988 product page does not mention “high-ESR”, but that part is covered by the “electrolytic capacitor” part.

- Ben

Hi Ben,

I installed a .1uF decoupling capacitor and a 100uF 50V high ESR capacitor on the bottom of the board very close and tested two A4988 Carriers (I was able to get another one) and they don’t get hot and are at ambient temperature. I was hesitant about testing the last one I had until I talked to the person in charge of the project and he said he didn’t care if I smoked another one and gave me another one to test. I do design work at home for companies that want micro-controllers designed to run their machines. Since I didn’t investigate the A4988 closely enough, I missed the decoupling cap too. I did have one on the +5VDC though.Thank you to everyone for helping me. The only reason I mentioned the scope was because I kept having to repeat that I had one, but from now on I will answer every question regardless. Sorry for mentioning it. I now believe it was rude considering I was asking for help.

Ken

Great, I’m glad to hear your remaining boards are working! Note that the board has its own smaller decoupling capacitors across VMOT and ground (one 4.7 uF one and two 0.1 uF ones), so you might only need the 100 uF one as protection from LC voltage spikes. No need to apologize for mentioning that you already said you had a scope; I should have seen it the first time (I think I read through your setup description too quickly as it didn’t seem to be relevant to the more immediately likely problem of the LC spikes).

- Ben