Need help understanding tantalum bypass capacitor requirements in DRV103H driver circuit

Thread Starter

rauty

Joined Apr 20, 2017
4
My background is in ME but in my job as a test engineer has increasingly required me to develop some EE skills as well (which I'm actually pretty happy about). As such, I'm a relative noob but hopefully not totally clueless. I've been designing a circuit to drive 8 solenoids on a 4-cylinder automotive engine. These solenoids cause the camshafts to switch between high and low lift profiles. There are 4 solenoids on the intake side and 4 on the exhaust side. I've selected a TI DRV103H to drive each of the solenoids (they draw about 1.5 A each). Using their 'basic connections' circuit works fine on the bench for driving one of the solenoids. An image of the circuit for each solenoid is shown below:
1588164831975.png
In the above circuit VS=13.8V, Intake2_In=13.8V (I've included a divider to lower it to 5V at pin 8), and C4 is a tantalum bypass capacitor, which they recommend using. I've chosen to use the higher (22 uF) value capacitor to make it handle up to 3A if necessary. I'm designing a PCB to drive all 8 of the solenoids in sequence. There will be 8 sets of the above circuit on the board, along with connectors for power, signal inputs, and driver outputs. My intent is to use the middle two layers of the board for power (VS) and ground planes. There will most likely be overlap between the solenoid events such that more than one solenoid can/will be active at a given time. My question is around the tantalum capacitor, which is effectively connected between VS and GND in each circuit. When one or more solenoids fire, they're going to see all 8 of those capacitors connected in parallel. Will that be an issue in this circuit? Should I select a different value for each capacitor, or use a different number of them, or will it be fine the way it is? I'm close to completing the design and would appreciate any advice on this. Let me know if you require any additional information.
 

OBW0549

Joined Mar 2, 2015
3,406
When one or more solenoids fire, they're going to see all 8 of those capacitors connected in parallel. Will that be an issue in this circuit?
No. They're just bypass capacitors; the more the merrier.

Should I select a different value for each capacitor, or use a different number of them, or will it be fine the way it is?
I imagine you'll be OK as-is. If it were up to me, though, I'd augment each 22 μF tantalum cap with a 100 nF ceramic cap in parallel, just for good luck.
 

Delta prime

Joined Nov 15, 2019
336
Thanks for the input. I'll keep that in mind
You have only just begun. ;)
Solenoids,BacK E.M.F .use reverse bise protection diode .high in-rush current to start
also produces hysteresis.capacitor in parallel with the coil will cause a current surge on turn-on, but there will be none without the capacitor because the coil's inductance will cause the current to ramp up as the magnetic field that builds up around the coil fights the change in current from the supply
FET driver per solenoid. This will isolate each solenoid from the others so that variations of the DC resistance of each so not cause uneven actuation as if they were all directly wired in parallel.by the way nice driver "Nevermind".:(
 

BobaMosfet

Joined Jul 1, 2009
1,108
My background is in ME but in my job as a test engineer has increasingly required me to develop some EE skills as well (which I'm actually pretty happy about). As such, I'm a relative noob but hopefully not totally clueless. I've been designing a circuit to drive 8 solenoids on a 4-cylinder automotive engine. These solenoids cause the camshafts to switch between high and low lift profiles. There are 4 solenoids on the intake side and 4 on the exhaust side. I've selected a TI DRV103H to drive each of the solenoids (they draw about 1.5 A each). Using their 'basic connections' circuit works fine on the bench for driving one of the solenoids. An image of the circuit for each solenoid is shown below:
View attachment 205694
In the above circuit VS=13.8V, Intake2_In=13.8V (I've included a divider to lower it to 5V at pin 8), and C4 is a tantalum bypass capacitor, which they recommend using. I've chosen to use the higher (22 uF) value capacitor to make it handle up to 3A if necessary. I'm designing a PCB to drive all 8 of the solenoids in sequence. There will be 8 sets of the above circuit on the board, along with connectors for power, signal inputs, and driver outputs. My intent is to use the middle two layers of the board for power (VS) and ground planes. There will most likely be overlap between the solenoid events such that more than one solenoid can/will be active at a given time. My question is around the tantalum capacitor, which is effectively connected between VS and GND in each circuit. When one or more solenoids fire, they're going to see all 8 of those capacitors connected in parallel. Will that be an issue in this circuit? Should I select a different value for each capacitor, or use a different number of them, or will it be fine the way it is? I'm close to completing the design and would appreciate any advice on this. Let me know if you require any additional information.
As OBW0549 stated, your configuration shows it used as a bypass cap. Here is what that means- a bypass cap does a couple of things- it isolates DC (DC can't pass between pin 5 and ground when the cap is fully charged, but AC can), and it acts as a mini instant power-source. A capacitor charges up and in a digital environment, there are current spikes all the time as current is needed in digital time-scales. Like instant now. A battery can't supply that- a battery is a chemical process- it's for massive dump long-haul, but not digital. A capacitor as a bypass cap fills that need by providing instant energy to the load, and the recharging in the down time. Having several in series, just means you have more capacity of current to dump instantly at any given moment of time, so it should be more 'forgiving/reliable'.
 

Thread Starter

rauty

Joined Apr 20, 2017
4
As OBW0549 stated, your configuration shows it used as a bypass cap. Here is what that means- a bypass cap does a couple of things- it isolates DC (DC can't pass between pin 5 and ground when the cap is fully charged, but AC can), and it acts as a mini instant power-source. A capacitor charges up and in a digital environment, there are current spikes all the time as current is needed in digital time-scales. Like instant now. A battery can't supply that- a battery is a chemical process- it's for massive dump long-haul, but not digital. A capacitor as a bypass cap fills that need by providing instant energy to the load, and the recharging in the down time. Having several in series, just means you have more capacity of current to dump instantly at any given moment of time, so it should be more 'forgiving/reliable'.
Thanks for the explanation. I was hoping they would just operate as 'batteries' in a sense, but just wanted to make sure there would be no other negative effects.
 

Analog Ground

Joined Apr 24, 2019
397
"Bypassing" is a bit of an art. You are going with T.I.'s recommendation which is good (22uF). Also you are using power and ground planes which are great. Use a 22uF for each DRV103. Locate them close to their associated IC with direct connection to the power and ground planes. Use thermally relieved pads if using through hole capacitors. For surface mount, use two or three vias located right beside each of the capacitor solder pads. No long traces. Same treatment for the DRV103 power and ground pads. Get them connected as short as possible to the power and ground planes. Sometimes it is a good idea to add a few larger valued capacitors sprinkled around the board in the area of the drivers. Again, it is more art than engineering without giving it a lot of thought. Here is a link explaining the issue of transient currents as things switch on and off.

https://www.allaboutcircuits.com/technical-articles/clean-power-for-every-ic-part-1-understanding-bypass-capacitors/
 

Thread Starter

rauty

Joined Apr 20, 2017
4
For surface mount, use two or three vias located right beside each of the capacitor solder pads.
Thanks for your suggestions. I do plan to use SMD components throughout except for the output headers. I have all the components in pretty close proximity to each other to minimize space. I will be using some vias to connect to the ground and power planes using short traces. Do I also need to place more of them next to the pads (but without a connecting leg)? Should they just go straight through the board, or connect to the ground plane? I usually use 20 mil vias, is that sufficient? How close to the pads should they be placed? Just outside the mask zone? Lastly, should the via pads be exposed or covered?
 
Top