16-channel 12V PWM output circuit

Go To Last Post
13 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I currently have a PCA9685 LED driver IC, which offers 16 channels of individual PWM control but with only 10mA source capability. I would like to control 16 12V outputs, each at up to 200mA (for activation of coil-based relays). Following on from this post, have there been any major updates in the industry regarding multi-channel N channel MOSFETs? I've searched through all the usual websites and haven't come across anything obvious. I'm somewhat space limited, so won't be able to manage 16 single or 8 dual-channel MOSFETs, and would rather look at 8- and/or 16-channel options (if they exist) before I look to use quad channel ICs.

 

If 8- and 16-channel MOSFETs aren't available, is there another option I could consider? The ULN2803A seems like a good option, but I'm not sure how I'd maintain a 12V output given the additional voltage drop?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Maybe these...but watch out for total allowed current & total power (heat) in one package

 

http://www.ti.com/lit/ds/symlink/tpl7407l.pdf

 

https://www.eenewsembedded.com/news/octal-dmos-fet-arrays-cut-power-loss-multi-channel-driver-circuits

 

look at ST semiconductor...they have a lot of multioutput automotive chips

 

ULN2803A ---bleh, you are correct...a lot of voltage drop!

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That looks great, thanks avrcandies! The DMOS FETs look like they fit the bill very nicely. They're rated to 500mA per channel, which is considerably more than I need as well. The TI option looks like it'd work as well, but I'm not sure why they opted for 7 channels...

 

I'll read through the DMOS datasheet in more detail, and will likely go ahead with this part. Thanks again!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

ULN2803A ---bleh, you are correct...a lot of voltage drop!

 

DMOS version ... TBD62083 from Toshiba.

#1 This forum helps those that help themselves

#2 All grounds are not created equal

#3 How have you proved that your chip is running at xxMHz?

#4 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand." - Heater's ex-boss

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

jars121 wrote:

... PWM control ... for activation of coil-based relays...

 

I'm curious as to your application. Relays aren't really known as being high-speed devices.

#1 This forum helps those that help themselves

#2 All grounds are not created equal

#3 How have you proved that your chip is running at xxMHz?

#4 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand." - Heater's ex-boss

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm curious as to your application. Relays aren't really known as being high-speed devices.

I've seen where PWM is used to control coil current, so you get a fast latch & then a low-current hold (by lowering the pwm)...perhaps that is it? 

https://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=PWM_and_Auto_Relays&DocType=CS&DocLang=EN

 

I'm curious as to your application

 12 relays to drop me a dozen different doughnuts

 

they have a whole bunch of drivers...look toward bottom of link   ...the rds  is somewhat a joke (2 typ, 3 ohms max), dip pkg not much good for hi-current anyhow, even if rds low.  Nowadays, even 0.1 ohm is not so good.

https://toshiba.semicon-storage.com/us/product/linear/transistor-array.html

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sat. Mar 9, 2019 - 03:52 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry, I should have added a little more detail. I would like PWM capability, as well as high enough drive to use coil-based relays, not necessarily at the same time :) E.g. simple ON/OFF control of coil-based relays, or PWM control of some other device. The 16 outputs are user-configurable, so I'm just trying to cover the bases.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quick datasheet question regarding the TBD62783 from Toshiba.

 

 

Am I reading the above correctly (using the FWG model as an example) that the maximum channel current is 400mA if only a single channel is used, which drops to just 160mA (per channel) if all 8 channels are used at 50% duty? It makes no mention of maximum per channel current if all 8 circuits are on with 100% duty? I.e. how do I know the worst case scenario?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Am I reading the above correctly (using the FWG model as an example) that the maximum channel current is 400mA if only a single channel is used, which drops to just 160mA (per channel) if all 8 channels are used at 50% duty? It makes no mention of maximum per channel current if all 8 circuits are on with 100% duty? I.e. how do I know the worst case scenario?

Yep, that's prob correct....it all boils down to heat (pun intended)...with an Rds of 2-3ohms....much heat can be developed.  Also all channels share one common pin, creating a current bottleneck/limit there.  

 

Why not use 12 of these:

https://www.mouser.com/datasheet/2/408/SSM3K56FS_datasheet_en_20140404-1151071.pdf

 

You can't say you don't have room, since you MUST have room available for copper heat dissipation.

 

It would be wise to use protested fet (protected against shorts,overtemperature, etc)  ....see Zetex for parts  search digikey for ZXMS & it will pop up a list

 

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sun. Mar 10, 2019 - 12:09 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think you're right. As much as I'd like to use two of the TBD62783 ICs, I think the current bottleneck and lack of any inherent over-temperature/over-current control makes it unfeasible. I'll look into some quad drivers in the meantime, otherwise 16 single drivers would certainly work, and as you've stated, I'd find the room!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've been doing some further reading on this today, and wanted to make sure I'm on the right track. Below is a schematic for one of the 16 outputs of the PCA9685 PWM LED driver:

 

 

  • Each PCA9685 output connects to an N-Channel MOSFET.
  • A resistor to GND (R2) offers some protection against floating PCA9685 output during reset.
  • A series resistor (R1) will have a footprint assigned on the PCB which will allow me to try a few different resistance values to account for the MOSFET gate charge (starting at 0 Ohm and working my way up).
  • Each driven device (which could be inductive or non-inductive) connects to the PCB via header pins (CONN1).
  • Each driven device is connected to a common 12V rail on the PCB.
  • A Zener flyback diode is placed in parallel with the MOSFET to provide some protection for inductive loads.

 

Does this look/sound ok? As for the flyback diode, is a Zener the obvious choice? If so, what would be the ideal Zener voltage for the diode?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You can leave room for the zener, but don't use it, generally.  The fet's zener should be able to handle  things ,except for extreme  situations.

You could specify avalanche rated fets, which  are even more ruggedized.

Why do you confuse yourself by using the wrong symbol...99.99% of  mosfets have the intrinsic zener

 

I'd use a 10k pull down  &  100-1000 ohm for gate if switching on-off like a light switch.  100k pulldown is not the best considering hot leakages.  A zener for the gate is good ESD protection (some fets have these built in)

If switching sharply at high speed (like you want to switch in 15ns), the gate charge can draw huge current spikes (and often uses a pulse-rated driver).   But for sloppy switching, no worries.  Sharp switching give low heat, but more  transients &  emissions for  FCC testing.

 

 

Sometimes you can add a diode/zener combo at a coil to pass current at shutoff.  Has the advantage that the flyback current is localized right at the coil.   But for low power situations, the mosfet zener can usually do fine.

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sun. Mar 10, 2019 - 04:19 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That all makes sense, thanks again for your help. The maximum switching frequency of the PCA9685 is ~1.5kHz, so I don't need rapid switching capability by any means.