ULN2803 or equivalent in SMD package? TI's???

I'd be glad to order you some samples and let you know what they look like when they come in (:

Possibly you're looking for a narrow (150 mils) body, and the 18 pin is a wide (300 mils) body?

Also, Toshiba has a surface mount version of the ULN2803. I don't have the part number though...

Tony

Thanks, but I would like to order PCBs this weekend if I can :)

Hi,

Either one will save me a boatload of space and a little money. More importantly, this is the last through-hole part in this particular design. I just finished changing all the electros over to SMD, now if I can get rid of that pesky Darlington driver, I will have a significantly cheaper PCB.

ULN2803AFW, 300 mil package, but it's more expensive than the TI part. I guess I should lay out for the Toshiba part and assume that TI's is identical. Actually now I come to think of it, that's very likely to be true. I think I need sleep.

You could always do both a 150 and 300mil patterns. I have not seen std SO packages not on 50mil pitch, but there are offerings in both

150 and 300 mils width - mainly for thermal reasons.

You could also look at SMD MOSFETS, they are lower loss than darlingtons, and so can give a smaller thermal layout ?

-jg

The only thing that worries me is pinouts. I don't feel comfortable assuming things. Chip mfrs do weird things for no apparent reason.

This is also a "mature" design that is just being respun for lower production cost, so I don't want to change it any more than necessary. I'm going to take a chance on rework and order a 250-piece run of boards. This component is the single item I'm not happy about; everything else, I have checked datasheets for all important characteristics, and I've physically checked components on a 1:1 printout of the artwork.

I don't really have a thermal problem though. The package does not become noticeably warmer than ambient during normal operation of the appliance. I'm actually using paired channels to drive four outputs, each output only needs to sink 120mA continuous at most (more typically,

60mA), and the switching speeds are low. There *are* some transient current spikes that can get up pretty high, which is why I want plenty of margin. But they are rare and don't significantly contribute to chip heating.

The other part I'm going to be looking for soon is an H-bridge IC equivalent to the LMD18200 or 18201 (the next thing I'm costing-down is the DC motor controller). That should be fun. I bet it will be some odd package with big heatsink tabs. OBTW, these are submarine modules in case you didn't guess :)

Alegro has taken them of the market (almost)

The 'W' almost universally means wide body, which is 300 mils. I see someone mentioned Allegro is dropping the ULN2803, (and TI is only doing the SMD one now ) - it could be devices like TPIC6B273 /

595 etc (which have 3 current grades) have replaced it ?

See

-jg

Right... However TI and Toshiba (and I think ST?) list them as current production without caveats.

The problem is finding a replacement with similar characteristics. I think that the target applications for these devices are being filled by ASSP microcontrollers with high-current drivers built in.

It's simple enough to build something out of discrete components, but it's easier to use an off-the-shelf part like the ULN28xx.

ULN2803 - like all darlingtons - drop around 1V. This can make them marginal in 5V systems, let alone lower voltages.

"Digital" transistors are very easy indeed.

--
Mike Page BEng(Hons) MIEE           www.eclectic-web.co.uk
Quiet! Tony's battling the forces of conservatism, whoever we are.