Implementing a MC33063AD

[Watth]

Hey every one.
MC33063AD is a buck/boost converter. http://www.ti.com/lit/ds/symlink/mc33063a.pdf
Being a noob, I am trying to copy the typical application example for a step-up converter into KiCAD.

In this case I wish to provide an output of 5V, from 1 to 3 AA cells.

There a two ground nets, since as I am trying to layout the thing on a one layer PCB, I discovered I 'd need to ad a jumper wire - symbolized by a 0 ohm resistor (R0).
So, can you confirm that it is correct?

Thanks a lot.

[Siwastaja]

I can see absolutely no reason for two ground nets in this case. Just connect everything to the same GND.

[Watth]

I can see absolutely no reason for two ground nets in this case. Just connect everything to the same GND.

It's needed for the routing.

[Siwastaja]

Why/how?

I'd say the only sensible way to route this is to use solid, low-impedance ground plane; when done with 2 layers only, with good via stitching. Avoid any length of traces when going to the ground plane.

I.e., you want to connect Ct1, C1, U1, C0 etc. to the same physical plane, as well as ever possible.

There is a layout example on the datasheet page 17.

[Siwastaja]

Ah, I think I understood. You are DIYing a single layer PCB and must use a jumper wire on the ground net. Yes, signifying the jumper wire using a "component" (specific jumper wire component, or 0R) is OK in itself.

However, in case of a ground connection in a switch mode converter, forget about it, it's the worst layout possible. Do it in dead-bug/3D style on bare unetched copper clad - super easy to do when the circuit is simple like this, and the result will be good because you have a very solid ground plane and low-impedance (short) access to it at every component location.

[Watth]

Yes, I didn't specify it was for a homemade PCB.  Sorry

I now consider using a double layered PCB.
Considering the schematics, do you think my values are correct?

[Siwastaja]

Suggestion:
If the point is learning PCB design, by all means design it, but indeed, do it with 2 layers; use the bottom layer as a ground plane; heck, it's most likely possible to draw layout this simple (+ mostly SMD) so that there are no traces on the bottom, whichs means you don't need to photoresist/expose it, just tape it over with plastic tape to prevent etching!

If you want to have the thing done and working as quickly as possible, deadbug it to a piece of unetched clad. I have had DC/DC converters working in less than 30 minutes that way, and the performance is great due to fairly good layout that is possible in deadbug, but impossible or at least tedious in single-layer PCB.

[Benta]

I'm at a loss as to why two GND nets are required for routing. Please enlighten me. I mostly use FreePCB and have never run into this issue. Is there a design rule that's set incorrectly?

[mariush]

Hey every one.
MC33063AD is a buck/boost converter. http://www.ti.com/lit/ds/symlink/mc33063a.pdf
Being a noob, I am trying to copy the typical application example for a step-up converter into KiCAD.

In this case I wish to provide an output of 5V, from 1 to 3 AA cells.

There a two ground nets, since as I am trying to layout the thing on a one layer PCB, I discovered I 'd need to ad a jumper wire - symbolized by a 0 ohm resistor (R0).
So, can you confirm that it is correct?

Thanks a lot.

My advice would be ... don't do it.

* The chip needs at least 3V in order to run, therefore you need at least 3 AA batteries to get it working. 2 rechargeable AA batteries would only give you about 2.5v, 2 alkaline batteries in the best case scenario would give you 3.1..3.2v and quickly drop close to 3v or below 3v. So basically, you need at least 3 AA batteries to make it work.

* The chip is ancient, has minimal internal protections, runs at low frequencies (<100kHz) , it has high internal power consumption (up to 4mA, datasheet doesn't even specify how much it uses when input voltage is below 5v),  and fairly inefficient, your converter would probably be 65-70% efficient.
You'd lose a lot of power just to keep the ancient chip working, then you'd waste power in the inductor, in the shottky diode, electrolytic capacitor etc

Due to being so old and running at such low frequencies, you need big inductors and big electrolytic capacitors - basically the whole circuit would take about as much room on the circuit board as an additional AA battery.
You'd be better off just going with 4 AA batteries, or 4 x AAA batteries if space is an issue.  4 AAA batteries (at around 1100mAh each) would probably behave like 3 AA batteries at 1600-1800mAh used  with such inefficient boost converter.

Go with 4 AA or AAA batteries ... rechargeable batteries will have 1.15v .. 1.35v , alkaline batteries will have 1.45v.. 1.65v  ... so you have  4.6v to 5.4v for rechargeable batteries and 5.8v to 6.6v for alkaline non-rechargeable batteries.
Whatever you power with 5v , probably works with as little as 4.8v so in reality you're going to use almost all the juice in rechargeable batteries. Don't worry about minimum voltage.
You only need to worry about maximum voltage.

You can basically use a very small, very simple to use LDO (linear regulator with very low dropout voltage) .. you can easily find such chips that have only about 25-50mV dropout voltage (the chip needs a voltage a bit higher than output voltage in order to output a stable set voltage).  Also, most LDOs will output a lower voltage just fine, if the input voltage is below the output voltage+dropout voltage, so for example if you configure a LDO to output 5v but your batteries output 4.9v, the LDO will happily output 4.9v - dropout voltage.

Just figure out the maximum current consumption of your circuit and pick a LDO that would be capable of providing that much current, with dropout voltage as low as possible, and which can take up to 7-8v (possible with 4 fresh alkaline batteries)
Most of these LDOs only need a couple of ceramic capacitors which barely take any pcb space, and the adjustable ones only need a couple of resistors to set the output voltage.
If the device uses a lot of current, I would probably use an adjustable LDO and set the voltage to my circuit to about 5.2v (or as much as I'm sure my chips would tolerate, most chips like microcontrollers can go up to 5.5v in perfect safety conditions so 5.2v wouldn't be a big deal... the reason for going above 5.2v would be to lose as little as possible of energy in the LDO.

For example, 5.2v in ( 4 x 1.3v = 5.2v) , 5v out .. basically have 96.1% efficiency  ...  6.2v in ( 1.55v x 4 for alkaline batteries) and 5.2v out = ~ 84% efficiency .. better than what this mc33063 can do

Here's examples of LDOs that would work here:

TPS77050DBVT  Fixed LDO Voltage Regulator, 2.7V to 10V, 35mV Dropout, 5Vout, 50mAout, SOT-23-5 : http://uk.farnell.com/texas-instruments/tps77050dbvt/ldo-fixed-5v-0-05a-sot-23-5/dp/2382990
NCP551SN50T1G  Fixed LDO Voltage Regulator, 12V in, 40mV Dropout, 5V/150mA out, TSOP-5 : http://uk.farnell.com/on-semiconductor/ncp551sn50t1g/ic-ldo-150ma-cmos-low-iq-smd/dp/1468772
TEXAS INSTRUMENTS  REG101NA-5/250G4  Fixed LDO Voltage Regulator, 1.8V to 10V, 60mV Dropout, 5Vout, 100mAout, SOT-23-5 : http://uk.farnell.com/texas-instruments/reg101na-5-250g4/v-reg-ldo-0-1a-5-0v-sot-23-5-101/dp/1207241RL
TPS76950DBVT  Fixed LDO Voltage Regulator, 2.7V to 10V, 71mV Dropout, 5Vout, 100mAout, SOT-23-5: http://uk.farnell.com/texas-instruments/tps76950dbvt/regulator-ldo-fixed-5v-sot-23/dp/2323601
BA050LBSG2-TR  Fixed LDO Voltage Regulator, 2.5V to 7V, 90mV Dropout, 5Vout, 150mAout, SMP-5 : http://uk.farnell.com/rohm/ba050lbsg2-tr/linear-reg-fixed-5v-0-28a-smp/dp/2343139
ADP3301ARZ-5  Fixed LDO Voltage Regulator, 3V to 12V, 100mV Dropout, 5Vout, 100mAout, SOIC-8 : http://uk.farnell.com/analog-devices/adp3301arz-5/ldo-0-8-0-1a-5v-8soic/dp/2094222
MC78FC50HT1G  LDO VOLTAGE REGULATOR, 5V, 120mA, SOT-89-3 : http://uk.farnell.com/on-semiconductor/mc78fc50ht1g/ldo-voltage-regulator-5v-120ma/dp/1254286
LP2980IM5-5.0  Fixed LDO Voltage Regulator, 2.1V to 16V, 120mV Dropout, 5Vout, 50mAout, SOT-23-5 : http://uk.farnell.com/texas-instruments/lp2980im5-5-0/v-reg-ldo-5-0v-smd-2980-sot235/dp/9779329

... and these are only the "fixed 5v output" ones, there's about 200+ adjustable output LDOs I filtered out before picking those above.

If you insist on having some boost dc-dc converter in your design, pick something more modern.. here's some suggestions:

LT1303CS8-5#PBF  Boost (Step Up) Switching Regulator, Fixed, 1.8V-10V In, 5V And 200mA Out, 185 kHz, SOIC  http://uk.farnell.com/linear-technology/lt1303cs8-5-pbf/ic-switching-regulator-155khz/dp/2238166
NCP1400ASN50T1G  Boost (Step Up) Switching Regulator, Fixed, 800mV-5.5V In, 5V And 0.1A Out, TSOT-23-5 : http://uk.farnell.com/on-semiconductor/ncp1400asn50t1g/ic-switching-reg-180khz-sot23/dp/1460673
NCP1402SN50T1G  Boost (Step Up) Switching Regulator, Fixed, 800mV-5.5V In, 5V And 200mA Out, 200 kHz, SOT-23-5 : http://uk.farnell.com/on-semiconductor/ncp1402sn50t1g/ic-dc-dc-converter-5sot2/dp/1652393
ADP3050ARZ-5  Buck, Buck-Boost, SEPIC Switching Regulator, Fixed, 3.6V-30V In, 5V And 1A Out, 240 kHz, NSOIC-8 : http://uk.farnell.com/analog-devices/adp3050arz-5/buck-5v-1a-8soic/dp/2074857
 TPS61032PWPR  Synchronous Boost Step Up, Switching Regulator, Fixed, 1.8V-5.5V In, 5V/4A Out, 700 kHz, HTSSOP-16 : http://uk.farnell.com/texas-instruments/tps61032pwpr/ic-sync-boost-conv-600khz-htssop/dp/1207333

again, these only fixed 5v output, plenty more that can be adjusted to 5v output using 2 resistors.   Most of the above go from about 200kHz and up, so you can use tiny surface mount inductors and ceramic capacitors , but the higher the frequency the more important it is to do proper routing (very small loops between chip and inductor, thick ground and voltage traces etc etc

[Watth]

Thanks all for your feedback.
Since I already have the MC33063AD, I'll still use them, but, as advised, on a double layer board. Their avantage is that they can handle up to 1.5A. I might use it for simple loads such as LEDs (I want to convert a USB lamp to be powered by 4 or 3 AA).
Thanks, mariush, for those suggestions.

[mariush]

Ok, thanks for answering, good to know you read my long message.

Also, remember that the 1.5A is the output switch current... it doesn't mean the chip can output 1.5A of current at any output voltage. Especially in boost mode, it doesn't means you can make for example a circuit that takes 5v in, and outputs let's say 12v at 1.5A  ... that's not what they mean in the datasheet by 1.5a switch current.
In real world, for a 5v to 12v, you're looking at something like 12v @ 0.25A maximum.