PCB shop for hobby projects?

P

Phil Graham

Honorary PhD
Mar 10, 2011
651
1
18
Atlanta, GA
Does anyone have recommendations for hobbyist PCB house? There are a whole bunch of options (e.g. ExpressPCB), but I welcome first hand experience. I'm putting together some motor controllers and/or a buck converter.

Nothing fancy: small 2 layer boards, silkscreen and soldermask, SMT or through hole, ability to panelize several boards.

Thanks in advance for any recommendations!
 
Re: PCB shop for hobby projects?

I've used these guys (in china) several times for prototype orders

PCB board manufacturing, PCB prototyping service & PCB assembling services - Gold Phoenix

It looks like they updated their website... Last time I used them it was a fixed price for all the boards that come out of one standard sized panel. You can select more options al a carte but basic 2 sided board is more than adequate for prototyping.

IIRC boards are sent from china by air, so turn around time is decent.

JR
 
Re: PCB shop for hobby projects?

I really like APCircuits for prototyping.. if you can use their pre-selected drill sizes you can get a board shipped out next business day (no mask or silk) then when you're ready to make a bigger order you can go with one of the slower but cheaper options.

Jason
 
Re: PCB shop for hobby projects?

Pad to pad does any size you want-single or up to 4 layer boards.

The price is totally variable on how many or how long you can wait. They can do 1 day orders-but the biggest price break is 10 working days.
 
Phil Graham said:
Does anyone have recommendations for hobbyist PCB house? There are a whole bunch of options (e.g. ExpressPCB), but I welcome first hand experience. I'm putting together some motor controllers and/or a buck converter.

Nothing fancy: small 2 layer boards, silkscreen and soldermask, SMT or through hole, ability to panelize several boards.

Thanks in advance for any recommendations!
Click to expand...

Seeed Studio Fusion PCB service. 2-layer, 5cm x 5cm, 10 boards for $10. Price increases a bit with size, but still can't beat it.
 
Re: PCB shop for hobby projects?

Phil,

I have used ExpressPCB for two projects with good results. I like their design tools: simple but not too simple. You can do ground-plane fills and pretty much all the big boy stuff with just a little effort. Not saying others are not as good or better, but this worked well and I'll use them again.

--Frank
 
Re: PCB shop for hobby projects?

Hey again, rather than starting a new thread, I'm going to ask a related question.

Before I go off and drive my high side FET gates with a bootstrap cap, does anyone have recommendations on an an IC targeted at providing 5V above V_nom to make sure the FET is fully open? Supply side will be batteries of indeterminate voltage (and recharging regeneratively), but somewhere between 24 and 48V most likely.

Thanks!
 
Re: PCB shop for hobby projects?

Phil Graham said:
Hey again, rather than starting a new thread, I'm going to ask a related question.

Before I go off and drive my high side FET gates with a bootstrap cap, does anyone have recommendations on an an IC targeted at providing 5V above V_nom to make sure the FET is fully open? Supply side will be batteries of indeterminate voltage (and recharging regeneratively), but somewhere between 24 and 48V most likely.

Thanks!
Click to expand...
I can find my way around a circuit but not sure exactly what you're looking for. I've see switching regulator ICs that have a charge pump boost circuit built in to generate a low current several volts higher bias supply.

If you are switching with a reasonable clock frequency its pretty straightforward to generate a small step up voltage supply with a few caps and a couple diodes.

JR
 
Re: PCB shop for hobby projects?

John Roberts said:
I can find my way around a circuit but not sure exactly what you're looking for. I've see switching regulator ICs that have a charge pump boost circuit built in to generate a low current several volts higher bias supply.

If you are switching with a reasonable clock frequency its pretty straightforward to generate a small step up voltage supply with a few caps and a couple diodes.

JR
Click to expand...

JR,

Thanks for the feedback.

Let me explain in more detail:

  1. The project/research I want to work on is battery charging/discharging control. The motor controller is a means to that end, as it provides for battery discharging and charging. I'm going to use brushed DC motors.
  2. I'm using International Rectifier's IR21844 half-bridge MOSFET driver (http://www.irf.com/product-info/datasheets/data/ir2184.pdf). Feed it PWM and it automatically pumps out a high/low pair with resistor-adjustable switching deadtime.
  3. Once I get the boards done and built, an Arduino (or Arduino-compatible) will provide the pulse and duty cycle control. I've got to derive power for the Aurduino regardless. On the motor side of the fence, I'm going to use an optical encoder to measure RPM directly, that way I can observe back EMF vs. RPM. The arduino will also take in V-I data from the battery side. I'll probably mount it on a bicycle or razor scooter.
  4. Most of the time I expect the unit to operate via active rectification, with PWM duty cycle changing to alternate between ramp to current limiting under acceleration (low FET) and ramp to voltage limiting during regenerative braking (high FET). Near the top of the charge curve the unit should also exhibit current limiting behavior.
  5. Once the batteries are near the top of their charge state, the high side FET might need to stay open for extended periods of time to act as a freewheeling diode. It is this usage state where I don't think the bootstrap is going to work, as the circuit might become charge depleted before the next low side cycle re-charges the bootstrap.
This project is not about me re-inventing the wheel on motor controllers, but rather about getting some practical circuit board layout experience, Arduino coding experience, and playing around with the battery and motor algorithms. At all points where I can use an off the shelf bit to get data into the Arduino, provide voltages, etc. I will.

Down the road I'd like to take what I've learned and apply it to a windmill/battery combination, or PV/battery combo in a practical home setting.
 
Last edited:
Re: PCB shop for hobby projects?

Phil Graham said:
JR,

Thanks for the feedback.

Let me explain in more detail:


[*]The project/research I want to work on is battery charging/discharging control. The motor controller is a means to that end, as it provides for battery discharging and charging. I'm going to use brushed DC motors.
[*]I'm using International Rectifier's IR21844 half-bridge MOSFET driver (http://www.irf.com/product-info/datasheets/data/ir2184.pdf). Feed it PWM and it automatically pumps out a high/low pair with resistor-adjustable switching deadtime.
Click to expand...
OK schematics help... 21844 has the charge pump build in for HS switch turning on.
[*]Once I get the boards done and built, an Arduino (or Arduino-compatible) will provide the pulse and duty cycle control. I've got to derive power for the Aurduino regardless. On the motor side of the fence, I'm going to use an optical encoder to measure RPM directly, that way I can observe back EMF vs. RPM. The arduino will also take in V-I data from the battery side. I'll probably mount it on a bicycle or razor scooter.
[*]Most of the time I expect the unit to operate via active rectification,
Click to expand...
I'm not sure I understand what "active rectification" means in this context. The high side mosfet switch has a body diode that will conduct whenever the motor voltage (back EMF?) is a diode drop higher than battery voltage.
with PWM duty cycle changing to alternate between ramp to current limiting under acceleration (low FET) and ramp to voltage limiting during regenerative braking (high FET).
Click to expand...
Again not sure I follow... I may need to re-eductate myself about brushed DC motors (magnet? field coil? ?) Under acceleration, on/off duty cycle will modulate battery energy applied to motor. For braking it seems voltage coming from motor could depend on speed like a generator in a car. The alternator in a car modulates an internal coil winding to modulate output current.
Near the top of the charge curve the unit should also exhibit current limiting behavior.
[*]Once the batteries are near the top of their charge state, the high side FET might need to stay open for extended periods of time to act as a freewheeling diode.
Click to expand...
How do you control the regenerative braking separately from the forward acceleration?
It is this usage state where I don't think the bootstrap is going to work, as the circuit might become charge depleted before the next low side cycle re-charges the bootstrap.
Click to expand...
bigger cap.. The circuit may be self-charging, so first on cycle may be weak but it will catch up, just like when motor first turns on.
This project is not about me re-inventing the wheel on motor controllers, but rather about getting some practical circuit board layout experience, Arduino coding experience, and playing around with the battery and motor algorithms. At all points where I can use an off the shelf bit to get data into the Arduino, provide voltages, etc. I will.

Down the road I'd like to take what I've learned and apply it to a windmill/battery combination, or PV/battery combo in a practical home setting.
Click to expand...

Lets NOT reinvent this wheel. Do you have a similar design that "inspired" you that we can look at to see how this has been done before? I am actually OK with reinventing wheels but first I like to look at all the old wheels so i don't reinvent any old mistakes.

Perhaps I just don't understand how your selected motor behaves, I am not a motor expert. I gave some positive attention to regenerative braking for buses and subways back a few decades ago when it was a new concept, and more recently for race cars where it is very promising (not so much for NASCAR but for racing that involves braking for turns). There should be lots of art out there to look at.

Sorry if you are already way ahead of me on this design.

JR

edit- electric car just did 200MPH over measured mile /edit
 
Last edited:
Re: PCB shop for hobby projects?

John Roberts said:
I'm not sure I understand what "active rectification" means in this context. The high side mosfet switch has a body diode that will conduct whenever the motor voltage (back EMF?) is a diode drop higher than battery voltage.
Click to expand...

Right John, you'll see these circuits with no high side FET, just a diode.

When the high side opens and the low side closes, depending on the controller voltage, current will flow through the low side back to the battery.

Here's a pretty good intro page if you want a quick refresher:
4QD-TEC: PWM speed control

How do you control the regenerative braking separately from the forward acceleration?
Click to expand...

You don't, as I'm sure you guessed :)

Forward acceleration is set by a duty cycle that raises the voltage above the present, instantaneous value of the motor's back EMF, and deceleration (via regenerative braking) is set by lowering the duty cycle below the present back EMF value. For instance, you could determine the EMF pretty well by comparing the duty cycle % vs. current flow from the battery. You can also correlate with the present RPM and acceleration vector.

You build "inertia" into the code in how it executes a delta in duty cycle. The delta increase in V_avg to behave as current limiting, and the delta decrease to behave as limiting on deceleration.

bigger cap.. The circuit may be self-charging, so first on cycle may be weak but it will catch up, just like when motor first turns on.
Click to expand...

I wonder if there's a way to make the software always drive the low side first without too much logic complications.

Lets NOT reinvent this wheel. Do you have a similar design that "inspired" you that we can look at to see how this has been done before? I am actually OK with reinventing wheels but first I like to look at all the old wheels so i don't reinvent any old mistakes.
Click to expand...

If I was going to buy off the shelf, it would probably something like this Roboteq 30V - 40A unit:
SDC1130 - Single Ch. 40A, Brushed DC Motor Controller | Brushed DC Motor Controllers

Sorry if you are already way ahead of me on this design.

JR
Click to expand...

No apologies, you're way ahead on all the useful, necessary outside electronics surrounding the half-bridge. I greatly appreciate the help and questions.
 
Last edited:
Re: PCB shop for hobby projects?

Phil Graham said:
Right John, you'll see these circuits with no high side FET, just a diode.

When the high side opens and the low side closes, depending on the controller voltage, current will flow through the low side back to the battery.

Here's a pretty good intro page if you want a quick refresher:
4QD-TEC: PWM speed control
Click to expand...
I think I get some of it now... The duty cycle (if frequency is fast enough) applies an average voltage to the motor. If this average voltage is more than the speed/demand of the motor it will draw current from battery and accelerate the motor. If average voltage applied based on duty cycle is less than voltage generated by motor, current flows back into the battery. So magic is in tweaking duty cycle for speed/load.
You don't, as I'm sure you guessed :)

Forward acceleration is set by a duty cycle that raises the voltage above the present, instantaneous value of the motor's back EMF, and deceleration (via regenerative braking) is set by lowering the duty cycle below the present back EMF value. For instance, you could determine the EMF pretty well by comparing the duty cycle % vs. current flow from the battery. You can also correlate with the present RPM and acceleration vector.


You build "inertia" into the code in how it executes a delta in duty cycle. The delta increase in V_avg to behave as current limiting, and the delta decrease to behave as limiting on deceleration.




I wonder if there's a way to make the software always drive the low side first without too much logic complications.



If I was going to buy off the shelf, it would probably something like this Roboteq 30V - 40A unit:
SDC1130 - Single Ch. 40A, Brushed DC Motor Controller | Brushed DC Motor Controllers



No apologies, you're way ahead on all the useful, necessary outside electronics surrounding the half-bridge. I greatly appreciate the help and questions.
Click to expand...

I am still a little fuzzy on this.. maybe i need to soak the idea in beer and look at it again later.

it seems easiest how much "go or slow" control is to measure current in/out of battery with a very low impedance series R (perhaps a piece of wire). A gas pedal that directly varies duty cycle could be adjusted manually based on how fast you want to go. Many microprocessors have multiple A/D converters available, perhaps detect current in low side of battery, with level shift to always be positive voltage. Utility of detecting RPM of motor depends on linearity of voltage output.

JR
 
Re: PCB shop for hobby projects?

John Roberts said:
I think I get some of it now... The duty cycle (if frequency is fast enough) applies an average voltage to the motor. If this average voltage is more than the speed/demand of the motor it will draw current from battery and accelerate the motor. If average voltage applied based on duty cycle is less than voltage generated by motor, current flows back into the battery. So magic is in tweaking duty cycle for speed/load.
Click to expand...

That's the guts of it.

it seems easiest how much "go or slow" control is to measure current in/out of battery with a very low impedance series R (perhaps a piece of wire). A gas pedal that directly varies duty cycle could be adjusted manually based on how fast you want to go. Many microprocessors have multiple A/D converters available, perhaps detect current in low side of battery, with level shift to always be positive voltage. Utility of detecting RPM of motor depends on linearity of voltage output.

JR
Click to expand...

I'll probably use a hall effect sensor, as I'd like to know the current direction.

Lots of thing that can be done there in software. I just want to get the controller up and running.
 
Re: PCB shop for hobby projects?

Phil Graham said:
That's the guts of it.



I'll probably use a hall effect sensor, as I'd like to know the current direction.
Click to expand...
Voltage drop across a sense resistor would indicate polarity of current.

Hall effect, is magnetic field responsive. Does that involve a one turn coil like a current probe? I guess current flowing in a straight wire generates a magnetic field.

I recall putting a hall effect electronic points replacement on my motorcycle back in the '80s. The problem with the magnet on a disc, was that you could run the two stroke motor backwards. Of course you couldn't crank it that way but stuff happens. The company sent a replacement disc with two magnets placed so it couldn't start/run backwards.
Lots of thing that can be done there in software. I just want to get the controller up and running.
Click to expand...

Yes, I am a convert to the dark (digital) side... you can do "almost" anything you can imagine.

JR
 
Re: PCB shop for hobby projects?

John Roberts said:
Voltage drop across a sense resistor would indicate polarity of current.
Click to expand...

Sorry, I was trying to make a bad joke about the Hall effect. I remember having to calculate which face the charge piles on given the direction of current flow. From Wikipedia: https://en.wikipedia.org/wiki/File:Hall_Effect_Measurement_Setup_for_Electrons.png

I found a page that shows that the sketch chinese controllers just use a big ol' 47uF in parallel with a little ceramic in their bootstrap circuit, and omit the gate pulldown resistor. I suppose that's one way to keep things working.
 
Phil Graham said:
Does anyone have recommendations for hobbyist PCB house? There are a whole bunch of options (e.g. ExpressPCB), but I welcome first hand experience. I'm putting together some motor controllers and/or a buck converter.

Nothing fancy: small 2 layer boards, silkscreen and soldermask, SMT or through hole, ability to panelize several boards.

Thanks in advance for any recommendations!
Click to expand...

I use a local pcb shop by one of our warehouses. They have a general stock of drill sizes that are included for a few hundred holes (I've never come close to half) and can usually do an 'after lunch tomorrow' turnaround on anything that will fit into a rack unit in small qty.

I haven't done any large orders as the orders sizes I do are in multiples of 2 and I usually only build/design pairs of things.

Just have a chat with the operations guy about what format/layers to send files and any sizing/design to keep down material waste and handling costs. Usually they will have a web or offline program to send them over the web.

It's not as cheap as the overseas, but I can't beat the time and distance when I get an idea at 3 in the morning. If I was doing large orders on proven boards, then that probably would change.

I've thought about doing my own, either small cnc or old fashion chemicals, but the reality is I don't have the time, let alone it's hard to find the time to get something finished in Eagle in the first place.


BRad
 
Re: PCB shop for hobby projects?

I've used Batch PCB in the past for some small oddly-shaped boards, but they probably aren't what you're after (they consolidate boards from multiple customers onto panels, allowing for small-run prototypes without large setup costs). For larger runs, Advanced Circuits has been a reasonable option and can do quick turnaround if needed.
 
Re: PCB shop for hobby projects?

Take a look at myropcb.com. I've used them for about 10 years with excellent results. I use them for every prototype and production board my company develops.
 
Top Bottom