At a rough guess, about $150/board at qty.250 ($37.5K), or about $125/board at qty.500 ($62.5k), plus shipping.
Cost based on:
- SMT
- assembled and tested
- assumes one one proto run
- same as proto board, but voltage regulators you specified in the thread
- screw power terminals
- you take care of tech support for the board, and any liability for your customers
- based on previous pricing I've received in the past for off-shore assembly and MOQ's
- re-testing in Canada
The Parallax proto board with an external switching regulator down to 9V (from 60 you would like) is a much cheaper solution. and is what I would recommend for you.
The $150 per board sounds okay, but the 250 of them for $37,500 is way out of my reach. Tubular reminded me about the blank Proto Boards for $4.99, I could build my own from them, but then you have the large 40 pin dip.
This is the reason people were critical of your asking Parallax for the change - they would never recoup their investment.
A good way of thinking of it is asking yourself how much you would have to charge to make a board and not lose your shirt on it. You have to account for development costs amortized over number of units, man power costs, overhead, and at least a small profit. Assembly costs are very significant, especially at lower volumes (which also amortize development costs over fewer boards).
The reason for 250 units is so that it is affordable to make it. Assembly costs only start to become reasonable at 250 units; and the parts need to be ordered in quantity... prototyping has to be paid for, testing time of proto's and production boards, pcbs produced, etc etc etc... and a bit of profit.
You will not find any company willing to make a semi-custom or custom board on a one-off basis for $150, as they would lose their shirts on it.. Some may do it at qty.100 if they can reasonably assemble that quantity in-house.
The picture is different, as others have pointed out, if you design the PCB yourself, have it produced, and assemble it yourself. That method is more affordable, if you have the time available, as it gets rid of MOQ's for parts and assembly, and instead of paying consulting rates,you can pay yourself your salary rate for the time required.
The $150 per board sounds okay, but the 250 of them for $37,500 is way out of my reach. Tubular reminded me about the blank Proto Boards for $4.99, I could build my own from them, but then you have the large 40 pin dip.
I can still produce my own board, however, I am still not convinced that Parallax would never recoup their investment. They sell Proto Boards for $24.99, so I am certain they would not be spending a $150 per board to have them made.
And I am still convinced that CNC and automation is a market they should try to enter.
It's not impossible to solder a SMD yourself, just not easy. You have two choices, 1) do it with a regular soldering iron or 2) machine a custom head to solder the whole chip in one operation.
1) You can solder a wire to each tiny surface mount pin by first tinning the wire and the pin then joining them in a seperate solder operaton. Tedious and slow but it works for one-off.
2) You can buy a complete set of soldering tools for $9K I've been told. Instead get a local machinist to cut you a piece of copper to fit over all surface mount pins at once, connect it to an iron, heat it, tin it, and you're ready to go. This may be somewhat better for more than one but less than a custom order.
Since you, Bruce, already work with CNC it should be easy enough for you to chip out your own soldering head and thread a bit of rod to attach it to your soldering iron or a new one that uses screw tips.
I've done my own PCB designs and cut them out with a desktop CNC using a Dremel engraving tool but that's only small enough for 1/10" pin geometry. I don't know how to get smaller without using photo-etch which is chemically messy, but still much cheaper than thousands of dollars, given CNC availability.
I know industrial control is a huge business,I have been in it for about 28 years now...
99% of the time 24VDC is available, thus the value of a board with a switching regulator capable of it 60VDC is limited, and not worth the increased BOM costs.
The current proto board is targeted at engineers and hobbyists for prototyping, not as an industrial board for embedded applications.
Parallax will have amortized development costs over thousands of boards - and they have in-house manufacturing capability. Not to mention being the processor manufacturer
Google the thread Parallax did on their development costs, you will find it interesting.
99% of the time 24VDC is available, thus the value of a board with a switching regulator capable of it 60VDC is limited, and not worth the increased BOM costs.
NOT mention that that Voltage need certificate for High Voltage - Class II.
And I am still convinced that CNC and automation is a market they should try to enter.
With that point, I agree. They should provide a simple set of Propeller circuitry and software that will work to make automation easier. As it is I'm having trouble learning what it takes to make a PAUSE command work to deliver a specific delay, how to drive a FET to make it go its whole range to 24V and how to get an RCTime command to work that works easily on a BS2. I'm trying the strategy of working to improve the Gold Standard http://www.parallaxsemiconductor.com/goldstandard
which I hope will lead in this direction.
Switching regulators capable of a 18-36V (2:1) or 12-36V (3:1 input range (i.e. 24V nominal) are plentiful in both isolated and non-isolated forms. It's not even necessary to provide the regulator with the PCB -- just the footprint so one can be purchased and soldered in if the user desires. But 60V? No, that's a highly specialized requirement and beyond the realm of reason to expect. It would be foolish for a company like Parallax to provide an input voltage range extending that high.
But the Proto Board can't handle that 24VDC, so you must by a seperate switching regulator if you want to use it If not you must supply somewhere around 9V
And just so you all know, this all started after spending a could days researching switching regulators. Of course that is what everyone said I needed.
It's not impossible to solder a SMD yourself, just not easy. You have two choices, 1) do it with a regular soldering iron or 2) machine a custom head to solder the whole chip in one operation.
Actually dead bug can be a **ch for the smallest devices, but there I found a work around. Last spring I finished up prototyping on a small board using SMD parts. I ordered some headers from Proto Advantage that allowed me to wrap the project on standard 1/10 in center board. To mount the devices, I could have had Proto Advantage do this as well, but a cheaper alternative was to simply obtain Zephpaste and air flow them myself. Please don't laugh at my process or think I am gaming you. I am not. My airflow device was modeled after an radio shack desoldering tool as seen on a u-tube video. Paste on the pads, tacky under the body to mount part and reflow. Only hard part was the 56 pin qfp, the center was designed to flow as well, and so could not put sticky under the chip to keep it imobile during reflow. Others have used the zephpaste or similar and toaster ovens (careful of the fumes).
But the Proto Board can't handle that 24VDC, so you must by a seperate switching regulator if you want to use it If not you must supply somewhere around 9V
[*]Change the power connector to a 2-position wire screw clamp
[*]Swap out the existing regulators for something like the LM2591HV-5.0 and LM2591HV-3.3
The gadget ganster has already resolved the first request, you can either power up via a standard plug or on board wire screw clamps.
idbruce, While I sympathize with what you are requesting I can also see the Parallax side of this request. There are just too many power supply voltages in use for various pieces of equipment to provide a board that supports them all. In my case the majority of the systems have either a regulated +5V supply or 24VAC from a class 2 transformer. I have also dealt with 12VDC, 12VAC, 48VDC, 48VAC, 120VDC, 240VAC, 480VAC, and 600VAC but thankfully those have been rare and there are transformers to convert the AC voltages to 24VAC.
When I have 5VDC available I simply short out the 5V regulator with a jumper. For any other voltage I use a transformer to convert it to 24VAC and have a small power supply/switching regulator that converts that to 5VDC. I would love to see a protoboard that has a 2 position screw clamp and no +5V regulator but I don't expect Parallax to provide that. One small piece of wire and 2 solder joints takes care of removing the +5V regulator from the circuit and a solder sucker and 2 pieces of wire replace the power jack.
There's a solution to everything, you just can't be afraid to look for it. I'm currently using the Propellor Platform, PP lcd modual and a PP protoboard along with parallaxs' Digital IO Board. All this is powered from 12Vdc and proved a bit much for the onboardd regulators. So I didn't use the included screw terms for power. I fed the 12Vcd to the protoboard through a series of 4 1N4002 diodes then into the Vin of the PP modual. Ta da , the 12vdc is now 8.2vdc and my regulators are very happy.
What's my point... You don't need a custom board from parallax. The boards they already make will work. Let's say there Propeller proto board. There is a place for a 3 pin header rigth by the 3.3v reg, one of the holes is lables Vin. Now get one of there 3x4 protoboards put your switching regulator circuit on it and feed its output into the Vin on the propeller protoboard, mount it to the bottom side and you got yourself one sweat HV board and for far less then it would cost you or parallax to make a new design. You would probably even have some room left over for some driver IC's.
I ran into the same problem when working with the BS2 proto board (Item 27120). I had to control a large set of relays capable of switching over 5000VAC for HiPot testing, and while the regulators on the BS2 were sufficent to run the low-level switching and other goodies, I had to design my own higher current power supplies to run the bigger set of relays that did the real work. I found, in my opinion this was the best option. I'm not saying there is not a market for it by any means, it would have been convieniant for me to have such an option with the prop, but it does not take much to design in something like this on your own, and move away from the "Hobby" sector, and develop your own circuitry. It certainly should take less time to do this than it would to rant in the forums about how much you need one made for you. This includes part research, CuCl etching, board population, and test. Don't get me wrong, it is a great idea, I just fear, like the others, this is a very specialized item that few would find the need for, even with the amount of people I read about here that do have "Industrial needs".
If you folks want to continue to hobble high voltage applications together by altering boards, or adding parts to the Proto Board to overcome it's HV weaknesses, then I am sure you will continue to do so for many years to come, because instead of supporting my idea, you are telling Parallax, "I am happy adding switching regulators, shorting voltage regulators, or adding extra boards to overcome this weakness". For those of you who do not support this suggestion, I hope you continue to do it YOUR way for the rest of your mortal lives. AS FOR ME I would prefer a seperate Proto Board for high voltage applications.
Sincerest apologies... I did not mean to upset you. I have read many fantastic ideas that you have, and did not mean to turn this into a flaming thread. I was just giving an alternate solution, as your request. I have a few DipTrace files that I drew up and recycle certain schematics from them for different applications, and designed a protoboard with all of the goodies in one spot for my base prototyping, including bigger regulators, RTC, TVout, option for battery or wall current supplies, etc, etc. Once I get a base design going, I then rip the designs I need out of my DipTrace files, and make up the custom board. They are professional looking, and when we start distributing my equipment out to our other plants, it is just a matter of sending my files out to the board house and have them made up so I don't have to monkey with making 100's of said item by hand. they are not "Hobbled" together by any means, and it is a solution that has worked very well for me. well, at any rate, here is a pic of my current rev proto board that I made up. YES it is on cheap perfboard, but again, this is just my proto stage, not anything that is going to be relied on in the field. If a solid solution were to be already available, I would not have put the time and energy into this setup, but, as it stands, I have no need for it since the "Basic" tools I have incorporated into it for my needs would still not be there.
You did not upset me, nor did anyone else. I just have my point of view, and everyone else has theirs. However, I do hope that the non-supporters do it their way for the rest of their lives, like I previously stated.
or adding parts to the Proto Board to overcome it's HV weaknesses
I think that's where you don't seem to want to hear what others are saying: most of us would not in fact add HV components to a Proto Board, we'd order a custom PCB though one of the various vendors providing that service. I find it fascinating that you'll cast molten aluminum to build a custom case, but won't make a custom PCB for one of your projects.
Custom PCBs for a finished product I can handle, but custom PCBs for prototyping is not for me. In my eyes, the Propeller Proto Board exists so that people may experiment and hopefully it will result in the sale of additional Propeller chips for Parallax in finished products. The same should hold true for a HV Proto Board.
In the next couple days, I will post a few photos to this thread of a real nice looking and functional product that I am working on. In this particular project, it would have been a wonderful thing if I had a Propeller HV DIP Plus Kit, but I didn't have one. Why not you ask? Because they don't exist. Now of course I can assemble my own Propeller HV DIP Plus Kit, but it would have been nice to have had all the parts in one kit without dealing with all the hassle of ordering all the seperate parts etc.... And this will be a product that I intend to commercialize. When you see it, you are gonna say, "That Bruce is a trouble maker, but he sure makes nice stuff"
Look forward to seeing it. I visited a local supplier yesterday who has some nice proto boxes, screw terminals etc. Doesn't quite fit a prop proto board unfortunately - wrong proportion, but propstick will be fine.
One afterthought - there are accepted pcb standards that dictate the minimum spacing between conductors at various voltage levels. There are rules for >30v, >50v etc, and it depends on the kind of coating present too. The proto boards have that nice ground plane mask that at 60V may or may not violate these rules. I think if they are deemed to be 'coated' it may be fine, but I'm not an expert in this area, It needs looking at
I don't know if you need a local supplier in Melbourne, but if you can order from Mouser, they sell Hammond and Budd boxes. You may want to look at these links.
Screw terminals are an extremely poor choice for industrial environments, especially when vibration is present. For individual connections, screwless, spring-loaded terminal blocks offer much more compliance, and a nearly zero possibility of a connection working loose from vibration. But why stop there? If you've ever had to swap out equipment from a control cabinet, you will soon learn to appreciate the advantages of mass-pluggable terminal blocks. Removing individual wires from the old and reterminating them in the new is yesterday's news -- actually 1950's news.
Comments
Just out of curiousity, what would an order like that run? Of course I understand it would just be a guesstimate.
Bruce
At a rough guess, about $150/board at qty.250 ($37.5K), or about $125/board at qty.500 ($62.5k), plus shipping.
Cost based on:
- SMT
- assembled and tested
- assumes one one proto run
- same as proto board, but voltage regulators you specified in the thread
- screw power terminals
- you take care of tech support for the board, and any liability for your customers
- based on previous pricing I've received in the past for off-shore assembly and MOQ's
- re-testing in Canada
The Parallax proto board with an external switching regulator down to 9V (from 60 you would like) is a much cheaper solution. and is what I would recommend for you.
The $150 per board sounds okay, but the 250 of them for $37,500 is way out of my reach. Tubular reminded me about the blank Proto Boards for $4.99, I could build my own from them, but then you have the large 40 pin dip.
Bruce
That is exactly my point, and that of others.
This is the reason people were critical of your asking Parallax for the change - they would never recoup their investment.
A good way of thinking of it is asking yourself how much you would have to charge to make a board and not lose your shirt on it. You have to account for development costs amortized over number of units, man power costs, overhead, and at least a small profit. Assembly costs are very significant, especially at lower volumes (which also amortize development costs over fewer boards).
The reason for 250 units is so that it is affordable to make it. Assembly costs only start to become reasonable at 250 units; and the parts need to be ordered in quantity... prototyping has to be paid for, testing time of proto's and production boards, pcbs produced, etc etc etc... and a bit of profit.
You will not find any company willing to make a semi-custom or custom board on a one-off basis for $150, as they would lose their shirts on it.. Some may do it at qty.100 if they can reasonably assemble that quantity in-house.
The picture is different, as others have pointed out, if you design the PCB yourself, have it produced, and assemble it yourself. That method is more affordable, if you have the time available, as it gets rid of MOQ's for parts and assembly, and instead of paying consulting rates,you can pay yourself your salary rate for the time required.
I can still produce my own board, however, I am still not convinced that Parallax would never recoup their investment. They sell Proto Boards for $24.99, so I am certain they would not be spending a $150 per board to have them made.
And I am still convinced that CNC and automation is a market they should try to enter.
Bruce
1) You can solder a wire to each tiny surface mount pin by first tinning the wire and the pin then joining them in a seperate solder operaton. Tedious and slow but it works for one-off.
2) You can buy a complete set of soldering tools for $9K I've been told. Instead get a local machinist to cut you a piece of copper to fit over all surface mount pins at once, connect it to an iron, heat it, tin it, and you're ready to go. This may be somewhat better for more than one but less than a custom order.
Since you, Bruce, already work with CNC it should be easy enough for you to chip out your own soldering head and thread a bit of rod to attach it to your soldering iron or a new one that uses screw tips.
I've done my own PCB designs and cut them out with a desktop CNC using a Dremel engraving tool but that's only small enough for 1/10" pin geometry. I don't know how to get smaller without using photo-etch which is chemically messy, but still much cheaper than thousands of dollars, given CNC availability.
99% of the time 24VDC is available, thus the value of a board with a switching regulator capable of it 60VDC is limited, and not worth the increased BOM costs.
The current proto board is targeted at engineers and hobbyists for prototyping, not as an industrial board for embedded applications.
Parallax will have amortized development costs over thousands of boards - and they have in-house manufacturing capability. Not to mention being the processor manufacturer
Google the thread Parallax did on their development costs, you will find it interesting.
99% of the time 24VDC is available, thus the value of a board with a switching regulator capable of it 60VDC is limited, and not worth the increased BOM costs.
NOT mention that that Voltage need certificate for High Voltage - Class II.
With that point, I agree. They should provide a simple set of Propeller circuitry and software that will work to make automation easier. As it is I'm having trouble learning what it takes to make a PAUSE command work to deliver a specific delay, how to drive a FET to make it go its whole range to 24V and how to get an RCTime command to work that works easily on a BS2. I'm trying the strategy of working to improve the Gold Standard http://www.parallaxsemiconductor.com/goldstandard
which I hope will lead in this direction.
-Phil
But the Proto Board can't handle that 24VDC, so you must by a seperate switching regulator if you want to use it
[*]Change the power connector to a 2-position wire screw clamp
[*]Swap out the existing regulators for something like the LM2591HV-5.0 and LM2591HV-3.3
The gadget ganster has already resolved the first request, you can either power up via a standard plug or on board wire screw clamps.
When I have 5VDC available I simply short out the 5V regulator with a jumper. For any other voltage I use a transformer to convert it to 24VAC and have a small power supply/switching regulator that converts that to 5VDC. I would love to see a protoboard that has a 2 position screw clamp and no +5V regulator but I don't expect Parallax to provide that. One small piece of wire and 2 solder joints takes care of removing the +5V regulator from the circuit and a solder sucker and 2 pieces of wire replace the power jack.
What's my point... You don't need a custom board from parallax. The boards they already make will work. Let's say there Propeller proto board. There is a place for a 3 pin header rigth by the 3.3v reg, one of the holes is lables Vin. Now get one of there 3x4 protoboards put your switching regulator circuit on it and feed its output into the Vin on the propeller protoboard, mount it to the bottom side and you got yourself one sweat HV board and for far less then it would cost you or parallax to make a new design. You would probably even have some room left over for some driver IC's.
That's my two cents.
Bruce
You did not upset me, nor did anyone else. I just have my point of view, and everyone else has theirs. However, I do hope that the non-supporters do it their way for the rest of their lives, like I previously stated.
Bruce
I think that's where you don't seem to want to hear what others are saying: most of us would not in fact add HV components to a Proto Board, we'd order a custom PCB though one of the various vendors providing that service. I find it fascinating that you'll cast molten aluminum to build a custom case, but won't make a custom PCB for one of your projects.
Custom PCBs for a finished product I can handle, but custom PCBs for prototyping is not for me. In my eyes, the Propeller Proto Board exists so that people may experiment and hopefully it will result in the sale of additional Propeller chips for Parallax in finished products. The same should hold true for a HV Proto Board.
Bruce
One afterthought - there are accepted pcb standards that dictate the minimum spacing between conductors at various voltage levels. There are rules for >30v, >50v etc, and it depends on the kind of coating present too. The proto boards have that nice ground plane mask that at 60V may or may not violate these rules. I think if they are deemed to be 'coated' it may be fine, but I'm not an expert in this area, It needs looking at
I don't know if you need a local supplier in Melbourne, but if you can order from Mouser, they sell Hammond and Budd boxes. You may want to look at these links.
http://www.hammondmfg.com/scpg.htm
http://www.budind.com/die_cast_boxes.php
You may also want to go to the homepages of these websites and navigate through the sites, because they have plenty to offer.
Bruce
-Phil
Bruce