It serves as a continuous source of amazement to me that so few engineers have an awareness of what they cost...
It serves as a continuous source of amazement to me that so few engineers have an awareness of what they cost...
Let's have a go.
Engineer earning $75k/yr - $6250/mo - $300/day - $37.50/hr - $.625/minute
Say a fiddly machine takes about 5 minutes every hour - easy to overlook. $3.125/hour - $25/day - $125/week - $500/mo - $6000/year - $30,000 over a 5 year course of using the machine. So that is the direct cost. The missed opportunity and distractions preventing the engineer from designing something new and useful are far harder to estimate than just the lost time.
After owning/operating a machine shop, I because accustomed to counting seconds and analyzing the small details of repetitive tasks. The best equipment and software wastes the least amount of time possible and that seems expensive until you add up the value of all the seconds, minutes, hours, days, weeks, of expensive labor saved.
rx8pilot: I agree with your "long tailed" approach. I've already upgraded from "build my own pick and place" to "buy one". And indeed if the next nearest competitor was 2x the price and had nicer software I probably would have found the money somewhere to get it. But the "real" machines cost at least an order of magnitude more money than I can afford... at which point it's no longer practical to have in the shop.
Every time I use the hardware on the Neoden 4 I'm impressed. But the software feels so close to being okay, but not quite. If a few other users wanted to go together for better software, I would gladly get a developer, write a spec and make it. After writing most of a complete PNP control software recently, I can tell you that it's not an overwhelming undertaking but does require some concentrated effort to get right. Let me know if anyone has serious interest in this kind of thing... perhaps Neoden will release something that puts the icing on the cake but honestly I'm not holding my breath.
As for the real cost of cheap solutions... sadly most of us don't have the option of buying the Rolls Royce right out of the gate, so the fiddling with a cheap machine is really an "investment" in hopefully earning enough to get something better.... or at least that's how I see it.
rx8pilot: I agree with your "long tailed" approach. I've already upgraded from "build my own pick and place" to "buy one". And indeed if the next nearest competitor was 2x the price and had nicer software I probably would have found the money somewhere to get it. But the "real" machines cost at least an order of magnitude more money than I can afford... at which point it's no longer practical to have in the shop.
Every time I use the hardware on the Neoden 4 I'm impressed. But the software feels so close to being okay, but not quite. If a few other users wanted to go together for better software, I would gladly get a developer, write a spec and make it. After writing most of a complete PNP control software recently, I can tell you that it's not an overwhelming undertaking but does require some concentrated effort to get right. Let me know if anyone has serious interest in this kind of thing... perhaps Neoden will release something that puts the icing on the cake but honestly I'm not holding my breath.
As for the real cost of cheap solutions... sadly most of us don't have the option of buying the Rolls Royce right out of the gate, so the fiddling with a cheap machine is really an "investment" in hopefully earning enough to get something better.... or at least that's how I see it.
I did not have the money to buy a fancy machine either. I was stuck with a very challenging decision - buy new/lowend or used/higher end. The decision was to go for the used machine which I expected would be a bear to get running and I was not disappointed. I spent around $5k on a broken machine, $4k on parts, and about a month fixing and learning how to use it. I traded nights and weekends to make up for my lack of money. So, clearly the upfront cost was higher (cash + time) but it is quite solid and far more capable than an N4 no matter how you look at it. It was a compromise that had a good result. There were some very scary moments where it could have ended poorly, so there was definitely some real risk that the project would fail. Now that it is up and running, the effort need to keep it there is tiny. It is fast and reliable. It can hold tons of parts up to 56mm tapes. I have over 100 parts setup in the machine at any given moment so it saves a LOT of time changing from program to program - almost no part changes needed. Tall parts, wide parts, small parts - 1005 passives if needed, cut tapes, tubes, 6 nozzle changer, side scan on the fly, etc. All time savers.
As for new machines, your are right, there is nothing out there in this range. Most machines (including mine) you would blast $10k on a small number of feeders. It is a tough transition for a business.
rx8pilot: I did look for a used machine for about six months and didn't find anything in my area. Being in Canada, it's rather complex and expensive to get anything from the USA... even a lot of people on eBay won't ship large items to Canada. Neodentech took care of everything and all I had to do was pay the local sales taxes when the stuff arrived.
I'd love to upgrade to a bigger machine later, but it would be a long-term project when I'm not relying on it for actual production. My DIY project based on the Liteplacer + my own software and surplus feeders kept dragging on so I finally just order the Neoden instead. I think I have it working well enough to start making real boards.
The major problem I've been trying to solve today is this T-962C craptastic oven that was part of the kit. It's the worst thing ever... I don't want to change this thread but if anyone has recommendations of a non-crappy oven that is plug and play, I'm ready to throw this one in the garbage. Like, does IR even work for PCBs? I got the oven to not burn itself by taking out all the flammable tape, but now it just cooks boards so that they get all bubbly with black epoxy goo oozing out. This is on the lead-free profile too! I put some instrumentation inside and it looked like some parts of the PCB was reaching 325C.
Finding a used machine requires a lot of luck and a lot of time. I was seconds away from buying one of the new Asian machines when my opportunity arrived on eBay - and it was local. I have not seen a similar deal since.
Glad to have a break once in a while.
rx8pilot:
The major problem I've been trying to solve today is this T-962C craptastic oven that was part of the kit. It's the worst thing ever... I don't want to change this thread but if anyone has recommendations of a non-crappy oven that is plug and play, I'm ready to throw this one in the garbage. Like, does IR even work for PCBs? I got the oven to not burn itself by taking out all the flammable tape, but now it just cooks boards so that they get all bubbly with black epoxy goo oozing out. This is on the lead-free profile too! I put some instrumentation inside and it looked like some parts of the PCB was reaching 325C.
Also without wanting to redirect this thread:
Reflow Oven
I'm considering a 'R350' from
www.torchsmt.com at around $5,750 USD, for what it's worth.
The major problem I've been trying to solve today is this T-962C craptastic oven that was part of the kit. It's the worst thing ever.
The Qinsi QS-5100 looks a bit like the T962C and its price is similar, but its built a bit better and has a controller which actually works. A number of people who have had bad experiences with the T962C seem to find these satisfactory. See
for a comparison.
rx8pilot:
The major problem I've been trying to solve today is this T-962C craptastic oven that was part of the kit. It's the worst thing ever... I don't want to change this thread but if anyone has recommendations of a non-crappy oven that is plug and play, I'm ready to throw this one in the garbage. Like, does IR even work for PCBs? I got the oven to not burn itself by taking out all the flammable tape, but now it just cooks boards so that they get all bubbly with black epoxy goo oozing out. This is on the lead-free profile too! I put some instrumentation inside and it looked like some parts of the PCB was reaching 325C.
Also without wanting to redirect this thread:
Reflow Oven
I'm considering a 'R350' from www.torchsmt.com at around $5,750 USD, for what it's worth.
Good day Thommo,
I cannot speak about the Torch R350, but have you looked at the Puhui T-960 (
www.tech168.cn) which is somewhat comparable? The T-960 has 5 heat zones (3-up 2-down) and works pretty well. Best of all it is a lot cheaper. I bought mine for $1150 US (shipping was an additional $200 to my closest Ocean port) just over a year ago. Overall the build quality is quite good.
Cheers,
Sam
I'd seriously consider a vapour phase setup.
rx8pilot:
The major problem I've been trying to solve today is this T-962C craptastic oven that was part of the kit. It's the worst thing ever... I don't want to change this thread but if anyone has recommendations of a non-crappy oven that is plug and play, I'm ready to throw this one in the garbage. Like, does IR even work for PCBs? I got the oven to not burn itself by taking out all the flammable tape, but now it just cooks boards so that they get all bubbly with black epoxy goo oozing out. This is on the lead-free profile too! I put some instrumentation inside and it looked like some parts of the PCB was reaching 325C.
Also without wanting to redirect this thread:
Reflow Oven
I'm considering a 'R350' from www.torchsmt.com at around $5,750 USD, for what it's worth.
Good day Thommo,
I cannot speak about the Torch R350, but have you looked at the Puhui T-960 (www.tech168.cn) which is somewhat comparable? The T-960 has 5 heat zones (3-up 2-down) and works pretty well. Best of all it is a lot cheaper. I bought mine for $1150 US (shipping was an additional $200 to my closest Ocean port) just over a year ago. Overall the build quality is quite good.
Cheers,
Sam
Hi, this is my first post.
I have
DDM Novastar LS60V and currently need more capacity, after following this thread for quite some time i finally decide to order the Neoden 4 - now waiting for the shipment.
I want to comment on the 5 mm component height limitation, in my experience for low volume production up to 1000 board job, better to place tall/big/heavy components by hand!
YES, because the machine (lower end machine) will throw that kind of parts very often - even messing already placed parts.
I will share my experience with this N4 when i get it.
-ichan
@Ichan Can't wait to see some feedback from you once the machine arrives!
Thanks for assisting us all and agreeing to share your experience Ichan.
When are you expecting to receive your new Neoden4?
I have DDM Novastar LS60V and currently need more capacity, after following this thread for quite some time i finally decide to order the Neoden 4 - now waiting for the shipment.
I will share my experience with this N4 when i get it.
I want to comment on the 5 mm component height limitation, in my experience for low volume production up to 1000 board job, better to place tall/big/heavy components by hand!
That's usually a case of using the right nozzle, which is why you really want an auto nozzle change facility
I tell the Neoden contact (Tony) to take his time preparing the machine to be the best in quality and professionally configured, because i will review it in here
.
The job is already waiting for the machine, i will review it doing a real job for at least 2000 panelized board (4 board / panel).
Let say they need 1 week to prepare the machine plus 2 weeks shipment + customs then i may receive it at about the end of 3rd week of march, so i hope i can start posts about it around the end of march.
The vacuum pump also a weak point of low end machine, even the LS60V - based on that experience i order the N4 with spare vacuum pump too.
-ichan
I want to comment on the 5 mm component height limitation, in my experience for low volume production up to 1000 board job, better to place tall/big/heavy components by hand!
That would kill me for sure. Just one of my boards has 10pcs 820uf aluminum caps - running a few hundred having to hand place those would be brutal.
Sometimes, hand placing is the only sensible solution though.. I've got an inductor that heavy and large. ( 22x22x14mm ).. The consequences of a drop are ruining an entire panel..
If you need to do manual placing of th smd elcos I would immediately switch to through hole for those components, i still think that the reflow process will affect the lifetime of electrolytics esp the smaller cases ofcourse but eveyone in the industry will probably tell me i am nuts.
That would kill me for sure. Just one of my boards has 10pcs 820uf aluminum caps - running a few hundred having to hand place those would be brutal.
I started as a fully hand placer, ten caps is easy for me...
If you need to do manual placing of th smd elcos I would immediately switch to through hole for those components, i still think that the reflow process will affect the lifetime of electrolytics esp the smaller cases ofcourse but eveyone in the industry will probably tell me i am nuts.
Me too prefer TH for elcos for two reason, smd elcos are easy to be pulled off from the board and they take much longer time to reflow than the small parts.
-ichan
I see another limitation of N4 when ordering it, which i think never talked yet.
They only over 8mm, 12mm, and 16mm reel feeders...
D2PAK in reel for example won't go with this machine.
-ichan
I started as a fully hand placer, ten caps is easy for me...
I too spent the last 2 years hand-placing many 10's of thousands of parts by hand - fine pitch and 0402's
. A quantity that drove me to near insanity. I did, however, get very very good at it. I made special tools, organization, software, etc. I really feel like I totally maxed out hand placing. When I got my machine, it had some troubles with tall parts not fitting in the feeders, so I placed them by hand. The 'few minutes' delay getting them into the oven multiplied times each board was bringing back bad memories and adding up to a significant amount of time. I modified the feeders to fit the tall caps and now it goes direct from P&P to oven. Heavy inductors too - just got the right nozzles and slowed the machine down.
I am thankful for my sharp assembly skills, but I only want to use them in an emergency. Taking the PCB from P&P to the oven is such a joy. Having enough feeders to have 6 PCB's setup in the machine at a time is also amazing. The stress level has plummeted.
I started as a fully hand placer, ten caps is easy for me...
I too spent the last 2 years hand-placing many 10's of thousands of parts by hand - fine pitch and 0402's . A quantity that drove me to near insanity. I did, however, get very very good at it. I made special tools, organization, software, etc. I really feel like I totally maxed out hand placing. When I got my machine, it had some troubles with tall parts not fitting in the feeders, so I placed them by hand. The 'few minutes' delay getting them into the oven multiplied times each board was bringing back bad memories and adding up to a significant amount of time. I modified the feeders to fit the tall caps and now it goes direct from P&P to oven. Heavy inductors too - just got the right nozzles and slowed the machine down.
I am thankful for my sharp assembly skills, but I only want to use them in an emergency. Taking the PCB from P&P to the oven is such a joy. Having enough feeders to have 6 PCB's setup in the machine at a time is also amazing. The stress level has plummeted.
Hi
With a bit over a half million dollars worth of gear humming away placing and the like, we still have parts that get put on by hand. Sometimes it's for one reason, sometimes for another. We try to avoid it (by design) on the big runs. It's almost inevitable on the under a hundred pieces stuff ....
Bob
@ichan: if you have a DDM Novastar LS60V why on earth would you want a Neoden4? I think you're going to be grossly disappointed. I'm actually looking at the DDM LE-40V because my confidence level with the Neoden4 based on discussions in this group isn't there. Can you elaborate on why you need a Neoden4 when you have the LS60V?
@ichan: if you have a DDM Novastar LS60V why on earth would you want a Neoden4? I think you're going to be grossly disappointed. I'm actually looking at the DDM LE-40V because my confidence level with the Neoden4 based on discussions in this group isn't there. Can you elaborate on why you need a Neoden4 when you have the LS60V?
Hi
.... yes butting in .... sorry about that.
Short runs.
Keep the big stuff running the high volume boards. Move the dozens level stuff over to a small machine. Setup / teardown / validation on a cheap machine wastes less expensive machine resources. If the people time is roughly the same and you can keep the big stuff running full bore, you are ahead. If you keep the little guy tied up half of the time while the big guy(s) keep(s) running much closer to full time, you are making a lot of money off that little machine. It also beats hand placing those many batches of a couple dozen boards / panels ...
Yes, that is *very* mix / situation specific. My mix / your mix / his mix / how many parts / what kind of parts / programming time / scrap / quality ... who knows.
To move to the end of the page. We have tried this. It's one of those things that makes sense when you are maxed out or heading that way. It dies in the next slow down. The idea never seems to come back when things pick up again.
Bob
@sedelman, In short i am a dissapointed user of LS60V.
I can tell anything about that machine if someone ask, because i have repaired it by myself four times already.
Anyway, i bought it when the company name is still APS Novastar not DDM.
@uncle_bob, well that is not on my case
, i will gladly swap my LS60V with two fully configured N4 if anyone interested...
-ichan