PDA

View Full Version : Board-level repairs for machines...



Liger Zero
04-16-2010, 11:51 AM
Friend of mine across town has a nice old Windsor injection molding press. He uses it to mold a product he developed, some sort of contraption for hunters.

Trying to find parts for this machine is proving to be a headache. I have found an identical machine in a crate in a university lab across the country for $100,000 but they want to sell the whole machine not just part it out. It's been sitting there since 1988 unopened uninstalled and unoperated. Bastards. :D

Anyway. The fault lies in one or more of the circuit boards. More or less the Windsor-specific control logic talks to these boards which are hydraulic control boards, heat controller boards and the like. Aside from the fancy Windsor brain it seems to me these boards could possibly be repaired...


...but by who? Are there any little shops or companies out there that take one or two control boards, troubleshoot to the component level and attempt to repair? If so who do ya'll recommend?

Weston Bye
04-16-2010, 12:08 PM
I used to do that for a living, or attempt to, with varying degrees of success. Even when I was successful, the effort was costly for the customer. Schematic diagrams help, indeed are nearly essential, but are not a guarantee of success when the board gets beyond a certain degree of complexity, particularly where programmable devices are involved.

Best I can recommend is find sombody who has worked on the particular brand before. They should have learned what to look for and know the shortcuts.

Too_Many_Tools
04-16-2010, 12:45 PM
Friend of mine across town has a nice old Windsor injection molding press. He uses it to mold a product he developed, some sort of contraption for hunters.

Trying to find parts for this machine is proving to be a headache. I have found an identical machine in a crate in a university lab across the country for $100,000 but they want to sell the whole machine not just part it out. It's been sitting there since 1988 unopened uninstalled and unoperated. Bastards. :D

Anyway. The fault lies in one or more of the circuit boards. More or less the Windsor-specific control logic talks to these boards which are hydraulic control boards, heat controller boards and the like. Aside from the fancy Windsor brain it seems to me these boards could possibly be repaired...


...but by who? Are there any little shops or companies out there that take one or two control boards, troubleshoot to the component level and attempt to repair? If so who do ya'll recommend?

The general answer is "No".

Most machine electronics are company-specific with no schematics available.

The best you can do is find board replacements...with the accompanying sticker shock.

FWIW...electronics obselence is the reason why most modern machines are sold cheap..like your friend's.

There is no free lunch.

TMT

fasto
04-16-2010, 01:12 PM
It can be done.
I have a 21-year old Hurco VMC. It's equipped with Fanuc servo drives.
I got it cheap because of a problem with the Y-axis servo drive. It would fault out immediately on power up.
I have a degree in electrical engineering, 20-odd years of experience, and a rather completely equipped electronics facility.
I was able to reverse engineer the servo drive control board, it's 11"x18" and four layers, with traces on the internal layers. After developing the schematic, I could devise tests to figure out what was wrong. I found two faulty components (a Fanuc isolation amplifier and a generic op amp) and replaced them. Now, the drive works great.
Why did I do this? I could get a new drive from Fanuc for $5000, or have mine "professionally" repaired for $4000, or buy a used drive for $2600. I ended up spending something like 80 hours troubleshooting the drive, and a total of $60 on components. For my 80 hour investment, I have enough info should the drive dare to act up again I can fix it. Was my work cost effective? Not for an industrial application where downtime is measured in thousands of $$$ per hour; for my application where the company is, well, slow at the moment it was cost effective.
Could you find someone to reverse engineer your system and fix it? Probably, but the 80 hour time is not unreasonable for a complex system, and the work would probably cost anywhere from $60 to $150 per hour.
If the boards in question are relatively simple, and it is possible to determine which ones are faulty, the process might be easier.
If you'd like some more help feel free to send me a PM, and I'll see what I can do.

Liger Zero
04-16-2010, 01:17 PM
FWIW...electronics obselence is the reason why most modern machines are sold cheap..like your friend's.


Indeed. The machine I have uses fairly standard Fanuc boards and packages. That's why I chose it over some other really nice pieces of equipment. Plus Fanuc (while expensive) can repair most of what they sell/have sold. Service after the purchase was a consideration for me.

His Windsor on the other hand, he got it second-hand years ago and aside from the occasional leak, blown fuse or heater-band replacement he's had no trouble with it. It runs three molds, on alternate weeks so he can make his product.

I tried looking up Windsor-Klockner but their site is down, and the phone number in his operation manual is long disconnected... There are dedicated IMM repair services around but I figured I would try sourcing boards or board repairs for him first.

Liger Zero
04-16-2010, 01:20 PM
It can be done.
If you'd like some more help feel free to send me a PM, and I'll see what I can do.

I will get the details tonight and fire off a PM. If not, then Monday.

Liger Zero
04-16-2010, 02:11 PM
Found them online. They are Windsor and they are owned by an Indian company. Got contact information and everything.

Too_Many_Tools
04-16-2010, 03:36 PM
Found them online. They are Windsor and they are owned by an Indian company. Got contact information and everything.

I would be interested in what it ends up costing to fix your machine.

TMT

Liger Zero
04-16-2010, 04:12 PM
I would be interested in what it ends up costing to fix your machine.

TMT

Based on a brief phone call, it may not be a board-level problem at all instead it might be a solenoid issue i was instructed to verify several outputs from the clamp control manifold.

I will do this tomorrow, yes on a Saturday.

If we get signal (main screen on!) at the specified voltages then it is indeed a board level issue. Lady on the phone says they still support this particular controller, as this model of machine sells very well in developing countries.

It consists of valves, sliders, toggle switches and a small alpha-numeric input panel with a three line LED readout. I've run these before they are very very simple to setup and operate.

Cost for direct replacement boards start at $4,500 for the smaller of the two boards, this is known as the CLAMP OUTPUT SEQUENCE DRIVER board and the other board is MANIFOLD INPUT BOARD NO CORE (front) board. That one is $4,200 for a replacement.


If it comes down to both boards being replaced they will cut us a deal, no specifics on that. I was told to check the voltages and call back Monday if they are within the specified ranges... that means its one board or the other.


Now if I could just get the plastic lumber guy out of his funk I could move forward on THAT project as well. He's in denial over the scope of the problems with his toy. Can't say I blame him, that was quite a wad he dropped on a dysfunctional machine.

boslab
04-16-2010, 05:12 PM
I had to sort a Hupfield injection moulder some years ago, give up with the old ttl board complete with relay logic, ripped the lot out binned it and replaced with a mitsibushi plc it was fairly painless, later added a HCI panel to make programming easy, you could even add your own graphics by importing bitmaps to make it look swish, it came out so well i was even pleased myself, which is rare, plc $150, hci 200, job done, the temperature controll aspect with a plc looks terrifying but is really easy, plenty on mr plc
mark

JoeFin
04-16-2010, 05:30 PM
Just finished repairing my Hurco MHP 3K.

Originally equiped with a Dynapath 20 and upgraded to a Dynapath 50 in 1997. The hard part was the machine rebuilder used the Dynapath 20 MTB I/O board and made some alterations to run it with the Dynapath 50 controler. I had to marry together 2 sets of drawings and make notes along the way as to the condition (True/False) on all the permissives and then match those up to the I/O Words displayed on the screen

Paul at Dynapath was a great help. Glen - not so much

hoof
04-16-2010, 07:17 PM
Liger Zero, Have you delt with a company called "Electrical South"? They have been around for some time but they were purchased by a company called the Schnitter group. ( not sure of the spelling). These are the floks that own Square D. Well I have found the ability and service has increased ten fold. Kind of pricey but they quote the repair first and most machine specific boards I've sent out come back working. Not affiliated with them at all but they have impressed me in the past few years,

Liger Zero
04-16-2010, 10:23 PM
I had to sort a Hupfield injection moulder some years ago, give up with the old ttl board complete with relay logic, ripped the lot out binned it and replaced with a mitsibushi plc it was fairly painless, later added a HCI panel to make programming easy, you could even add your own graphics by importing bitmaps to make it look swish, it came out so well i was even pleased myself, which is rare, plc $150, hci 200, job done, the temperature controll aspect with a plc looks terrifying but is really easy, plenty on mr plc
mark


I've heard of this being done, they sell entire retrofit kits just for molding presses.

Never heard of a Hupfield injection molding machine though.

Bruce Griffing
04-16-2010, 10:30 PM
If I were faced with this problem, I would first check the DC power supplies for both voltage and waveform (look for excess ripple). If that is OK, I would then replace all of the electrolytic capacitors on both boards. If the boards still failed, I would probably do relays next. The total cost of the components will be small compared to the replacement cost, and there is a good chance for success.

If you take nice photographs of the boards (both sides) and post them, we can make suggestions as to probable suspect components.

Liger Zero
04-16-2010, 10:39 PM
I was just discussing this with someone! Apparently caps dry out and go bad or something.

Question is board type. I've worked on single layer boards. Multi-layer boards no. Anything I need to know prior to firing up my fine-tip iron?

Bruce Griffing
04-16-2010, 10:47 PM
You have to be careful not to overheat the board. Use a heat setting just above what you need to melt solder around a lead. I use a solder sucker to remove the solder, but you have to be very careful as these broadcast tiny bits of solder that can short traces on the board. If you can reach the leads and have a replacement device, cut the leads so you can easily remove them one at a time. Electrolytic capacitors can sometimes be removed by "rocking". You heat one lead with side pressure on the top of the can. This will allow that lead to be partially pulled out before the other lead holds things up. Then do the same to the other lead. Etc.

Liger Zero
04-16-2010, 11:25 PM
You have to be careful not to overheat the board. Use a heat setting just above what you need to melt solder around a lead. I use a solder sucker to remove the solder, but you have to be very careful as these broadcast tiny bits of solder that can short traces on the board. If you can reach the leads and have a replacement device, cut the leads so you can easily remove them one at a time. Electrolytic capacitors can sometimes be removed by "rocking". You heat one lead with side pressure on the top of the can. This will allow that lead to be partially pulled out before the other lead holds things up. Then do the same to the other lead. Etc.

So nothing special compared to standard boards then? :)

Paul Alciatore
04-17-2010, 12:56 AM
So nothing special compared to standard boards then? :)

No. Actually multilayer boards are usually harder to damage with soldering heat than single or double sided ones. They are probably just made better.

What Bruce said is right on. Only other tip would be to use extra solder while trying to rock a capacitor out. It will allow more heat transfer in a shorter time thereby lessening the chance of board damage. Then suck the solder out of the hole after removing the part's leads.

Bruce Griffing
04-17-2010, 07:40 PM
One other tip. Photograph the board from many angles before removing any parts. Note the polarity of the electrolytic caps in advance as well.

besser
04-18-2010, 02:08 AM
With much success I have rebuilt many mobile phones and find the NO. 1 problem is the effects of the environment.

1: Wash the board (Can of electrical cleaner (I ultra sonic but it's a little over the top
2: Reflow the whole board of solder (heatgun remelts the joints
3: refresh the power supply

I find this fixes so many phones and old computer boards that I am convinced, dust, humidity and vibration are the killers of circuits. I'd then start checking components. A dry joint can stop you in your tracks and you start thinking it's the components when its not.

Good luck

JMcTool
04-18-2010, 03:24 AM
With much success I have rebuilt many mobile phones and find the NO. 1 problem is the effects of the environment.

Very frequently true. One of the most common homebuilt computer debugging solutions was to reseat the connectors, and failing that get out a pencil eraser and clean the little "fingers" at the edge of the board that go to connectors. Solves a huge number of problems.

Jim

ptjw7uk
04-18-2010, 04:24 AM
Biggest problem with modern electronics is the very low power used such that even the slightest rise in contact resistance will cause problems.
In the old days it used to be heat causing IC creap as the device was used this effect has mostly been cured by the practice of soldering most of the components to the board so that if it works from day one then it will normally work for years.

peter

Weston Bye
04-18-2010, 07:32 AM
Failure, for no apparent reason, may be caused by tin whiskers, particularly where vibration over time is involved.

http://en.wikipedia.org/wiki/Whisker_(metallurgy)

Made more common with the imposition of RoHS standards. The lead in solder suppressed whisker growth. Remove the lead and you now have built-in obsolescence.

Besser's step 2 would eliminate the whiskers and buy more time until they began to grow again.


With much success I have rebuilt many mobile phones and find the NO. 1 problem is the effects of the environment.

1: Wash the board (Can of electrical cleaner (I ultra sonic but it's a little over the top
2: Reflow the whole board of solder (heatgun remelts the joints
3: refresh the power supply

I find this fixes so many phones and old computer boards that I am convinced, dust, humidity and vibration are the killers of circuits. I'd then start checking components. A dry joint can stop you in your tracks and you start thinking it's the components when its not.

Good luck

Evan
04-18-2010, 08:02 AM
A major problem with boards that contain power control components is the wave soldering of both small component leads and large connector leads simultaneously. It is extremely common for the heavy connector leads to be poorly soldered. They may look fine on casual inspection but under a 10 power glass a hair fine crack may be seen in the solder encircling the connector pins. I would rate this as the number one problem on the wide range of power supply and power control boards that I used to service that were of the same vintage as that press. When I was trouble shooting such a problem I would first verify it with a magnifier and then resolder every connector pin on the board. I recommend you try that as it is easy and cheap and has a good chance of solving the problem.

Another thing to check is for blown back emf diodes in the solenoids. If they are running on unfiltered DC (usually 24 volts) then there is often an integral diode in the solenoid to prevent latching of the control circuit. If the diode has gone open the solenoid will have exactly the same resistance in both directions. It can be made good with an external diode at the solenoid connector.

JoeFin
04-18-2010, 08:50 AM
A major problem with boards that contain power control components is the wave soldering of both small component leads and large connector leads simultaneously. It is extremely common for the heavy connector leads to be poorly soldered.

Yep that was very common back in day

I would suggest he start off verifying the power supply voltages 1st

2nd make darn sure all the permissives are TRUE or ON

3rd clean and inspect the boards for connectivity just as you have suggested

and then move on to component diagnoses and or replacement

Most important of all as far as component replacement is concerned is ARE the chips still available.

Liger Zero
04-18-2010, 02:15 PM
I checked the solenoids on the clamp manifold, the clamp motion switches and the interlocks on the machine using a simple VOM meter. All is well there. Next I traced the wires, no shorts, everything is connected.

Tomorrow I don't have a lot of time to dedicate to this, I will go over the boards under a magnifier. If nothing jumps out at me as being obviously blown or cracked (and I will check the diodes) I will tell my friend to buy new boards.