ROLAND VP-540 Will not power up

[netsol]

update: I found two sorted components, will order and install. see how it goes

i am curious, what was shorted?

for the layer to layer connections, the best thing is to have the board put through one of those ovens the repair depots use for the nvidia chipsets that go intermittent
i no longer have access to one

these were the best way to do the repair without risk of damage to the board

THE PROBLEM IS, the roland service manual is sorely lacking, it givies you enough info to troubleshoot to the point where you would replace the board, not do a component level repair

Totally agree, I also have a degree in EE and circuit design and 9 times out of 10 it is slogging through the circuit and testing and diagnosing

Probably the best, most versatile piece of test equipment of the last 50 years was the Huntron Tracker (1000 or 2000)

it gave you the ability to go along a power supply buss and it could sense proximity to a dead short. (along with Many other capabilities. Kind of a curve tracer on steroids)
you could move along the 24 volt source and tell if you were getting closer or moving away from a short.
you could compare 2 transformers & find the one with shorted windings.

they still turn up on ebay from time totime.
could be the most useful piece of equipment you ever buy, if you have the patience to become familiar with it

 

[Adam Edge]

i am curious, what was shorted?



Probably the best, most versatile piece of test equipment of the last 50 years was the Huntron Tracker (1000 or 2000)

it gave you the ability to go along a power supply buss and it could sense proximity to a dead short. (along with Many other capabilities. Kind of a curve tracer on steroids)
you could move along the 24 volt source and tell if you were getting closer or moving away from a short.
you could compare 2 transformers & find the one with shorted windings.

they still turn up on ebay from time totime.
could be the most useful piece of equipment you ever buy, if you have the patience to become familiar with it

what if main 42v rail is ok down to 5v and 3v but no signal from cpu to make secondary 42v active which powers up the machine after main switch is pressed?

 

[Adam Edge]

update: Replaced Regulator L4960 as well as AIC1086. Both were shorted. Now I get LED on servo board also. Close but no cigar! I'm thinking main chip/cpu

 

[cornholio]

for the layer to layer connections, the best thing is to have the board put through one of those ovens the repair depots use for the nvidia chipsets that go intermittent
i no longer have access to one

these were the best way to do the repair without risk of damage to the board

THE PROBLEM IS, the roland service manual is sorely lacking, it givies you enough info to troubleshoot to the point where you would replace the board, not do a component level repair

Roland has/had the best and most detailed service manuals of all suppliers in this market. There is no way to troubleshoot boards like these in depht without specialized test equipment. (Think of FPGAs or microcontrollers with programmable pins... is it input or output?)
I don't do component level repairs on these boards, except the obvious ones.
Certainly not on a customers site.
If you need to charge hourly rates... it may be cheaper to replace a board.(As long, as it's available. This particular printer was introduced in 2007... and replaced by the VPi less than two years later)

 

[damonCA21]

update: Replaced Regulator L4960 as well as AIC1086. Both were shorted. Now I get LED on servo board also. Close but no cigar! I'm thinking main chip/cpu

Did you test them for shorts off the board as well? Also when you remove them you need to test the pads to see if the shorts disappear. If they do then the component was at fault, if they don't then the component is most likely fine and the short is coming from something else connected to the component.

 

[Adam Edge]

Did you test them for shorts off the board as well? Also when you remove them you need to test the pads to see if the shorts disappear. If they do then the component was at fault, if they don't then the component is most likely fine and the short is coming from something else connected to the component.

of course, both shorted off board and shorts gone after replacing. I wouldn't come to that conclusion otherwise

 

[damonCA21]

More investigation then! This is what I meant when i said it's never as simple as just replacing a couple of components. You need to know why those components shorted, and what else is connect to them in the signal path now

 

[netsol]

Roland has/had the best and most detailed service manuals of all suppliers in this market. There is no way to troubleshoot boards like these in depht without specialized test equipment. (Think of FPGAs or microcontrollers with programmable pins... is it input or output?)
I don't do component level repairs on these boards, except the obvious ones.
Certainly not on a customers site.
If you need to charge hourly rates... it may be cheaper to replace a board.(As long, as it's available. This particular printer was introduced in 2007... and replaced by the VPi less than two years later)

your right. while i have fixed a few system boards I WOULD HATE to pay anyone the hourly rate for the time i wasted on this

 

[netsol]

you could CAREFULLY apply switched power with a constant current/current regulating power supply.
i am completely unfamiliar with your model. what are the switched power voltages? 12VDC & 5VDC?

i hesitate to give this type of advice, because you can do more harm than good, but, you could apply power upstream from the CPU & see if you get any pulses/signs of life when you do so

 

[Adam Edge]

you could CAREFULLY apply switched power with a constant current/current regulating power supply.
i am completely unfamiliar with your model. what are the switched power voltages? 12VDC & 5VDC?

i hesitate to give this type of advice, because you can do more harm than good, but, you could apply power upstream from the CPU & see if you get any pulses/signs of life when you do so

I'm actually thinking to swap out the main chip with known working one from my other board with ethernet drop out issue. Image if it gets fixed and after all that work the new board has Eth problems too lol that would piss me off but I'm optimistic as the new printer is in good condition and not used much at all. Once the board went it was left unused for many years.

Does anyone know if the carriage lock sensor is controlled by a particular chip. Both main board and sensor were dead when i got the machine so I'm thinking could be related

Switching voltage is variable from 5V to 40V in step down configuration with a soft start

 

[Adam Edge]

Lets play bingo. Everyone pick a component and I'll swap it out with working board. What prize should the winner get? All entries pays $10 and the winner gets it all?

 

[cornholio]

I'm not sure, if it's a good idea to ramp up the power supply voltage. There are several voltages generated on this board. When I remember correctly, there is one big switching regulator that creates 5V from 41V and 5 or so linear regulators creating voltages lower than 5V. It's even possible, that there are negative voltages around, but these are generated on the servo board as far as I remember.
So when the switching regulator starts up with a much lower input voltage, than designed for, it needs more current and possibly kills it again, if no protective circuit is implemented.

 

[cornholio]

Head lock is cm5 and goes through IC6 (74LVC541) and finally trough latches into IC30A.

 

[Adam Edge]

Head lock is cm5 and goes through IC6 (74LVC541) and finally trough latches into IC30A.

thanks for looking it up

 

[netsol]

I'm not sure, if it's a good idea to ramp up the power supply voltage. There are several voltages generated on this board. When I remember correctly, there is one big switching regulator that creates 5V from 41V and 5 or so linear regulators creating voltages lower than 5V. It's even possible, that there are negative voltages around, but these are generated on the servo board as far as I remember.
So when the switching regulator starts up with a much lower input voltage, than designed for, it needs more current and possibly kills it again, if no protective circuit is implemented.

i am not sure either. it certainly WAS a winning strategy with calcomps, years sgo & with rolands as new as the sc545-ex. ( i still print on one quite a bit) life was siimpler back then & LSI was more resilient.

in those days finding a service manual was a rarity. remember only a few years ago, a tech couldn't scan a service manual to pdf & post it online. powering a printer's circuits with a current limiting power supply was the only option, much of the time.

 

[damonCA21]

Swapping the main chip is a good way to end up with two dead boards I would only really consider it as a last resort. Removing it can be useful to see if the shorts then vanish, but that will only confirm the chip is bad not that the board will work properly with the other chip installed.
Swapping main chips is a bit of a crap shoot unless you know 100% the other one was fully working, and as we have mentioned, 99% of the problems with these boards are physical connections rather than bad components.
At least you will get plenty of practice soldering though!

 

[Adam Edge]

"
Given that the mainboard's power-up relies on a signal from the main chip to enable the second 42V input, the issue could be in the power-up control circuitry around the main chip or in the signal path that manages this activation. Here are some targeted steps to investigate further:

1. Power-On Signal Path​

• Check for the Power-On Signal: Trace the pathway from the main chip that should send the enable signal for the second 42V input. This line may involve a MOSFET, transistor, or relay that acts as a switch.
• Inspect Switching Components: Test any transistors, MOSFETs, or relays that the power-on signal would activate. A fault here could prevent the mainboard from enabling the second 42V input.

2. Main Chip Supply and Functionality Check​

• Confirm Main Chip Power Supply: Ensure the main chip has all necessary voltages on its power pins, particularly if it requires both primary and secondary voltages (like 3.3V and 5V).
• Check Reset and Clock Signals on Main Chip: The main chip’s reset line should be in the correct state (typically high, but refer to the datasheet if available), and the clock signal should be active. An inactive clock or stuck reset line would prevent the chip from sending the enable signal.

3. Enable Signal Source and Driver ICs​

• Inspect Driver Circuitry: Sometimes, an enable signal is managed by a driver IC, such as a buffer or a simple transistor circuit. Measure continuity and resistance across any relevant driver ICs and their connections.
• Compare Signal Voltages: Compare voltage levels on the enable signal line between the faulty and working boards. The faulty board may have a missing or weak signal due to a faulty driver or passive component like a resistor or capacitor on the line.

4. Firmware/Boot Issue​

• Firmware Corruption: In some cases, if the firmware or internal software fails to initialize correctly, the main chip may not execute the power-up signal sequence. Try re-flashing the firmware if you have access to the equipment or contact Roland support for a firmware file.
• Diagnostic LEDs or Indicators: If there are any LEDs on the board that typically indicate boot-up stages, observe whether they’re responding as expected. This may help confirm if the main chip is even attempting to boot.

5. Thermal Analysis​

• Check Main Chip Temperature: A thermal camera or careful touch test on the main chip may reveal if it’s overheating or remains completely cold. Either of these could indicate that the chip isn’t functioning as expected."

Regards
HAL-9000

 

[damonCA21]

Ai again basically it is just telling you to check everything

 

[Adam Edge]

Ai again basically it is just telling you to check everything

not really dude, it's being more specific than you have been lol. It actually knows what it's talking about and pointing me in the right direction after only a couple of inputs

 

[damonCA21]

And yet you still haven't fixed it....