Roland VA-3 boot issue

[LateLesley]

 can't for the life f me find C69 on the diagram for the jack board (I'm assuming this is on the jack board?) C72 should have 12V on it. The -9.7V on C70 sounds sus too. It might be a bad ground, or bad measurement. According to the diagram C70 is to do with a circuit for the pedal input.

It might be worth naming which board you are talking about, as a few of them have the same component numbers, confusing things.

So maybe just preface them MAIN, JACK, PSU/AMP, CONTROL, XPGS4 etc. so we know which board you are referring to. But C69 on the jack board is a mystery.

[Audiorepair]

It took me a while to find C69 and C70.

They are on page 17, upper right around CN9.

The Nomenclature is a bit messed up on this PDF.  What is marked as +12v and -12v on page 17 is actually +Vcc and -Vcc.
The raw 12v from the power supply on the jack board is called 12v on page 29, then it becomes VA, then it is turned into Vcc using a transistor circuit Q1 on page 25.
-Vcc comes from the negative switching circuit and 7912. 
The voltages on C69 and C70 are marked 12v and -12v but are actually +Vcc and -Vcc.

What a mess.

[LateLesley]

Ah! so they were on the main board. That's kinda what I suspected, when I couldn't find C69 on the jack board. I did find a C70 on the jack board too though, making me think I was just missing the cap somewhere. Seeing C69 and C70 in the main board, it makes more sense.

So yeah, they should have +12V and -12V on them. One thing I will say @frisc0, is just check the ground connections between the boards are intact, sometimes they can pass grounds via the chassis, and if you have the boards out, it can disconnect these. SO make sure they are intact, and that you have the meter connected to a good ground before measuring voltages.

A floating ground could mess up all sorts of stuff.

And if you aren't getting voltage in the right place. just start from the source, and measure components/connections along the path, till you find where it's getting lost. Sometimes tracks on boards can be cracked, or blown out. You've already seen that damaged cap C88, so maybe check for any other damaged caps connected to the chassis, from other boards. But since voltage is being lost somewhere, that is a definite thing to investigate. It could explain the IC6 heating too, if voltage is being dropped somewhere, might be something shorted on the main board. I'd measure the voltage on the connector with the main board disconnected, then try again with it reconnected, and if it drops dramatically, something on the main board is pulling those power lines down.

Thanks @Audiorepair for pointing me in the right direction of those caps. :-)

[frisc0]

Thanks guys and sorry to leave you searching without naming the PCB... 

Good tips again with tracing the missing 12V..., I did some tracing already, but the schematics are rather err... schematic, so I need some time to trace things down. Good tip on the ground, I did have the boards unscrewed from the chassis indeed, so I'll take that into account!

[frisc0]

I'm an idiot. I measured voltages on C69 and C70 on the mainboard via the top of the caps, don't know what one actually measures then? I know now that it's not the actual voltage over the cap... :|

I checked the flatcable from the jackboard to the mainboard, the -12V is a solid -12V (okay, -11.96V), but the +12V is somewhere around +8.7V. Tracing the +12V towards the PSU I see the +12V from the external adapter dropping a volt or so on the PSU already and on the jackboard again. Without the mainboard attached it reads +9.4V. Would this be normal? Probably not something that causes the boot issue since the 12V lines aren't used in the digital circuit I think?

So it's probably now up to checking the reset lines as audiorepair suggests (and still trying to find why IC6 on JB and IC1/R1/R11 heat up that much...).

[LateLesley]

I would suspect that the +12V should be near to it, 8.7V sounds way too low.

2 options.

1) The external adapter is faulty, and not able to supply the load of the Keyboard.

or

2) Something on the +12V line is drawing way too much power, and pulling it down.

It might just be worth testing the adapter, as these do fail. Load it up and see if it can maintain the load. (Car bulbs can be good for load testing it.) See if the voltage drops down when it's loaded.

If it checks out fine, then its off to trace everything running off the +12V.

[Audiorepair]

Probably not something that causes the boot issue since the 12V lines aren't used in the digital circuit I think?

Well actually they are.

pg17 top left.


They use op-amps run from +12v -12v supplies to drive the bender and mod signals etc to the CPU.
Should one of these 12v supplies fail, the op-amps would go full DC and try to input +12v or -12v to the CPU.

They have diodes in place to prevent anything going above 5v, but still, they are mixing 12v and 5v circuitry.


I would make sure you fix all the obvious 12v problems first, and then tackle the boot issue if that doesn't fix it.

[frisc0]

I would suspect that the +12V should be near to it, 8.7V sounds way too low.

2 options.

1) The external adapter is faulty, and not able to supply the load of the Keyboard.

or

2) Something on the +12V line is drawing way too much power, and pulling it down.

It might just be worth testing the adapter, as these do fail. Load it up and see if it can maintain the load. (Car bulbs can be good for load testing it.) See if the voltage drops down when it's loaded.

If it checks out fine, then its off to trace everything running off the +12V.

Followed your suggestion and checked the 12V adapter with a car bulb. Could only find a 21W bulb and put the trusty multimeter in between, which measured a steady 1.75A without drops. The adapter is a 2A adapter and seems to be able to deliver that.

So tracing the 12V is next then. Any clues as to where to begin, what would be the suspect components? I'll start with disconnecting everything from the PSU and start measuring there, then moving 'up' to the jackboard and mainboard next. And I guess it's only the (unregulated) +12V that's suspect and not the (regulated) -12V?

[LateLesley]

OK, IC47 on the main board and surrounding caps (page 17, C-5 grid reference) IC26/27 + caps (Page 18 N-11/12)

IC10 +caps on the D Beam board (Page 26, D-28) and D215 (D-22)

IC2 on the LCD Control board (Page 24, G-28)

I think thats it. Maybe check Zener D1 on the jack PCB (page 25, E-26) and Q1 (M-15)

[frisc0]

OK, IC47 on the main board and surrounding caps (page 17, C-5 grid reference) IC26/27 + caps (Page 18 N-11/12)

IC10 +caps on the D Beam board (Page 26, D-28) and D215 (D-22)

IC2 on the LCD Control board (Page 24, G-28)

I think thats it. Maybe check Zener D1 on the jack PCB (page 25, E-26) and Q1 (M-15)

Thanks LateLesley for these tips!

Checked the LCD control board, the negative voltages (incl the -21V on TP1) are good, the +12V is too low, around +8.7V. The LCD control board is powered via the mainboard, which in turn is powered via the jackboard.

I checked on the jackboard, zener diode D1, there is 4.08V over it, where specs say it the zener voltage should be 5.1V if I understand correctly. Unsoldered it and measured with the ohm meter. It measures 25Mohm in the 'right' direction, O.L. the other way around. Would this diode be faulty?

How is it used in this circuit, as a voltage regulator?

The adapter delivers 12.3V without load. On the PSU the 12V line is 11.5V, when only the jackboard is also connected the voltage drops to 10.8V. This 0.7 drop is the magic 0.7 silicon number... coincidence?

[Audiorepair]

Checked the LCD control board, the negative voltages (incl the -21V on TP1) are good, the +12V is too low, around +8.7V. The LCD control board is powered via the mainboard, which in turn is powered via the jackboard.

There is little point pondering why a zener circuit may not be producing what you think it should, after you have explicitly stated that the +12v supply is too low at around 8v.

THAT is your problem.


Get the 12v supply working first, and forget everything else for now.

[LateLesley]

Yeah something's drawing too much. IF you ran a car bulb at 1.75A without a power drop, and then it's dropping lots on the keyboard, something is pulling WAY too much power. 

I went back to the circuit diagrams, And started looking at the PSU setup.

So here's what I'd like you to check if possible.

First, on the Power & Amp board, with the power off, Check the resistance across R1 and R2. For R1 you should see around 1.1 ohms. It's a 2.2 ohm resistor but in parallel with R11 which is the same. R2 I think should be 10 ohms??

If that checks out, power it up with the jack board connected, And measure the voltages in these few places.

Across D2
Across the power switch. (Just eliminates a faulty switch)
Across R1
Across R2
Across C14
Across C3

On the jack board

Across C117 (near IC19 i think, going by the schematic)
Across C77 (Should be near IC11)
Across C72
Across C56

The reason for this is to find out where the voltage is dropping a lot, and figure out which circiut is drawing too much power, to narrow down where the fault is.

Nice to knows are if IC1 on the Power board (should be a 3 pin regulator on a heatsink TO-220), and IC19 on the jack board (another 3 pin regulator) are getting warm quick at all.

That hot IC6 could be in question too. But get those measurements back, and well see if we can start narrowing it down.

Audiorepair is right, we do need the 12V stable, but that might be hard if something is sucking enough current to draw it down in voltage. These measurements, especially across R1 and R2, might help narrow it down, while checking everything is healthy on the power and amp board.

[frisc0]

First of all I must say a big thank you for spending so much time helping me out here, I have all the confidence these efforts will pay off

Here are my measurements:

Adapter without load: 12.3V
Adapter with 21W car bulb: 1.75A and 11.7V
R1, unsoldered: 2.3 ohms
R11, unsoldered: 2.3 ohms
R2, in circuit: 10 ohms

With jackboard NOT connected:
Across D2: 0.76V
Across the power switch: 11.27V
Across R1: 5.5mV
Across R11: 5.5mV
Across R2: 0mV
Across C12: 11.46V
Across C13: 11.46V
Across C14: 11.46V
Across C3: 4.95V

With jackboard connected:
Across D2: 0.76V
Across the power switch: 11.27V
Across R1: 244mV
Across R11: 244mV
Across R2: 0.48V
Across C12: 10.82V
Across C13: 11.33V
Across C14: 11.05V
Across C3: 4.95V

On the jackboard:
Across C117: 9.71V (near IC19 i think, going by the schematic)
Across C77: 5.01V (Should be near IC11)
Across C72: 10.82V
Across C56: 9.72V

Does this point into some direction?

Again many thanks in advance for your help and analysis!!!

[Audiorepair]

Short the power switch, if it has volts across it it is not a working switch.

[LateLesley]

OK, not really. As AudioRepair mentioned, when I said measure across the switch, I meant between the two connections, not between switch n ground. But I managed to figure out it's probably OK.

I went through the circuit, and it doesn't seem too horrendous what you are measuring. It seems it has been designed like that.

As far as I can work out, there's 200mA on IC1 in the Power.Amp Board, and about 48mA on the jack board, and the voltages seem consistent with that. It seems VCC on the jack board will drop down to  9V (I simulated part of the circuit around Q1, cos the 1V drop concerned me, but it seems right)

So I suppose the next step would be to connect main, Then give me the voltages over R1, R2 , C12, C14, and C72 C56 on the jack board. Lets see if that increases the power consumption a lot. ( Thats why I get you to measure across R1 and R2, I can work out the currents on those power lines from there. )

Also, I'm not ruling out your 12V adapter, the way it's dropping voltage, it could be weak, though the bulb test seems to discount that a little.  I wonder if you have another 12V adapter with enough current to try??

[frisc0]

Short the power switch, if it has volts across it it is not a working switch.

What I did was unplug the power switch from the PSU board and measure the voltage between the pins that the switch would short when connected. I thought that was wat LateLesley meant and I think she confirmed that .

This evening I'll connect the main board and report my measurements. I must say that with only the jack board connected to the PSU, IC6 on the JB already gets hot (although not as hot as with the MB connected as well).

Meanwhile I asked around for another 12V 2A+ adapter, unfortunately I don't have any of these (I do have laptop adapters but they are 15-20V which I think will be too much for this keyboard?).

To be continued...

[LateLesley]

Actually, AudioRepair was right, I was wanting you to confirm, with the power on, IE the switch closed, that there was no voltage across it. It's a quick check for a bad switch. You weren't meant to disconnect the switch, I wanted to know if there was any voltage across it while it was in place and on. Sorry!! 

As for other adapters, yeah, laptop ones will be too high a voltage.

I look forward to the next test results. That hot IC6 is annoying me, but since you're getting the right output voltage, and the current consumption isn't too nuts (I think - though maybe thats why there was 200mA running through IC1 on the power board, maybe there is something wrong on the -12V somewhere - maybe check the op-amp ICs for any hot ones??).

Maybe just do the finger test on IC 2,7,9,10,14,18 on the jack board, before adding the main board, to see if any of them are getting hot. In fact, baybe just check IC11 too, and any transistors you see, like Q1,Q10,Q11. I'm beginning to think that the hot IC6 might be telling us something else is wrong.

[frisc0]

Actually, AudioRepair was right, I was wanting you to confirm, with the power on, IE the switch closed, that there was no voltage across it. It's a quick check for a bad switch. You weren't meant to disconnect the switch, I wanted to know if there was any voltage across it while it was in place and on. Sorry!! 

Ah... makes sense

There is 2.4mV across the switch, should be 0V I guess...does that indicate a fault?

[frisc0]

So I suppose the next step would be to connect main, Then give me the voltages over R1, R2 , C12, C14, and C72 C56 on the jack board. Lets see if that increases the power consumption a lot. ( Thats why I get you to measure across R1 and R2, I can work out the currents on those power lines from there. )

Ok, with the main board connected, here are the voltages:

PSU
Across powerswitch: 4.8mV (hmm...?)
R1: 0.68V
R11: 0.68V
R2: 0.62V
C12: 10.42V
C13: 11.12V
C14: 10.35V

JB
C72: 10.43V
C56: 9.20V
C117: 9.20V
C77: 5.01V

With the main board connected IC6 is getting hotter and also IC1 on the PSU gets warm incl the heatsink and also R1 and R11 warm up (with only the JB connected there is hardly any heat on the PSU components). Apart from IC6, no other components on JB or MB get hot.

[LateLesley]

OK....

A couple of mV across the switch isn't too bad. Switch contacts do have resistance, and will drop something, if it was 0.5V, we'd start looking at it. 5mV, not so much. 

Right - 0.68V across R1 = 618mA being drawn on the 5V rail - I actually don't think that is too unreasonable.

And you still have about 9V on the +12V line - again something I'd seen as an acceptable drop considering the circuit.

Time to add the LCD board, and see what changes.

IC6 would get hotter, more current drawn on the -12V by adding the main board. I am wondering if there is a fault, or it was designed to run hot.

[frisc0]

Time to add the LCD board, and see what changes.

Ok, with the LCD board connected, here are the voltages:

PSU
R1: 0.77V
R11: 0.77V
R2: 0.66V
C12: 10.32V
C13: 11.06V
C14: 10.22V

JB
C72: 10.32V
C56: 9.05V
C117: 9.04V
C77: 5.01V

And with the inverter board also connected:

PSU
R1: 1.31V (hot)
R11: 1.31V (hot)
R2: 0.66V
C12: 10.04V
C13: 10.81V
C14: 9.39V

JB
C72: 10.04V
C56: 8.78V
C117: 8.78V
C77: 5.01V

So, maybe everything normal after all? That would be back to square one and looking at a digital issue... 

I'm mentally preparing for a 'case lost'...

I'll see if I can get my hands on another power adapter with higher specs, to rule it out.

Update: I found a 12V 3A power supply from and old Asus EEEpc, but, unfortunately, same results as with the other adapter...

[LateLesley]

ok, I think we might need to look at the inverter board, as it seems to add a significant jump. It's going from ~700mA on R1/R11 to 1191mA, thats about a 500mA jump.

Ok scratch that, it seem normal, according the datasheet. https://datasheet.octopart.com/CXA-M10A-L-TDK-datasheet-10491771.pdf

Seems it can draw 0.6A - 1.0A which then fits in line with your figures.

I think all the power seems to be operating as intended, given the circuit.

I think we are into a digital issue, maybe just verify the supply voltages to all the digital ICs on the main board, and check them all for hot spots??

[frisc0]

Back in the digital domain...

Supply voltages are good, ~5V

Nothing gets hot on the MB/XP board

Checked several reset lines, all ~5V

Shorted these reset lines to GND and watched the LCD for action:

MB CPU: system reboots, LCD goes gray and after 3 seconds goes yellow/orange, without anything on the LCD (same behaviour as when turning on the device)
XP CPU: system reboots, same
LCD ctrl: system reboots, same
FDC: no action

It does not seem to be forced in reset, something else is halting the boot process. I wonder if I can do some additional measurements to trace down the issue or whether this goes into ROM/Flash failing, which would be end of story...

[frisc0]

I noticed the adapter cable easily disconnects/connects and I can imagine that that might cause issues if this would happen during a memory write action. Would there be a possibility to 'reset' / 'reflash' the flash memory on this board? And then not via the key-combi mentioned in the manual since it does not respond to that, but maybe on the IC itself, by applying the right voltages to the right pins?

[LateLesley]

I think this might have to be the way you go. You probably need an EPROM programmer.

From the manual, I think this is the flash chip (IC20 on the Main Board, Listed on page 6 of the manual) - https://datasheetspdf.com/pdf-file/62817/SharpElectrionicComponents/LH28F160S5T-L70A/1

You could probably get a programmer like the TL866II, and program it using that.

http://www.autoelectric.cn/en/tl866_main.html

You will need an adapter though, a TSOP 56.

Ouch! I just looked up the adapter and it was $130!!!...

OK, there's a couple of other programmers, which seem to come with the TSOP56 adapter, for cheaper.

https://www.aliexpress.com/item/32812908596.html?spm=a2g0o.productlist.0.0.15591808vDCcPp&algo_pvid=3369ab48-2d96-4f97-bb5a-8a20dc909118&algo_exp_id=3369ab48-2d96-4f97-bb5a-8a20dc909118-39&pdp_ext_f=%7B%22sku_id%22%3A%2264559555259%22%7D

And actually the other one isn't cheaper with the adapters. it's cheapest option seems to be the T56 + 12 adapters, which has the tsop56  https://www.aliexpress.com/item/32812908596.html?spm=a2g0o.productlist.0.0.15591808vDCcPp&algo_pvid=3369ab48-2d96-4f97-bb5a-8a20dc909118&algo_exp_id=3369ab48-2d96-4f97-bb5a-8a20dc909118-39&pdp_ext_f=%7B%22sku_id%22%3A%2264559555259%22%

It's gettin kinda pricey. Sorry.

I wonder if it would be worth just getting a new flash chip, chucking it in, and seeing if it would be ready to program then? I mean flash chips do fail, so maybe having a fresh one in, it might just fire up ready to program? it might be the cheaper option.

https://www.ebay.co.uk/sch/i.html?_from=R40&_trksid=p2334524.m570.l1313&_nkw=lh28f160+tsop&_sacat=0&LH_TitleDesc=0&_odkw=LH28F160&_osacat=0
 
I'm just spit-balling now, cos i'm at the limits of my knowledge. Maybe other peeps can pitch in and help.