Ensoniq VFX SD mainboard troubleshooting

Pulling out your hair? Don't know what to do or where to go? Ask in here.
Forum rules
READ: VSE Board-Wide Rules and Guidelines

If your Help request has been solved, please edit your first post in order to select the Image Topic Icon to let others know your topic has been solved.
cutoffres
Junior Member
Junior Member
Posts: 153
Joined: Fri May 15, 2009 2:24 pm
Location: Europe

Re: Ensoniq VFX SD mainboard troubleshooting

Post by cutoffres » Wed Feb 08, 2017 6:07 pm

Ok, i think i understand: The VRES must be connected to a cpu to "exist", so when i unplug the mainboard from power supply board what i measure is not the VRES line but the LM2926 output voltage. Do i understand right?

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Wed Feb 08, 2017 7:41 pm

cutoffres wrote:Ok, i think i understand: The VRES must be connected to a cpu to "exist", so when i unplug the mainboard from power supply board what i measure is not the VRES line but the LM2926 output voltage. Do i understand right?
Not exactly. The VRES doesn't need a CPU, the CPU needs a VRES. VRES is always generated externally to the CPU.

The LM2926 has two output voltages, but both are +5VDC. The difference is that the first is to power the digital chips with a steady and constant voltage. This turns them on and keeps them running. The second voltage is VRES which is designed to flip the CPU (and some other chips) in/out of reset mode. It's usually pegged at +5VDC but it can't come on right when power is applied because the CPU needs time for the OTHER +5VDC to power it up before it can be told to boot. You can't boot something that isn't even powered up yet, right? Another factor is that the crystal/clock needs time to stabilize before boot can occur. So there needs to be a delay before VRES jumps from 0V to 5V, although the delay is typically immeasurably short with a digital multimeter.

So one output from the LM2926 should ALWAYS measure +5VDC unless the power switch is off. The other output from LM2926 should almost always measure +5VDC EXCEPT for a very short delay when power is first applied, or when there is a fault condition (as it seems to be in your case).

If VRES is stuck low then the CPU is constantly being told to stay in reset mode. If VRES never goes high then the CPU will just sit there waiting for the "all systems go" signal to hit its reset pin which is ultimately supplied by LM2926. In this case, it's never going to occur until you find out why LM2926 isn't outputting +5VDC on its VRES pin.

LM2926 is either broken and needs replacement, or there's a fault condition causing it to keep VRES at zero volts.

cutoffres
Junior Member
Junior Member
Posts: 153
Joined: Fri May 15, 2009 2:24 pm
Location: Europe

Re: Ensoniq VFX SD mainboard troubleshooting

Post by cutoffres » Fri Feb 10, 2017 7:00 pm

I've replaced C21 the fault delay cap without success.


I've scoped the LM2926 OUTPUT VOLTAGE and DELAYED RESET OUTPUT when power ON:
Mainboard plugged to the Power supply board:
OUTPUT VOLTAGE PIN 2 goes +5V
DELAYED RESET OUTPUT goes +0,3V with a 450ms delay

Mainboard unplugged from the power supply board:
OUTPUT VOLTAGE GOES +5V
DELAYED RESET OUTPUT goes +5V with a 450ms delay

As you state that the delay should be around 40ms and i find a 450ms delay or my time mesurement are not good or the LM2926 is not good, or both.

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Fri Feb 10, 2017 8:03 pm

cutoffres wrote:I've scoped the LM2926 OUTPUT VOLTAGE and DELAYED RESET OUTPUT when power ON:
Mainboard plugged to the Power supply board:
OUTPUT VOLTAGE PIN 2 goes +5V
DELAYED RESET OUTPUT goes +0,3V with a 450ms delay

Mainboard unplugged from the power supply board:
OUTPUT VOLTAGE GOES +5V
DELAYED RESET OUTPUT goes +5V with a 450ms delay

As you state that the delay should be around 40ms and i find a 450ms delay or my time mesurement are not good or the LM2926 is not good, or both.
I don't think the delay being higher is really a problem because the CPU needs a minimum time to power up but there is no max. The CPU doesn't really care when it comes out of RESET so long as the clock and voltage being supplied to it are stable when it does.

So the VRES voltage on the PSU *does* reach +5VDC, but only when not connected to the mainboard? If it ever reaches 5V then the LM2926 should be good. If it's stuck constantly at or around 0V when connected to the mainboard then there is probably a fault on the mainboard.

Earlier posts made me think that VRES was always low and never reached 5V.

Assuming VRES sticks low only when the mainboard is attached, we should also test VRES with the mainboard connected by itself, so that the display/keypad and keyboard are not connected to the mainboard. If there is a fault on one of the other boards then it could cause a fault to seemingly be coming from the mainboard. If the fault is truly isolated to the mainboard then we need to trace the VRES line to see where it is going bad.

Here is where I start running out of great ideas. But you could try taking out ALL the socketed ICs that need RESET, like the CPU, RAM, ESP chip, etc. at once. See if VRES is able to stay at 5V. If so, keep putting back ICs one-by-one until VRES is stuck low again.

Lifting a leg of R19 is another way to isolate part of the circuit from the RESET line. If that gets VRES to 5V then try connecting R19 again and lifting either the collector or emitter of Q2. That might narrow it down a bit.

Like I said, I start becoming useless at this point. I'm don't have any great solutions for fault isolation. I'm sure there's a more graceful way to ohm it out, but...

Ashe37
Moderator
Moderator
Posts: 3908
Joined: Mon Jan 19, 2009 3:43 pm
Real name: Unpronounceable
Gear: Ensoniq SD-1/32,SQR,VFX,ESQm
Virus Indigo, M3-61 , MS2000BR, Volca Bass
Emu XL-7, Matrix 6r
TG-33, K3m, Blofeld, Micron, Mopho, BS II, JV-1080
Band: Eridani V
Location: Central VA

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Ashe37 » Sun Feb 12, 2017 12:53 pm

Rasputin wrote:and in fact there's a virtual VFX (even if not fully complete) already floating around out in the ether somewhere.
Where? i haven't seem it.

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Sun Feb 12, 2017 7:02 pm

Ashe37 wrote:
Rasputin wrote:and in fact there's a virtual VFX (even if not fully complete) already floating around out in the ether somewhere.
Where? i haven't seem it.
Okay, this is the Ensoniq SD-1 but same difference.



I think I've seen a specifically VFX version somewhere, but it is purely an academic difference.

Ashe37
Moderator
Moderator
Posts: 3908
Joined: Mon Jan 19, 2009 3:43 pm
Real name: Unpronounceable
Gear: Ensoniq SD-1/32,SQR,VFX,ESQm
Virus Indigo, M3-61 , MS2000BR, Volca Bass
Emu XL-7, Matrix 6r
TG-33, K3m, Blofeld, Micron, Mopho, BS II, JV-1080
Band: Eridani V
Location: Central VA

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Ashe37 » Thu Feb 16, 2017 9:01 pm

I'll believe it when i see it. The UVI stuff is just fixed samples of patches they programmed, with artbitrary filters, not an emulation.

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Thu Feb 16, 2017 9:54 pm

Ashe37 wrote:I'll believe it when i see it. The UVI stuff is just fixed samples of patches they programmed, with artbitrary filters, not an emulation.
Umm, well, I'm not sure why you think it's so far fetched. MESS/MAME already have drivers for the proprietary Ensoniq chips so all you need is the ROM dumps and you've got a virtual SD-1/VFX. And I already have the ROM dumps (not just the obvious firmware, but the waveforms) and so do some other people apparently.

The CPU, RAM, UART, FDC, etc. are all run-of-the-mill. The display/keypad MPU is so simple in operation that it's only necessary on real hardware. There's nothing important locked away in there.

:dontknow:

It doesn't really matter, I guess, because the VFX series is pretty much so undesired that they're at just about giveaway prices still.

Ashe37
Moderator
Moderator
Posts: 3908
Joined: Mon Jan 19, 2009 3:43 pm
Real name: Unpronounceable
Gear: Ensoniq SD-1/32,SQR,VFX,ESQm
Virus Indigo, M3-61 , MS2000BR, Volca Bass
Emu XL-7, Matrix 6r
TG-33, K3m, Blofeld, Micron, Mopho, BS II, JV-1080
Band: Eridani V
Location: Central VA

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Ashe37 » Thu Feb 16, 2017 10:11 pm

Except, MAME emulates the OTTO but not the ESP chips..... Also, the usual fail is people not setting it up to that the software can scan a wavetable with dissimilar sample lengths. Doesn't matter, it will be a free plugin anyway.

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Fri Feb 17, 2017 4:16 am

Ashe37 wrote:Except, MAME emulates the OTTO but not the ESP chips..... Also, the usual fail is people not setting it up to that the software can scan a wavetable with dissimilar sample lengths. Doesn't matter, it will be a free plugin anyway.
Ensoniq ESP driver for MAME:
https://github.com/mamedev/mame/blob/ma ... es5510.cpp

You can think VFX emulation is c**p all you want though. I never claimed it was a suitable replacement, I just said it was proof-of-concept that someone has already made considerable strides toward those ends. I stand by my claim that the VFX has no special properties that necessitate the survival of the actual hardware.

Seeing as I own one and I've taken many hours developing this thread, I clearly find current value in the hardware even if just for the physicality, convenience, and nostalgia of it. But if you're arguing that nothing can currently take its place 1:1 then, yes, I agree.

Ashe37
Moderator
Moderator
Posts: 3908
Joined: Mon Jan 19, 2009 3:43 pm
Real name: Unpronounceable
Gear: Ensoniq SD-1/32,SQR,VFX,ESQm
Virus Indigo, M3-61 , MS2000BR, Volca Bass
Emu XL-7, Matrix 6r
TG-33, K3m, Blofeld, Micron, Mopho, BS II, JV-1080
Band: Eridani V
Location: Central VA

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Ashe37 » Fri Feb 17, 2017 10:02 am

I don't think its c**p, i will believe it when i see it. I'd love to have a functional one rather than moving the giant heavy keyboard. Patch compatability would be a plus. The thing is, there are other limits where the filter needs to work and sound the same and that isn't just a matter of having the samples off the ROMs or an emulation of the sound chip. I talked to Digital Sound Factory at NAMM 2014 about this kind of thing for a long while , because he does have the Ensoniq samples along with the Emu ones, but there's a long complex and convoluted explanation as to why he has made sample sets instead of building emulations.

Of course, the other reason no one has done one, especially not a commercial one, is that Creative can be... harsh about their IP.

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Fri Feb 17, 2017 12:00 pm

Ashe37 wrote:I talked to Digital Sound Factory at NAMM 2014 about this kind of thing for a long while , because he does have the Ensoniq samples along with the Emu ones, but there's a long complex and convoluted explanation as to why he has made sample sets instead of building emulations.

Of course, the other reason no one has done one, especially not a commercial one, is that Creative can be... harsh about their IP.
Interesting. I would have liked to have overheard the conversation.

Licensing is always an issue, yeah. But seeing as SQ8L is a much lauded VST and had the same obstacles (plus having to emulate an analog filter) then it's reasonable for VFX to happen successfully in the freeware domain. Whether or not someone actually does is anyone's guess, but considering the VFX has a digital filter then it should be even more possible to recreate than the ESQ series.

I look at it something like this: The Yamaha TX16W was recreated in software about as seamlessly as you can get, but you know if there was a software version of the Ensoniq Mirage that people would argue the filter wasn't creamy enough (or <insert complaint here>).

You're right in that there are barriers to it ever happening in a full-fledged (commercial?) product that is 100% satisfying to end users (h**l, the Roland D50 and Korg Wavestation software versions still don't satisfy everyone!) but from a technical standpoint, the VFX series seems very possible to me. But that's all just amateur conjecture. Maybe there's some hidden stumbling block I'm not aware of--which makes the NAMM conversation all the more intriguing to me.

Ashe37
Moderator
Moderator
Posts: 3908
Joined: Mon Jan 19, 2009 3:43 pm
Real name: Unpronounceable
Gear: Ensoniq SD-1/32,SQR,VFX,ESQm
Virus Indigo, M3-61 , MS2000BR, Volca Bass
Emu XL-7, Matrix 6r
TG-33, K3m, Blofeld, Micron, Mopho, BS II, JV-1080
Band: Eridani V
Location: Central VA

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Ashe37 » Fri Feb 17, 2017 12:20 pm

Its mostly that each step in the wavetables on the VFX series has a variable length, that is a specific stumbling block he mentioned.

cutoffres
Junior Member
Junior Member
Posts: 153
Joined: Fri May 15, 2009 2:24 pm
Location: Europe

Re: Ensoniq VFX SD mainboard troubleshooting

Post by cutoffres » Wed Mar 15, 2017 5:41 pm

So back to this Ensoniq VFX SD, I summarize where we were:

This VFX SD was faulty: No boot , even nothing on the screen at power up...

Power supply and mainboard where faulty.
Power supply has now been serviced and give perfectly the needed voltages.

On the mainboard there was a problem with the VRES voltage that was wrong.
After troubleshooting the mainboard (with rasputin help), i've suspected the MC68681 DUART being faulty. Now i've replaced the MC68681 DUART the VRES voltages are good :)

The VFX SD is still faulty with the same -no boot- fault.

So i have double checked all voltage:
Mainboard NOT CONNECTED to the power supply: All voltage are good.
Mainboard connected to the power supply: All the voltage are good except for -12 ANALOG

So i may have one or several component that are faulty on the mainboard -12V ANALOG line (a lot of components to check).
I'm going to continue and find the faulty component. Maybe there is is a perfect culprit on the -12 ANALOG line?

Rasputin
Senior Member
Senior Member
Posts: 808
Joined: Thu Jul 21, 2016 4:52 am

Re: Ensoniq VFX SD mainboard troubleshooting

Post by Rasputin » Thu Mar 16, 2017 4:08 am

cutoffres wrote:So back to this Ensoniq VFX SD, I summarize where we were:

This VFX SD was faulty: No boot , even nothing on the screen at power up...

Power supply and mainboard where faulty.
Power supply has now been serviced and give perfectly the needed voltages.

On the mainboard there was a problem with the VRES voltage that was wrong.
After troubleshooting the mainboard (with rasputin help), i've suspected the MC68681 DUART being faulty. Now i've replaced the MC68681 DUART the VRES voltages are good :)

The VFX SD is still faulty with the same -no boot- fault.

So i have double checked all voltage:
Mainboard NOT CONNECTED to the power supply: All voltage are good.
Mainboard connected to the power supply: All the voltage are good except for -12 ANALOG

So i may have one or several component that are faulty on the mainboard -12V ANALOG line (a lot of components to check).
I'm going to continue and find the faulty component. Maybe there is is a perfect culprit on the -12 ANALOG line?
While everyone likes to think a PSU recap and the keyboard service bulletin fix will get 99% of broken VFX machines repaired, unfortunately that's not the case. As you've found out, the VFX can definitely be quite a pain to repair. Too bad it's not fixed yet, but I'm happy to see you're still fighting the good fight. You just have to remember that it CAN be fixed!

I really wish they would have put that UART in a socket on that thing. Would have been SO helpful.

On the mainboard, toward the right side about halfway up, there's a -5V regulator [79L05] that touches the -12 line. Remember that it's pinned differently from the 78L05 though. What are you measuring there?

There's also TP1 which connects to a couple 50V capacitors (possibly labelled C31, C32 on your board). That might be worth investigating. I personally would approach it by trying to find a way to lift a leg on something in the -12 line to isolate certain parts of the analog circuit to see if the problem goes away.

Post Reply