Ensoniq DP4 - won't boot up

[Enzo]

OK, so the basic display board computer is working, that directs us at least initially to the communication interface.

[audiopete]

Good that's what I thought as well. There is a hex buffer near the cable I assume that's part of the communication network.

Just no luck.....my 50" Panasonic plasma died yesterday after less than 2 years.....this one replaced a year and a half old Samsung LCD that developed a bad panel......what happened to TVs that lasted? Now I have to go through the Panasonic service manual to try to fix the 7 blinking lights of death.....ahhhh!

[Jazz P Bass]

Scan board?

[audiopete]

That seems most likely. I had a nice talk with Panasonic today and my case was elevated "up the chain". We'll see...basically I want to know if they stand behind a 2 year old $1000 TV crappin the bed.

If only the Ensoniq was so easy....

I fixed a Jomox drum box with some good owner inflicted board damage from cap job gone awry, and the original problem of a shorted LM7805, and there is no support from that one man operation either. Makes me wonder if I would want to buy a synth for not small dollars and have a headache if it has a problem. At the least you have to ship to Germany and probably wait 6 months.

[audiopete]

Ok quick update - been through most of the logic chips with a logic probe and none are dead or obviously malfunctioning. I ordered a couple NOS Motorola 68B03P MCUs and I will swap out the main processor and the 8 MHz crystal oscillator and go from there...... maybe investigate the eprom further, if necessary. The processor is an 8 bit design from 1979.

I can't really tie up my bench with this for days so I have to go at it here and there when I can. Gives me time to sleep on it and do some more research into digital circuits.

[audiopete]

Ok quick update - been through most of the logic chips with a logic probe and none are dead or obviously malfunctioning. I ordered a couple NOS Motorola 68B03P MCUs and I will swap out the main processor and the 8 MHz crystal oscillator and go from there...... maybe investigate the eprom further, if necessary. The processor is an 8 bit design from 1979.

I can't really tie up my bench with this for days so I have to go at it here and there when I can. Gives me time to sleep on it and do some more research into digital circuits.

[audiopete]

So another update (because I hate dead/inconclusive threads), been through most of the logic circuits and some signals are flowing and some are not. It appears that the microprocessor, ROM, RAM and communication interface chips are all functioning, as checked with scope and logic probe. Without a schematic it's hard to trace signal flow. I have verified about 60% of the logic chips are functioning according to their spec sheets, but I have some issues like stalled clock signal/logic signals on a d type flip flop and a missing output (no hi or lo signal) on a hex buffer chip (all other buffers on that chip working fine), even though I've swapped the chip out. I need to trace where the signal goes and why it's gone.

I bought another DP4 needing repair, but its front panel and operating system are working fine and the menus all come up. The new unit is not passing audio in or out, so I'm hoping it's a blown regulator or something easy on the +-15VDC. I haven't received it yet, so I'm only guessing at this point. With some luck, I should be able to compare the working logic circuit to the dead one and figure out where to go next. I can now isolate whether the problem is with the mainboard, front panel board, or both. At the end, I hope to have two functioning DP4s and learn some more about repairing these buggers...

Thanks again to all who helped.

Cheers

[audiopete]

Update - The second DP4 arrived and after some quick tests I discovered the -15VDC rail was down, the fuses corresponding to this were blown, and there was a short from -15VDC to ground. After removing the TL072/TL074 audio opamps, I still had a short so I started checking other components. One by one I started lifting decoupling caps and luckily I found a shorted cap. I put in a new 0.1uF ceramic, installed new opamps because they're cheap, and all is well.

Almost......

Now all my panel LEDs are lit up and there is big DC on my outputs. Knowing the ribbon cables are basically shite, I think I shorted or damaged the cable/front panel. I will swap the cable and if that isn't it, I will start exploring the panel board. On it goes.....

[Jazz P Bass]

Thanks for the update.

It's always good to get an Ensoniq going again.

[audiopete]

Ok second unit all good. The problem was the analog ribbon cable (I assume it was shorting at the connector(s) and sending supply voltage down an audio line or two) so I disassembled the press fit connector, snipped off about 3/4" of ribbon cable and press fit the connector back on the fresh section of ribbon cable. I did this to both ends and it is working like a charm now.

I also confirmed that the issue with the original DP4 is on the mainboard, as the front panel and ribbon cable for that unit works fine with the good unit. Now I can concentrate on the mainboard of the non functioning unit and compare the working one to the bad one.

[audiopete]

Thanks for the update.

It's always good to get an Ensoniq going again.

Yes I agree, especially when there are no schematics Actually, the DP4+ schematics available are pretty helpful for most parts of the circuit. The architecture is different using different opamps to basically do the same thing (no quad opamps) and different processor/communication architecture. A lot of the component labeling is different but it isn't too hard to translate to the DP4. I think the DP4+ is SMD components, so I guess Ensoniq decided to redesign some things using the newer components available. I last used one on a mix session back in 93 or so and remember being very impressed. The 4 in 1 is the best feature and easily the best value in an FX unit. This one is going in my rack once I get the other one fixed up.

[audiopete]

Update - After shelving this for awhile, I have it back on the bench and I have been tracing the circuit. I have a bad clock signal that runs the ADC, DACs and the programmable logic sequencer and related logic. I think the gate array chip does this logic sequencing in the DP4+ schematics, which I have attached. The analog schems seem to match up, except component numbers are different in places. The digital schems match except for the gate array and location of the display board mpu. It looks like it is on the digital board on the DP4+, but my DP4 has it on the display board. I know the display board is good because it works with my working DP4.

Anyway, I am at the point of tracing the DACLK signal back to source from the ADC (page 9 of analog board schems). It is coming in at a much slower frequency than the working unit shows, and it is corrupted. It looks like a square wave but the 5V level falls maybe 2V from rising edge to falling edge. I am going to disconnected the clock line from the ADC and see if the signal improves. I am also going to look at the zener diode, caps and resistors on that line.

The schems show DACLK coming from the gate array (U45 - page 3 of digital board schems). I assume this is a custom chip for Ensoniq. The DP4 has a PLS159AN programmable logic sequencer and supporting 7400 series logic. I am going to trace back and find out where DACLK originates from. I'll update when I have more.

[MatanaTeknik]

Hi,

First post here..

Im in the middle of fixing a DP4 with the same problem as yours. Did you ever manage to fix it?
On a Dutch forum a guy had the exact same issues and amnaged to fix this by swapping the 68B03 CPU chip.
Did this also work for you?
And by any change to you happen to have a cpu chip left?

thnx

[audiopete]

Hi,

First post here..

Im in the middle of fixing a DP4 with the same problem as yours. Did you ever manage to fix it?
On a Dutch forum a guy had the exact same issues and amnaged to fix this by swapping the 68B03 CPU chip.
Did this also work for you?
And by any change to you happen to have a cpu chip left?

thnx

Hi, I haven't fixed it yet, but I have gotten to know it a lot better I am waiting on a Sony CX20018 DAC because I suspect a bad DAC is the problem. All clock signals go back and forth but the DAC for channel 3&4 has an oscillator crystal driving it. I lifted the bitclock (BCLK) feed from the ADC and looked at the DAC wordclock (WCLK) out, and it is still present and corrupt - meaning it doesn't need a clock from ADC - it is generated internally in the DAC from the crystal. I have to assume it is the master clock. I also assume the crystal is good but I only have a 20MHz scope, so I can't see it. I hope so because it is 73.728MHz and I cannot find a replacement anywhere. I fear I would have to build a 5V cmos active clock and interface it in place of the crystal. I have replaced the 68B03 CPU, the I/O chip for the CPU (68something), the CPU on the display board and a few logic chips, all to no avail. That's when I decided to bite the bullet and start documenting the clocks with continuity tester and datasheets. With the schems for DP4+, plus other clock designs from some Ensoniq keyboards, I was able to start figuring things out and which way they flow. I also have a second working unit to compare to, so that helps. I know my display board is good, for example.

Do you have a scope or logic probe? There are lots of reasons why you could have this issue, so you need to troubleshoot and start rulings things out. What have you checked and how? Confirm continuity of your ribbon cable first, make sure your CPU is running, make sure reset circuit is working, clocks are proper, data moving between RAM, ROM and CPU, proper voltage on all chips etc etc.

If you need parts I can help you with that. The first thing I did was recap the entire unit to rule out bad caps.

[audiopete]

So the sage continues here.....

- discovered that the interface chip for the CPU to the display board that I replaced before was bad and messing up CPU Enable signal. I changed it again and some of my bad clocks are now crisp and clear.
- still a bad , smeared WCLK signal from the A/D converter. changing the converter did not fix it. I next suspect the crystal oscillator circuit on the CX20018 so I've ordered the tiny load caps, coupling cap and inductor and I am going to change out those components. In comparing to my working unit, the crystal circuit clock signal seems to be fuzzier than the good one, so perhaps it isn't in spec anymore. If this doesn't fix it, the next thing I am going to do is pull and test every 74 series logic chip. With a socket, if the chip tests good, it can go back in. I picked up a TOP2008 USB eprom programmer which also tests all logic chips. I've gone through some of my stock and I have found bad parts, so it's worth it. When I fix this one I can put the last version firmware in both units via 2 new burned eproms each.