Alan,
You must live somewhere the average house price, is not $800,000.
Standard bench service rate, in this particular part of the world, was $45 an hour, thirty years ago.
Car mechanics are well over 120 per hour, now.
I'm just sayin'..........
Gregg Sheehan
If one person has delusions, we call them psychotic. If, however, 1.5 billion people have delusions we must apparently call them a religious group, and respect their delusionary state.
--- On Wed, 5/27/09, Alan Probandt <***@yahoo.com> wrote:
From: Alan Probandt <***@yahoo.com>
Subject: Re: [YamahaDX] Re: DX7 is not booting up at all!
To: ***@yahoogroups.com
Received: Wednesday, May 27, 2009, 5:37 PM
To check ripple, set the scope on AC input; 1 volt per division. Put the probe on the +5V, the +15v line, or the -15V line, NOT the main power 120VAC line. Ripple will look like the top of an airplane wing. It shouldn't be more than a quarter volt peak-to-peak.
As a microprocessor electronics technician, I disagree with the idea that someone should get $200 to 'look' at a an old system that costs $50. A skilled person should be able to find obvious and 'easy' stuff in about a half hour of probing if they have a schematic. They should charge $20 for that first half hour. They should speak a report as the work into a speech-to-text program and give the text file either as a printed report or as a file on a flash disk to the customer if they can't fix the problem in a half hour. They should charge no more than $40 an hour and no more than twice what they paid for the parts they replace.
With the economy collapsing it's time people started being realistic about what they charge for the services they provide. I know that everyone disagrees with me, but that's my opinion. Thank you.
--- On Wed, 5/27/09, Lars Larsen <creatorlars@ gmail.com> wrote:
From: Lars Larsen <creatorlars@ gmail.com>
Subject: Re: [YamahaDX] Re: DX7 is not booting up at all!
To: ***@yahoogroup s.com
Date: Wednesday, May 27, 2009, 10:56 AM
Good, thanks for the advice. It's definitely apparent that I'm not getting valid signals on AD0/D0, AD2/D2, AD3/D3 but the others look okay.
I removed IC25 & IC26 as suggested. This seemed to make a lot of sense, since the left LED digit would give random values upon power up -- and sometimes change to another random value when I heard the reset "click". Whereas the right one would always show "8." This didn't fix the issue, though. But I can order replacement chips as suggested when/if I get the rest of the circuit operational.
I am seeing a high pulse at the Reset pin of the main uProcessor shortly after startup, then it stays low. So I don't think it's constantly resetting itself or anything. Something somewhere has to be crashing the data bus on those three lines. Power looks good to me -- how can I check for ripple? Caps show no signs of bulging/leaking. I don't guess there's an easy way to isolate parts of the circuit to check for faults. Awful lot of chips to replace one by one -- don't think I'd want to go that far, unless I had a good guess what was wrong.
I agree that the amount of work put into debugging a circuit like this is worth the money for the expertise -- but, as you've said, not worth it for the DX when I can wait and get a new one sometime for under 200 easily. I paid 50 for this one, the guy told me it just needed a new battery (obviously just trying to get rid of it.)
Any other ideas/guesses greatly appreciated! I'll keep poking in it as I continue to think of stuff to test/try.
Thanks,
Lars
On Wed, May 27, 2009 at 12:28 PM, Gregg Sheehan <***@yahoo. com> wrote:
Alan's right. Use the scope. For digital work, a DMM goes about as far as the power supply for a troubleshooting tool.He is also right about the specific voltage levels of data and address signals. They are rarely perfect 0 VDC and 5VDC. This is where experience will tell you that a good data line, for example is typically 3.5 volts high( or higher), and conversly "off" will not be perfectly zero but around.7 (transistor threshold).
All this yapping and typing, is refreshing my memory ( pun intended). DX UPC troubleshooting 101----make sure the reset circuitry is functioning. Monitor the reset line with scope- power up and after a short delay, there should be a state change and the reset line will then settle to whatever it's normal running state is. (Hi/ low? -I forget- the schematic will say- "reset" with the "not"line over is active low, run state is high) Note that this happens a short time AFTER the supply voltages come up.
Another point of failure- the D/A converter and D/A buffer. I will look at a schematic and see if anything else pops up. Is the power supply clean?- ripple will cause many headaches. Pooched filter caps were common.
Re: the guy charging 200.00. It sounds like a lot, but imagine spending an afternoon doing this with no resolution. You only make what you bill out. I'm fixing an OB-8 right now- same type of problem- it will be expensive, but these go for huge money right now- almost approaching new list price- 3-4K$, so it's worth it, to spend 200-500 on repair. Not so our beloved DX boards, unfortunately. It's like all the cheap throw away Chinese audio appliances, n'est pas?
Gregg Sheehan
If one person has delusions, we call them psychotic. If, however, 1.5 billion people have delusions we must apparently call them a religious group, and respect their delusionary state.
--- On Tue, 5/26/09, Alan Probandt <alan_probandt@ yahoo.com> wrote:
From: Alan Probandt <alan_probandt@ yahoo.com>
Subject: Re: [YamahaDX] Re: DX7 is not booting up at all!
To: ***@yahoogroup s.com
Received: Tuesday, May 26, 2009, 11:06 PM
Gee, I wish that I could get $200 just to look at a vintage synth. Most likely this technician is just saying that to avoid dealing with vintage microprocessor gear. I'm reminded of this when I looked at the DX7 schematic. The 63B03 CPU is a clone of the Motorola 6803 that was popular for a period of time in the late 1970s. This was the CPU in the first computer that I ever bought, the Radio Shack MC-10 (mini-color computer) in 1984. I was a student then and remember reading about signal lines like 'E' that are unique to this CPU family. But it has been a long time.
IC32 is a data bus transceiver that strengthens the data lines between the CPU and all the memory chips. It's a 74LS245, not a 254. It definitely should have clean signals (0 to +4.5v) and very active signals on the pins of the right side of the chip (pins 11-18). Actually any voltage level on the Data pins (11-18) that is above +3.5V will be a valid logic 1 signal. Anything below 0.7v is a logic 0. Anything between +3.5 and 0.7 is an invalid logic level and should never be seen on the data bus. Pin 19 will show about 1.18MHz and it should be a stable waveform on the scope, although probably not a square wave. Having IC32 be bad would definitely muck up the system. All the data lines on the right side of the chip should show constant activity. If their is no activity check the corresponding data/address pin opposite it for activity. IC32 allows data on one side of the chip to flow (and be amplified) out the other side when pin 19 \E is
low. The direction of the data, either from the CPU to the RAM and sound chips, or from the ROM to the CPU depends on the logic of the DIR pin 1.
If several of the data pins on this chip are definitely not showing any signal, then it should be replaced. I'll catch the devil for saying this, but you can substitute a 74HCT245 if a 74LS245 is impossible to find. Get a very small pair of diagonal clippers and cut each pin of the bad chip right at the body of the chip. Leave the other end of the pins soldered to the board. Solder the pin tips of the new chip to the clipped pins of the old chip. Power up and check that the data bus signals are getting through the new chip and to the memory ICs. Don't use the frequency meter to test if the signals on IC32 are good, use the scope.
Looking for the perfect gift? Give the gift of Flickr!
__________________________________________________________________
Be smarter than spam. See how smart SpamGuard is at giving junk email the boot with the All-new Yahoo! Mail. Click on Options in Mail and switch to New Mail today or register for free at http://mail.yahoo.ca
