Sunday, January 30, 2011

CVP-7 repair and modification

Most exciting this week was the CVP-7, an ancient Yamaha digital piano, which I got for free and was able to sell for 120.00.

This unit had a broken circuit board. But not surprisingly, when the circuit board was repaired, there was still no sound produced.

Although at first I doubted that I had stitched up the board properly, I used another pearl of wisdom I have developed in my years of fixing musical gear: if the power supply is okay, and there is no sound but there is power, check the VOLUME CONTROL.

Volume controls are always in accessible spots, so the user can get to them easily. The pessimist will note that this also means they are easy to drop things on, easy to spill things on, and easy to wear out.

So, I always find it useful to check the volume pot or fader, when there is no sound.

The easiest way is with an ohmmeter. Check for the fixed resistance by finding the two terminals between which the resistance doesn't change when you move the control. If the pot is still in circuit, that resistance may be less than the value of the pot... but if it is more, this is a definite sign of trouble. This means the pot is broken.

Now check for the changing resistance formed by connecting one of the leads to the fixed resistor, and the other lead to the wiper. This should change, although as mentioned before, the numbers read in circuit may be less than but NEVER MORE than the value of the pot.

SO, naturally I found a bad volume pot on the Yamaha.

I replaced it, and the unit works fine.

There was one last thing: I added a line in jack.

That's because this thing is from 1986 and the piano patch sounds awful.

This way, someone could use a midi controller to make the sounds, and patch it back into the unit, and use the piano's built in amplifier and effects.

To do that, I cut the trace that feeds the op amp just before the volume fader.

There is a left and a right, so I cut them both, and used a double pole switch that the user could use in order to toggle between the internal sounds and the line in sounds.

In retrospect, a pot would have been better, so that the user could blend the two sounds if desired.

(The traces I cut are around IC15 on the AN board. They are shown in the schematic I posted here in a picture. When you do this, be careful to leave the resistor and cap in place which connect terminal 1 and 2 and 6 and 7 of the op amp! Leave those connected to the op amp, and do not cut them! Otherwise, it will sound terrible). I also added a coupling resistor of 150k in series with each of my lines in. This is necessary. Why a 150 K? I used a pot and turned it until I got a good sounding signal.

When you do an experiment like this, use a pot, and turn it until it sounds good. When you find the value, use your ohmmeter to see exactly what value that is, and then use that resistor. You could also install a trimmer.

For my ground, I grabbed a ground from connection 8 pin 4. I used that as a ground on my line in jacks, and I used it as a ground on the shielded cabling I used to make the connections. THere's a lot of digital noise in that thing. Make your wires as short as possible, put your input jacks close to the "an" board, and shield all your cabling. Keep the wires away from the digital boards, unless you want to hear a high pitched wine in the background all the time. I use old patch cords for shielded cables.

So now, I was able to sell the unit as fully working, and with the ability to have it's sounds upgraded by using a computer or midi module, or another keyboard, to generate the tones.

Saturday, January 22, 2011

One that got away

This particular unit I acquired really cheap. It's a digital guitar effects processor. The listing said that no data showed up on the screen. That was a gamble, but I felt it was the power supply. The listing said it was on "all the time", which probably meant the filter caps were worn out.

Right away, I found a low 5 volt rail. A bad filter cap feeds ac into the rail, and makes it show up as less than 5 volts. Filter cap replaced.

Now it works, kind of. It turns on, and it consistently says on the screen 01 No DSP. I felt for sure that meant the digital signal processor components were not registering with the micro controller.

I suspected software. I reloaded the firmware, and suddenly, it turned on! It worked.

Being a thorough individual, I let it sit for an hour or so, and went to try it again. "No DSP". Back to that problem.

Now at this point I began to suspect the connections on a socketed IC. I cleaned it and tried again. No change.

I did battle with it for an hour, focusing on reflowing the solder around the various IC's. I'd seen issues like this, intermittent digital problems, almost always caused by a necessary reflow of solder around surface mount components, especially when lead free solder may have been used.

No real success.

This was one that was better left unfixed. This happens; at times we have to realize that a repair is not feasible.


Saturday, January 15, 2011

Filter Cap issues in a KORG DDD-1, ROLAND TR505, Alesis MicroverbIII

The This Korg DDD1 blew fuses one after the other. This often indicates a bad rectifier. The rectifier tested okay, but I noticed that there was no impedance when ringing through the filter capacitors with the multimeter.

I isolated the power supply board (unplugged everything from it), and the symptom did not change.

It was only a matter of time before the bad filter capacitor could be located.

I check filter caps with a continuity tester. In general, they should show continuity for a second or two, then the resistance should increase, until no continuity is shown.

If you are checking them "in circuit", then of course they may seem to be conducting slightly, but that is just a result of other components near them. It's important to see that curve, from low impedance to high impedance, which is a capacitor charging up. You should always be able to get it to beep for a second.

After replacing the filter caps, I also replaced the rectifier diodes.

I felt like it was possible that a leaky diode put more strain on the capacitors. This may not have been necessary, but it is cheap and easy to do.


On the TR 505, which also didn't power on, I was able to easily repair the traces around the power input jack.

This is a very easy repair.

This is a repair man's dream: Easy to spot, easy to fix, and there are no lingering doubts that you may have not addressed the whole problem!

On this Alesis unit, I encountered a whole different situation.

It powered on, but there was plenty of noise on the outputs.

I started with the power supply again, because this noise sound like 60 cycle AC noise bleeding into the system.

This time I used the scope, and sure enough on the negative 12 volt rail, there was plenty of AC leaking in.

What do we blame here?

1) The diodes. The rectifier diodes may have failed and are letting AC into the circuit. 2) A shorted regulator, or a short downstream on the -12volt rail. Why? I have noticed that when there is too much of a load on a rail, the filtering potential of the filter caps is defeated, and they don't have the capacitance to "smooth out" the rail, because the rail is drawing too much current.

How do we check that? Do we lift a leg of the 12 volt regulator and stick our ammeter in series with the load, to see how much current the rail is drawing?

We can. I don't. I just put my finger on the regulator.

If it doesn't burn my finger within a second, then it's not drawing too much current.

Finally, option 3 is a blown filter capacitor.

That means there is no filtering going on, and this was the case.

Using my continuity ringer, I noticed that there was NO BEEP, not even for a second, no matter which way I connected my leads to the cap, and no matter how long I waited. With a 300 uF cap there should be a beep.

So I replaced the cap, and all OK: smooth power rail with the oscilloscope, nice clean sound coming out of the outputs.

Walking Away From Non-Feasible Repairs

There have been successes and failures in the first week.

I purchased three drum machines for 50, was able to fix two of them with recycled parts within an hour, and they should bring in at least 150 dollars for the two of them.

I acquired a USB interface for less than 50 dollars, replaced the USB connector and did some soldering, and was able to sell it for 140 dollars.

I have acquired for free a Yamaha keyboard, and repaired that quickly, and anticipate getting 200.00 or so for it.

I repaired a dj mixer which cost 50, and anticipate selling it for 125.00.

I bought a monitor for 10 dollars, fixed it, I thought, but it failed after I put it through a few hours of testing . This is the beauty of this business model: I can walk away from an item after it becomes apparent that it's too difficult to repair... and there is no need to call a customer and tell them "sorry, your item died on the bench".

Monday, January 10, 2011

Initial Startup

With the coming of the new year, 2011, I have embarked on a new business venture. I intend to acquire damaged or broken musical devices, repair them, and resell them on Ebay.

Why do I think this is a good idea?

For five years, I repaired electronic musical equipment for a profession. I was an authorized technician, doing warranty and non-warranty work, for Yamaha, Roland, Korg, Nord, and a number of other manufacturers. My living was made exclusively through diagnosing and repairing various types of musical equipment: primarily keyboards, but also mixers, digital recorders, drum machines, and the occasional studio monitor or amplifier.

The repair process involved on one hand the technical work of disassembling and diagnosing equipment. On the other hand, it involved extensive interaction with customers.

Every piece of gear that came through the shop was attached to a customer- a person who required initial estimates, a time frame, updates when the course of the repair was changing- much of the day was spent discussing the situation with the customer.

In this new model, the customer is eliminated from involvement all during the repair process. If a potential repair is determined to be unfeasible, it can be discarded without the necessity of explanation.

This saves time.

There is no time line to establish and conform to. The equipment is not needed by Friday night for the customer's gig happy hour gig.

At the end of the repair, a fair price for the equipment is assured through Ebay, without negotiation.

Also because of the nature of Ebay, the time of warranty can be established and easily enforced.

Unsuccessful repairs will still result in the acquisition of parts, which can be sold on Ebay or used in future endeavors.

All these factors contribute favorably to this business idea.

The drawback is that broken gear needs to be found and purchased cheaply. This would have been more difficult in the time before Craig's List and E bay, but the digital revolution has made this a much easier task.

Finally, my years of experience repairing this equipment helps me to discern what is easily fixable and what is not, enabling me to avoid purchasing any major headaches.