Delta 1010 Clock Mod

[cuibono]

So I now strongly suspect that the bad breakout cable was the original source of my troubles.

....aren't likely to be the cause of the problems you are currently having.

I agree with Kevin - I doubt the smd caps are your problem (and it is very likely you will create more problems by trying to remove them).

For the most part, electronics work in a logical manner - so the question is, how did your breakout cable cause the noise in your two output channels? Did you inspect the cable at all? It seems like a continuity tester might easily tell you something valuable (either a short/open or no fault at all). It is very possible there is nothing wrong with your breakout cable, and the problem lies elsewhere. Generally, something needs to be repeated to draw conclusions about it. None the less, you should get out your scope, and trace where the noise is coming from. That will be the most productive way of fixing the problem.

Get out your meter and your scope. Do measurements, get hard, repeatable data, then act - otherwise, youre working in the dark.

 

[mtl777]

What makes it hard if not impossible to use a scope on the 1010's D/A board is because the separate A/D board sit's on top of it and blocks access to it. The two boards are separated by standoffs and have a 25-pin header connection to each other. Also, due to relay switching, the breakout box won't power up unless it is connected to the PCI card via the breakout cable and the PC is on. In other words, what I'm trying to say is that the unit and interconnections must be assembled completely in order to power-on the breakout box. In that state, the D/A board is inaccesible because it is underneath the A/D board. That is why I am resorting to component elimination logic methods in trying to fix it.

Regarding repeatability, I have repeatedly checked that whenever I used the bad breakout cable my mouse would malfunction. Replacing that cable with a good one makes the mouse work fine again. It may not seem related to the problem, but the mere fact that that bad cable causes my mouse to malfunction makes the cable highly suspect. If it can mess up the mouse which is not directly connected to it, how much more the D/A board which is its first connection on the breakout box.

 

[cuibono]

I have repeatedly checked that whenever I used the bad breakout cable my mouse would malfunction.

Ah, a very good start. I realize the board may be hard to get at. You probably don't want to go probing in cramped places, because if you short something out, it might not be pretty.

But where there is a will, there is a way. I highly encourage you to find a way to probe the analog outputs. It is only way to 'know' what you are doing.

Good luck...

 

[mtl777]

Good luck...

Thanks! I really need that.

Quick question... The SMD caps all look the same. Are they really different values or are they just reading differently because they were measured in-circuit?

 

[kevinkr]

Thanks! I really need that.

Quick question... The SMD caps all look the same. Are they really different values or are they just reading differently because they were measured in-circuit?

Probably not, the 0.9uF - 1.15uF are probably 1uF 20% tolerance, the 0.44uF is probably a 0.47uF, etc. That said it is extremely unlikely that these caps are the culprit in the problems you are having - and before you even consider replacing them you need to figure out what they are doing. Try to determine whether or not they are supply bypasses or coupling caps - if bypass caps its unlikely they have anything to do with this problem. You should probably be suspecting op-amps or maybe even some of those new caps you installed. You might want to find out whether or not the manufacturer will service your hacked cards..

 

[mtl777]

Is it common practice of manufacturers to make unmarked SMD caps that look exactly the same but are of different values? Why would they make it hard for troubleshooting?

 

[kevinkr]

Yes, ceramic caps traditionally are unmarked, resistors smaller than about 0805 are usually unmarked due to lack of space.

All of this stuff goes through incircuit and functional testing when manufactured and the factory techs have schematics, assembly drawings and other information to troubleshoot assembly line fall out so markings aren't needed.

Authorized service people will have the same information to allow them to fix things as well.

A lot of this stuff is considered uneconomic to repair even at the factory and if found defective and also sufficiently low cost may simply be scrapped as the more economic approach.

Have you tried to get the service information online. Try the Turkish Denom site if still around and just google for information. I have found service manuals for a lot of things this way.

 

[mtl777]

Man, it's so hard to find a service manual for the 1010!

I have another question though. What better diodes would you recommend to replace the four 1N4001 rectifier diodes in the 1010's PSU? Would the UF4001 be good? ...

http://www.vishay.com/docs/88755/uf4001.pdf

In case you need to know first how the 1010's PSU works, it uses a 9 VAC wall wart. There are two separate half-wave rectifier/doublers which convert the 9 VAC to DC and then double it. One of the rectifier/doublers works on the positive phase of the 9 VAC to produce a doubled positive DC voltage. The other rectifier/doubler works on the negative phase of the 9 VAC to produce a doubled negative DC voltage. The resulting doubled +/- DC voltages are then sent as input to the +/-15 VDC regulators, which then convert this input to a regulated +/-15 VDC output supply for the op amps.

The four 1N4001 diodes are the ones next to each other (D1, D2, D3, D4) in this picture:

An externally hosted image should be here but it was not working when we last tested it.

Here is a crude diagram I made of the PSU layout, with the red lines depicting the connections (it's pretty rough and incomplete, just enough to give you an idea):

An externally hosted image should be here but it was not working when we last tested it.

 

[mtl777]

Now that the holiday rush is over, I hope someone would have the time to answer my question in my last post. Anyone, please?

Thanks!

 

[kevinkr]

Yes the UF4001 which is a fast recovery rectifier would be fine - for better reliability I would probably go with the UF4002 as it has a higher piv rating.

I'd also replace D5 and D12 - the ones feeding the 5V regulators..

Have you resolved the noise issue?

 

[mtl777]

Just the answers I am looking for! And you even answered what would have been my next question - replacing the D5 and D12 diodes. Thank you so much!

BTW, what does "piv rating" mean, and why is higher piv rating better?

I haven't fixed the noise issue yet. Been so busy during the holidays. I'll do further testing this weekend.

 

[kevinkr]

PIV = peak inverse voltage, it's the reverse voltage rating on the diode which you do not want to exceed under any circumstances. UF4002 has a PIV rating of 100V which is the minimum I use under any circumstances (except for very low voltage high current applications where I will use Schottky diodes, usually with rather low piv ratings) - the higher PIV rating provides additional assurance that distant lightening strikes and other line borne transients don't take out the rectifiers, and usually there is little or no difference in cost at hobbyist quantities.

 

[mtl777]

Speaking of Schottky, I was considering two brands of MBR1100 for the 1010's PSU -- the On Semiconductor with typical forward voltage of 0.47V and junction capacitance of 77pF versus the International Rectifier with typical forward voltage of 0.6V and junction capacitance of 35pF. Considering the tendency of the 1010's power supply to sag especially at high loads, I think that the On Semi's lower forward voltage would be good, but the difference is not much -- only 0.13V. On the other hand, the On Semi has more than twice the junction capacitance of the IR. I don't know, is higher junction capacitance good or bad? Which of these electrical characteristics is more important to consider for this application, and consequently, which of the two brands would be the better choice?

Thanks!

 

[mtl777]

I wish someone could answer my last question now.

Anyway, I have another question. Browsing at Newark, I'm seeing that ceramic SMD caps are only available under the heading "Ceramic Multi-layer". Here's one example:

http://www.newark.com/70K9187/passives/product.us0?sku=KEMET-C1210C105K5RAC-TU

Is ceramic multi-layer the correct type of SMD cap for my 1010? Sorry for my ignorance but what does multi-layer mean? Is it really the only type available for ceramic SMD?

Thanks!

 

[kevinkr]

Yes, SMD ceramics are pretty much all MLCC but they are very rarely bad, that is unless they crack, or are subjected to severe electrical stress beyond their ratings... Make sure you buy X7R or NPO/COG dielectric for best stability as you don't know what was originally specified for those locations.

 

[mtl777]

Thanks for replying! Are SMD resistors as durable as SMD ceramics and very rarely bad also?

 

[kevinkr]

Thanks for replying! Are SMD resistors as durable as SMD ceramics and very rarely bad also?

They are subject to all the standard overload mechanisms any type of resistor is, however you can usually determine by comparison with another identical circuit which if any are bad. ESD and excessive voltages are common causes of smd resistor failure beyond the issue of obvious over dissipation. (i.e. running continuously beyond their design ratings.) Cracking due to excessive board flexing is another fairly common failure mode. (This sometimes manifests itself as crackling/rustling noises in audio circuits if the resistor does not go immediately open. Don't get your hopes too high however..)

 

[mtl777]

Thanks. I checked all the SMD resistor values on the defective outputs and they all seem to be fine (based on what their markings say). Though a few of them would take longer to read and "settle" the value, like 5 to 7 seconds. Is that an indication of a faulty resistor on its way out?

I have fixed the low output level problem of output #4! It turns out that one of the legs of an opamp was slightly raised and not touching the solder pad very well. I just heated and put pressure on it with a soldering iron to make it touch the solder pad. Now the output level is back to normal!

The only problem remaining is the noise on output #5. I will try to resolder the opamps related to this output and see what happens. What kind of solder can I use for surface mount? Is plain 60% tin / 40% lead OK? Or what about silver bearing solder (62% tin / 36% lead / 2% silver)? And what is the recommended soldering temperature? Can I make do with my Weller WTCPT soldering iron with 700 deg.F tip as long as I do it very quickly? Thanks again.

 

[mtl777]

Probably not, the 0.9uF - 1.15uF are probably 1uF 20% tolerance, the 0.44uF is probably a 0.47uF, etc. That said it is extremely unlikely that these caps are the culprit in the problems you are having - and before you even consider replacing them you need to figure out what they are doing. Try to determine whether or not they are supply bypasses or coupling caps - if bypass caps its unlikely they have anything to do with this problem. You should probably be suspecting op-amps or maybe even some of those new caps you installed.

At long last, the noise problem on output #5 has finally been fixed!! It was the op-amps that went bad. I replaced two op-amps.

Man, after agonizing so long, I feel so happy and relieved that this problem has finally been fixed. Thank you so much everyone for all the good advice on this forum especially Kevin's suggestion to check the op-amps! Thanks, man!!

For soldering the op-amps, I used very fine 0.015" 63Sn/37Pb solder with a PTJ7 1/64" tip on my Weller WTCPT. This was running at 700 deg.F but I made sure to work it quick and not linger too long. It worked great!

 

[mtl777]

Now that my problem is fixed, let's get on with the mods!

My next question is regarding the booster caps, C14 and C17, in the voltage doubler/rectifier part of the PSU. I'm currently seeing VLSOM brand 470uF 63V for these caps. Have you ever heard of VLSOM caps? Are they any good? I would like to try replacing them with ultra low ESR, high ripple current rated KZE series caps from United Chemi-con. Do you think this is a good idea?

I want to be careful, though, because according to this thread, the original poster JohnR warns against using bigger values for these caps, thus:

"don't just blindly change any of those PSU caps either; you'll end up with fried regulators unless you put _serious_ heatsinks on them (I machined mine out of solid alloy, but I have access to engineering equipment)."

Why is he saying this? Is it because using bigger booster caps will cause the rectifier output, which becomes the input voltage going into the regs, to become higher, thus making the regs hotter? I wonder if this is true? If so, then to be safe I won't use bigger caps. I will just replace with same value but better quality KZE caps. However, is there a danger that "better quality caps" might mean that these KZE's are much more efficient -- and therefore they are actually equivalent to a higher value than the caps being replaced? I've heard of people replacing ordinary caps in motherboards with lower value Sanyo OS-CON's because the OS-CON's are so efficient that a lower value OS-CON is equivalent to a higher value of the ordinary cap. So, do you think I should play safe and replace those stock VLSOM caps with lower value KZE's instead?

I would greatly appreciate your advice. Thanks!