D5500 Power Supply--Complete Rebuild From Scratch

epatsellis · Mar 13, 2009

ewww, while the bifurcated sockets are OK, machine pins are more reliable...

ic-racer · Mar 13, 2009

epatsellis said:
ewww, while the bifurcated sockets are OK, machine pins are more reliable...

Agree. What I was thinking is that I have 4 of these boards and that in the future, component swapping for troubleshooting will be possible.

ic-racer · Mar 14, 2009

I got all the new components on the board last night. I'm still waiting on a 14 pin socket, then I'll be set to try it out.

ic-racer · Mar 16, 2009

epatsellis said:
ewww, while the bifurcated sockets are OK, machine pins are more reliable...

I misunderstood this when I first read it. This is a machine pin socket.

ic-racer · Mar 18, 2009

Here it is, almost ready.

ic-racer · Mar 18, 2009

Only reason the board is not ready to go is that I discovered a slight error in my selection of the replacement chip for the 'Dual NOR driver' UDN-5714M.
This component is not available from Mouser, CEI or Digikey, and I'm still looking for them. I thought the 75454b would be a replacement and ordered them, but on examining the datasheets, there is a discrepancy at pin #6.

ic-racer · Mar 18, 2009

While waiting to source the above parts I did an overhaul of my 3 working heads. Now they all zero in on the filtration quickly and are very stable. I replaced the electrolytic caps in all 3 boards. The biggest improvement in stability was to set the 'offset voltage' in the color probes. The manual contains many errors in the procedure and for years I could not figure it out. I finally sat down with the schematic and a 'test head' and figured out all the errors in the shop manual and was able to set the value. Looking at the corrections in my manual, it looks like I could compile quite a long list of errors. Perhaps I'll put these together and post them if there is interest.

One of the heads also had a bad rubber wheel which I repaired.

Coincidentally, there was a post yesterday on the Large Format forum indicating that new rubber wheels are still available from Omega (number 800-777-6634). I'll have to check that out for spares.

ic-racer · Mar 18, 2009

I had previously posted that the board that fried light bulbs had a cold solder joint on the connector pin to the light enters the board. I discovered the same condition on another board also. Looks like this is a common failure point. I re-soldered this connection on all the boards.

Now I know how that was causing the lamp failures. The high resistance connection caused the RMS voltage sensor to ramp up lamp voltage until it arced across the connection. This illuminated the lamp with a high voltage, and the system is not quick enough to ramp the voltage down again and the bulb blows.

ic-racer · Mar 18, 2009

I realized that used care in removing all the old components (just in case some would be hard to source), so I cleaned up the legs of the three old UDN-5714Ms and popped them in the sockets and it checked out find.

So, at least I proved to myself that a complete component-for-component rebuild is possible and affordable. Components were about $30.

ic-racer · Mar 25, 2009

As I was fixing these heads up I noticed the bulb socket contacts were black from oxidation. I spent about an hour polishing the contacts and applying DeOxit. I saw a number on the socket "QLV-1" and looked it up. It turns out this is a readily available socket. I will have to pick up one as a spare.

ic-racer · Mar 25, 2009

I also found a place that has the TD62479P (a cross reference for the SN75479N )

Not too bad, $5 each. http://www.utsource.net/

I had a quote from another place that wanted $20 each, minimum order of 10.

ic-racer · Apr 8, 2009

Just an update on this thread. I was finally able to do the transaction for the Dual NOR Gate components. It turns out the place is in China, and the parts should be arriving shortly. Shipping and transaction fees brought the price to $10 per chip. That is actually not that bad, as the RMS CONVERTER chip cost $13 itself.

Too bad I could not get these from a USA source. I found the chips at 3 USA sources but all 3 businesses were holding the chips for ransom to the tune of $250 to $500.

Testing the old components removed from my re-furbished board has shed insight on the initial failure of the board 9 years ago. I'll know for sure once the new chips arrive and will post the answer after I double check things.

ic-racer · Apr 28, 2009

Ok, an update here. I ordered three of the TD62479P chips from China. These chips have the same logic function as the original, but for some reason (input or output impedance) they did not work.

On a whim I called Omega/Satter and I was very surprised to find that they actually DO carry the replacement UDN-5714Ms chip for the D5500! What an odd-ball thing, as they don't have any other D5500 parts (besides the rubber roller wheels).

I ordered the chips and they should be arriving shortly.

ic-racer · Apr 29, 2009

Conclusion:

I put the 3 new UDN5714m chips in the board and .... it works perfectly! So, as Murphy would have predicted, the 'most difficult to locate' chip was the culprit! This finding does lend some closure to the fact that at the start of the project, the only part on the board with visible damage was the burned resistor that carried current to the UDN5714m chips.

Where to go from here? Well, in the never-ending-circus-ride-of-perfection, I see that the lamp voltage in this newly re-built board is absolutely rock solid. On the other boards, the one-tenths digit of the voltmeter jumps around a little when setting the lamp voltage. I ordered multiples of many of the parts, so I think I am going to go ahead and re-build another of my 4 boards. Probably the one with the least-steady lamp voltage.

ic-racer · Dec 15, 2009

OK, so that first board that I did has been in service for about a year and is working fine. I had planned on doing this total re-build to two boards and I just completed the component removal of the second board yesterday.

So, the goal is to have a 'brand new' power supply in each of my main heads.

Of course realizing that this type of rebuild is a little bit of an 'odd-ball' thing. Usually the defective component would be located and replaced. In this case, though, since the power supply lives up in the head with the 250W light, I wanted to replace ALL components on the PC board, so that they would likely outlive me. Also, the cost to replace every component (except the transformers) is only about $35.

Component removal went much faster this second time. I had no intention of saving any of the ICs or other components as they are all currently available new or NOS from USA suppliers. I just clipped the legs, removed the component, then de-soldered the legs. This allowed me to maintain the integrety of nearly all the 'through-hole' copper traces that connect the top and bottom of the board.

ic-racer · Dec 25, 2009

I powered up the rebuilt power supply board today and it works perfect.

Closing thoughts on the D5500 are that one of the most important things to do to ensure stability of the closed loop is to make sure the contacts between the mixing box and the power supply are clean. I have been using Deoxit with good results.

The signal carried through that connector is very low voltage and is easily destroyed or corrupted by just a little oxidation on the contacts. The symptom is that the motors keep 'hunting' and moving the filters slowly during the exposure.

ic-racer · Mar 13, 2022

Component lists, re-uploaded.

ic-racer · Mar 13, 2022

Pre and post pictures.

ic-racer · Mar 13, 2022

Board stripped of old components, ready for new.

albada · Mar 13, 2022

Based on what I've read in this old thread, I suspect that you are (or will be) the only person in the world with a working D5500.

ic-racer · Mar 15, 2022

I think an Arduno based controller for the three servo motors in the D5500 head would be no more complex than that required to control 3 batteries of colored leds that so many people have already designed.

albada · Mar 16, 2022

For mono work, such an Arduino controller could have a "Grade" knob on it letting you set grade directly.
I'm trying to tempt you.

D5500 Power Supply--Complete Rebuild From Scratch

Attachments

Privacy & Transparency

Privacy & Transparency