Capture Memory and Power Supply confusion.

[eaaron]

So after almost a year of playing our new to us ADC-9300, two annoyances finally got to me last week and now I could use a hint or two.

Annoyance number one:
Since I setup this console I have had a blank stop draw-knob(would be Vibrato if used) in the Swell division that loses it's mind. After the organ is powered up for 10+ minutes the capture action solenoid starts to miss-fire. When hitting cancel or any general or swell div preset it will randomly pop out. Not a operational issue as it doesn't do anything(...yet...) but has been driving me nuts so I want to fix it. I pulled the draw-knob assembly out and figured out how the unit works. I am hoping it is just a bad transistor on the push side that starts leaking when it gets warm and causes the knob to pop when the power supply pulses. I planned on replacing both push and pull transistors, and would just replace the "flip-flop" IC for good measure but the IC has been discontinued and worrie that the IC might be the issue and not the transistors. Does anyone know what would be a direct replacement? The part is CD54HC74FX which is a custom of the CD54HC74F. The X just meant the factory cut off pin 1, guess Allen saved 5 seconds assembly doing that. I would guess I could use any CMOS Filp-flop with the same pinout and specs but if there is a know replacement I would prefer that.

The second annoyance:
This one I though was going to be much worse than it turns out. I have had an intermittent issue when recalling presets or using a reversible where the audio would stutter if used while playing a sustained chord. I thought this might be a capture supply issue as the pedalboard lights would dramatically dim when it would happen and thought the cages might be drooping due to current draw. After measuring every PSU output in the console I found that wasn't the case, actually I will say that this ADC's supplies are as rock solid as they come. Turns out the issue is with the Swell to Great coupler and Choir to Pedal coupler tabs, it seams that when the capture pulse happens and the tabs do there little jiggle those particular tabs are pulling in just enough to make the reed switch chatter, making it sound like the audio was stuttering as the voices would chatter along with it. Seeing as I have never taken an Allen tab out I was just looking for some advice as to best way to get at it. Also if anyone has had this issue in the past, what their solution was? New felt stops? Adjust the magnet position? or something else. The lamp issue was a bad connection at the male Edison connector plugged into the distro box.(Lights still do the Allen organ solenoid dim, but not as bad)

Lastly, for my own knowledge, In troubleshooting these I mapped out the entire capture system solenoid system. This 9300 has 2 capture power supplies. PS1 powers and pulses the Choir, Great, and Solo divisional draw-knobs, the 4th output is unused. PS2 powers and pulses the Pedal, Swell and tab rail, the 4th output is unused. From what I can tell each PSU has 4 outputs with individual triggers. The triggers are held high(+5VDC) and pulsed to ground to trigger the solenoid moves after the capture scanner runs it's matrix to setup the direction. From what I have found in my brief search of this forum, the capture board should only pulse the division that is in the recall. For example a general piston should pulse all outputs. A swell division piston should pulse the swell, PS2-2, and rail for couplers, PS2-3. And a reversible should just pulse the rail. Which makes sense to me as why pull power to all solenoids when you can target an area. However all the inputs on the power supplies are tied together. There are 2 wires coming off the USCP-2 board, one to each power supply, but they both fire on every piston no matter what division. Just wondering if I am understanding this correctly.

Thanks

-Erik

[Hamman]

Can't really answer your questions, However, jbird will probably be along in short order to read and give his ideas. Would like to see some pics of this organ though! :-)

[Admin]

Is this a two or four level capture system?

[eaaron]

Admin - This is a 4 level capture, USCP-2 board.

Hamman - There are some pics of my testing and install in the General Chat area. http://www.organforum.com/forums/sho...-years-without

-Erik

[Admin]

doing that. I would guess I could use any CMOS Filp-flop with the same pinout and specs but if there is a know replacement I would prefer that.

Lastly, for my own knowledge, In troubleshooting these I mapped out the entire capture system solenoid system. This 9300 has 2 capture power supplies. PS1 powers and pulses the Choir, Great, and Solo divisional draw-knobs, the 4th output is unused. PS2 powers and pulses the Pedal, Swell and tab rail, the 4th output is unused. From what I can tell each PSU has 4 outputs with individual triggers. The triggers are held high(+5VDC) and pulsed to ground to trigger the solenoid moves after the capture scanner runs it's matrix to setup the direction. From what I have found in my brief search of this forum, the capture board should only pulse the division that is in the recall. For example a general piston should pulse all outputs. A swell division piston should pulse the swell, PS2-2, and rail for couplers, PS2-3. And a reversible should just pulse the rail. Which makes sense to me as why pull power to all solenoids when you can target an area. However all the inputs on the power supplies are tied together. There are 2 wires coming off the USCP-2 board, one to each power supply, but they both fire on every piston no matter what division. Just wondering if I am understanding this correctly.

Thanks

-Erik

Since all the inputs are wired together, I suspect that Allen may have been repurposing the PSU they used for the prior DM system. There were indeed separate outputs from the DM to board for each division. There was logic that decided that it needed to pulse all outputs when two or more divisions were involved in the change. Couplers and backrail tabs could only be actuated from the General Pistons.

[jbird604]

The first thing to try on your coupler tab issue is to regulate the little felt stoppers that limit the up and down movement of the assembly. On the rocker tabs, if these are not properly adjusted, you can look down the row and see a tab out of alignment in either the on or off position, so you might be able to spot it right away visually.

The felt plugs must be rotated just like screws, as they are inserted into threaded holes for that purpose. Not an easy task, but it can be done. It should eliminate any jumpiness of the couplers.

[eaaron]

Thanks John, I will take a climb back there tomorrow night before rehearsal to try and adjust them.

As far as the capture PSU I was thinking there may be some oddities with this console being built at the end of the ADC era, and that it has 5 well endowed divisions. To be honest I have been thinking of attempting to create a capture memory system to add to what is there. I just need to map out how and where recall and storage happens, and how that is getting transmitted though the stop matrix. Plus add the bonus of converting stop recall to midi.:-> With 3 other organists rotating though regularly and only 4 memory levels, one of which doesn't work, it can be frustrating to not be able to let everyone have their own memory. Plus we had one visitor for a wedding that wiped all memories and put what he wantedX-(, some of which was down right bizarre, and I had to spend an hour between services re-programming.

[eaaron]

Ok so I got the two coupler tabs adjusted and no longer have the coupler dropping out when the capture pulses. :-) Those felts were the trick, the swell to great was completely out and the tab arm was hitting the back of the magnet which also explained the mysterious click from time to time, a bit finicky to get that one back in but not too bad.

On the drawknob.... that is a different story. After changing out the push/pull transistors it made it through 3/4 of a mass. Looks like the IC on the board is bad also. Once the chip heats up it goes wacky and just throws the stop whenever it wants. I put the board on a bench supply and after about 20 minutes the outputs controlling the transistors start bouncing wildly.

I cannot find a source for the 54HC74F chip as it is obsolete. (Well I found one source that wants $20 a piece! Yikes!) Doing some cross reference it looks like the 74HC74 has the same specs for operation voltages and switching, just a faster switching setup 21ns vs the 54's 25ns. This looks like it should work well, but I was wondering if anyone has had any experience with replacing these. It is only a $.52 experiment if it doesn't work. Unless someone has a red flag on this I plan on giving it a try.

-Erik

[jbird604]

From what little info I could glean about the 74xx vs the 54xx chip families, the only difference is in operating temperature range, with the 54xx series being specified for "extreme conditions" such as military applications. I have no idea why Allen would've used them in drawknobs. Online info also seems to indicate that today's 74xx chips have such excellent temperature characteristics that they are every bit as hardy as the 54xx chips of the past.

If the replacement chip has exactly the same functionality, identical pin-out and the same operating voltages, I don't see why it wouldn't work. The slight difference in switching time should not matter, especially down in the nanosecond range.

You really don't have much to lose, since the knob is unusable as it is, and it's a blank one anyway. As long as you do the soldering with care and make absolutely sure that you don't damage the little pc board or create any solder bridges, I don't see how you could hurt anything. And if it doesn't work, you've only lost 52 cents. You could then proceed to order that $20 part.

Disclaimer: any advice I give over the internet is worth exactly what it costs you, and liability is limited to the refund of said fee!

[eaaron]

I have replaced the Flip-Flop on the offending knob and all looks well. It was nice to work on a good old fashioned DIP chip as opposed to a SMD type. Don't need the microscope to check for shorts. :-B

The capture recall is working properly across the entire console. Between a wedding tomorrow and Sunday I will have a definite yea/nay as to the issue being resolved.

As a reference to anyone in the future there are 2 ICs available the CD74HC74N and the CD74HC74E. The CD74HC74E is the direct replacement to the CD54HC74F. The E designator is the 125 C version, the N designation the 85 C version. Other than the slightly faster response time they are identical to the 54HC74 and actually shared a datasheet from TI.

-Erik

[jbird604]

Thanks for that info. I don't do a lot of IC replacements any more, but I always found the mind-boggling array of similar-looking chip numbers very confusing. I know that one 74XX chip has the same functionality as another 74XX (where XX is the same two numbers), but the letters before, after, and between the numbers are Greek to me (actually Latin).

When I'm needing to replace an IC, I usually resort to seeking out whatever is available on Mouser with the same digits, then feverishly researching the prefixes and suffixes, trying to determine if my 74LSXXN can be replaced with some 74HCXXE or something like that. Most of the time the replacements that I can find on Mouser will work, but I can recall a time or two when they did not work, possibly due to some difference in voltage tolerances.

On certain 70's model Rodgers organs, in the "time-share" circuitry, there are IC's that are obsolete and NLA except from scalpers, and I have paid $30 or more for a single IC at times just to get the right part for the job. There could be modern-day equivalents for some of them, but I have been unable to determine this.

But I'm always glad to pick up new info, and this bit about the 54XX being replaceable by the 74XX"E" series is something that I will remember (as long as my memory cells keep on working).

[geoelectro]

The 74Cxx versions are even harder to find. I looked on eBay and found some 74C151 IC's that were shipped from China for only $2 shipping. I bought 15 of them. I wasn't able to find them in any "normal" supply channel.

Geo

[eaaron]

So after an excruciating long wedding today that knob is behaving and staying where it should.:-) I will keep an eye on it, but hopefully that was it. Now I just have to decide what to do with that drawknob:devil:, vibrato or divisional MIDI, or something more devious....



ICs and transistors, specifically in high gain power amplifiers, tend to be the hardest thing to source. The real trick I have found is OEMs order a chip with specific requirements in high bulk from an chip manufacturer that get's a custom part number. This is normally based on something similar the manufacturer has in their chip catalog already, just changing a rating, or switch speed, or the max Vcc voltage. After the bulk run, which typically holds the OEM for a year or two, the chip manufacture decides to create a new part with the better specs some OEM paid the R&D for, slap a new number and make the original base(and lots of times the OEM custom) chip obsolete so they only have to produce one chip style instead of 3.

From a business standpoint it is forward thinking. From a design standpoint it works as you have what you want with better specs, and from a replacement of old parts it works too, except for finding the cross reference. I have found that TI has a great online catalog of spec sheets, that most of the time will have overlapping, or series based data so if a chip has been improved with a different number, it can be traced. TI also has had a version of most RCA, ON semi-conductor and ITE chips, so you can work backwards and sideways to find a number.

geo - those 74Cxx series are a royal pain. If I remember the 151 is a multiplexer. That entire series has disappeared and everyone is making the same thing in a SMD package, and not the good old fashion DIP. We had to get one of those adapter cards that converts a SMD to through hole for an older keyboard at work. Took us a month to figure out the exact part though.