SSlmixed.com Forum
Tech discussion => 4K,6K,8K Series => Topic started by: Cavern Sound Sudios on January 20, 2023, 01:29:42 PM
-
SSL 4000 E/G
My Total Recall Logic Card CF82E88 keeps burning out the 10R resister (R6)
When the S9E ( the 10 way ribbon that goes from the 88 card and daisy chains from bucket to bucket ) and the B3 ( 6 way square on the 88 card ) are both connected the voltage across the resistor (when all buckets and the master section are powered up) is 1.18v and everything works fine
When all the buckets are off and only the master section is powered up the voltage across the R6 reverses polarity and measures at -3.3v and the resistor burns out.
This scenario is repeated even when all the channels are out of the console and all the cards are removed from the master section
-
If you have the wider keyboard, the Recall driver circuitry is ached to the bottom of the keyboard PCB which maybe where your problem is. If it is the shorter keyboard, the Recall driver circuit is on a small circuit board under the center section. Maybe an opto isolator?
-
There should be negligible voltage across that resistor, calcs indicate you have more than 100mA flowing in the 1 volt condition, and 300 mA in the 3 volt condition. Does this happen on every card in every module? Sounds to me like you have a short there or a partial short. Pull the card and set it up on bench supplies.
-
Thanks guys
Only happens on the 82E88 card,
I have it on the bench now, I will put 18V DC into it and see what i get across that resistor
Standby
-
18.5 volts into the 88 card on the bench and i get 1.3 volts across the resistor
when its In the console i get 5.4V across it
-
Every time i drop a channel in to a bucket ( with the bucket switched off) the voltage across that resister rises by 0.08v
So that times 56 channels = 4.48v across that resistor which burns it out, incidentally when i turn all the buckets on with all the channels in the voltage drops to 0.7 and everything is fine, everything on the console has been removed and disconnected except the 88 card and 1 channel
Any ideas where the short might be ?
-
I dont have that schematic here, but it sounds like you are somehow backfeeding all the modules. I take it this is a single card, not the one fitted in each module? Look for a bussing path from the card to all modules. Also, it seems like the card is drawing a lot, more than 100 mA by itself. This could indicate a fault on the card.
-
All 56 multiplex cards on all 56 channels gives me the same result across the resistor on the 88 total recall card
-
OK, so this is one master card on the total recall. Read my email again and understand what I am saying.
-
Yes 1 x total recall master card
I have replaced all of the CN17 and 4016 chips on the 88 card, I will replace all
the transistors today
-
I think you are approaching it the wrong way looking at the voltage across the resistor alone.
There is a high chance there is a short somewhere between the +18V and +11V rails. The TR along S9E works on these two logic levels.
This issue could be in my experience one of a few causes.
A blown BC212 transistor giving a dead short across it. Have they been replaced, and if so, ensure that if they have been replaced with BC212B, in lieu of the original and unavailable BC212L transistors, that they are rotated 120° CCW to match the correct pinout to board layout. Common mistake.
One of the 10way S9E ribbon cables has a short in it. check all links between boards. remove sequentially and see if this resolves it. As I understand, your issue is only on the left hand side of the console (1-24).
There is a short due to some foreign matter in the back plane edge connector. check closely.
Some photos might assist.
Anthony.
-
Thanks mate
A blown BC212 transistor giving a dead short across it. Have they been replaced, and if so, ensure that if they have been replaced with BC212B, in lieu of the original and unavailable BC212L transistors, that they are rotated 120° CCW to match the correct pinout to board layout. Common mistake.
This i haven't checked, I will look at that today
One of the 10way S9E ribbon cables has a short in it. check all links between boards. remove sequentially and see if this resolves it. As I understand, your issue is only on the left hand side of the console (1-24).
Issue is on both sides master ( CH 1 -56) , i have replaced and moved around all of the S9E cables, including the Y split coming from the 88 card
-
This is what you need for the BC212B transistors.
(https://static.wixstatic.com/media/bb0400_3c5c6a69c6c249878cd056d39c4eb5a8~mv2.jpg)
(https://static.wixstatic.com/media/bb0400_63341ed699a640b3819256fc6a83d3af~mv2.jpg)
Anthony
-
We really need a mod so the multiplex cards on each channel receive their 11v when the master section is powered up regardless of the buckets being switched on or off
There is no short on the console
The 88 card is functioning as designed
I will have to replace the 10ohm resistor on the 88 card with this
https://www.jaycar.com.au/rxe075-ptc-fuses-speaker-protection/p/RN3460
I have a 0.08v rise in the voltage across R6 for every channel that is inserted when the buckets are switched off
-
Surely the multiplex cards should get their 11v from the module they are on?
If not, mod them so they do.
-
For those of us following along at home, it might be good to know the function of R6; i.e. on what rail. I don't know about the '88 card, but on the G series '323 card, there are 1K pull-down resistors that will hold each TR line in S9 to 11 volt logic low, regardless of the ribbon being connected to the buckets. I trust all the voltages are present on the BICC and directly on the CMOS chips?
-
They do, but when the bucket is switched off and the 11v isn’t being supplied the multiplex cards pull too much from the 88 card, do you know if the 11v is supplied to
all of your buckets when only the master section is switched on ?
-
When you switch a bucket off, all supplies are removed.
-
Thanks i really appreciate the feedback, please shout out any suggestions
The 11v rail in each bucket is only present when the buckets are switched on ( thanks for that clarification)
The R6 is located on the Total Recall Card and it is a 10Ω fuse for the +18V supply to the 88 card, it provides positive power to the T1 & T2 chip
NOTES:
The 88 card receives its power directly from the PSU passing through the master toggle On/Off switch
(There is is a direct cable to the 88 card from the master section toggle ON/OFFF switch)
The card has +18 -18 11 & Ground via a black 6 way connector
I can completely remove my master section from the console and every bucket PCB and the 88 card still receives the +18 -18 and the 11v
When the S9E ribbon cable that feeds the 88 card to the buckets (daisy chaining from bucket to bucket) is unplugged or plugged into the 88 card everything is fine without any channels in the console.
When the S9E is plugged in and all channels are fitted into the buckets everything is fine as long as the bucket power is on and the Multiplex cards are powered up
When the S9E is plugged in as you add each channel to a bucket ( bucket power switched off) the voltage across the R9 increases by 0.08v ( 56 channels x 4.48v )
Two experienced technicians and myself have all checked the 88 card and it seems to be working fine, all the chips have been replaced and all the transmitters tested, everything passed successfully as working
If i remove the 88 card and a bucket PCB from the console and bench test i i get the same results
I have also tested it with another persons channel strip and also with someone else's multiplex card
I will replace all of the transistors on the 88 card with the ones Anthony suggested and then the card has just about been completely rebuilt.
My total recall card i have is CF82E88 and my multiplex cards are CF82E03
I could mod the 88 card by replacing the R6 with https://www.jaycar.com.au/rxe075-ptc-fuses-speaker-protection/p/RN3460
I really need another CF82E88 Total Recall card to test against
In shot - The multiplex cards are causing the 88 card to draw too much voltage across the R6 whenever the multiplex cards are not powered up
I need a coffee :)
-
The thing to do is establish what pin on the ribbon is pulling the current. You could do this by studying the schematics, or by making a little adapter by the master card which allows you to disconnect the wires one by one. Easy to make with a couple of sockets, some wire, and a particular type of prototype pcb. You can get a prototype pcb that is laid out as two interlocking grids which bus every second pin hole. This allows you to put two headers on and all the same pins get connected. Then, in the middle of this, you could put in a bunch of links. You can then disconnect one or all pins at will. And its only a 10 pin header. Start by disconnecting aal the obvious suspect pins, and then connect one at a time. Once you know which line is drawing the current, you can probably modify each multiplex card. Some work to be done!! Send the bill to SSL for not designing it properly!
-
Update on this issue..
The S9E cable sends the address line logic down to the module back planes, utilising logic low (+11V), and logic high (+18V).
The reference voltage for logic low is not sent down the S9E line but is instead picked up off the +11V from each module / bucket.
When the bucket is turned off, the logic low is floating, creating a 55mA draw on the 88 card.
When six buckets are turned off, that adds up to 330mA, equating to over 1W on the 10ohm fusing resistor.
I tested it on my console, and the 88 card's fusing resistor started heating up, pulling 330mA.
I'd never noticed this before as I keep all the buckets on all the time.
Anthony.
-
Can you post schematics of the two cards? Presumably the address lines go into logic, probably cmos? Maybe you could put diodes on the lines or some resistance?
-
The address line are pulled down to +11V at the CF82E88 card end via a 1K resistor , then go through a current limiting resistor (15K) at the CF82E03 multiplexer end, and in to some 4000 series logic, also running at the same logic levels (+12V / +18V).
I was thinking sending +11V down the spare pin 10 of the S9E, which goes direct to pin 10 of the S21E in to the multiplexer card, and then either a pull down resistor on each line or diode as you say.
Hassle is that you would have to do that on every card in the console....or just keep all the buckets on. Not a big deal when a 48 channel pulls less than 500W using a switchmode.
Anthony.
(https://static.wixstatic.com/media/bb0400_ecef238b58e045fc884ed689f8828f74~mv2.jpg)
(https://static.wixstatic.com/media/bb0400_c401001ebca840a7ab6c166bb50058d8~mv2.jpg)
-
Two things. Theres a set of resistors shown on the input to the multiplexer card, change this value up? Or second, just put a switch on the driver card to switch the 18 volts off? Or put a current limiter (2 transistors) in series with the 10 ohm, set above the normal operating current. That one would be set and forget, and only 4 parts!
-
Looking deeper, it appears the G series Total Recall Bus Driver addressed this issue by using a 1.2ohm thermistor in place of the 10ohm fusing resistor.
https://au.element14.com/epcos/b59910c0130a070/ptc-thermistor-1-2ohm-80v-th/dp/3407757?ost=b59910c0130a070 (https://au.element14.com/epcos/b59910c0130a070/ptc-thermistor-1-2ohm-80v-th/dp/3407757?ost=b59910c0130a070)
Anthony.