Studer B67 fix explained

I explained two posts ago that the problem with this tape machine was IC 3, an SN75462 that was put on backwards and therefore fried. That IC was responsible for interpreting logic from the transport and operating the tape sensors’ locking solenoids. Here’s the data sheet for the SN75462, and I can’t say I understand it all, but the gist of it is that it has two NAND (or AND) gates followed by an NPN transistor whose collector is taken to a pin on the actual chip, so it’s not necessarily connected to Vcc powering the chip. This is called an open collector, and here’s a very good explanation of what it is, and what are some applications. Here’s a little drawing of what this looks like in this particular case (can also be seen in the data sheet):

open collector

 

Here’s the portion of the schematic that shows IC3 in the circuit, the anode of the zener (D59 in this case, but all of them do) goes to ground.

IC 3 circuit

Now, when I was measuring the voltage at pins 3 or 5 of IC3 (after it’s been replaced with a functional one), what I saw was that it was “delivering” +24V when it was supposed to deliver +5V (Vcc), and that was very confusing because how could it put out such a high voltage when Vcc is 5V? Not only that, the +24V was identical to what I was seeing on the other side of the relay (that should have tipped me off..)

The answer to this is the open collector but also a KVL of the circuit. I was thrown off before because someone mentioned off-hand that the reverse diode pulls up the voltage, and that sent me on the wrong path for a while. I finally understood what’s going on when I drew the NAND gate + the transistor + diode + relay together:

open collector + relay

Here’s how I understand it. When the transistor is ON, the collector is pulled to ground through the transistor, so current flows through the relay. When the transistor is OFF, no current flows through it so the collector is free to being pulled up. What does it mean, though? KVL shows it. There’s +24V on one end of the relay, but no current flows through it (the transistor is OFF and the zener is reverse biased), so there’s +24V on its other end. That other end is the collector, and that’s why I was seeing the exact same voltage on both ends of the relay.

Now, one might ask why use the zener at all? Pull it out of the circuit and you get the same behavior. However, when you open the switch connected to a relay, the relay shoots out a spike of high voltage. Without another path to ground, that voltage will fall on the transistor and burn it. Instead, the zener starts conducting as soon as the voltage across it is higher than 30V. A regular diode reverse biased won’t work because it won’t recover from being pushed to its breakdown region. And of course, a forward biased diode will always conduct so current will always flow through the relay.

Studer B67 Mk II problem fixed*!

(*almost)

I was in the middle of writing a progress post and then I went to probe around IC3 in the transport and realized that IC3, was put in the other way around. That is, pin 1 is where pin 8 should be, etc. I flipped it around and it is now fixed! Here’s a more detailed explanation of what I’ve done in the past few weeks:

First a recap of the problem: Upon turning the machine ON, only one motor will be responsive to the transport controls. For instance, pressing PLAY would make the take up motor spin but not the supply. Then if I spun the roller on the right clockwise, the take up motor would turn off and the supply would turn on and start spinning. Spin the roller counterclockwise and now the take up would run and the supply motor would shut down. Even weirder was that disconnecting the J5 connector from the pre-divider board caused the machine to operate correctly.

So when I picked up troubleshooting the machine again I decided to go over the counter’s schematic to try understand better what’s happening with QP-DIR1 and QP-DIR2. To be honest, I’ve done that before, but this time I also graphed the waveforms along the way from where the signals come into the board, and their way to becoming Y2-FORW and Y2-REVS. I also keep these drawings with the manual for future reference. That didn’t tell me much other than that when the roller on the right is spinning clockwise, Y2-FORW goes HIGH and when it’s spinning counterclockwise it’s going LOW and Y2-REVS goes HIGH. I also realized (might have before as well) that the pre-divider carries these signals FROM the counter TO the transport.

Advice on the tapeheads.net reel-to-reel forum directed me to the tension sensors, so I started learning the circuit comprised of ICs 4, 1, 6, and 2. That made sense to me because that circuit is responsible for generating the pulsating signal that controls the motors. My understanding of how IC4 is used is that it is some kind of a comparator. So depending on YAN-TT1 and YAN-TT2 it’ll go positive or negative. That’s how the lower half of IC4 (YAN-TT2) was acting and that made sense. However, the voltages from YAN-TT1 weren’t enough to cause the comparator to work the right way. Instead it went from +12V to something like -2V. (By the way, that’s still how it is!) I thought that since this voltage is dependent on the displacing of the tension arm, that maybe its mechanical settings are out of wack.

So I took out the left sensor and adjusted it per the manual. I put it back in the machine and now both left and right sensors were locking in place. This one made me scratch my head, but it led me to start figuring out what’s controlling the solenoids of the two sensors. This stuff isn’t mentioned in the manual, so it took a bunch of probing and continuity tests to realize that the solenoids are fed the unregulated +24V and the respective outputs of IC3 (SN75462). I mentioned that to a friend and he said that IC3 goes bad often, so I put in an order for replacements. Meanwhile I went to check the voltages at the inputs and outputs of IC3 to see if it’s working right and that’s when I realized that it was put on backwards. I flipped it and now not only are the sensors not locked, but BOTH motors are responsive to the transport controls!

However, the sensors should lock when the machine is stopped. A quick check of the voltages in and out of IC3 showed that it’s not functioning right – it’s a NAND gate but when both inputs voltages are high (IC7 pin 7 is HIGH meaning the machine is stopped, and then a signal derived from pin 7 also HIGH) I get +24V at its output, but it should be 0V. It’s a good thing I ordered some SN75462 so I’ll drop a replacement in and see what’s up.

Update on the Studer B67.

I believe I mentioned before that I isolated the problem to the counter board. I don’t know anymore if that’s true, but by probing around it, I realized that the flip-flop comprised of IC3 isn’t flipping or flopping. From all the probing, my understanding of the schematic, and the help of online friends I realized that the tape sensor is in charge of of activating the flip-flop. When the roller turns one way the flip-flop latches in one position, and when it turns the other it latches in the other position.

The manual states that QP-DIR1 and QP-DIR2 are approximately 90 degrees apart, and I realized that depending on the direction of rotation, either signal should be leading. What I was seeing instead is that QP-DIR1 was always leading.

I pulled out the sensor board and powered it with my bench supply. I got 0V  at QP-DIR2 (R13), which is weird because I was seeing it at the counter board. So I went probing around the phototransistor (DLQ2). At the collector the voltage was swinging between 4V and 5V, DLQ1 was reading similarly.

DLQ2’s emitter is connected to ground through a trimpot and also to the base of Q5. So I checked the measurement at the base. I was seeing 0V. I tested the trimpot and it was basically shorted to ground. I figured it’s bad so I pulled it out. I tested it outside the circuit and it worked fine. I put it back in, and same thing. I paused and thought for a bit, and realized that it’s possible that the base-emitter junction is short and that’s why the trimpot was reading 0 ohms. I pulled the transistor out and bingo! That junction is short (as well as the collector-base junction). Surprisingly, I have a few BC237Bs on hand, so I put one in.

Since I had the board out, I replaced C2, a 10uF 25V Frako. Since it’ll go short one day, now is a good time to replace it (and I had a replacement).

I also cleaned the board and the weird yellow residue that was on the roller, so now it’s ready for the new rubber ring I got a few months ago.

I installed the sensor back in the machine and now the flip-flop works as intended. With the roller turning CCW IC3 pin 6 is HIGH and pin 3 is LOW. Turning it clockwise makes IC3 pin 6 go LOW and pin 3 HIGH. Good.

However, the problem of just one motor spinning isn’t gone yet. Now what happens is that one motor spins when I pressed PLAY/FF/RW, but turning the roller in the appropriate direction turns on the other motor and turns off the first. So for instance, if the supply is working and I turn the roller clockwise, the take up will start running and the supply will slow down to a stop. Similarly if the take up is running and I turn the roller CCW. In this case the supply will start running and the take up will slow down.

The trimpots (R10 and R19) control the duty cycle! I have to figure out how to adjust them to get 50% duty cycle.

Studer B67 recap: Part VIII

I just cleaned and recapped the counter board. The underside was filthy, so I thought maybe something got short because of the dirt.

I reinstalled it in the machine (after I finished putting the new caps in), turned it on, and it still remains that only one motor is responsive, but also the problem where spinning the roller causes the supply to stop working is back – but with the take up! So the take up is still responsive 80% of the power ons, but now if I spin the right hand roller by hand, that causes the take up to stop running and the supply to start spinning. When the supply is responsive, spinning the roller doesn’t make it stop. Huh.

Going to replace that chip now.

More Studer B67 work trying to figure out what’s going on.

I don’t remember what prompted me to do that, but I decided to turn the right hand roller guide by hand when the supply motor is the responsive one. Turns out, it turns off the supply motor and turns on the take up motor.

My friend instructed me to test the MOVE voltage at the right sensor PCB.

I taped both tape lifters in the UP position (to trick the machine that there’s there’s tape threaded, I believe). Then I checked the voltage between J1-2 and J1-1 in these different scenarios:

Pressing PLAY and turning the roller by hand:  +4.78 VDC.
Machine in STOP and turning roller by hand: +4.78 VDC.Machine in STOP and roller is still: +0.03 VDC.

FIY these three voltages are what should be expected (essentially +5VDC when there’s tape “moving” and 0 VDC when no tape is moving).

Then I turned to checking the voltages on IC8 on the transport board. This is a pain because I had to remove the transport again.

All voltages are with respect to pin 7 which is ground.

Pin 6 was constantly at -0.53V regardless of the machine being in PLAY, STOP, and the roller spinning or still.

In STOP mode, pin 4 measured 4.31 VDC when the roller was spinning and 0.12 VDC when the roller was still.

In PLAY it dips to 0.12 VDC and quickly goes up to 4.23 VDC (roller not spinning). Spinning the roller makes it go down to 0.12 VDC.

Still not sure what this all means. I’ll update when I do!

Studer B67 running/spinning motor problem – fixed!

Straight to the point: The casing of the take up motor’s power transistor was making contact with the top chassis. I uninstalled it off, and there was a piece of fuzz on one of the transistor’s legs going all the way up to the casing. I cleaned it up, reinstalled the transistor, tested that there’s no continuity between the transistor’s casing and the chassis, and turned the machine on. The motor wasn’t spinning anymore! Can’t say if it was the fuzz that caused it, but reseating the transistor fixed it.

So how did I arrive at that? I’m not yet fully proficient in electronics, so what’s following here might be incorrect, but my prodding and testing was guided by my friend who’s a Studer tech, so maybe not!

Anyway, in my last post I detailed the measurements I took on both motor control boards. The readings on the supply board are normal, the readings on the take up board weren’t, particularly, the voltage between pins 5 and 10. On the other hand, the voltage on pins 7 and 10 was within the tolerances of what would come from the transport, so that reduced the problem to a short that happens somewhere around the motor (as opposed to the transport). I started checking for continuity at the transistor socket, and I noticed something peculiar. On the take up board, I had perfect continuity between pins 5 and both the transistor’s casing and the chassis. Same thing with pin 6. Pin 5 is the collector and 6 is the emitter, and it didn’t make sense that they should be at the same voltage. Also, that’s not what I got for those pins on the supply motor’s board. My friend suggested to check if the casing of the transistor is touching the chassis, and that was it.

I should mention, however, that now there’s a different anomaly. I turned the machine on and followed the manual’s initial checkout procedure (p. 2/17, section 2.6). When I pressed play, and only the take up motor was responsive. I turned off the machine, tested some connections on the supply board, turned the machine on, and now the supply was responsive but not the take up. Turned it off again, tested some connections on the take up, turned it on, and now the take up is running and not the supply. What the hell? Hopefully this one will be easier to chase down.