Hi John
I should have made it clear that on my PLL all three were way out, and I had never touched them. All three times were too short. I have noticed this on SP-10 MKII's as well.
I don't understand the ageing process though, but age they do and in my view it is an essential adjustment on the II/III.
It would be good to know if yours were out and in which direction ?
Thanks
Dave
I should have made it clear that on my PLL all three were way out, and I had never touched them. All three times were too short. I have noticed this on SP-10 MKII's as well.
I don't understand the ageing process though, but age they do and in my view it is an essential adjustment on the II/III.
It would be good to know if yours were out and in which direction ?
Thanks
Dave
I've no idea as I tried to set VS voltage first, which effects the sync position settings.
I didn't have any luck with the offset voltage. I made a divider to get 40mV from my bench supply and wired it in to the EC and ECR lines using the same polarity as their temp solution using the built-in 12V supply. The motor doesn't run (I don't know if it's supposed to) when you hit start, and I read 0mV across 32V and VCS on CN102. I noticed on the MK2A you effectively disconnect EC and ECR from the control circuit when doing this, so I tried that as well - also 0mV. In either case moving the pot didn't do anything. After this I measured the output of IC104 at pin 7 - moving the pot does, of course, adjust this voltage. I strongly recommend anyone working on one to measure the voltage here referenced to 0V before adjusting VR101 so you know what the start point was.
The schematic shows the running voltage at pin 7 of IC104 as 11.8V, so I set it to that, and re-set the sync positions for all speeds. After doing so I noticed that the platter would over-shoot on speed changes and take a second or two to drop back and lock. The same behavior was seen by applying significant drag with my finer and releasing it quickly. Sometimes it'd overshoot and "bounce" against lock and drop back again and then lock.
I adjusted VR101 up a bit and readjusted sync which improved it, so I iterated through this process a few times until it stopped bouncing. I ended up with 11.95V on IC104 pin 7 at the end, and I had to reset middle voltage when done. I measured all the other operating voltages on the schematic and measured waveforms and they were all close enough that I'm not concerned.
The good news is everything appears to be running fine and I can no longer feel any odd vibrations. At the very least, I'm confident that all the circuits are working as they're supposed to and anything else that may need done is minor, most likely adjustments. I'd love to know how to properly set the offset voltage, and get confirmation that the way I set VS voltage is correct.
The other bit that needs addressed is Q260 on the control board. This regulates a 20V line to 11.5V that runs the strobe LEDs and brake solenoids. While the platter is spinning it's no problem, but with it stopped the wedge brake is engaged and Q260 gets extremely hot in a hurry. The only thing to check is the diode across the solenoid, which I doubt is bad. Seems like this may be a design flaw. I've seen pictures on amp8's site where most have had this transistor relocated to the heatsink. I may do that later on.
Also, I noticed a pulsing noise on the position detection return lines when the table wasn't running that went away when I disconnected CN202, which are the lines for the LEDs and solenoids. I can't tell if the noise is present with the table running, but the current draw would be very small without the solenoids engaged.
Last, one of the drive transistor screws was stripped out at the heatsink, so I had to use an almost-too-short screw and thin nut. Luckily I did't have to drill out the cavity in the main heatsink for this modification to clear.
Currently everything is buttoned back up and I have it spinning away at 78 to bed the thrust bearing in as it's running on a new surface. This is about all I can do until I can get it in a plinth so I can measure wow and flutter. It's been one hell of a learning experience.
I didn't have any luck with the offset voltage. I made a divider to get 40mV from my bench supply and wired it in to the EC and ECR lines using the same polarity as their temp solution using the built-in 12V supply. The motor doesn't run (I don't know if it's supposed to) when you hit start, and I read 0mV across 32V and VCS on CN102. I noticed on the MK2A you effectively disconnect EC and ECR from the control circuit when doing this, so I tried that as well - also 0mV. In either case moving the pot didn't do anything. After this I measured the output of IC104 at pin 7 - moving the pot does, of course, adjust this voltage. I strongly recommend anyone working on one to measure the voltage here referenced to 0V before adjusting VR101 so you know what the start point was.
The schematic shows the running voltage at pin 7 of IC104 as 11.8V, so I set it to that, and re-set the sync positions for all speeds. After doing so I noticed that the platter would over-shoot on speed changes and take a second or two to drop back and lock. The same behavior was seen by applying significant drag with my finer and releasing it quickly. Sometimes it'd overshoot and "bounce" against lock and drop back again and then lock.
I adjusted VR101 up a bit and readjusted sync which improved it, so I iterated through this process a few times until it stopped bouncing. I ended up with 11.95V on IC104 pin 7 at the end, and I had to reset middle voltage when done. I measured all the other operating voltages on the schematic and measured waveforms and they were all close enough that I'm not concerned.
The good news is everything appears to be running fine and I can no longer feel any odd vibrations. At the very least, I'm confident that all the circuits are working as they're supposed to and anything else that may need done is minor, most likely adjustments. I'd love to know how to properly set the offset voltage, and get confirmation that the way I set VS voltage is correct.
The other bit that needs addressed is Q260 on the control board. This regulates a 20V line to 11.5V that runs the strobe LEDs and brake solenoids. While the platter is spinning it's no problem, but with it stopped the wedge brake is engaged and Q260 gets extremely hot in a hurry. The only thing to check is the diode across the solenoid, which I doubt is bad. Seems like this may be a design flaw. I've seen pictures on amp8's site where most have had this transistor relocated to the heatsink. I may do that later on.
Also, I noticed a pulsing noise on the position detection return lines when the table wasn't running that went away when I disconnected CN202, which are the lines for the LEDs and solenoids. I can't tell if the noise is present with the table running, but the current draw would be very small without the solenoids engaged.
Last, one of the drive transistor screws was stripped out at the heatsink, so I had to use an almost-too-short screw and thin nut. Luckily I did't have to drill out the cavity in the main heatsink for this modification to clear.
Currently everything is buttoned back up and I have it spinning away at 78 to bed the thrust bearing in as it's running on a new surface. This is about all I can do until I can get it in a plinth so I can measure wow and flutter. It's been one hell of a learning experience.
Sometime between 5 and 6 hours of running today the problem is back. Not as bad, but I can feel sporadic corrections and see a little bit of movement in the strobe. I'll see what it does tomorrow, and I've a replacement for Q260 on the way in case the noise from that 12V line is causing issues.
I'd really like to know how to set the offset voltage properly - I think that's part of the issue.
I'd really like to know how to set the offset voltage properly - I think that's part of the issue.
In regards to offset voltage, I think I figured out the issue. I believe EC and ECR lines should be disconnected from the control board per the MK2A manual which occurs when the test connector is removed which disconnects EC and ECR. The MK3 manual tells you to connect 0V and 12V, though if you're using an external supply for the 40mV only 0V needs to be connected. That's where it stops, but I think the issue is the STP1 line. With the motor on, this line should be 6V, and is the only line between control and drive that tells drive it should... drive. If the wife is working tomorrow I'll give it a try.
My replacement for Q260 arrived today as well, so I'll replace that and move it to the heatsink.
Dave,
If you get yours back on the bench, I'd love to know the following:
1) AN660 pin 13 referenced to 0V - voltage for 33, 45 and 78.
2) AN660 pin 7 referenced to 0V - voltage for stop, and 33.
My replacement for Q260 arrived today as well, so I'll replace that and move it to the heatsink.
Dave,
If you get yours back on the bench, I'd love to know the following:
1) AN660 pin 13 referenced to 0V - voltage for 33, 45 and 78.
2) AN660 pin 7 referenced to 0V - voltage for stop, and 33.
Hi John
It's Church this morning then an afternoon out. But Monday I'm free (I think)
Yesterday was chainsaw day, a once a year treat to tidy the garden !
However I have been thinking about this. When I first received the MK3 it was almost as new and the bearing wear confirmed this. There was a slight noise on 78rpm and I put this down to the wedge brake that was arguably just rubbing. Seemed to have fixed it, at the time...........
However having rebuilt the PSU the PLL was out, so adjusted this and VR101 for stability. Everything is fine, except at 78rpm the same very slight juddery noise is occasionally there, just like when it arrived. I now know this is an issue and the correct set-up needs to be established.
I'll try to get the voltages requested later today. I think though that I need to sit down and analyse the circuit properly, clearly the official service manual is not right.
.
It's Church this morning then an afternoon out. But Monday I'm free (I think)
Yesterday was chainsaw day, a once a year treat to tidy the garden !
However I have been thinking about this. When I first received the MK3 it was almost as new and the bearing wear confirmed this. There was a slight noise on 78rpm and I put this down to the wedge brake that was arguably just rubbing. Seemed to have fixed it, at the time...........
However having rebuilt the PSU the PLL was out, so adjusted this and VR101 for stability. Everything is fine, except at 78rpm the same very slight juddery noise is occasionally there, just like when it arrived. I now know this is an issue and the correct set-up needs to be established.
I'll try to get the voltages requested later today. I think though that I need to sit down and analyse the circuit properly, clearly the official service manual is not right.
.
Last edited:
AN640G
The concern is that it's stopped working, woops !
I bought 4 on eBay, it says Panasonic and they often list them AN640G Integrated Circuit | eBay
The concern is that it's stopped working, woops ! I bought 4 on eBay, it says Panasonic and they often list them AN640G Integrated Circuit | eBay
Did ya happen to measure some voltages before you blew it up?
My idea for offset voltage didn't work - still 0. I can get a measurement with it running and moving VR101 does affect it, but it settles back to a certain voltage per speed. Too low and the voltage difference oscillates, too high and the motor starts fighting itself.
The circuit and measurement points are similar to middle voltage which I can set.
My idea for offset voltage didn't work - still 0. I can get a measurement with it running and moving VR101 does affect it, but it settles back to a certain voltage per speed. Too low and the voltage difference oscillates, too high and the motor starts fighting itself.
The circuit and measurement points are similar to middle voltage which I can set.
Last edited:
I tried to reset offset voltage again, empirically. If VR101 is set too low or high it'll lose lock observing ST2. Like what I did watching the strobe last time, I did it this time watching ST2. I edged it up just to the point where the notch in ST2 would snap back after retarding the patter without overshoot. However, this is damn close to the point where the motor will start fighting itself.
I may break down and fire up my NIB MK2A. It's still in the factory bags untouched, so I could get some good measurements from and unused, and untouched table. Electrolytics that have sat for over 30 years is a bit of a concern, and I really don't want open it, but it may be the only way we can measure this circuit as it's supposed to be.
I may break down and fire up my NIB MK2A. It's still in the factory bags untouched, so I could get some good measurements from and unused, and untouched table. Electrolytics that have sat for over 30 years is a bit of a concern, and I really don't want open it, but it may be the only way we can measure this circuit as it's supposed to be.
Last edited:
Okay....
6L6 and I spent a couple hours on the phone going over this whole offset voltage setting. Jim concluded that R147 is current sense, Q113 and Q114 are voltage shift, on through the opamp to the current sense line of the AN640G. To me that means we're calibrating the current to the motor from the P transistors for a given error voltage. It also mean the platter has to be spinning for this to work.
Measuring the table on at 33, EC and ECR are ~7V with a voltage difference between them of ~.22mV. Based on that it seemed unlikely that disconnecting EC and ECR from control and putting 40mV on the would do anything. So, I tried again, leaving all the lines hooked up and injecting 40mV in to circuit, as I had tried before.
To my utter amazement, though sluggishly, the platter began to spin. VR101 is a VERY touchy setting and requires a lot of settling time, and I noticed that a little bit low or high and the platter won't move on its own. It's most likely this is why it didn't work for me before.
I let the platter spin up and everything stabilize before attempting to adjust the control. This took a good five minutes, and you have to be patient as it will over or under shoot by 1-2mV and then come back very slowly. In a nutshell each tweak of the pot takes a few minutes to make sure it stabilized. And, again, crazy sensitive adjustment.
It took a good half an hour to get it set at around 10mV +/- 0.25mV and verified over two power cycles. I also noticed when running at 33 the last go around I had ~8.5mV across 32V and VCS, and after setting the offset I had ~7.5mV. The ~ is because it moves around a good +/- 0.5mV. We're back to no vibrations, all speeds good, and quick lock during speed changes and quick release of load from the motor.
I've drilled and tapped the heat sink to relocate Q260, so I'll do that now, and then button it up again for another burn-in round tomorrow. Fingers crossed.
6L6 and I spent a couple hours on the phone going over this whole offset voltage setting. Jim concluded that R147 is current sense, Q113 and Q114 are voltage shift, on through the opamp to the current sense line of the AN640G. To me that means we're calibrating the current to the motor from the P transistors for a given error voltage. It also mean the platter has to be spinning for this to work.
Measuring the table on at 33, EC and ECR are ~7V with a voltage difference between them of ~.22mV. Based on that it seemed unlikely that disconnecting EC and ECR from control and putting 40mV on the would do anything. So, I tried again, leaving all the lines hooked up and injecting 40mV in to circuit, as I had tried before.
To my utter amazement, though sluggishly, the platter began to spin. VR101 is a VERY touchy setting and requires a lot of settling time, and I noticed that a little bit low or high and the platter won't move on its own. It's most likely this is why it didn't work for me before.
I let the platter spin up and everything stabilize before attempting to adjust the control. This took a good five minutes, and you have to be patient as it will over or under shoot by 1-2mV and then come back very slowly. In a nutshell each tweak of the pot takes a few minutes to make sure it stabilized. And, again, crazy sensitive adjustment.
It took a good half an hour to get it set at around 10mV +/- 0.25mV and verified over two power cycles. I also noticed when running at 33 the last go around I had ~8.5mV across 32V and VCS, and after setting the offset I had ~7.5mV. The ~ is because it moves around a good +/- 0.5mV. We're back to no vibrations, all speeds good, and quick lock during speed changes and quick release of load from the motor.
I've drilled and tapped the heat sink to relocate Q260, so I'll do that now, and then button it up again for another burn-in round tomorrow. Fingers crossed.
.
If we take a step back, both yours and mine suffered(s) the same issue. Mine both before and after changing the PCU capacitors. I don't think this is a coincidence at all.
I'm waiting to do some dynamic checks and this time will make sure there are no invisible small metal screwdrivers underneath the control PCB that was flattened out onto the bench !
R147 is definitely a current sense, but why ? What part of the PLL does the CS line perform ? We aren't looking at current or power, we should be looking at speed and phase. However for a separate issue, I wonder what the waveform and frequency is on R147 when the turntable is locked at 33rpm ?
With the PLL locked, I would expect EC (3 sync pots & 3 relays, incorrectly marked as a duplicate ECR on page 24) and ECR (phase control pin 7 IC203) to be stable, and hence if unplugged would be out of lock and possibly all over the place ?
I did say previously that VR101 (post #39) was very critical, and it is, but why ?
I am struggling to see how the MK3 works. As you know I have successfully duplicated the MKII completely, with modern "easy to get for ever" components and a big PIC to do all the logic. I fully understand the MKII, but the MK3 defeats my understanding.
.
If we take a step back, both yours and mine suffered(s) the same issue. Mine both before and after changing the PCU capacitors. I don't think this is a coincidence at all.
I'm waiting to do some dynamic checks and this time will make sure there are no invisible small metal screwdrivers underneath the control PCB that was flattened out onto the bench !
R147 is definitely a current sense, but why ? What part of the PLL does the CS line perform ? We aren't looking at current or power, we should be looking at speed and phase. However for a separate issue, I wonder what the waveform and frequency is on R147 when the turntable is locked at 33rpm ?
With the PLL locked, I would expect EC (3 sync pots & 3 relays, incorrectly marked as a duplicate ECR on page 24) and ECR (phase control pin 7 IC203) to be stable, and hence if unplugged would be out of lock and possibly all over the place ?
I did say previously that VR101 (post #39) was very critical, and it is, but why ?
I am struggling to see how the MK3 works. As you know I have successfully duplicated the MKII completely, with modern "easy to get for ever" components and a big PIC to do all the logic. I fully understand the MKII, but the MK3 defeats my understanding.
.
Last edited: