Bottlehead Forum
Bottlehead Kits => Legacy Kit Products => Paramount => Topic started by: Roger on March 03, 2012, 02:17:11 AM
-
Hi,
I've just finished soldering my Paramounts. All resistance measurements taken and found to be ok. Plugged one of the amps to AC, and took the voltage measurements, and everything was ok (284 volts on pin B6 after reg had kicked in and 445 volts on pin A2). I know this is a bit low, but I use a 240 volt power transformer on a 230 volt power grid. Plugged the other amp to AC, and the LEDs on the B side of the regulator don't light up. Measure 350 volts on B6 and 496 volts at A2 (on the four pin socket). I have checked all my solder joints and reflowed a couple of them.
How do I proceed with the amp that doesn't light up on the B reg side? Any ideas?
Roger
-
By the way, I got the hum down to 1mV on the amp that works properly. Is that acceptable with 2A3 as the output tube (and to be used with Avantgarde Duo Grosso with a sensitivity of 107 dB)?
Roger
-
I have Paramours (2A3) and I get them down to 1mV and 1.7mV. The Paramour is less sophisticated than the Paramount.
As for your problem with the second amp. Verify the orientation of the LEDs. If backward they won't turn on. Voltage in that part of the circuit, B side, would be high because of the lack of conduction. The schematic doesn't say if the B side feeds the diode string or if the B side is the plate load for the driver half of the tube. So I'm not sure either.
496V is less than 10% high for the A2 expected of 464; that shouldn't be a problem. 350V on B6 is not regulating, and the LEDs would indicate that.
Your problem is somewhere on the board. You might need to lift it to see if you missed any solder pads.
-
Sounds like the regulator half (B6) is working. Without the manual in front of me, my guess is the gain stage half (B1) is not. Check your connections to B1, B2 and B3 from the PC board and check the jumpers on the PC board to be sure they are connecting the correct pads.
-
Thanks to both of you. I'm in Norway and it's already saturday evening. I'm half way through a six-pack, so I'll check again tomorrow.
Roger
-
Doc B, is it possible to have connections wrong (ref. your comment above), and correct resistance measurements?
Roger
-
Yes there are probably some points in the measurements where a good reading won't necessarily point out every flaw. Have you tried replacing the tubes tubes with the ones from good amp? That might be an easy test to start with. I do need to correct my previous comment about pin B1. That was the plate of the second triode of the 12AT7 used in the old Paramount. The plates of the 5670 triodes are B6 and B4.
I am working from the photo of the soft start board on the web site. The C4S load on the B side is the load for the shunt regulator. So what is happening right now is the Zener string is regulating to 350V, but the C4S loading the shunt reg and hybrid shunt reg are not working. The Zener string will give that 350V reading even if the tube is not working - that is why it is there, to regulate until the tube warms up and the hybrid reg takes over. Check to see if the tube is heating up, and if you have around 6.3 volts DC across pin B1 and B9. If the tube is heating up, check the jumper connections on the PC board next. It's fairly common to forget one or miswire one.
-
Yes there are probably some points in the measurements where a good reading won't necessarily point out every flaw. Have you tried replacing the tubes tubes with the ones from good amp? That might be an easy test to start with. I do need to correct my previous comment about pin B1. That was the plate of the second triode of the 12AT7 used in the old Paramount. The plates of the 5670 triodes are B6 and B4.
I am working from the photo of the soft start board on the web site. The C4S load on the B side is the load for the shunt regulator. So what is happening right now is the Zener string is regulating to 350V, but the C4S loading the shunt reg and hybrid shunt reg are not working. The Zener string will give that 350V reading even if the tube is not working - that is why it is there, to regulate until the tube warms up and the hybrid reg takes over. Check to see if the tube is heating up, and if you have around 6.3 volts DC across pin B1 and B9. If the tube is heating up, check the jumper connections on the PC board next. It's fairly common to forget one or miswire one.
I'll check it tomorrow. Thanks.
Roger
-
By the way this is unrelated to the issue at hand, but I believe there may also be a typo in the voltage measurements regarding terminal 13, which I believe is actually connected to ground and should be 0.
You might also check the transistors on the B end of the board for shorts, by measuring across each possible pairing of the transistor leads, left to center, left to right, right to center. With the test leads one way you should see a very high reading and with the test leads swapped on the transistor leads the other way should be maybe around 700-1500 ohms. If you see a very low reading like 50 ohms the transistor is shorted and needs to be replaced. I doubt this is the problem, but it's worth checking.
Argh, without the schematic in front of me I am losing track of which end is which on the PC board. Looks in the photo like the B end is actually the load on the gain stage, not the load on the shunt reg. That doesn't really change any of my suggestions for sorting it out. You might try adjusting the blue trimpot to about the same value as the one on the good amp and see if that changes anything. The amp should work no matter how the trimmer is set, but again, it is something easy to try.
-
By the way this is unrelated to the issue at hand, but I believe there may also be a typo in the voltage measurements regarding terminal 13, which I believe is actually connected to ground and should be 0.
You might also check the transistors on the B end of the board for shorts, by measuring across each possible pairing of the transistor leads, left to center, left to right, right to center. With the test leads one way you should see a very high reading and with the test leads swapped on the transistor leads the other way should be maybe around 700-1500 ohms. If you see a very low reading like 50 ohms the transistor is shorted and needs to be replaced. I doubt this is the problem, but it's worth checking.
Argh, without the schematic in front of me I am losing track of which end is which on the PC board. Looks in the photo like the B end is actually the load on the gain stage, not the load on the shunt reg. That doesn't really change any of my suggestions for sorting it out. You might try adjusting the blue trimpot to about the same value as the one on the good amp and see if that changes anything. The amp should work no matter how the trimmer is set, but again, it is something easy to try.
I have switched to the tube from the working amp. No change. The A side of the reg is for the driver, and the B side is for the regulating half of the 5670. It is the B side of the reg that doesn't work.
Roger
-
Found the problem, I think. Had forgotten to solder the wire from the reg board to pin B8 on the nine pin socket. Will do that tomorrow and check again. Wonder how I missed that. Just hope I haven't damaged anything because of this.
Roger
-
Found the problem, I think. Had forgotten to solder the wire from the reg board to pin B8 on the nine pin socket. Will do that tomorrow and check again. Wonder how I missed that. Just hope I haven't damaged anything because of this.
Roger
Couldn't wait. Did the soldering with a beer and a cognac beside me. Everything lit up nicely. Music tomorrow?
Roger
-
The amps are up and playing, and they sound good. It sounds a little on the bright side of neutral, but I guess they need some hours to open up. The amps are dead quite. In a few weeks it will be time to starts some tube rolling.
The system is VPI Classic with AT OC9/MLII, Trigon Advance phono pre, Electrocompaniet EMC1 Up CD player (which I haven't used for years. I prefer vinyl), BorderPatrol Line 1 and Avantgarde Duo Grosso horn speakers.
Roger
-
Good job, glad you got it sorted out. Yes the amps are a bit bright at first. They will come into better balance as they break in.
-
Good job, glad you got it sorted out. Yes the amps are a bit bright at first. They will come into better balance as they break in.
Here in Norway, we are already at sunday evening (9 hours time difference). I have played for about 8 hours today, and the amps are already much better than right after the first power up. Looking forward to more music from my system. It sounds great now. What amazes me is that they are dead quite on 107dB horns.
Roger
-
By the way this is unrelated to the issue at hand, but I believe there may also be a typo in the voltage measurements regarding terminal 13, which I believe is actually connected to ground and should be 0.
You might also check the transistors on the B end of the board for shorts, by measuring across each possible pairing of the transistor leads, left to center, left to right, right to center. With the test leads one way you should see a very high reading and with the test leads swapped on the transistor leads the other way should be maybe around 700-1500 ohms. If you see a very low reading like 50 ohms the transistor is shorted and needs to be replaced. I doubt this is the problem, but it's worth checking.
Argh, without the schematic in front of me I am losing track of which end is which on the PC board. Looks in the photo like the B end is actually the load on the gain stage, not the load on the shunt reg. That doesn't really change any of my suggestions for sorting it out. You might try adjusting the blue trimpot to about the same value as the one on the good amp and see if that changes anything. The amp should work no matter how the trimmer is set, but again, it is something easy to try.
Is there a typo in the manual. It says that terminal 10 should read 233V. Terminal 10 is the negative side of the 47uF/450V cap (which is connected to ground on the schematic). I measure 0V in my amps.
A new problem arose yesterday. I had played with the stock tubes for about a week and everything has been working fine. Changed the driver to Bendix 6385 and the output to NOS (1959) RCA 2A3. I cleaned the tube pins before plugging them in, and inserted the tube a couple of times to ensure good contact. No sound out of the amp. Measured the voltages (only the B+ and the input to the 2A3), and the voltages on the output of the 6385 were high (on both amps), so I adjusted down. I was not able to get them low enough to achieve 2A3/3. Changed back to the stock tube. No sound out of one of the amps.
Any idea how to troubleshoot this problem?
Roger
-
Yes, every terminal in Paramount ending in 3 or 8 is screwed to the top plate. And again, yes, the negative side of the cathode bypass cap is grounded... somewhere I don't see it on the picture in the manual. So you are spot on with both measurements.
Biasing is specific to the driver tube. So that might be a problem. Did you change the driver before the output tube? Did you leave the same drivers in when going back to your stock tubes? I think you see where this line of questioning goes. You are troubleshooting pretty well so far.
-
If everything is showing proper voltages LEDs are lighting up, etc., and you have no sound it indicates that there is an open in the signal path. Suggest you trace the signal path, looking most closely at what was disturbed when tubes were changed. It could be a loose tube socket, or something may have worked loose that is attached to one of the sockets. If the amp was going back and forth between the bench and the listening room there might be a connection that became loose when connecting and disconnecting interconnects or speaker cables.
What usually happens to me is it turns out to be the interconnect cable rather than the amp...
-
Yes, every terminal in Paramount ending in 3 or 8 is screwed to the top plate. And again, yes, the negative side of the cathode bypass cap is grounded... somewhere I don't see it on the picture in the manual. So you are spot on with both measurements.
Biasing is specific to the driver tube. So that might be a problem. Did you change the driver before the output tube? Did you leave the same drivers in when going back to your stock tubes? I think you see where this line of questioning goes. You are troubleshooting pretty well so far.
I changed both the driver and the output tubes at the same time. Changed back to the 5670 and the Sovtek 2A3 supplied with the kit. No sound out of the right channel amp. Muffled sound out of the left channel amp.
Roger
-
If everything is showing proper voltages LEDs are lighting up, etc., and you have no sound it indicates that there is an open in the signal path. Suggest you trace the signal path, looking most closely at what was disturbed when tubes were changed. It could be a loose tube socket, or something may have worked loose that is attached to one of the sockets. If the amp was going back and forth between the bench and the listening room there might be a connection that became loose when connecting and disconnecting interconnects or speaker cables.
What usually happens to me is it turns out to be the interconnect cable rather than the amp...
I have been over both amps several times, and I can not find anything being loose, but I'll check again and again and again....until everything is working.
Plugged in an SS amp in my system, and everything works fine (same cables, same preamp etc.)
-
You could try running some signal through the amp while you test various points in the circuit where that signal should be present, to see if you can locate where it is stopping. Set your DMM to AC volts and attach the black test lead to signal ground at point like that negative terminal of the bypass capacitor that we had the typo about. With the amp on and some relatively loud and steady music signal going into the input jack check the center pin of the input jack, then the side of the .1uF coupling cap at the grid of the 300B, then the side of the parafeed coupling 3.3 uF cap that feeds the output transformer. You should see low level AC volts at the input jack that vary with the music, for example maybe a range of 0 to 1 volts. At the .1uF cap this should be a bit higher, maybe up to 20V. At the parafeed cap it should be even higher, maybe up to 150V, so be sure to set your meter to at least 200VAC for that measurement. It should go without saying that one needs to be very careful taking these measurements on a live amp. Make sure it is sitting in a stable position and keep one hand in your pocket while probing the test points. None of those points should have much DC voltage on them, but many nearby terminals will have.
If you can find where the signal stops we can narrow down the places you need to look to find the issue.
-
You could try running some signal through the amp while you test various points in the circuit where that signal should be present, to see if you can locate where it is stopping. Set your DMM to AC volts and attach the black test lead to signal ground at point like that negative terminal of the bypass capacitor that we had the typo about. With the amp on and some relatively loud and steady music signal going into the input jack check the center pin of the input jack, then the side of the .1uF coupling cap at the grid of the 300B, then the side of the parafeed coupling 3.3 uF cap that feeds the output transformer. You should see low level AC volts at the input jack that vary with the music, for example maybe a range of 0 to 1 volts. At the .1uF cap this should be a bit higher, maybe up to 20V. At the parafeed cap it should be even higher, maybe up to 150V, so be sure to set your meter to at least 200VAC for that measurement. It should go without saying that one needs to be very careful taking these measurements on a live amp. Make sure it is sitting in a stable position and keep one hand in your pocket while probing the test points. None of those points should have much DC voltage on them, but many nearby terminals will have.
If you can find where the signal stops we can narrow down the places you need to look to find the issue.
I have a function generator and a scope. I just need a few BNC to RCA adaptors to use the test equipment to test the amps. I'll order some and do some measurements. Will take a few days though. Meanwhile, I have other amps I can use.
Roger
-
Excellent, the generator and scope will make the job much easier.
-
Hmm, confusing.
I have done some AC measurements today with a Tektronix MSO2000 scope and a TTi function generator. I have done tests with various levels of sine waves, frequency was 1kHz.
With the function generator set to 2V p-p I measured 31.2V (62.4V p-p) at B4 (anode of 5670), the same at the OB side of the 300K resistor (which it should be) and the same at A3 (grid of 2A3).
When measuring at the output transformer side of the 3.3uF parafeed capacitor I get a disturbed 4.0 to 6.0V reading.
Measured the heater voltage between A1 and A4 to be 2.2V DC.
I then did some measurements with a higher AC voltage in. The gain in the 5670 again showed to be approx. 31x.
I got the same disturbed 4.0 to 6.0 V reading at the 3.3uF cap.
I resoldered all the connections on the four pin socket, the 22ohm resistors, the hum pot, the plate choke (PC2), the output transformer (OT1), the two cathode resistors and all the wires leading from these points to the terminal boards. No change.
I'm running out of ideas for what to do. Any ideas?
Roger
-
Did the same measurements with the other amp. Same 2V p-p, 1kHz sine wave in. Measured 31.4 V (62.8V p-p) out of the 5670, and 146V out of the 2A3. So everything is fine here for now with this amp.
I suspect a bad capacitor (parafeed) in the other amp. Comments?
Roger
(forgot to mention that I measured 10.6V AC at the output of the output transformer in the functioning amp)
-
Roger,
Doc suspects an open in the signal path. You can give it a signal and trace where it disappears. Since some will have a DC on the signal AC couple the input to your oscilloscope. That way you only see the 1k Hz signal.
It comes in on the center conductor of the RCA jack. Measure there then measure at pin 3 on the 5670. The output of the 5670 is on pin 4. For 2A3 that exact signal should appear at tube pin 3 on the 2A3. And the output of the 2A3 is on pin 2. That goes through the parafeed capacitor. This signal should be seen on the output transformer Terminal 1 (red wire). Of course the output of the transformer is taken away on a red wire too. Since I don't know your speaker impedance I can't tell you which lug. But that red wire goes to the red (+) lug of the speaker outputs.
Or you can start by touching up all the connections mentioned. There will be several that I didn't mention in between but you should be tracing the wires between the connections/terminals as well.
-
Roger,
Doc suspects an open in the signal path. You can give it a signal and trace where it disappears. Since some will have a DC on the signal AC couple the input to your oscilloscope. That way you only see the 1k Hz signal.
It comes in on the center conductor of the RCA jack. Measure there then measure at pin 3 on the 5670. The output of the 5670 is on pin 4. For 2A3 that exact signal should appear at tube pin 3 on the 2A3. And the output of the 2A3 is on pin 2. That goes through the parafeed capacitor. This signal should be seen on the output transformer Terminal 1 (red wire). Of course the output of the transformer is taken away on a red wire too. Since I don't know your speaker impedance I can't tell you which lug. But that red wire goes to the red (+) lug of the speaker outputs.
Or you can start by touching up all the connections mentioned. There will be several that I didn't mention in between but you should be tracing the wires between the connections/terminals as well.
Hi Grainger,
I'm not sure if you read my posts or not. I have just measured my amps with a function generator giving signal in to the amp, and I have done a bunch of measurements with a scope. Everything measures correctly up to and including the input to the 2A3 tube. On the parafeed cap I get crappy measurements.
The scope has been set to AC measurements (so DC is filtered out), and everything measures as expected up to the 3.3uF cap, as mentioned above. All connections to the four pin socket and to the end of the amp have been resoldered.
Roger
-
Of course a bad capacitor is a possibility - for instance, one with a broken connection inside the cap. If you have access to a capacitance meter you can check it (disconnect one side); otherwise swap it for th one in the good amp.
Another possibility is an error in wiring the output transformer secondary. If the secondary is shorted, that would present only the DC resistance to the 2A3, reducing the output substantially. You can disconnect the transformer from the parafeed cap to check.
-
Of course a bad capacitor is a possibility - for instance, one with a broken connection inside the cap. If you have access to a capacitance meter you can check it (disconnect one side); otherwise swap it for th one in the good amp.
Another possibility is an error in wiring the output transformer secondary. If the secondary is shorted, that would present only the DC resistance to the 2A3, reducing the output substantially. You can disconnect the transformer from the parafeed cap to check.
Hi Paul
Thanks for the advice. Checked the output transformer connections, and they are correct (16 ohm, my speakers are 18 ohm). My DMM can't measure capacitance, but when I measure it's resistance I can see that it charges from the battery in the DMM.
Roger
-
Roger,
Since I don't own a Paramount I skimmed over the voltage readings. My bad! The magnitude of the voltage is inconsequential. I didn't follow your tracing the signal from input to output.
The fact that the amp worked after you found an open trace/solder joint means nothing was defective to begin with. This is a very good thing!
The quick and easy way to check if you have a bad parafeed cap is to take the one from the "good" amp and put it in the "bad" amp. If that solves the problem look at the outer cover of the "bad" capacitor for damage. These PP Film caps are very, very rarely bad on installation. And changing a tube should never damage the cap.
That is why I'm betting on a missed connection. That is why my last suggestion was to touch up all the solder joints along the way.
-
Roger,
Since I don't own a Paramount I skimmed over the voltage readings. My bad! The magnitude of the voltage is inconsequential. I didn't follow your tracing the signal from input to output.
The fact that the amp worked after you found an open trace/solder joint means nothing was defective to begin with. This is a very good thing!
The quick and easy way to check if you have a bad parafeed cap is to take the one from the "good" amp and put it in the "bad" amp. If that solves the problem look at the outer cover of the "bad" capacitor for damage. These PP Film caps are very, very rarely bad on installation. And changing a tube should never damage the cap.
That is why I'm betting on a missed connection. That is why my last suggestion was to touch up all the solder joints along the way.
Thanks for the advice Grainger. I have already been over all the solder connections from where I found the problem up to the end of the amp. Everything has been resoldered. I have also checked for wrong connections, but I can't find any. So, I'm running out of options here. I can always try to change the caps, but I have never experienced a film cap being bad.
Roger
-
Then...
. . . The fact that the amp worked after you found an open trace/solder joint means nothing was defective to begin with. This is a very good thing!
The quick and easy way to check if you have a bad parafeed cap is to take the one from the "good" amp and put it in the "bad" amp. If that solves the problem look at the outer cover of the "bad" capacitor for damage. These PP Film caps are very, very rarely bad on installation. And changing a tube should never damage the cap. . . .
That is the only way I see to nail down the problem. I've been kit building since 1965.
-
Found the problem. A failed 2A3. A lot of trouble for a trivial problem. But, now everything is working fine, and I learned a lot about this kit during the troubleshooting.
Roger