Bottlehead Forum
Bottlehead Kits => Legacy Kit Products => Paramount => Topic started by: Brent on March 17, 2019, 08:16:44 AM
-
I've been running my Paramounts for five or six years now with no issues, and using 6SN7 drivers with adapters for a couple or so years, also without apparent problems. About a week ago, at power up, one channel gave a loud static-y sound. Other channel powered up normally. Music played back through both channels, but was distorted and lower in level in the channel with the power up noise.
I powered down and thought I'd identified a bad 6SN7 as the culprit, as swapping the driver and adapter from the functioning amp to the problem amp seemed to eliminate the distortion. Not having an extra 6SN7 on hand, I decided to put the original 5670s back in both amps to run until I could get another 6SN7.
Things got interesting while adjusting plate voltage with the stock 5670 drivers. One was fine, but the problem amp is stuck at 208 or 209 volts at OB and will not adjust voltage with any number of turns of the adjustment screw on the 10Kohm trim pot.
My first guess is that there is a problem with the trim pot, but I wanted to check here first to make sure there isn't some other problem. I measured voltages for the bad channel as follows:
T9: 208 or 209
T19: 208 or 209
OB: 208 or 209, will not adjust with trim pot
OA: 240
B side Kreg: 7.07
A side Kreg: 2.48
A1: 70-71
A2: 430
A3: 0
A4: 65
Everything else is within spec according to the manual. Will replacing the trim pot take care of this, or do the voltages suggest another problem? Possibly worth mentioning is that I've moved with the past 3 years, and wall voltage in this house varies from 114.5V or so to 123V. My BeePre buzzes at around 115V, so I'm running the BeePre, Paramounts, and Eros off an AVR that switches taps while I'm listening sometimes.
Please let me know if there are any other measurements that would help identify the problem, and please let me know if it's obvious what I need to replace. Luckily, I am able to run my speakers off my SEX 2.x, so things aren't too bad while this is getting solved.
Thanks for any suggestions you might have!
Brent
-
A side Kreg: 2.48
Does this voltage change at all when you turn the pot?
IIRC, you can check that the pot's resistance is changing if you measure the DCR between the "X" pad and "bReg".
For what it's worth, it sounds like your 10K trimmer is turned all the way down to 0 ohms, and you need to go the other way.
-
Paul,
You’re right: the pot was the need all the way down. If I turn t the ther way, I can raise the voltage at OB. Raising OB to 229v raises kreg to 3.09v.
So I guess it’s not the pot. Thoughts?
-
Your OA voltage is also whacky. It should be roughly 300.5V or so DC, not 240.
I would strongly suspect you have a broken wire or a poorly soldered joint on your C4S board. I know that you're thinking "but it's been working for years", and that isn't really a factor, especially if you've been rattling down the road during your move.
Another thing that comes to mind is perhaps a popped zener diode. The easy way to check this is to heat up the easiest leg of one of them to get to and pull it out and away from the terminal strip.
-
Thanks for the reply. I’ll check joints and wires first, then the diode, and report back.
-
Replaced all the wires on and from the board and reflowed solder joints. What would I be checking with the diodes?
-
Replaced all the wires on and from the board and reflowed solder joints. What would I be checking with the diodes?
The easy way to check this is to heat up the easiest leg of one of them to get to and pull it out and away from the terminal strip.
I should have added that you can disconnect one end of one diode in the string, then power on the amp and see if your OA voltage pops up to 300.
-PB
-
Unfortunately, OA is still reading low with the Zener string disconnected.
-
Put the zener string back in the circuit, then disconnect the wire leaving KregB, then recheck.
-
When I re-attach the Zener string and disconnect KregB, I still get 210 volts at OB and 246 at OA.
-
Pull the 5670, then check again.
This will probably come down to removing the C4S board from the amp and giving all its joints a healthy reheating, as well as inspecting the 9 pin socket thoroughly.
-
Okay, I get 350 volts at OA and OB with the driver tube pulled. I’ll thoroughly and patiently reflow the joints on the board. Hopefully that will work
Thanks for your help!
-
I finally was able to resolder the components on the shunt reg board. Upon testing, I am still getting abnormal behavior. OA starts at 345v, then drops to 300v for a moment before dropping rapidly to 243v or so.
Is there a particular component that would cause this? The board is getting a bit torched at this point--I probably have my iron too hot. I am open to replacing the board entirely, if BH still stocks it. One of the SII boards looks like it may be the same. I'm happy to get parts elsewhere, but I might seek help selecting the right ones if anything is unclear to me.
I appreciate any additional thoughts on what's wrong here.
Thanks,
Brent
edited to add: the LEDs on the A side of the board are dimmer than the LEDs on the B side. I can't tell if that's a symptom or an indicator of the cause of the low voltage at OA. Seems like it starts up normal, though, before the voltage drops, so the LEDs are probably just a symptom.
-
If you disconnect the wire at Kreg b that heads to the 9 pin socket, that will disconnect the HV regulator. If you do that, what voltage do you get at OA?
Another test I would do is measure the DC resistance between each pair of legs on each TL431. If your meter will do the continuity beep, that is a good setting to use. You may have a 431 that has bitten the dust and needs to be replaced.
-
With the trimmer turned all the way down and B side kreg disconnected from the 9 pin socket, I get 235v at OA.
Resistances for A side 431
r to A. 5.01 Kohm
r to K. 0.5 ohm
A to K. 5.01 Kohm
Resistances for B side 431
r to A. 2.48 Kohm
r to K. OL at 2M
A to K. OL at 2M
Those are clearly. Hopefully I got the measurements right. Does clarify things at all?
-
Interesting, so your 431 chips seem OK, and disconnecting the regulator still doesn't bring OA up, which is somewhat unexpected. My next suggestion would be to check for shorts between pins of each transistor on the B side C4S. By process of elimination, this seems to be the problematic area.
All of the parts are still in stock and a replacement Paramount Soft Start half kit can be made for you if that's helpful. The flat rate repair service is also an option.
-
I’m not getting shorts between any transistor pins on either A or B side.
If the problem can be isolated to the board, maybe it’s simpler just to replace it at this point. FWIW, I’ve resoldered all the terminals around and connected to the 9 pin socket too.
-
Yes, the B side C4S appears to have some issues. If I had your amp in for repair, I would just build and install a new PCB.
-
I’ve sent an email to order the half kit. I’ll follow up when I’ve installed it just to confirm it fixed the issue. Thanks again for your help.
Brent
-
I received the half kit last week and installed it a couple days later. Voltages are normal and the amp is back in business. Woo hoo!
I have a couple of observations about the newly installed board. First, both amps used to 'squeak' a few seconds after turning them on, about the same time the normal background hiss came up (my speakers are about 100dB, so the hiss is easily within normal operation for the amps). The amp with the new board doesn't squeak anymore. Second, regarding that background whitish noise, it's noticeably quieter on the repaired amp.
Both of these observations have me curious. The only visually obvious parts difference was that the new board had a plastic PN2097A instead of the metal can 2N2097A transistor that came with the amp. Would this account for either or both observations? I recall reading on this forum about different batches of LM431 correlating with better and worse noise levels, so it could also be that (despite no obvious visual differences there, though I didn't compare part numbers on the LM431s).
I ask because one of the reasons I was previously using 6SN7s in place of the 5670s was the reduction in self noise from the Paramounts. It wasn't terrible, but with 100dB speakers it was noticeable when things were quiet. However, if I could get both amps as quiet as the one I just repaired, I don't think I'd bother with 6SN7s. There is still a bit more gain than I need with the stock 5670s, but that's not such a problem with the BeeQuiet on my BeePre ;D
Thanks as always!
Brent
-
The 431s that we got that were bad have "WS" written on them. I don't think the PN2907 should be any quieter than the 2N2907, but it could be!