Advice on Parabee and failing C4S

[banjopicker]

One of my Parabees started producing scratchy distorted output at a reduced volume level. Once I isolated it to the amp (and ruled out the tubes) I flipped it over and saw that one of the two LEDs on the C4S was not lit.

Obviously the LED should be replaced, but is there anything further upstream that I should be looking at as a cause of the C4s failure? The manual shows 6x 100uf 450v electrolytics in the power supply that Doc has suggested degrade over time. Could old electrolytics on the PSU cause a cascading failure of the C4S? Would it be wise to replace the electrolytics while I have the soldering iron out? I can read a schematic but I in no way understand electronics theory.

As much as I would love to send this in to Doc for a factory refurb, it isn't an option financially, and these units are proudly DIY anyway (with a lot of community help). I am on my way out of town so can't get the amp on the bench with a multimeter before I leave, but I would like to get prophylactic parts order in so I can get right into it when I get back. Would appreciate advice on the most productive way to get the music flowing again.

Part of a complete Bottlehead system: Seduction, Foreplay 2, Parabee, and Straight 8s -- working pretty well after all these years, and a lot of fun.

Thanks!

[Paul Birkeland]

Obviously the LED should be replaced

No, unfortunately that's just not how these things work.   There are three wires entering the C4S board, one is ground, one is B+ in, and one is plate out.  You'd want to post DC voltage measurements of all of those points.  It's possible that you need to debug the board, but it's equally possible that there's something else going on in the amp that's causing the issue.  I definitely would not pull out the board and replace the LED, as that's unlikely to resolve your issue.

Could old electrolytics on the PSU cause a cascading failure of the C4S? Would it be wise to replace the electrolytics while I have the soldering iron out? I can read a schematic but I in no way understand electronics theory.

This depends a little bit on what's in there now, but generally this is a good idea at the 20 year mark.  If you happen to have black Panasonic snap-in caps, I would leave them alone for another decade.

These are my personal Paraglows that I built out of a few dead amp carcasses I collected over the years and this is a service I offer for legacy product owners.

[banjopicker]

No, unfortunately that's just not how these things work.

Thanks for the reality check. That means I will have to wait for two weeks before I can look at it properly or order parts. I still have my manual, so I assume you are suggesting that I check resistances and voltages against the checkout values in the instructions. Will do when I get back and post any anomalies.

Thanks!

[banjopicker]

Trip canceled so able to test the Parabee with a multimeter.

All the resistance measurements looked normal for a right side Parabee.

On the voltages I got around 330v for terminals 1 and 2, and then the voltmeter wouldn't settle down to give a reading for 15, A1, A2, A4, and B6. B8 showed .004v

Is there a way to test the C4S in-circuit?

Thanks for your help

[Paul Birkeland]

On the voltages I got around 330v for terminals 1 and 2, and then the voltmeter wouldn't settle down to give a reading for 15, A1, A2, A4, and B6. B8 showed .004v

Do you have a 300B installed?

You'd want to have good voltages on A1/A2/A4 before even considering the possibility that there might be an issue with the C4S board.

[banjopicker]

Yup, both tubes were installed when checking the voltages. I am able to switch tubes between the blocks and the problem doesn't move with the tubes. The multimeter responds as expected for terminals 1 and 2 but then after that voltages aren't stable enough for a reading and the display flashes quickly as it tries to keep up. I have the audio input shorted to ground and the negative of the multimeter clipped to the same bus.

Thank you.

[Paul Birkeland]

A2 should show a very, very consistent reading. 

A1/A4 having a proper voltage would indicate that the 300B is drawing appropriate current.

The absence of stable voltage on A2 is worth focusing on, as this tells me that there is likely an issue with the power supply in that amp.

[banjopicker]

I doublechecked the voltages this morning and this is what I got:

1     365
2     366
15     flashes but can read 80, and 99
A1     2.9
A2    flashes but can read 78, and 94
A4    flashes but can read 64
B6    flashes but can read 45, and 50
B8    .005v

In this test setup with the input grounded both leds are lit on the C4S, further reinforcing your diagnosis that the C4S isn't the problem.

Attached are some pictures for a visual inspection.  The .47 uF cap hiding behind the C4s looks a little rough but I don't see any obvious failures.

I have not been able to find any posts relating to troubleshooting the Parabee's power supply. My understanding that the most failure-prone parts should be the capacitors but I don't know how to isolate and test them, and I certainly don't know enough theory to be poking around without guidance.

Thanks so much for your help.

[Paul Birkeland]

The big choke mounted to the bottom side of the chassis plate is your plate choke.  What DC voltage do you get on each terminal of the choke?

If you are using a bunch of clip leads to measure voltages, you're going to want to slip the boots off the clip leads and solder them so they are more reliable.  That may explain the inconsistent voltages you're seeing.

[banjopicker]

I am a total dumbass. When I checked the voltages my dmm was set to AC the whole time. You specifying DC for the choke tipped me off. For future reference: if flashing readings doublecheck ac/dc.

These readings should make more sense:

1     365
2     366
15     69.7vdc
A1     70.8
A2    422
A4    70.9
B6    212
B8    - millivolts

The choke readings were 422 for the one connected to the capacitor and 444 for the other one.

The bad readings seem to be 15 and B8.

Thanks, this is a great help.

[Paul Birkeland]

OK, so the 300B is working well, that's good. 

I don't have a paper copy of the Parabee manual, so I have no idea what terminal 15 connects to or what it's supposed to be.

Instead, there are three terminals on the C4S board.  One is grounded and will show 0V, then there is one that sits next to Q2 (the bigger of the two transistors) and is connected to its middle leg. We will call this the output. 

Then there's another terminal that feeds the smaller of the two resistors and the LED's, that's the input terminal.

Can you check the DC voltages on the input and output terminals? 

Is the driver tube glowing?

[banjopicker]

Where you have been leading me with the troubleshooting made me suspect the 6N1P. It was glowing fine and I had swapped it between amps, but the trouble stayed in the right amp. I realized that if I have had trouble in the past (which I don't remember, but there have been many moves over the last fifteen years) I would have switched the tubes because the amp is designed to use only half the tube. If I did that with a failing tube in the past it would make sense that switching the tube between the amps wouldn't eliminate the tube as the problem, and fooled me into looking elsewhere.

My spare tubes and my amp have been in different places, so I had to wait to test it out. Reunited and placed back into the system with a new Russian 6N1P installed, everything plays beautifully. But then I put the original tube back in and it still worked. So now I think there must have been either a short in the amp that ended because of the poking around or the 6N1P has a physical failure that was "fixed" by all the poking-jostling. So at the moment I have no clear answer to what was failing.

I will go back and test all voltages again sometime soon, but I am not free to spend time on it at the moment (I am not where my system is). I am interested that one amp may "consume" the 6N1P so much faster than the other--maybe the voltages will shed some light on that. I am also interested in hearing what I should look/listen for before I should set some time aside to swap the electrolytics.

Thanks so much for your patience and help. Sorry to involve you with what appears to be user failure. Your original warning that the C4S was NOT the obvious culprit saved me from trying things that would have led me down an unproductive rabbit hole. Will update when I have more info.

[Paul Birkeland]

6N1Ps are pretty nice sounding tubes, but they are not the longest lived things in the world.  They are super duper dirt cheap though, I'd just get 20 of them and put a new pair in every year or so.

[banjopicker]

Of course, since swapping the 6N1P's back again didn't change anything again, it was just wishful thinking on my part for me to believe that the problem had resolved itself.

It worked for a little while, then while I was listening, there was a pop in the affected channel, then it went back to the same low-vol distorted output as before.

So, setting up to get the measurements you suggested, I connected the input and multimeter to the common ground at terminal 16. I then set the dmm to DC. Turned the amp on with tubes installed. Only one led lights on the C4s. The instant I touch the "I" (input?) with my meter lead the other LED immediately lights and stays lit.

I turn off the amp. Figuring out that the G on the C4S is likely ground, I attach the ground of the multimeter to the lead labeled "G" on the C4S. I turn the amp back on. Both LEDs are still lit. I take these measurements:
I    302v
O   210v

I am in between locations so I am testing the amp away from its system but I am assuming that everything works properly as long as the LEDs are lit, which they currently are, but won't know for sure until I put it back in my system early next week.

Any ideas for a root cause?

[Paul Birkeland]

The instant I touch the "I" (input?) with my meter lead the other LED immediately lights and stays lit.
I    302v
O   210v
Any ideas for a root cause?

There's a loose connection in the amp.  Physically moving it restores the connection, then the amp goes back to working.