Carvin Belair Repair/Tweak

[AZJimC]

I came across a deal on a Carvin Vintage series amp. The objective is to first get it playing, then tweak it for better performance.  I pulled the chassis, and immediately saw that my task would be more than tube swapping.

The ends of the chassis are open, and rats had gotten in, made a home, and left lots of rat pee everywhere. The power supply resistors were totally toasted, along with the circuit board underneath. I pulled the boards out of chassis, and hosed everything out, and gave the boards a shower, then lay them in the Arizona sun all afternoon to bake it all dry. I pulled the resistors, and cleaned the charred board material away, and filled the area with super glue. Fired the amp up after replacing the resistors, and the new ones began to get very hot.... so I investigated further. A look at the backside of the board showed that the trace coming from the last power resistor, R41, had sparked across to the nearby ground. The interesting thing is that this burned board had formed a "carbon" resistor, valued at 281R. I actually couldn't believe that until I scrapped the carbon away with a pointed micro file. I remeasured resistance to ground, and it was over 10m ohm. Cleaned it better, and the resistance was to high to measure.

Now, I fire the amp up, and it works. A few bent tube pins that were slipped outside the grips had to be straightened, and sockets retentioned.

This thing sounds like it has potential, but I'm getting lots of crackle, and hiss. Going over the signal path with the scope, I found that the hiss grew with each stage. First suspect, the 1/4w 220KR plate loads. I swapped a those for 220K mil spec 1/2 watts. Silence began to be quieter.

As I continued looking, I found an ultra sonic oscillation going on starting at V3A. I thought, "Oh, rat piss!". V3 is right beside where the resistor had formed on the board, so I pulled the board, and dug even deeper, then filled the area with epoxy. There was no evidence of further burning, so I was just being sure. It had no effect on the oscillation I was seeing.


I found quite a few interesting things during the work I already have done, so I'll be typing up the story so far in the next few posts.

[AZJimC]

At this point, the amp was working, and the only symptom of the oscillation was a bit of "hair" on certain notes. D's really brought out a static, any D, even in the higher registers. For now, I'll ignore the hairy notes and get involved in the tweaks. There are a lot of tweaks for this amp on the net, hasserl.com, I think, has a lot of info. One thing nearly every review mentions about this series of amps is the rather anemic sound of the "soak" or distortion channel.

After playing a while, I was in agreement, the channel sucks. The first thing to do? Remove the clipping diodes, and replace them with a 470k grid leak. That woke the channel up, and might be enough for some folks, but since Carvin was getting their crunch from the diodes, there is a lot of wasted gain available. The popular mods talk about adding cathode bypass caps V2A-V2B. He recommends a 4.7uf on both cathodes, but I've heard his mods on youtube, and I'm partial to more of a scream than a growl, lets say, so I swapped V2A cathode, R10, which is 4K7R, to 2K2R, and tacked a .68 cap across it.  I don't think I'll need one across R13. (1K5R on V2B).

This thing is sounding pretty killer now, except the that annoying "hair" on the clean channel. (I verified it's there on the crunch channel too, Just not too noticeable with distortion.)

Now I need to attack this crud sound, cause I can already hear the two Celestion Vintage 30's under attack from this sweet sounding amp. except for that hair, and a nasty crunch on rare occasion.

[AZJimC]

So, I'm thinking, The crackle comes out on both channels, so I start at the tube that is common to both channels, V3A. A couple things are going on at this stage... The Channel relay is selecting from two inputs, so likely not both contacts messed up on a relay at the same time, the same sounding way, so I move on in the circuit. Next, there is an effects loop, and it uses switching jacks to bypass when nothing is plugged in...... AHA!!! I plug in a jumper, and bingo, the loud crackle, under load, and D notes, is gone! I spray some deoxit in the jack, and jerk a plug in and out like a lonely adolescent, and no more crackle! Now I'm happy, because I know how hard some of those can be to find. Cheap parts, and they always ask me, "Why not just buy the new micro surface mount technology, heavy metal screamer amp at Metal Center rather than having one built?"  Cheap parts, that's why.

I hook the scope up to V3a Plate, and that dang oscillation is still there, well, I was hoping....

I mess with a few things, and notice that turning the Presence all the way up causes the oscillation to get into the audible range like a high pitched chirping. I think, is the OT wired backwards? Unlikely, this is a factory amp, and layout prevents swapping the leads, without untying the wires, and the power tube board is marked, BRN and BLU. Idiot proof?? Never underestimate the resourcefulness of an idiot!.

Head scratching... schematic squinting, surely it couldn't be in the layout... or could it? From front panel going toward back, there is the preamp board, then power tube board, then a board on rear that has hookups for rear jacks, and the rear mounted presence control. So, I think... what the heck could be something that would have a high enough impedance to catch a signal coupling via wiring? In checking the layout I see a short two wire cable that caries the drive signal to the output tubes, and it's close to nothing... but, right across the chassis is a flat cable going from the middle of the preamp board to the rear board. I lifted this cable a bit, and the oscillation faded away, now I'm getting somewhere. This cable carries signal for send/return jacks, as well as NFB... likely the return input is getting a signal somewhere... this cable crosses the plate wires on output tubes at a right angle, and lifting it a bit stops the oscillation, sooo, I untied the flat cable, and began to twist... the whole thing into a roll. Plugged it back to the board, and still could see a tiny signal on the scope when it neared the brown wire. Makes sense, since at least one side of the Phase Inverter signal has to be a positive in relation to the return jack signal, which feeds into V3A. I carefully routed the cable to miss the brown lead, and even when laying on the blue lead, there is no more gratuitous ultrasonic wave, and the presence works as expected, without chirping.


Steady headway, and this amp is really sounding good... so, I'll continue with the mods.

[AZJimC]

Moving on, I determined to add a PPIMV to this amp, it's quite loud with the 4xEL84 outputs, and I just like a master to control how hard I'm hitting the output tubes. There is plenty of preamp distortion available, so, with the master, there should be quite a palate of sounds available, and at a volume suitable for most any setting.

One note, I did swap the resistors in the drive channel's tone stack to 100K for R15, and 47K for R16.

After looking at the popular PPIMV mod online, I have opted to do it a little differently. The way he is doing it is to pull out C36 and C37, the coupling caps, and replace them with double the capacitance, since they will be in series his way. He uses 4 caps, because a single pair in this position would allow bias voltage on the pots, and turning the grounded pots down would also short the bias voltage to ground. To these caps, he connects to the dual 1MA pots, two to the end, and two to the wipers. This works, and changes the circuit very little, but still adds a 1M path to ground to the original circuit. Granted, this is a very tiny difference overall, but it is my understanding that the LarMar MV is intended to replace the bias carrying grid leaks with the pot, so when it is turned up, the circuit is the same as stock, so I'll do it that way. No caps to swap then, and I have devised a simpler hookup.

I locate the two 220K resistors on the board, R37 & R38. I noticed that the two are joined at one end, the end that connects to bias supply. I have a 500K dual pot, so I will need to place this pot across a 470K to be close to the original resistor values. I will replace the two 220K resistors with 470K, and wire the pot across them. Since one end is linked on the resistors, I can use a single wire to the ground end of both gangs the pot, and two wires to the other end, from each resistor. A jumper from each High end of the pots to the wipers will complete the hookup. The control will be mounted in the popular position, replacing the channel select switch. I have a mini carling 1/4" mount SPDT. a single throw is all that is required, since the select switch operates the relay. It has to be in the open position, CH1, for the footswitch to work.

There is room enough to drill a 1/4" hole and mount this select switch above the input jack. That's the plan.  I'm considering also adding LED's to show which channel is selected, but the objective right now is to get the MV in place.

So, there's the plan, I'll be back....

[AZJimC]

Okay, I've got the MV soldered in, and first thing I notice is that the vol goes from nothing to a bunch in a few degrees of rotation. The answer would likely be to use a 250K dual pot, and 2M2R resistors to replace the the two 220K on board. I'll suffer, since there is still a lot of adjustment left after full vol, adjusting power tube drive.  Now to the channel select switch.

This is becoming a very nice and versatile amp.

[Willabe]

 

[AZJimC]

Thanks Willabe,
after starting with rat piss, it's getting to be a cool amp. I had to share the mess I found.

Back to the task, I have a word of warning... if you work on one of these, get a rubber screw driver to tighten the knob set screws. They're crap, and crack in half at the beginning of tightness. Now I have to get a full set of knobs, but I needed another for the master vol anyway. By the way, if you ever want to put a pot in a switch hole, get some of the isolating washers doug has. They fit on a pot, and keep it centered in a 1/2 inch switch hole. One on the back side of the panel, and nobody needs to know You've got a half inch hole. Now I'm tired, and it's friday night... who wants to join me for karaoke and "rejuvenating" beverages? I'll lament the demise of a few knobs, and awake refreshed and ready to order some new knobs. Seriously, if you're anywhere near Las Vegas, it's worth the trip... I promise a good jam on request as well.... g'nite folks... Back soon,

Jim

[Willabe]