Carvin Legacy 100watt - low heater voltages

[six el six]

I have 2.6v ac on the EL34s, 2.8v dc on V1,2,3 and 2.95 v ac on V4,5.

The amp is not putting out the power it should. It's not terribly low but it's not 100 watts either.

All other voltages appear reasonable.

The amp is humming intermittently with the volume turned down.

Could the low heater voltages be bringing the amp power down like this?

[sluckey]

I have 2.6v ac on the EL34s

What do you measure BETWEEN pin 2 and pin 7 of an EL34?

[six el six]

What do you measure BETWEEN pin 2 and pin 7 of an EL34?

[/quote]

5.3v ac.

I should add, the amp starts to act up after it's been on a few minutes. When I first fired it up this morning it was fine for about 10 minutes. Then the volume started to drop. It occasionally makes a humming sound even with the volume controls down all the way.

[six el six]

Posted voltages.

file 13360

phase inverter voltages wonky?

The pins on the circuit board are brittle and the circuit board shows plenty of signs of being hot for extended periods. Heat problems?

http://www.el34world.com/charts/valve/ValveData.php?e=view&f=13360

Thanks folks,
SIX

[sluckey]

What do you measure BETWEEN pin 2 and pin 7 of an EL34?

5.3v ac.

That's too low. Should be 6.3vac. Either the PT filament winding is putting out a low voltage or something in the filament strings AC or DC is dragging it down. Disconnect the PT filament wires from the board (or whereever they actually connect) and measure the voltage across the unloaded winding. What do you get?

[drgonzonm]

Just curious, the heaters on the tubes aren't being fed from one filament circuit are they?  

The schematic I found shows two filament circuits, with the 12A_7 being fed a rectified and filtered DC.  


Never mind, I looked at the the jpg, and saw the separate circuits. 

[six el six]

What do you measure BETWEEN pin 2 and pin 7 of an EL34?

5.3v ac.

That's too low. Should be 6.3vac. Either the PT filament winding is putting out a low voltage or something in the filament strings AC or DC is dragging it down. Disconnect the PT filament wires from the board (or whereever they actually connect) and measure the voltage across the unloaded winding. What do you get?

Unloaded the filament winding reads 6.56 vac.

Would a bad cap in that filament rectifier circuit drag down the voltage in the entire filament/heater circuit?

[sluckey]

Would a bad cap in that filament rectifier circuit drag down the voltage in the entire filament/heater circuit?

Yes. Diodes could too. Can you disconnect the ac feed to the diodes? Maybe at points B2 and B3? Then reconnect the PT filament winding so that it only feeds the big tubes. Now measure the voltage across pins 2 and 7 of a big tube.

The above test will give you an idea if the PT is OK. If the voltage is approx 6.3VAC, the PT is likely OK and the problem is likely diodes or caps in the dc supply.

[six el six]

I'm getting 4.9 vac  between pins 2 and 7 with the PT disconnected and ONLY the 6.3 filament winding feeding points H3bC and H3bD on the power tube pcb.

But just reading the unloaded filament tap it's 6.5vac.

What gives?

[six el six]

Interestingly, the of the two brown secondaries listed on the schematic, one is in  fact orange, the other is clipped at the transformer i.e. it's not even attached to anything.

The orange (8v) lead goes to B7 or D11 (however you see it).

[sluckey]

I'm getting 4.9 vac  between pins 2 and 7 with the PT disconnected and ONLY the 6.3 filament winding feeding points H3bC and H3bD on the power tube pcb.

But just reading the unloaded filament tap it's 6.5vac.

That's not good news. The PT is not able to even supply the 4 EL34 tube filaments. It could possibly be a bad EL34 but not likely. With it connected like this and monitoring the filament voltage, pull one EL34 and see if the voltage increases. If not, pull another. Etc, etc, until there are no EL34s plugged in.

If it appears the PT is bad, then if you have another PT that can supply enough current for 4 EL34s, temporarily wire it in. If the filament voltage now reads correctly for the AC string and the DC string, the Carvin PT is bad.

[sluckey]

The orange (8v) lead goes to B7 or D11 (however you see it).

That is the ac input to the 5VDC power supply for the relays. Since only one wire is connected, the rectifier is operating as a half wave recto. It'll work as is but would probably work better if both wires were connected. Not part of the filament problem though. Has anyone been hacking this amp?

[six el six]

That's not good news. The PT is not able to even supply the 4 EL34 tube filaments. It could possibly be a bad EL34 but not likely. With it connected like this and monitoring the filament voltage, pull one EL34 and see if the voltage increases. If not, pull another. Etc, etc, until there are no EL34s plugged in.

If it appears the PT is bad, then if you have another PT that can supply enough current for 4 EL34s, temporarily wire it in. If the filament voltage now reads correctly for the AC string and the DC string, the Carvin PT is bad.

1. I've confirmed the EL34s are good. I test power tubes and 12ax_ types individually in a modified champ circuit, not a tube tester.

2. Monitoring the filament voltage, I pulled one el34 at a time and sure enough the voltage rose. When all el34s were removed, voltage between pins 2 and 7 had risen to a little over 6.3 vac.

3. The only PT I have around that can supply that much current (4 x el34 = 6 amps) is in a beat to hell and back Marshall Major I can't seem to find time to restore. I'm reticent to disconnect the secondaries on that PT because of the way that amp is put together.

4. I'm getting a Hammond 115/12.6v 20 amp PT so I'll have something to quickly do this troubleshooting procedure in the future.

Is there another good test I can do on this Carvin's PT to be sure it's bad?

That is the ac input to the 5VDC power supply for the relays. Since only one wire is connected, the rectifier is operating as a half wave recto. It'll work as is but would probably work better if both wires were connected. Not part of the filament problem though. Has anyone been hacking this amp?

I don't think anyone has mucked around with this amp. Other than what I described (clipped secondary lead), it looks bone stock. Maybe Carvin had a reason for this? Problem they addressed?

[sluckey]

Is there another good test I can do on this Carvin's PT to be sure it's bad?

Well, you could leave the EL34s unplugged and use a small 6.3V transformer just to test the small tubes and the dc filament circuit. I suspect that's all fine though. Just looks like the Carvin PT has give up on the filament winding.

[drgonzonm]

I looked at the circuit again, discovered that all the tubes are being heated from the filament windings.  (You can see B2 tapped from one of the leads that are going to the EL34's.  I would verify that that the diodes on the B3-B2 taps are all good.  (See my post earlier in the thread). 

Are both sets of 100 ohm resistors R27--R28 and R78--R79 necessary?  I know the artificial taps are used to reduce hum. 

Since this amp uses pcb boards, and it appears there are several.  I would also assure that the connections are all good and there is no oxidation inside any of the pins.  (Copper oxide is a semi-conductor.

[sluckey]

One other thing to try... Reconnect all filaments as normal. Now use some insulation piercing meter probes to pierce the PT filament leads close to the PT body. If you don't have such leads then make a small nick in the heater leads insulation near the PT with a utility knife, exposing the bare wire. Measure the voltage at the pierced leads. If the voltage is still low the PT is bad. If it is normal (approx 6.3) then you have a bad connection somewhere in the string. Look at all the connectors, board traces, solder joints, etc.

[six el six]

One other thing to try... Reconnect all filaments as normal. Now use some insulation piercing meter probes to pierce the PT filament leads close to the PT body. If you don't have such leads then make a small nick in the heater leads insulation near the PT with a utility knife, exposing the bare wire. Measure the voltage at the pierced leads. If the voltage is still low the PT is bad. If it is normal (approx 6.3) then you have a bad connection somewhere in the string. Look at all the connectors, board traces, solder joints, etc.

How approximate is approximate? It reads 6.11vac right at the PT.

Is that definitely bad?

I've had a heck of a time troubleshooting just the preamp board's filament circuit. I've got .3 vac on preamp tubes 1,2, and 3. How ac is still present there I can't figure out

I'm inclined to believe that the PT is bad just because it couldn't properly power the power tubes. I mean, all secondaries were disconnected from everything except the power boards filament hookups where the molex connector attaches.

Thoughts?

I looked at the circuit again, discovered that all the tubes are being heated from the filament windings.  (You can see B2 tapped from one of the leads that are going to the EL34's.  I would verify that that the diodes on the B3-B2 taps are all good.  (See my post earlier in the thread).  

Are both sets of 100 ohm resistors R27--R28 and R78--R79 necessary?  I know the artificial taps are used to reduce hum.  

Since this amp uses pcb boards, and it appears there are several.  I would also assure that the connections are all good and there is no oxidation inside any of the pins.  (Copper oxide is a semi-conductor.

Diodes check good and I use an analog and a digital meter to check. The caps in the filament rectification circuit test good as well, all 3 2200uF read fine. I also don't see any signs of oxidation on the molex connector pins.

[sluckey]

How approximate is approximate? It reads 6.11vac right at the PT.

Is that definitely bad?

I'd say that's good news. You still got about 5.3 right across the EL34s, right? Start looking for a bad/loose connection, solder joint, etc. You may be able to connect your meter right at the tube socket and start wiggling connectors or wires going into connectors. Resolder any solder joints in the filament string. Since your ac string and dc string are both low, I'd start at the PT and check all connections up to the point where the ac and dc strings go their separate ways. If you have to pull the board(s) to resolder connector pads just go ahead and reflow all that are associoted with the filaments.

You may get lucky like plexi50 and find a connector or fuse or something that's hot. A good connection will not be hot but a loose connection is a resistive connection and current flowing thru a resistive connection will cause heat.

[six el six]

I've reflowed the solder on the heater pins ON TOP of the circuit board and heater voltages have increased to the point where the amp is putting out much more power.

I'm going to reflow the solder joints on the underside tomorrow when I get up and hopefully this amp will be "cured".

Thanks to everybody.

[plexi50]

Did you find your problem?

[six el six]

Did you find your problem?

I believe sluckey and drgonzonm's advice to check the pcb/pins for oxidation and all heater conncections/pins was a good call. After reflowing the solder on the all heater pins voltage increased significantly as did output power.

I'll confirm once I get time to reflow the underside of the boards.

[drgonzonm]

Give the troubleshoot kudos to Sluckey, he has the life experiences when it comes to tubes and electronics.  Most of my experience was hanging around the instrument shops at the mines and talking to the techs.  I know whose knowledge is most valuable, the guys and gals who's life depended on their skills. 

[six el six]

Conclusion:

It was a PCB problem. Resoldering/flowing all the heater pins brought the heater voltages up to a reasonable voltage.

All three boards had to be reflowed/soldered including the board that holds the connections for the amp's rear inputs.

I replaced the 2200 6.3 v caps in the dc filament supply. AC on the heaters in the preamp's dc tubes  dropped by ~50% - from ~.25v ac down to ~.125vac .

Thanks again to everyone who helped.