An issue with the ECL86 reverb amp. Help

[Bob-I]

I've completed the wiring of this amp using the schematic attached.

Overall it doesn't sound bad. I have some hum that I'm tracking down and some funny buzzy tone that's a concern. Take a look at the schem and see if you see anything that could be causing this. It's not PO, I've checked all over with my scope and I dont' see anything that looks like oscillation to me. I'm thinking that the cheep filters may be a cause but I'm open to suggestions.

The reverb sounds nice, effective and smooth but I'm concerned about the tube. I have a 1.5K screen resistor but I'm getting some odd voltages. The screen is higher than the plates by 20V and the plates are being dragged down to 131V from a 360V source. I've read that the pentode side of this tube is similar to an EL-84 but it's not acting like it.

Suggestions?

[skyboltone]

Lift the two 22uF caps in the screen/plate circuit and see what happens to the voltages. They just don't seem natural there. Additional filtering? In the case of the plate circuit I would think that the tube would see it as a load on that side of the resistor.

[VacuumVoodoo]

There's obviously something wrong that defies Kirchoffs laws:

4.04V cathode voltage on ECL86 pentode with 180 Ohm cathode resistor says plate+screen current is ca 22mA which seems about right but, assuming that everything is wired as per schematic and V1 stays at 360V:

You have 210V voltage drop from V1 to screen on 1.5k+2.2k, this says that screen current is 43mA and 230V drop from V1 to plate which suggests pretty high plate current. Assuming 500 Ohm primary resistance that gives over 100mA! It makes no sense, with this high plate & screen current the ECL 86 should glow like a beacon and it would surely sag your V1 supply a lot.

Measure voltage drop across 1.5k from V1 to Reverb tranformer/22uF to get actual plate current. Recheck your other measurements too.

Could this be a really badly leaking 100uF cathode decoupling cap that makes the 180 ohm cathode resistor look much smaller for dc?

[Bob-I]

Lift the two 22uF caps in the screen/plate circuit and see what happens to the voltages. They just don't seem natural there. Additional filtering? In the case of the plate circuit I would think that the tube would see it as a load on that side of the resistor.

The first is to decouple the reverb from the rest of the ckt, the second is the screen filter. I can't see the ckt without them.

[Bob-I]

There's obviously something wrong that defies Kirchoffs laws:

4.04V cathode voltage on ECL86 pentode with 180 Ohm cathode resistor says plate+screen current is ca 22mA which seems about right but, assuming that everything is wired as per schematic and V1 stays at 360V:

You have 210V voltage drop from V1 to screen on 1.5k+2.2k, this says that screen current is 43mA and 230V drop from V1 to plate which suggests pretty high plate current. Assuming 500 Ohm primary resistance that gives over 100mA! It makes no sense, with this high plate & screen current the ECL 86 should glow like a beacon and it would surely sag your V1 supply a lot.

My thoughts exactly. The tube doesn't glow so the voltage drop is somewhere else. Maybe a bad cap?

Measure voltage drop across 1.5k from V1 to Reverb tranformer/22uF to get actual plate current. Recheck your other measurements too.

Could this be a really badly leaking 100uF cathode decoupling cap that makes the 180 ohm cathode resistor look much smaller for dc?

Or a leaky folter cap. thx

[Bob-I]

You have 210V voltage drop from V1 to screen on 1.5k+2.2k

oops, the 1.5K is a 15K. I'll swap that out and take some more measurements.

Also, I'm getting some acring between terminals on the tube socket, plate to heater and plate to cathode. I'll heatshrink these and see what happens.

I'm thinking that I just need to up the screens resistor, maybe a 2.2K and get that screen voltage in line. Then I can tackle the ground loop problem.

[Bob-I]

No help with the heat shrink, still arcing.

The biggest issue is the hum. Any thoughts?

[VacuumVoodoo]

Arcing ECL86 is an unusual situation. New socket or recycled "vintage" stuff that might have had beer spilled over it?
Is the resistor in series with transformers primary really 2.2k or maybe 22k? It would explain low plate voltage.

Is it reverb humming? You need to identify the source of hum. How is reverb tank positioned relative to PT and choke? Where is ECL tube placed? Your reverb recovery stage has flat gain all the way down to almost DC. Not needed. 220nF cathode cap is enough, ca 3.3 - 5.6nF input cap with 220k from grid to gnd will give you reasonable LF cut-off.

[Bob-I]

Arcing ECL86 is an unusual situation. New socket or recycled "vintage" stuff that might have had beer spilled over it?
Is the resistor in series with transformers primary really 2.2k or maybe 22k? It would explain low plate voltage.

I'll have to try a new socket. It's a ceramic that I had here, can't remember the source.

Is it reverb humming? You need to identify the source of hum. How is reverb tank positioned relative to PT and choke? Where is ECL tube placed? Your reverb recovery stage has flat gain all the way down to almost DC. Not needed. 220nF cathode cap is enough, ca 3.3 - 5.6nF input cap with 220k from grid to gnd will give you reasonable LF cut-off.

No, the reverb isn't humming. I've taken out the reverb tube and disconnected the tank and I still have the hum. The ECL is between the CF and the PI. Thx for the advise on the recovery stage.

[VacuumVoodoo]

No, the reverb isn't humming. I've taken out the reverb tube and disconnected the tank and I still have the hum. The ECL is between the CF and the PI. Thx for the advise on the recovery stage.

How long is the wire from reverb pot to mix tube grid? It's not shielded, is it? Disconnect it from mix tube grid. Still got hum?

[DiademRoyale]

I hope that I am not out of line for asking this, but are you using an ECL86, or ECL83? The two tubes are not interchangeble. Your discussion has been pertaining to an ECL86, but the schem drawn shows an ECL83. The schematic appears correct for an ECL83 to me, but definitely NOT for an ECL 86. Just $.02 ........MDW

[VacuumVoodoo]

I hope that I am not out of line for asking this, but are you using an ECL86, or ECL83? The two tubes are not interchangeble. Your discussion has been pertaining to an ECL86, but the schem drawn shows an ECL83. The schematic appears correct for an ECL83 to me, but definitely NOT for an ECL 86. Just $.02 ........MDW

I think ECL83 type designation on the schematic is an error that followed from an earlier discussion. But if the tube in the socket is in fact an ECL83 it would explain some of the problem but not all of it.

[Bob-I]

DiademRoyale, VacuumVoodoo is correct, that tube is an ECL86. What's wrong with this schematic? Maybe the source of my acring is something incorrect in the schem.

I found the source of the hum.... damnit. A BAD TUBE IN V1. Brand new EH 12AX7 was bad causing the hum. 4 hours of troubleshooting and dummy me didn't think to swap out the tube!! I even took the OT off and moved it away.

[DiademRoyale]

Sorry- I should have been more concise- the pinoout is correct for ECL86, componet values are correct for ECL83-pin-out for ECL83 is different as are voltage and load requirements- enough so that it would cause problems. Are you using an 86 or an 83? Also, try using the artificial
ground trick for the
filament supply and see if that helps calm down the hum somewhat. Use the same point as your power tube cathodes use for gronding the two 100
ohm resistors- lengthen the resistorconnection but keep it very close to the chassis and cross any signal path wires encountered at right angles if necessary. ECL-83, 86, 6AN-8, 6BM_8, etc are, by virtue of being two amps in one envelope, prone to being more sensitive to noise problems brought on by discrepancies with their gain structuring and filament noise. Good luck. Maybe something here will help.... MDW.......... DiademRoyale

[DiademRoyale]

Sorry- I should have been more concise- the pinoout is correct for ECL86, componet values are correct for ECL83-pin-out for ECL83 is different as are voltage and load requirements- enough so that it would cause problems. Are you using an 86 or an 83? Also, try using the artificial
ground trick for the
filament supply and see if that helps calm down the hum somewhat. Use the same point as your power tube cathodes use for gronding the two 100
ohm resistors- lengthen the resistorconnection but keep it very close to the chassis and cross any signal path wires encountered at right angles if necessary. ECL-83, 86, 6AN-8, 6BM_8, etc are, by virtue of being two amps in one envelope, prone to being more sensitive to noise problems brought on by discrepancies with their gain structuring and filament noise. Good luck. Maybe something here will help.... MDW.......... DiademRoyale