SF twin reverb 135 watts needs refurbishing.

[rockman627]

I'm caught up with my workload so my boss is digging into the vault. We have a master volume twin reverb that sounds mediocre even with 2 JBLs. I am more familiar with standard twin reverbs, but whatever. I guess I'll do the standard routine: get rid of the ground switch, replace the filter cap section, the rectifier/bias section, the power tube screen grid resistors and grid stoppers. Anything else I should consider? Should I get rid of the rear outlet?
twin_reverb_sf_135_schem.pdf

[rockman627]

some pics

[Helmholtz]

Does it give full clean power?

Please describe "mediocre sound".
Do all controls work?

[rockman627]

This had three 5881s and a 6L6GC. I found another Sovtek 5881 and put it in. It hummed a lot until I adjusted "output tubes matching" in the back. I would guess this balances the bias between the two pairs of output tubes, but how do you set the bias level? Is it the 33k resistor feeding the output tubes matching pot?
Furthermore, this appears to have a humdinger. The heater wires go from V10 to V1, but it skips V2. A separate set of wires go from V1 to V2 and the humdinger connects to V2. But there are 2 wires from the pot to nowhere. What's going on there?

[rockman627]

Pots generally work. A tiny bit of scratchy in 2 pots. Fully cranked, it is less loud than a hot rod deluxe fully cranked. No brightness on vib channel. It sounded dull on the floor, but now that it's on my desk it has a fairly even frequency response. On both channels maximum bass occurs at 5 and nothing happens after that.

[glebert]

I would imagine it might need new power tubes. Did you measure idle current? BTW, I've been meaning to ask the forum what the "proper" way to determine idle current is for ultralinear amps.

[rockman627]

I thought I would straighten out the a.c. wiring, filter caps,rectifier and bias circuit before I measured plate voltage.

[The Dude]

FWIW: I've always thought the MV Twins were mediocre sounding even when all working correctly, but that's just my opinion.

[rockman627]

I sure could use a layout diagram for this. Anyone got a link?

[The Dude]

All I could find is this blurry/less than adequate/non-original layout, but it may help you.

https://www.tdpri.com/data/attachmen...2e25813eea.jpg

[Helmholtz]

BTW, I've been meaning to ask the forum what the "proper" way to determine idle current is for ultralinear amps.

No difference when measuring cathode current. Cathode current always is the sum of plate and screen current.

The OT resistance method will give plate current + a fraction of the screen current - still good enough.

[nevetslab]

This had three 5881s and a 6L6GC. I found another Sovtek 5881 and put it in. It hummed a lot until I adjusted "output tubes matching" in the back. I would guess this balances the bias between the two pairs of output tubes, but how do you set the bias level? Is it the 33k resistor feeding the output tubes matching pot?
Furthermore, this appears to have a humdinger. The heater wires go from V10 to V1, but it skips V2. A separate set of wires go from V1 to V2 and the humdinger connects to V2. But there are 2 wires from the pot to nowhere. What's going on there?

I just serviced one of these in the past month. It did, at least, have a set of recent J/J 6L6GC Power Tubes, though not a Matched Set. I have a Compu-Bias meter with four Octal Bias Probes that I use all the time. Install all four probes into the tube sockets, then load the tubes back into them, and, using two tubes at a time, I power up the amp and check the outside pair V7 & V10, mark down their readings, then V8 & V9 tubes, doing the same before making any adjustments. 4 ohm dummy load or the amp's speakers connected. No signal, of course. As there is NO Bias Level Adjustment, I looked to see what they read.

They were sitting in the 22-24mA range, which I find on the low side. On the bias balance pot there is shown a 33k resistor. Not what I found.... instead a 10k resistor shunting that wiper to ground to tweak the bias level. I connected my decade box across that 33k resistor and trimmed it down to 100k across it which raised the plate current up to around 33mA. Three tubes measured 33mA, fourth one to 29.5mA, so soldering a 100k resistor across the existing 10k resistor did the trick. Adjusted the balance pot to even out the current spread.

I could use a distortion analyzer to optimize the bias level but over time, I've normally set them to around 30mA and the amps have sounded nice and healthy. Simply adjusting bias to a prescribed voltage doesn't account for what plate current you're dialing in, and the tube's Transconductance values really will land that current level. Some folks will measure the DCR across the plate and Center Tap, and adjust current for each pair of tubes using that mV reading across the DCR readings. I can do that also with my DC Current Probe. I find the bias probes far more convenient. You just can't leave them in place when driving a load, as it will kill the sense resistor in the probe.

Not sure on the humdinger findings....other's are addressing that.

[glebert]

I haven't seen a 135 MV schematic with a layout. This is the schem and layout that is closest to my 100W MV Twin, maybe it is similar to the 135W.

[rockman627]

The schematic gives no value for the diodes. I have a few in4007 diodes I'll use for the rectifier. Can I use them for the bias voltage supply?

[rockman627]

Nah, the 100w version doesn't match too well. The 135w version is fairly close but I can't make out any text. Well, it's going to be slow.