Re-tubing my Audio Research Ref 110

In reverse order, the slaved tube is behind the adjusted tube, and per the manual will have the higher V number so from left to right your output tubers are numbered V5 with it's slaved tube V7 behind it, V9 with it's slaved tube V11 behind it, V6 with it's slaved tube V8 behind it, and V10 with it's slaved tube V12 behind it. 
https://audioresearch.com/wp-content/uploads/2021/06/REF110_Manual.pdf

Remember what I told you above, you have some badly mismatched tubes and likely a faulty tube. My advice would be to try to return the KT120's and stick with 6550's from a reliable source-I recommend Jim McShane, Upscale and TC Tubes in that order. 

I am sorry you are having so much trouble. I would never recommend that a newbie to tube amps buy an amp like the ARC 110 without some brick and mortar support, a local friend who knows, or a local reliable tech. 

Problem with retubing on ARC Ref 110!

New matched tubes installed and starting biasing (after waiting 15 mins with amp on, speakers connected but no source selected.

First tube set at 65mv, slaved unit measured 66

second tube set, first at 65mv, slaved at 67

third set, first at 65mv, slaved at 73

fourth set, first showed up as -73. Adjusted to -65 (wouldn't adjust below -20), slaved unit created error on Fluke ('OL' and little red triangle with a lightning symbol lights up)

Any idea what's wrong? Am I doing it wrong or faulty tube, faulty amp?

BTW, is the slave tube the one nearest the front of the amp or nearest the back or is this irrelevant?

Great!  Many thanks

Although not true of house wiring, on your amp black is ground and yellow is DC and likewise you will use your black Fluke probe to black and the Fluke red probe on yellow. 

Thanks all. And when connecting the multimeter Does it matter which lead goes to which point (yellow & black points on the ARC)?

No need to call ARC. Yes, you can substitute the 120's for the 6550's*. 

And yes, as Audphile1 said, you want to disconnect the source (music should not be playing) when setting bias. You don't need to take all those steps the first guy who responded specified. To prevent any wildly off bias, wait ten minutes with the amp on and speakers connected but with no music playing and then do the first bias setting. And yes, your amp comes with one pot adjusting one tube with another "slaved" to that tube. So one bias pot for two tubes. The bias pot sets one tube to a specific setting and then the slaved tube is merely expected to be within a specified range. Don't sweat it. After that initial setting done to prevent any "flare ups" of potentially badly out of whack bias, just play music for an hour or two and kick back and THEN stop the source and check bias again resetting as necessary That is all you need. Not all that other stuff. After 20 hours or so you can check again. If your tubes won't bias to spec, you need to rotate the slaved pairs up to three times with different combinations and if they still won't bias to spec you got an unacceptably matched set (mismatched set). 

* KT120's are polarizing. They have grunt and extension but in the opinion of many lack the soul and midrange presence of good 6550's/KT88's. But you have already bought the KT120's so enjoy them. They are a good tube. Nothing to lose sleep over. 

I would talk to to Audio Research, technical support. They are great. Personally I always replace with exact duplicate tubes… but that is me. I cherish the Audio Research sound and simply wish to perfectly maintain it. But, that is me. 

Do the KT120 just drop in as a direct replacement for 6550c?

Did you contact ARC to confirm this and to find out what the KT120 should be biased to? 
 

Other than that, @jasonbourne52 provided the steps. Just have you source(s) powered off when biasing to be on the safe side. 

Speakers connected - check. Multimeter connected - check. 0.65 VDC - check. The four 6H30's can go into any of their respective sockets. After this setup play some music for 30 minutes. Keep the meter connected. Then check the meter reading and re-adjust to 0.65 VDC. Play some more music again for 30 minutes. Check meter and re-adjust if necessary. Then, for the third time, play some music for 60 minutes, check meter and re-adjust if necessary. I recommend checking the bias again after 24 hours use. Then after seven days to see if it has remained stable.