Confusing about matching tubes in ref40, pls advice.
 

Hi,
Approaching 4,800 hours on Ref 40. Time to replace 6h30 and second set of 6550,
Here are simple questions.
1. Do two of 6H30 in power supply unit have to be matched pair?
2. Do two of 6550 in power supply unit have to be matched pair?
3. What do eight of 6H30 in preamp unit have to be? 2 matched quad set? Or 4 matched pair set, or no need to be matched?

Thank you.

Audio Research says they all have to be matched. I just ordered a set from Audio Research when I had to change them out a few months ago in my unit.

it's best to match.
do you mix, or match the tyres on your car?
no need to compromise with reference gear.

Hmm.. ... What I mean was.. All of them will be the same model. And a matched pair here means two tubes of the same model that have very closely matched electrically, selected by measurement of many tubes. In some application, it is a must to have a matched pair or matched quad. But for Audio Research pre amp, I don't know do I have to pay more for matched pair or matched quad or not. Two of 6H30 in power supply unit are for separated channel, worked independently, so it could be possible that both do not have to have exactly electrical value. That s what I mean. And in this context, I do not use matched quad tires, all are of same model but have not been tested and selected to have exactly the same measurement. Thanks anyway for giving nice suggestion. So still waiting for the answer for this. Thank you. Toga

no problem, i understand what you mean.
was just having fun with the tyre reference.
what i was trying to say is, that tube matching usually only costs a few dollars more.
i have jadis power amp and sonic frontiers pre and have always opted for matching.
not sure whether it applies for arc pre, so i will leave it for somebody else to answer.

cheers!

FastEddy... Welcome! :wave:

Don't be too quick to dismiss the tire reference; on high performance cars like BMW, Porsche and domestics like Corvettes the tires do need to be matched for best performance, and are generally replaced in pairs unless a manufacturer has changed formulation in which case all four often require replacement.

Sure, you don't need to, but then the car won't perform at the level it was designed to.

I don't know much about tubes electronics.
I understand that, for example, two KT120 in each channels of ref75 must be matched pair because the single bias setting for each channel controls both tube biasing. So this is required due technical reason. But for pre amp, where there is no user adjustable of any kind for tubes, and there are four 6H30 per channel worked together. Also two in power supply unit, but I am sure they work independent for each channel. To make this simple I can just order a replacement tubes from ARC which they will write a no. On each tube then I just insert each tube into sockets that have the same number. Anyway, for this time I just want to try something like EH 6h30pi, I have to order ten of them. So here is why I must know how to order them correctly.
Do 4 tubes per channel are 2 groups of matched pair for balanced circuit, or all 4 are matched quad. That is always confusing to me.

Tube matching tolerance is tight, within a decimal. But I will not argue more about tire matching. It's a different matter. In two channel music system we do need two channels to be exactly matched as much as possible, unlike car where sometime wheel alignment for left and right can be different for some reasons. But enough about car because it is time to replacing tubes.

It s time to call ARC i guess.
Toga

I'm no expert at this, but on the ref75, there is a separate bias setting for each kt120.

Toga,

I measured all the 6h30 tubes I got for my REF40 and selected very close matches in transconductance for the amplifying tubes (those immediately after the volume control) and used the remaining as cathode followers or in the power supply.
I discarded all tubes that had differences in trans conductance values between triode sections of the same tube of more than 5%. Most of them were closer than 2%.