|
Register | FAQ | Members List | Calendar | Gallery | Today's Posts | Mark Forums Read | Search |
WAD General For discussions re World Audio Design |
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Valve matching with a KiT6550
How important is valve matching with a KiT6550?
Over the years I am lucky to have collected a few sets of matched valves, and over the same years, one valve from each set except one has failed. Red glow, no output, sparks and I even dropped done and smashed it. Doh! So each set has a value printed or written on it, the sets from Hotrox have two values. For example: SED Winged C two at 77mv, two at 76mv. One valve with no output. TungSol 6550 4 at 83mv, one glows red. TungSol 6550 two at 76, two at 68mv. This set works fine. TungSol 6550 4 at 58mv. TC 7.0. This set works fine. TungSol 6550 4 at 68mv, TC 7.4, one smashed. TungSol 6550 4 at 56mv, TC 6.7, one dead no output. Can I use two matched pairs to safely make up a quad using the remaining working valves? For example use two of the remaining TungSol 6550s 68mv in the left channel, and two of the remaining TungSol 6550s 56mv in the right channel? Would there be a channel imbalance that I can hear? Anything else to be aware of? Thanks as always, John |
#2
|
|||
|
|||
Re: Valve matching with a KiT6550
John wrote
How important is valve matching with a KiT6550? - It is important "to a degree", but do not get paranoid about it, as a few mA difference in anode current will most likely not be noticeable. Can I use two matched pairs to safely make up a quad using the remaining working valves? - Yes, but, and there always is a but - measure cathode volts on the 6550s to check that are not drastically out of balance - valve to valve and pair to pair. Imbalance valve to valve will probably show up as 100Hz hum in the L/S due to imbalance of current through the optx. Pair to pair see below. Would there be a channel imbalance that I can hear? Anything else to be aware of?, - I would doubt it, because we tend to listen to music/speech rather than constant tone. You could probably get up to 2dB difference in levels before it became noticeable and possibly only then on speech if you knew for certain that the voice was coming exactly from the centre. As an example - I once had my L/S swapped over L to R and didn't notice it for some time as some orchestras have violins the R some on the L. It was only when I played a known pop group did I notice it was wrong. The brain is a powerful thing and will compensate for all sorts of L/R differences John |
#3
|
|||
|
|||
Re: Valve matching with a KiT6550
Hi John and John, just seen this a bit late but will post it anyway
Try it by all means but watch the anodes for redness and listen for increasing humming caused by run-away of a poor pairing. The problem is they are aged by varying amounts and even when new may have been tested on different machines, at different times of the day with different mains voltages, and different characteristic settings, by different people. A quad/pair should be relatively similar to each other at the time of testing but I wouldn’t mix them later without re-testing. Ideally, go through them all in one sitting then pair them up considering if you want pairs to be from the same manufacturer etc. With your amp matching pairs in each channel should sound fine so long as the 2 pairs aren’t wildly different or aged. Yours look like AVO numbers but they could be from KT88 or 6550 settings or a mix of both. On AVO characteristic settings KT88 nominally gives 57mA and 6550 gives 87mA but I found KT88s can vary 40mA to 70mA from bulk new stock for example so never buy un-matched. I look for 10% match on current and similar slope mA/V. |
#4
|
||||
|
||||
Re: Valve matching with a KiT6550
Quote:
|
#5
|
|||
|
|||
Re: Valve matching with a KiT6550
Hi John and Richard,
Thank you for your comprehensive replies; reassuring and I have some testing to do. Presume I can do the usual voltage checks from the instructions? Last time I did this (earlier this month: replaced R35-R38) with a matched set from 2011 the anodes (pin 3) were 451, 452, 452, 452 volts for valves 6, 8, 5, 7 respectively. Pretty close. Cathode pin 8 was: V6 pin 8 49.9v V8 pin 8 47.3v V5 pin 8 42.7v V7 pin 8 45.7v So I can test like this for the remaining matched pairs. John |
#6
|
|||
|
|||
Re: Valve matching with a KiT6550
Quote:
Good shout though, |
#7
|
|||
|
|||
Re: Valve matching with a KiT6550
John, yes voltage checks in the amp are important and can be a useful pointer to valve condition but they won’t show the mA/V slope which tends to go off with age.
A valve tester sets the valve up with characteristic Anode, Screen, and Negative grid voltages then checks the mA current and mA/V slope at those settings, KT88 Anode 250V Screen 250V Neg Grid 20V 57mA current and 8mA/V slope 6550 Anode 400V Screen 225V Neg Grid 16.5V 87mA current and 9mA/V slope I use KT88 settings for both valve types as (as those voltages are available on the AVO160) and both valve types give very similar results. Your valves with 2 values may be showing current and slope results. I too have a matched set of 3 Svet winged C here, 2 marked with 50 7 and the other 54 7 (checked on KT88 settings) and if I get another I’ll check them all together... |
#8
|
|||
|
|||
Re: Valve matching with a KiT6550
I have just read this by Richard, - I too have a matched set of 3 Svet winged C, - Last week, I had a better tidy up and found a single Svetlana winged C, I think I advertised it on the forum a long time ago but had no interest, so, last week during the tidy up it went into the bin, sorry Richard, if only I had known you could have had it with plesure . Bob
|
#9
|
|||
|
|||
Re: Valve matching with a KiT6550
Thank you for the additional info Richard - and I started looking in to valve testers. I don't have one so could be an interesting addition to the toolbox. I figure if it can help me save the purchase of a set of 6550s (currently £360-ish at Hotrox) it's worthwhile...
|
#10
|
|||
|
|||
Re: Valve matching with a KiT6550
Quote:
Thanks Bob I know you would, not to worry |
Tags |
valve matching |
|
|