Tube preamp for Pass Labs amp


I am considering changing around my system and I am interested in knowing what (tube) preamps would be a good match to Pass Amps. Any suggestions? Less than $3000 would be nice new or used.
aronsss
i'll second Lector zoe which pairs marvelously with my Pass 250 (and my Lector cdp 7)
Atma Sphere MP1 works great with my Pass xa60.5's - Mvaudio -

The Atma Sphere MP1 is my dream since long time .. and I guess it's superior to the Lector ZOE .. Mvaudio
But Aronsss wrote under 3000 $USD and I don't think you can find an MP1 for less than 3000 $.

I'm happy to read Icorem uses the Lector ZOE on his Pass Labs 250 .. I'm using the ZOE on my Parasound Halos JC1's and it just replaced the Pass Labs X1.
But to replace the good quality of the X1 I needed to add the 6H30-DR to the ZOE.
So I suggest Icorem to try these fantastic tubes on the already super good Lector ZOE to get the nirvana with few dollars.
Just put not less than 250 hours on them!

My Cent!
Audiofeil, I'll bow to your experience, but the Pass amps appear to be less than ideal candidates for many tube pres, when one considers the the low input impedance and the preference for balanced inputs. Both of which would seem to disqualify many more tube than ss candidates. Perhaps you'd like to share more with us?
Unsound, almost any tube preamp manufacturer has to contend with transistor power amps that have a lower input impedance. Even the old Dyan Pas 3X was an example of that- the X designated that it was set up to work with amps with an input impedance of 10K ohms.

And of course these days there are tube preamps with balanced outputs too. The one we make has been in production longer than any other - and it has direct-coupled outputs so the 10K input impedance of transistor preamp is no worries at all.
Atmasphere, I know some pres have the the prerequisites, but it seems as though there are a lot more tube single ended pres with high output impedance's than ss ones. The Pass amps are particularly low input impedance, and the design seems to be particularly built around maximizing the use of balanced inputs.