5899 subminiature tube Vari-Mu

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Dmichel123

Well-known member
Joined
Nov 19, 2015
Messages
142
Location
Virginia
While searching for possible 6386 tube subs, I came across the 5899 remote cutoff pentode. The data/curves looked good to me, so I picked up a few and did some experimenting. I was able to tie two of them together and insert them directly in the 6386 socket of my Sta Level. I didn't take any measurements or do any scientific testing, but they seem to perform better in the circuit than the 5670 that the previous owner sold with it.

I ran the Sta Level for at least an hour, set for heavy gain reduction. I didn't notice any issues at all. The 5899s were pretty hot after that, may need a little air between them.

Anyway, I'm thinking of developing a socket adapter that would functionally work like a T-bar from Manley. I intend on doing more research/testing before designing and/or building the adapters. Any input would be greatly appreciated!!
 
Haven't heard of 5899 before, it might be easier to find matches with dual triodes like 6BC8, 5670 and similar tubes that are known to work well instead of 6386.
What's wrong with 5670? Seems like it gets a lot of use in this application, along 6BA6 and other pentodes which sometimes require trying more tubes for a match.
Try choosing easily available tubes, maybe even JJ 6386 because it will most likely be available for long time and works well. It could also be cheaper, compared to number of tubes and time spent to get a match during life of the unit.
Another thing is that many tube compressors used this tube, so why not choosing one and have it for all of them. Makes things so much easier...
 
6386 is too expensive.  5670 is not remote cutoff, distorts on heavy GR. There is a reason the T Bar mod sells.... The 6BA6 performs very well as a 6386 sub. BUT, If you're building a Fairchild or something with many Vari-Mu stages in parallel, heater current adds up fast and you need a very heavy, expensive PT...
 
I meant to come back and provide an update here and forgot.  I ran some analysis on the Sta Level while it was still on my bench, but forgot to save the results..... Smh. I do remember that the two 5899s outperformed the 5670 in deepish GR. Triode connected 5899s were at around .4% THD and the 5670 was at around 1.1%. I have designed a couple PCBs that would work with these: http://m.ebay.com/itm/1set-9pin-tube-converter-adapter-kits-for-el84-12ax7-ecc83-12au7-6gh8a-6u8-ecc82-/
One of the PCBs adapts a 6386/5670 socket to two triode connected 5899s, the other to four 5899s.
 
Looks like the plates on the 6386 see 230V in the Fairchild, but the 5899 is only rated to 165V max (or at least that was the max of the test? The tube data sheet wording seems ambiguous).

How have the tubes held up?
 
5899 tube data - https://frank.pocnet.net/sheets/138/5/5899.pdf

6386 tube data - https://frank.pocnet.net/sheets/142/6/6386.pdf
 

Latest posts

Back
Top