JE simple 45 SE grid leak resistor question

[thafred]

For some time now I´m into tweaking the circuit (mostly just changing parts) of my JE Labs Simple 45 amplifier and I very much love the music coming out of that little amp.

Latest tweaks, which also made the greatest impact on sound quality, included battery Bias for the 6SL7 and changing the 45´s cathode bypass caps from cheap 100µF electrolytics to 100µf/160V Audyn Caps...wow, what a difference in sound!!

I´m contemplating buying new tubes for my beloved little amp since my only pair of old RCA 45´s aren´t in the best condition and I don´t really dig the sound the TJ 45 mesh tubes put out (sounds very beautiful but misses out on "micro" PRAT the RCA 45´s are capable of!). Problem is that getting NOS 45 or 245 is really impossible in Europe and I´m afraid of having them shipped across the Atlantic. This leaves the EML45 as the only alternative as it´s easily available where I live. (shouldn´t sound too shabby from what I read on the net)

Browsing the EML homepage and EML45 datasheet I found that they want a 100K grid leak resistor for the tubes instead of the 470K i have in the circuit right now. (also found a post on AA by JAC stressing that the value must be that low for the tube to have any usefull lifetime)
my question is if this would pose a problem for my 6SL7 driver tube? It surely is harder to drive the grid if the GLR is 100K instead of 470K?
Do I just loose a little gain from changing the grid leak resistor to 100K or would I change more than that? (like bias points..ect.)
How does the grid leak resistor value influence the sound of the amp?

my amps shematic is according to simple45 (only change are the 1,2V NiMh cells floated off ground by 220R to give 1,54V cathode bias for 6sl7)

Your Help is much valued!!
thank you,
Fred

[Palustris]

I haven't tried the EML45 yet but I just read someplace that one user felt that it was better than the TJ and the NOS. I certainly can't confirm that although I have the AVVT 2A3 and 300B and find them excellent.

Regarding the change in load for the 6SL7, you are correct that it will reduce the load. You can get a good prediction of this if you know how to draw load lines. The following grid resistor is in parallel with the plate resistor so with the 470k GR the full load on the 6SL7 is 82,456 ohms approx. With the 100k GR it is 50,000 ohms approx. That is indeed a huge difference.

There are several practical solutions. First is to get a grid choke for the 45. That should make a substantial improvement in sound because most of the DHTs start to draw grid current before the signal reaches 0V on the grid. A grid choke gives a very low DC resistance to ground allowing the tube to drain off the excess electrons that collect on the grid while presenting a large AC load to the 6SL7.

The second solution that I highly recommend requires a lot more work. That would be to direct couple the driver to the OP tube grid. If you think the cathode resistor was constricting the sound, you won't believe the transparency afforded with direct coupling. It is night and day with respect to RC coupling. For more information on direct coupling Google "2A3 monkey". For a good driver for the monkey try the ECC99.

Another solution is to use a different tube than the 6SL7 that can drive the load more easily. That depends on whether you use a preamp or not. You need 55V minimum to drive the 45. If you have a preamp with enough gain you can use a 6SN7 but you will need almost 4V input to get full output from the 45.

[20to20]

Browsing the EML homepage and EML45 datasheet I found that they want a 100K grid leak resistor for the tubes instead of the 470K i have in the circuit right now.

Might want to double check that 100K spec. It is 100K for fixed bias on NOS 45 and 1M max for cathode bias. You may have seen the specs for fixed bias. You have cathode bias, correct?

[thafred]

you´ve been a great help Palustris, thanks alot!

the solution with the grid chokes sounds very interesting.

Would a choke like the Lundahl 1670 work well? 1670.pdf
Which values for the choke would you recommend otherwise?

I´ve seen the Monkey schematic before and love the idea of direct coupling driver and output tube! (someday I design a circuit like that but I need to learn more about old Mr. Ohm first

Atm I´m using a slightly modified PS1 directly into the amp (stepped attentuator) as a source and the signal is very weak so I need all the gain I can get (for me it´s more than enough in with 97db speakers) I love having as little parts as possible between source and speaker so I don´t want a pre amp for now.

[thafred]

Originally Posted by 20to20

Might want to double check that 100K spec. It is 100K for fixed bias on NOS 45 and 1M max for cathode bias. You may have seen the specs for fixed bias. You have cathode bias, correct?

I have cathode bias but I think in case of the EML 45 it applies for both. Maybe someone knows for sure? don´t want to damage that (..much too..) expensive tube!

[20to20]

Originally Posted by thafred

I have cathode bias but I think in case of the EML 45 it applies for both. Maybe someone knows for sure? don´t want to damage that (..much too..) expensive tube!

Their data sheet lists typical values with 50v on the grid. They didn't seem to address usage with cathode biasing. But if they think their tube is a direct replacement for NOS they'd have to consider a K biased user. No reason for such limited data. Pretty tube. Pretty price. Thumbs down.

[mach1]

Problem is that getting NOS 45 or 245 is really impossible in Europe and I´m afraid of having them shipped across the Atlantic.

Don't worry. I have had many DHTs shipped to Australia with no problems. Just use a reputable seller who will package them adequately.

[Palustris]

OK, I checked the EML data sheet and it clearly states: Grid resistor max 100k ohms.

So I would consult Jacmusic since they sell both the tubes and the grid chokes. They may suggest a better driver too, especially if you are using a preamp.

You are correct that setting up a 'monkey' does require a lot of juggling to get all the voltages right and ohms law is the key to getting it to all work out correctly; that and being able to draw load lines.

[thafred]

Yesterday I E-Mailed this question to Jac van der Walle and he replied that it is no problem to use 500K or greater with current production tubes and that the values were given for the old production run of EML 45! He told me that they kept the value as a safety measure, i guess it´s because there are lots of older EML tubes around.

Following my inquiry he even changed the datasheet on the emmission labs website. very neat :-)

As I allready know he is a pleasure to deal with, this guy will definately see some business with me again (allready bought my Lundahl OT and TJ45 at his shop back in 2005) For now I´ve placed the EML 45 on my X-Mas wishlist...let´s see how much my family loves me

Mach1: thanks for your opinion! Maybe I should just order the 245 from USA as I find this tube the most "sexy" looking of them all and even if I get the EML, sooner or later I will be lusting for a NOS UX245, even if it´s just for the coolnesfactor of it :-) .. I hope the seller doesn´t confuse Austria with Australia thou (experienced this several times as I was traveling in the US!...you get t-shirts here in vienna that say "no kangaroos in Austria" ;-) )

Palustris: I will eventualy order the grid chokes anyway since you made me curious how they can improve the sound of my AMP! thanks for all your help.
I plan to buy some Yamamoto teflon UX4 and Octal sockets too at JAC....