Tubelab SE: Removing MOSFETs? - diyAudio
Go Back   Home > Forums > Commercial Sector > Vendor Forums > Tubelab
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Tubelab Discussion and support of Tubelab products, prototypes and experiments

Reply
 
Thread Tools Search this Thread
Old 18th May 2016, 08:21 PM   #1
jdrouin is offline jdrouin  United States
diyAudio Member
 
Join Date: Feb 2012
Location: Tulsa, OK
Default Tubelab SE: Removing MOSFETs?

Has anyone tried removing the MOSFETs in the Tubelab SE? If so, how did it sound afterward?

I'm thinking of trying it to see if it puts out more of the tube "sweetness."
  Reply With Quote
Old 22nd May 2016, 09:06 PM   #2
w5jag is offline w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
I think you haven't gotten any responses because, for most people, including myself, the powerdrive circuit is what sets the TSE apart from more conventional amps. This is well documented on the Tubelab site.

Removing the mosfets presents some problems - the very real possibility of damaging the pcb in the process, and then all the experimentation required to reconfigure for their loss. You might be in a better position than most to try this, as your mosfets are already off board.

I think your question might be better phrased as: "Has anyone tried grafting a Tubelab front end unto a conventional DHT final?" because that's pretty much what you have after the powerdrive is removed. Presumably Tubelab has already done that and rejected it, as evidenced by the existence of the TSE.

I thought some about this while I was mowing today, and here's a pic of my proposal to graft a Tubelab front end on a conventional DHT final, thus implementing what is probably not that good an idea to start with.

I only see a couple of reasons to pursue this: I could use a utility amplifier for some purposes around the shack, maybe at the office. I have all the parts I need on hand, so long as we are talking a mono block.

I have an abundance of five and eight pin sockets, but only one four pin on hand, so I'm thinking of going with a five pin for the power tube socket, and use the four pin for the rectifier. Then I could use 46 or 47 and play around with DH pentode or ultra linear, thus potentially adding to my knowledge base in the process to make the project less like tilting at windmills.

Or, I could use the four pin for a 45 or 2A3, and use a conventional octal rectifier.

I'm thinking the front end to graft on would be the SSE front end, mostly because I don't want to fight with making a 5842 stable. I've got a seven pin socket shown, so the tube would be a 6AB4 (12AT7 equiv) or 6C4 (12AU7 equiv). Or a nine pin could be mounted, and just use half the tube, thus making it easier to move to a 5842 in the future. Cathode bias for the final. I think my 2.5 volt tranny has the suds to do DC if I want to go that way.

No guarantees this might happen. Just thinking. Lots of other stuff I need to do worse. Looks like it is feasible at no cost other than time.

Comments and criticisms welcomed.

Win W5JAG
Attached Images
File Type: jpg WP_20160522_15_36_24_Pro.jpg (564.6 KB, 814 views)

Last edited by w5jag; 22nd May 2016 at 09:11 PM.
  Reply With Quote
Old 23rd May 2016, 04:05 AM   #3
zman01 is offline zman01  Bangladesh
diyAudio Member
 
Join Date: Feb 2011
Location: Dhaka
Quote:
Originally Posted by jdrouin View Post
I'm thinking of trying it to see if it puts out more of the tube "sweetness."
Jeff,

If your output transformers have 3k or 3.5k secondaries (in addition to the 5k you are running at IIRC), try it out. Your 2nd harmonic should go up, giving it a warmer, "tubey" sound.
  Reply With Quote
Old 24th May 2016, 05:20 AM   #4
jdrouin is offline jdrouin  United States
diyAudio Member
 
Join Date: Feb 2012
Location: Tulsa, OK
Thanks for the thoughts, w5jag. I meant no disrespect to George's design in asking about removing the MOSFETs. The idea isn't to "improve" upon it, but rather to see how the sound can be changed by making modifications.

The reason I asked about it is that I've been studying more 300B circuits and have noticed that they tend either to have (1) no CCS chips or transistors (i.e. the stuff published in Sound Practices during the early 1990s), (2) a traditional circuit topology but with a constant current source, or (3) a hybrid design (like the TSE) that uses ICs judiciously for providing consistent voltage and current regulation or ideal loads.

I'm curious to hear and learn about all of them, so I figured that the TSE could (temporarily) serve as an example of type (2) if it were possible to remove the MOSFETs and make minor adjustments to run it safely.

I'm planning to build a pair of classic 6SN7-->300B mono blocks anyway -- either the J.C. Morrison 300B SET from the Fi Primer or the JE Labs one -- not because of something lacking in the TSE but because I want to explore what can be done with the 300B. I like the sweet-sounding "romance" of those older designs, and it would be nice to have that to listen to as well as the refinement and power of the TSE.

Since the TSE was my first amp build, I can't really comment on your proposed grafting of a Tubelab front end onto a DHT final. But I don't see why it wouldn't work well given that George's gain stages are so stable. If you're less knowledgeable about DH pentodes, then maybe a Tubelab front --> DHT build would be a more logical first step. Proceed to the more familiar unknown and then, when that has been tackled, go for the pentode version.

Also, FWIW, my MOSFETs are still on-board; it's the 300B heater regulators that are now off-board, as I use one per channel.

zman01: My OPTs have 5K primaries and 8ohm secondaries. I was thinking of using them to breadboard a classic 300B circuit as described above but not sure if the 5K primary would eliminate too much of the 2nd order harmonics I'm looking for. I'd love a pair of 3K MagneQuest FS-030 or 3.5K Lundahl 1623 but won't have the budget for something like that for quite some time to come.

Last edited by jdrouin; 24th May 2016 at 05:25 AM.
  Reply With Quote
Old 24th May 2016, 05:47 AM   #5
tomchr is offline tomchr  Canada
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Calgary
There's no real harm in trying. Pluck R14 and R25 (20 kΩ, 3 W) off the board. They're easy to de-solder. Then short the MOS from gate to source (pin 1 to pin 3). You can to that with a short piece of wire soldered to the pins directly. Just be careful that you don't short to the drain (middle pin) by accident as that will let the smoke out of your components for sure.

Do note that the bias voltage for the output tube will decrease by about 5 V when you short out the MOSFET, hence, the bias current for the output tube increase rather dramatically. Thus, I strongly suggest turning the bias pot to the minimum bias current setting before making the modification. Readjust the bias to the desired bias point after the modification. Do the same when you revert back to the original SE circuit.

I performed this experiment during the development of my DG300B. When driven hard, the blocking distortion that George describes on the PowerDrive page was clearly visible on the o'scope connected at the amp output. The speakers made some rather rude noises and the woofer cones moved quite violently until the amp recovered.

If you want more tube sweetness, I suggest using a cathode follower rather than a source follower. It's the same circuit topology, just a tube instead of the MOS transistor. E88CC makes a good cathode follower. Just be careful not to exceed the Vgk, Vak, and Vbk specs of the tube. I use a cathode follower in my DG300B.

Tom
__________________
Modulus-86: 40W/8Ω @ 0.000061% THD. HP-1: 3W/20Ω, 450mW/300Ω @ 0.000032% THD. DG300B and Other Tube Circuits.
Neurochrome : : Audio - www.neurochrome.com - Engineering : : Done : : Right

Last edited by tomchr; 24th May 2016 at 06:03 AM.
  Reply With Quote
Old 24th May 2016, 05:29 PM   #6
w5jag is offline w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
Quote:
Originally Posted by tomchr View Post
.... There's no real harm in trying. Pluck R14 and R25 (20 kΩ, 3 W) off the board. They're easy to de-solder. Then short the MOS from gate to source (pin 1 to pin 3).

Tom
Tom, if you had posted that about eight hours earlier, I probably would have tried it last night. Wife and child were out of town, so I spent the evening stripping a dumpster dive Eico HF 32 down for iron and parts. I'm supposed to be reworking a closet, so if I skip another night, bloodshed ( mine ) might be involved.

Quote:
...I'm planning to build a pair of classic 6SN7-->300B mono blocks anyway -- either the J.C. Morrison 300B SET from the Fi Primer or the JE Labs one -- not because of something lacking in the TSE but because I want to explore what can be done with the 300B. ....

Jeff, if you are getting into building, you need to put the Claremore hamfest ( usually February or March ) on your to do list - lots of good tubes and parts to be found there. A couple of years ago I dragged home a sackful of micanol base 6SN7's and 6SL7's for a buck per. Also, Joplin which is usually in August. Again, lots of good tubes and parts to be found. WBOSND was there last year selling tubes.

Jeff's question has really got me interested in making a simple hybrid DH utility amp, though.

Crazy question: I intend for the amp to be mono, but I see stereo inputs as being a useful feature, since almost everything, including some dual band radios, has stereo output these days. Here's the crazy question: use the SSE front end with one half of the 12AT7 ( or 12AU7 ) taking each channel, then apply one channel to one grid of a 46, and the other channel to the second grid of a 46.

Will I get a normal mono output mixed from both channels? Try it and see what happpens?

Win W5JAG
  Reply With Quote
Old 25th May 2016, 04:14 PM   #7
w5jag is offline w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
Thanks for the idea, Jeff! Let the experimenting begin.

I cut the left side of the "chassis" in about an hour and a half last night. I'm not really a fan of cake pan construction - the metal is flimsy and harder to work with than a real chassis, imo. About all it has going for it is that it is cheap, and Wal Mart has a decent selection. Instead of a punch, I used a nibbler and file for the tube socket. I was afraid a punch would leave the flimsy metal so wavy I'd never be able to beat it back out.

I cut the chassis to accept either transformer pictured. Both are about 250-0-250, but the smaller one lacks a 5 volt secondary. The bigger iron could be useful for bigger tubes, and I want the ability to accomodate those with minimal hassle.

I haven't cut the right side yet - other than the location for the electrolytic cap, I'm undecided about the type of rectifier tube ( or solid state ) and the front end tube type.

Power junk will go on the rear, line in and speaker out on the front.

I'll make a new thread once it gets closer to fruition so as not to further pollute Jeff's thread.

Win W5JAG
Attached Images
File Type: jpg WP_20160524_21_39_09_Pro.jpg (601.1 KB, 761 views)
  Reply With Quote
Old 26th May 2016, 09:06 PM   #8
jdrouin is offline jdrouin  United States
diyAudio Member
 
Join Date: Feb 2012
Location: Tulsa, OK
Well, given your initial trepidation, Win, I didn't expect you to try something like this, so I feel amused, impressed, and guilty all at the same time!

I will definitely follow your build thread with interest. And thank you so much for the tips on the hamfests, which sound really exciting. Claremore is really close by, and I'll try to find other ones too.

tomchr: thanks for the suggestion on how to do it. Given the chassis and layout I have right now, I don't think I could short the MOSFETs and add another pair of tubes as cathode followers. But I'll keep thinking on this.

The TSE is so good at what it does, that I'm thinking I'll leave it as-is (except for some minor capacitor tweaks) and just do new point-to-point builds. I might do an SSE as point-to-point monoblocks, since I like monoblocks and love the idea of a utility amp for tube rolling and UL/feedback/rectification experimentation, or a headphone amp, or a Morrison 300B with 3K Transcendar OPTs (they seem to be still in business), or a 300B/2A3 amp with Magnequest 2.5K OPTs and switching for heater voltages, or, or, or...

... Joplin hamfest in August, eh?

Last edited by jdrouin; 26th May 2016 at 09:10 PM.
  Reply With Quote
Old 3rd June 2016, 03:50 PM   #9
w5jag is offline w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
Quote:
Originally Posted by jdrouin View Post
Well, given your initial trepidation, Win, I didn't expect you to try something like this, ....

WTH, since the little bit I know comes from making mistakes, why not rig up an inexpensive flexible test mule, make a few more, and learn something? Maybe something useful or interesting will come from it.

With a hand selected 12AT7, the SSE front end should have no trouble swinging enough volts to easily drive a 45 at 45 B+ voltages, so I'm not sure I'm even going to fool with it. Still undecided. The 46 and 47 are much less popular, so I'll probably start with those.

I'm fairly far along on it, btw, the tube rectifier ( 6X5 ) power supply is installed and putting out all the filament voltages, and unfiltered B+ DC.


Win W5JAG
  Reply With Quote
Old 3rd June 2016, 04:39 PM   #10
diyAudio Member
 
Tubelab_com's Avatar
 
Join Date: Jul 2005
Location: West Virginia panhandle
Along the way to coming up with the PowerDrive and the TSE, I built a 45 based amp without mosfets. At low volumes the sound was pretty similar to a mosfet enhanced TSE. The mosfet or a good cathode follower (high Gm pentode) allows the 45 to be driven closer, and into the positive grid region without distortion or blocking. Given that the 45 is only good for 1.5 to 2 watts, most systems see this region far more often than you realize unless you have really efficient speakers (97db+).

The 45 is actually rated for A2 or AB2 operation and a good pair will produce about 20 watts in AB2 push pull.

Although not specified, the 300B responds to A2 and AB2 P-P, but doesn't provide much more power. I can get 10 watts in A2 SE and 30 watts in AB2 P-P.

Most of the "tubeyness" found in a SE amp, and especially a DHT SE amp, comes from the higher levels of 2nd harmonic without the higher order harmonics being too high. This comes from the interplay between the driver tube, output tube, OPT and speaker system. In any two stage amp some of the 2nd harmonic from the driver tube will be cancelled in the output tube since they both have slightly "bent" transfer curves but are operated out of phase with each other.

Removing the mosfets may (or may not) affect this since the dynamic load on the driver tube is no longer constant. The total amp gain will go down as a result of the increased load on the driver, and may not be sufficient if it is close now. Some CD players can not drive a 300B to clipping in some TSE's depending on OPT and speakers.

Quote:
I'm thinking of going with a five pin for the power tube socket,
If you can find some 307A's go with the 5 pin socket, and leave options available for more B+. Triode wired they make great 300B's and will eat 400 volts.

I have been experimenting with a pentode driving a pentode wired 307A. It has promise, but that's all I want to say about it now.

The CCS loaded 12AT7 will drive a 45 to clipping. It won't drive a 300B to clipping from a CD player. There is not enough gain.

A CCS loaded mosfet buffered (PowerDrive) 12AX7 will work, and some 12AX7's are capable of remarkably low distortion numbers at fairly high output voltages in this manner.....some just roll over and distort.
__________________
Tubelab, it's 5 year mission. To explore strange new tubes, to seek out new circuits and topologies, to boldly go where no tube has gone before......
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Tubelab SE zxx222 Tubelab 172 4th September 2016 12:53 PM
Mosfets for Tubelab SE? mctavish Tubelab 54 9th January 2015 12:18 AM
F/S Tubelab SE stonehome Swap Meet 14 25th December 2014 10:56 PM
Tubelab SE vs Simple SE Pez Tubelab 25 29th October 2011 02:16 AM
Need Help with Tubelab SE lth1 Tubelab 9 1st October 2010 09:06 PM


New To Site? Need Help?

All times are GMT. The time now is 03:16 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2
Wiki