Go Back   Home > Forums > > >

Tubelab Discussion and support of Tubelab products, prototypes and experiments

Reply
 
Thread Tools
Old 19th October 2020, 04:30 PM   #1
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
Default 3D printed SSE parts and another DH SSE

A few years back a forum member posited a query about a MOSFET-less TSE, and over the course of a lengthy thread, an SSE with directly heated final amplifier tubes emerged.

Tubelab SE: Removing MOSFETs?

The experimental DH SSE that emerged ( 2A3 finals ) has worked without any component failures or other issue for the last three years. The TSE and TSE II have nothing to fear from that amp, but it is a nice and very simple beginner DH amp nonetheless.

The chassis - a 12 x 10 inch piece of scrap plywood, plucked from a trash pile ( literally ), bugs me. Likewise, the exposed voltage, while acceptable at the time, due to changing circumstances is now unacceptable.

Another thing that has bugged me was the level of damage caused to the SSE PCB to make a DH amp. It's terrific for experimenting, but it's not for the faint hearted builder. Also as part of this thread, will be an attempt to build a DH SSE using readily available, directly heated, tubes that does not damage the PCB in any way - all modifications will be fully reversible.

I hate metal chassis working, even though it is the best choice for an amplifier, imo. I have zero woodworking skills, so that's out. Recent acquisition of some 3D printers has caused me to consider whether or not a 3D printed chassis can successfully support a tube power amplifier, specifically, the experimental DHT SSE and unmodified SSE amps.

Printing a breadboard or complete chassis may or may not work. It may be that the heavy power supply will need to be separated from the main chassis. It may be as simple as a plastic printed plate emulating the breadboard piece of wood, or maybe mounted on an inverted metal chassis, or a nice wood base. Heat from the tubes and power transformer may be a deal breaker. It may be a complete waste of time, plastic and bandwidth.

To start somewhere, I made a template of the SSE PC board, and printed it out, and it looks pretty good, but I don't have a board here at the lake to match it up with and make sure all the holes correspond satisfactorily. It may need some minor tweaking. I made a quick 10 x 12 x 2 inch model chassis ( the same size as my Hammond chassis / cage based child resistant SSE ), with the SSE template embedded, just to get an estimate of the economics. At my current filament price for high temp PLA, the 12 x 10 x 2 chassis would be about ten bucks at 50% infill, and will decline some as more holes are added for transformers, meters, etc. So, it looks like the economics are favorable.

I am still in the middle of another project that has turned out to be more difficult / time consuming than I had anticipated; there will a lot of trial and error involved here; and I can only print the big parts at my office where I have the big printer, so this project will play out over some length of time just like its predecessor, but if you are looking for a reason to build another SSE ( and who isn't? ) this could be that reason.

EDIT 11/27/2020

I have made the pcb template public on TinkerCad, so it can be downloaded and copied for any purpose. It is tagged under "tubelab" and "w5jag" so a search for either tag will bring it up.

This way anyone can download it, ungroup the parts in TinkerCad, and tweak it as needed for their personal printer, or modify it for any other use. I will do likewise with any other models made for this project.

I did this because while it fit to the pcb like a glove when printed on my big 500mm Creality, it was off a bit when printed on my box of parts printer. I suspect the calibration on the factory built Creality is accurate, whereas my box of parts printer is not at this time. So this way folks can tweak it as needed for their printer.
Attached Images
File Type: jpg DH SSE.jpg (190.7 KB, 357 views)
Attached Files
File Type: zip SSE pc board template.zip (6.9 KB, 17 views)
File Type: zip proto sse chassis.zip (5.2 KB, 10 views)
__________________
Win W5JAG

Last edited by w5jag; 27th November 2020 at 08:17 PM.
  Reply With Quote
Old 19th October 2020, 09:18 PM   #2
Tubelab_com is offline Tubelab_com  United States
diyAudio Member
 
Tubelab_com's Avatar
 
Join Date: Jul 2005
Location: West Virginia panhandle
I bought a 3D printer several months back, but due to changing circumstances, it's still in its shipping box.

I therefore have no experience with PLA, but my first Lexan TSE was not really the first...the first was made from some dumpster clear plastic which was "clearly" not polycarbonate, because it melted. OK, it didn't really "melt" but sagged pretty bad after only a few weeks of heat from the tubes.

I ran 45's, NX-483's and 300B's in that amp back then. It has seen only NX-483's in the last 12 years or so.

The Lexan version still looks the way it did when I made it nearly 15 years ago.
__________________
Tubelab, I blow stuff up so that you don't have to.
  Reply With Quote
Old 19th October 2020, 10:29 PM   #3
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
I have some clear PLA, but I have not tried it yet. I see that Push Plastic has some of the "high heat" PLA in clear, so I may pick up a spool of it:

Push Plastic High Heat PLA 3D printer filament

I have a spool of the grey high heat PLA, but I haven't tried it since I'm not to that point yet in the other project. The high heat stuff may or may not be necessary here - I'm going to try the ordinary PLA first and see how it works.

For just a breadboard, I don't think heat will be an issue, but it might be for a full enclosed chassis like your Lexan amps, or a 3D printed cage, without some consideration to ventilation.

We'll find out!
__________________
Win W5JAG
  Reply With Quote
Old 19th October 2020, 11:58 PM   #4
rayma is offline rayma  United States
diyAudio Member
 
Join Date: Apr 2011
Quote:
Originally Posted by w5jag View Post
I hate metal chassis working, even though it is the best choice for an amplifier, imo.
I have zero woodworking skills, so that's out.
Same for me, and a good option is a laser cut flat metal top plate, on a prebuilt wood base.
Or a local person could build the wood base, lots of woodworkers around. Some examples:
Laser-cut metal parts Shipped Fast | Instant Quotes
HWCHAS1310AL Hammond Mfg. Enclosures | Hawk Electronics
Kalos Wood Amplifier Cases for Valve Tube and Solid State

Last edited by rayma; 20th October 2020 at 12:02 AM.
  Reply With Quote
Old 22nd October 2020, 06:42 PM   #5
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
I tweaked the template a bit to give a better correspondence around the tube sockets, and made a partial print of the first five or six layers of the template.

This one looks usable - the holes for the tube bases may or may not need to be opened up a bit for more ventilation, and some ventilation for the 10M45's might be needed.

I tried to edit the first post, but apparently attachments cannot be added or removed, although it looks like the text can still be edited.
Attached Images
File Type: jpg SSE pcb test print 10_22_20.jpg (150.3 KB, 281 views)
Attached Files
File Type: zip SSE PC board template 10_22_20.zip (7.6 KB, 10 views)
__________________
Win W5JAG
  Reply With Quote
Old 25th October 2020, 04:27 PM   #6
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
The predecessor 2A3 SSE was characterized with a Heathkit AA-1 Audio Analyzer ( assembled by me from an unbuilt kit four or five years ago ), but I am going to try to use something a bit more modern on this build.

I got a thing called a Red Pitaya ("RP")

Red Pitaya - Stemlab swiss army knife for engineers

to use in another project, and never really considered using it for audio since I have the aforementioned new AA-1, and a companion HD-1 THD analyzer and Leader audio generator, but it looks like the RP might be useful for audio equipment as well. If it works, a shelf full of equipment can be replaced with something about the size of a pack of smokes.

The first screenshot is a 1 volt ( 7dBm ) 1 KHz FFT of the channel one signal generator output looped back into the channel one input, and displayed on the FFT spectrum analyzer function. It looks like about 0.05 - 0.06% THD to me, so should be plenty usable here. I measured my cheap hamfest scrounge Leader generator at about 0.04% with the HD-1, so this seems comparable to my old analog stuff, with the additional advantage of seeing a spectrum distribution instead of just a % number. THD is not hard to calculate, and there are online calculators that can do it from raw numbers.

The 1KHz sine wave looks nice as well. The other shots are 60 Hz, 6 Khz, 18 KHz and 19 KHz, all at 1 volt loop back, for two tone IMD testing. I don't understand the value of two tone IMD testing with 18 and 19 Khz tones ( maybe someone can explain it to me ), but the RP can do it, so I guess I will try it. The AA-1 has onboard 60 Hz and 6 KHz tone generators, but I usually used the leader to inject 200 Hz as a low frequency test tone, since the little OPT's I use do not handle 60 Hz very well.

The AA-1 is actually a very useful all in one piece of test equipment even in this era, if you can find a good one at a low price.
Attached Images
File Type: jpg 1 Khz 1 volt 0.05 THD.jpg (99.8 KB, 252 views)
File Type: jpg 1 KHz 1 volt scope shot.jpg (79.7 KB, 252 views)
File Type: jpg 60 Hz 1 volt FFT.jpg (93.8 KB, 252 views)
File Type: jpg 6 KHz 1 volt FFT.jpg (95.6 KB, 15 views)
File Type: jpg 18KHz 1 volt FFT.jpg (86.6 KB, 15 views)
File Type: jpg 19 KHz 1 volt FFT.jpg (85.9 KB, 20 views)
__________________
Win W5JAG
  Reply With Quote
Old 28th October 2020, 04:32 PM   #7
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
One of the things I had not given any previous thought to is the potentially poor manufacturing tolerance of cheap Chinese tube sockets. I grabbed three octals and one nine pin at random, stuck them in an SSE board, and while hard to tell from the parallax, they do not perfectly center in the outlines on the PCB. The socket stuck in the rectifier position is a particularly bad offender.

Regardless, I think there is enough tolerance in my template to be unaffected by this, although depending on the sockets, it may or may not look perfectly centered when viewed directly from above.

When passing the template over the tube sockets, it looks like it is sized to permit adequate ventilation around the sockets, so long as the template is below the upper surface of the octal tube sockets. Heat rises, so high heat PLA may not be necessary here.

For mounting above the upper level of the tube sockets, it fits the standard octal base and modern coin base like a glove, although heating of the PLA, even high heat PLA, may be problematic for this type of mounting. Fat bottom tubes like 6550 would need a much larger opening.

Tentatively, I think the most recent template is good to go for now.
Attached Images
File Type: jpg poor tolerance.jpg (161.3 KB, 50 views)
File Type: jpg IMG_20201028_094501837.jpg (171.8 KB, 85 views)
File Type: jpg IMG_20201028_094615502.jpg (162.0 KB, 86 views)
File Type: jpg IMG_20201028_095550671.jpg (190.4 KB, 72 views)
File Type: jpg IMG_20201028_095708840.jpg (179.1 KB, 68 views)
__________________
Win W5JAG
  Reply With Quote
Old 21st November 2020, 10:36 PM   #8
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
When considering how to use directly heated tubes on the SSE, something it was clearly not intended to accommodate, my first thought was that you canít get there from here. The 20 - 20 hindsight reality is that you can, easily, and anyone who wants an entry level, directly heated amp, should be able to produce a very good one with minimal effort and expense using an entirely unmodified SSE board.

I am not a tube historian by any means, but after De Forrest added the third element to the vacuum diode, a base / socket system with only four pins was entirely adequate for the vacuum tubes of that era, one pin being required for the plate (anode), one for the control grid, and two for the filamentary cathode. Thus, most directly heated tubes exist on a four pin base. As more electrodes were added to the tubes, and indirect cathode heating came to prevail, the number of pins on the base rapidly expanded, first to five, then six, seven, and finally by the mid Ď30ís or so, the eight pin octal base was well established.

The SSE uses the eight pin octal socket, therefore, to accomplish the goal of making a directly heated amplifier without any destructive modifications to the stock PCB, the first iteration of this project will be confined to directly heated tubes with an octal base.

The type 2A3 triode tube evolved into a 6 volt variant, the 6A3. The 6A3, when mounted on an octal base, became the 6B4G. NOS and new production 6B4G is readily available. 2A3 tubes have been proven to work satisfactorily with the existing SSE front end, provided that some component values are changed to reflect the lower B+ voltage that will be used, thus there is really no additional tinkering to be done to make a 6B4G SSE.
__________________
Win W5JAG
  Reply With Quote
Old 21st November 2020, 10:41 PM   #9
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
The SSE board uses the common 7AC octal tube pinout, The 7AC format assigns the pins as follows:

1. No connection, or metal shell
2. Heater
3. Anode ( plate )
4. Screen grid
5. Control grid
6. No connection
7. Heater
8. Cathode and suppressor grid ( or beam forming plates )

The SSE further incorporates a jumper between pins 1 and 8 making it compatible with the 8ET pinout ( EL34 / 6CA7 ), which brings the suppressor grid, or beam forming plates out separately to pin 1.

The 6B4G uses the 5S pinout. 5S connections are as follows:

1. No connection
2. Filamentary cathode
3. Anode ( plate )
4. No connection
5. Control grid
6. No connection
7. Filamentary cathode
8. No connection

The only conflict between 7AC and 5S are pins 2 and 7. 7AC uses these pins for the tube heater, while 5S uses these pins for the combined cathode and heater of a directly heated tube. Other than the pin 2 and 7 connections, the 6B4 is a drop in tube for the SSE.

In practice, the conflicting uses of pins 2 and 7 is of no consequence. The SSE is a cathode biased amplifier. The directly heated SSE will also be cathode biased. Cathode biased directly heated power tubes, so far as I know, require a separate filament transformer or winding for each power tube, preferably with a center tap, thus the SSE filament circuitry, for the power tubes, is of no use to us anyway, and our directly heated tubes need to be isolated from it.
__________________
Win W5JAG
  Reply With Quote
Old 21st November 2020, 10:46 PM   #10
w5jag is online now w5jag  United States
diyAudio Member
 
Join Date: Feb 2007
Location: Arkansas
The captain obvious way of doing this is to just take the tube socket pins for 2 and 7, and bend them up and away from the PCB. All of the other tube socket pins will be soldered to the PCB as normal. The isolated pins on each tube socket will be used as solder points for the leads from that tubes filament transformer; the transformers center tap will be soldered to pin 8, thereby establishing the connection point for the cathode bias. To ensure there is no contact between pins 2 and 7 and the PCB, the socket can be elevated slightly, or some electrical tape or other insulating material can be applied to the exposed pads for pins 2 and 7.

2.5 to 3K would be the usual OPT load, but 5K would work just fine. The cathode resistor will be 750 ohms.

A 300 - 0 - 300 power transformer with a 200 ma current capability should be satisfactory. A 6.3 volt filament winding at about an amp is needed for the 12AT7, and a 5 volt, minimum 2 amp winding is required for the rectifier tube. Each 6B4G tube will require its own 6.3 volt center tap, minimum 1 amp, transformer for its filament / cathode. All of these windings can exist on one core, indeed, the transformer I am using for my 2A3 SSE has the high voltage, 5 volt, and separate 2.5 volt center tap filament windings all on one core. I still have to provide a small filament transformer for the 12AT7.

Conceptually, thatís pretty much it- a really simple project. Other than the 6B4 tubes and a couple of small center tap filament transformers, a DHT SSE is not materially different than building a regular SSE. The component changes needed to the front end are discussed in the linked thread above that begat the original DHT SSE, and Iíll identify them again in a later post.
Attached Images
File Type: jpg IMG_20201121_153227742.jpg (127.1 KB, 54 views)
__________________
Win W5JAG
  Reply With Quote

Reply


3D printed SSE parts and another DH SSEHide this!Advertise here!
Thread Tools

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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
3D Printed CLD Speakers? Patrick Bateman Multi-Way 16 12th August 2020 06:39 AM
My 3D printed ToneArm Tankpopper Analogue Source 24 9th December 2019 03:13 PM
3D Printed Chassis Jennifer G Parts 67 30th March 2016 02:46 AM
2nd in a Series - Re-ribbon, new transformer and 3D printed parts for MXL R144 guitartoys Construction Tips 0 24th January 2016 03:34 AM
PARTS PARTS PARTS!!! Infocus Powerview 820, triplet, ballast, light, power supply!!! marioinla Swap Meet 1 31st August 2004 05:22 PM


New To Site? Need Help?

All times are GMT. The time now is 12:54 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio
Wiki