Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Instruments and Amps Everything that makes music, Especially including instrument amps.

Submini 6088 and 5672 PP amp, please critique
Submini 6088 and 5672 PP amp, please critique
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 19th April 2019, 02:22 AM   #1
sahazel is offline sahazel
diyAudio Member
 
Join Date: Mar 2019
Default Submini 6088 and 5672 PP amp, please critique

I've been experimenting with making low-power guitar amps using subminiature, direct heated tubes. I've gotten to a design that sounds decent and can produce reasonable bedroom practice volumes, and I figure this is a good point to stop and solicit some feedback.

What I have is two 6088 gain stages, a 6088 anode follower paraphase inverter, and two 5672s in push-pull. Originally I had two parallel 5672s operating SE, but I find the PP setup can get a clean sound at a slightly louder volume. All tubes of the same type have their filaments in parallel, and they tap a series chain of tubes, resistors, and diodes to get the desired voltages for bias. The output transformer is a Hammond 125A.

The power supply, not shown in the schematic, is currently a 44v CT transformer, bridge rectified to produce 60vdc, 30vdc pulled from the center tap, and then an arrangement of two LM317 regulators producing 45v and 7.15v.

Please note, this is my first foray into tube amps, and I only just started learning about analog electronics a few months ago, so I may well be confused about a great many things. Also please excuse my schematic skills this is my first one!

Some questions:
  • Any glaring mistakes or strange choices here? I don't really know what I'm doing.
  • Relatedly, am I going to destroy my tubes running them like this?
  • When tube data sheets specify a typical plate voltage, is that the source HT voltage measured before the plate resistor, or the quiescent voltage measured at the plate?
  • How should I be choosing plate resistor values? I've been doing it experimentally with a sine wave generator and oscilloscope, just trying to get the maximum clean volume at ~350hz, but I have only the vaguest understanding of the implications for bias points and current flowing through the tubes.
  • Speaking of current, I assume the current flowing through the tubes winds up on the filaments, and thus with my setup there's going to be some amount of interaction between all the tubes. Right?
  • The 6088s want 80k load impedance. I understand what the load is at the output transformer, but I have no idea how to work out / adjust what the impedances between stages are.

Next I'm thinking about switching the gain stages to 5678s, since I think those can get a bit louder than 6088s.

Any thoughts or suggestions would be appreciated!
Attached Images
File Type: png Schematic_6088-5672-PP-amp_Sheet-1_20190418183117.png (72.4 KB, 125 views)
  Reply With Quote
Old 19th April 2019, 09:28 AM   #2
Thomasha is offline Thomasha
diyAudio Member
 
Join Date: Jun 2012
Hi,

nice layout, only five tubes and probably a nice 200 mW punch.

I see you're using a series string scheme like I used in my last builds. I usually decouple the intermediate stages with large capacitors to ground the signal and avoid feedback. I see that you only used one capacitor, you could give it a try and see if it improves your sound.

Interesting approach of using the diodes in parallel with the heaters, allows you to bypass the extra current of the cathodes while dropping the right voltage between stages. If you use the 5678 you can drop the resistors, cause they require 2x the current of the 6088.

What I'm missing is the bias of the first stage through a 1M resistor, after or before R8, so that you have a ground reference there. With the negative side of the filament at 1.25v you can check the load line for the first stage at the datasheet. It will give you an idea of the size of the plate resistor.

R8 is also very low, normally you find something in the 33-68k range to cut some radio frequencies.

Second stage is also has a -1.25v bias, could be biased same as the first stage (same load line) or you could get more gain out of it. The first stage has the limitation that it gets microphonic at some point. You can try increasing the plate resistors to get more gain. Datasheet says 30k at 45V.

As specified at the datasheet you won't burn them because they can't draw enough current. Just don't short the high voltage to the filaments and you're fine. The bias of the 5672 also looks ok, so no worries there.

Answering your questions about the HT, it should be before the plate resistor. Assume the tube is drawing 0mA, so that the plate voltage is the voltage at the HT node.

Since you're using the filament voltage as the bias point, you can only change the gradient of the load-line, so the try-and-error approach is OK. It will give you the real ideal bias for specific tube you're using. The datasheet is not as good when it comes to those hard to read regions. I would test it with 800hz or 1000hz, more in the middle of the guitar frequencies.

Could make a video to show how it's sounding? That would be nice.

Congratulations for your first build! Direct heated tubes are a bit more tricky to get right, and there is little or none information about running the filaments in series. The advantage is that 60v is safe enough to play with.
  Reply With Quote
Old 19th April 2019, 09:46 AM   #3
mondogenerator is offline mondogenerator  England
diyAudio Member
 
mondogenerator's Avatar
 
Join Date: Sep 2009
Location: GB
Following with interest 😀

I have only just started out learning with tubes/valves and I am on a similar road to discovery.

I'm not brave enough to try DHT just yet, although I have some 1j29b (1sh29b or 1zh29b) and 1p24b which i can try soon!
__________________
I must be a child of good times... the scariest thing I've seen done thus far, has been connecting a cathode bypass capacitor reverse polarity...Then rushing to cut the HT!
  Reply With Quote
Old 20th April 2019, 07:59 PM   #4
sahazel is offline sahazel
diyAudio Member
 
Join Date: Mar 2019
Quote:
Originally Posted by Thomasha View Post
Interesting approach of using the diodes in parallel with the heaters, allows you to bypass the extra current of the cathodes while dropping the right voltage between stages.
My main goal with the diodes was to ensure I won't put more than 1.4v across the filaments. I originally had another two in parallel with R3, but removed them in an attempt to get closer to the right voltages, and I thought the sound improved a little bit, so I left it that way.

Quote:
Originally Posted by Thomasha View Post
What I'm missing is the bias of the first stage through a 1M resistor, after or before R8, so that you have a ground reference there.
Ah, right. I've added one and it doesn't seem to have much effect, but I have had stages drift out of bias when I've forgotten these before. I don't 100% understand how that works and why a ground reference is necessary. Do you?

Quote:
Originally Posted by Thomasha View Post
R8 is also very low, normally you find something in the 33-68k range to cut some radio frequencies.
Added a 47k. No apparent change, but I suppose this makes it more robust in less friendly RF environments. I guess this forms a low pass filter with the tube's plate-to-grid capacitance? Good to know.

Quote:
Originally Posted by Thomasha View Post
Second stage is also has a -1.25v bias, could be biased same as the first stage (same load line) or you could get more gain out of it. The first stage has the limitation that it gets microphonic at some point. You can try increasing the plate resistors to get more gain. Datasheet says 30k at 45V.
What I was finding was that with the second stage plate resistor above 15k, I had the phase inverter clipping the high rail long before the second stage clipped its high rail, almost regardless of the PI plate resistor value. Based on your comments I did some experimentation, and ultimately discovered that the tube I had in the PI stage was particularly bad this way. Swapping a different tube in, I can get the second stage much closer to clipping before the PI clips.

In general, it seems all stages are much closer to clipping at cutoff than at saturation. Makes me wonder if I should try biasing the grids at 0v despite what the datasheet says.

I'll see about making a video.
  Reply With Quote
Old 22nd April 2019, 03:21 AM   #5
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: Maine USA
Quote:
Originally Posted by Thomasha View Post
...What I'm missing is the bias of the first stage through a 1M resistor, after or before R8, so that you have a ground reference there....
Yes, ALL grids should know where to hang. This one near zero V. In *minimalist* guitar amps we sometimes rely on the guitar having DC conductivity, and this is generally true. For passive guitars. Not active axes. Not fuzz-boxes.

Agree a 1Meg is a good idea. You do not need it, until the day you connect a no-DC-conductivity source, and then the grid floats up/down out of its happy zone, the amp goes numb.
  Reply With Quote
Old 23rd April 2019, 03:57 AM   #6
sahazel is offline sahazel
diyAudio Member
 
Join Date: Mar 2019
Very helpful comments, PRR. You've filled in whatever was missing in my understanding so that ground references now make sense.

After experimenting, I found that setting the negative sides of my 6088 filaments at ground did indeed allow for more clean headroom than setting them at 1.25v, despite the data sheet specifying a typical -1.25v bias for the grid. I don't understand why, but I wonder whether the data sheet characteristic values are for pentode operation, whereas I have them triode-strapped. Doesn't make any sense when I try to draw load lines, though. Even after this change, all 6088 stages still clip first at cutoff rather than saturation.

Anyway, here's a short video of what I've currently got: YouTube

I'm also attaching an updated schematic. Basically I took some suggestions, moved the filament bias for the 6088s, and switched from using negative feedback to hold the PI to unity gain to just voltage-dividing the input. I found the negative feedback approach made PI clipping more abrupt, while just reducing the input made it act more like what was going on at the other rail in the 2nd gain stage.
Attached Images
File Type: png Schematic_6088-5672-PP-amp_Sheet-1_20190422204020.png (77.4 KB, 92 views)
  Reply With Quote
Old 24th April 2019, 02:39 AM   #7
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: Maine USA
Biasing too close to zero makes a low impedance at the grid. In the intended application (very minimal radios) this was bad for performance. "Low" may still be hundreds of K-ohms, which is often acceptable in audio. And we sure are not worried about melting the tube at low voltages with large plate resistor. So just do what works.

Anyway you *are* negative bias. You have small tube filaments on 0V and +1.25V. Grids at 0V. On-average, the filament cathode is 0.625 above ground; grid is -0.625V from average of filament. Yes, small positive grid drive causes some grid current at the low end of the filament. But everything in there is tiny, so the current is tiny.

Last edited by PRR; 24th April 2019 at 02:42 AM.
  Reply With Quote
Old 27th April 2019, 09:17 AM   #8
sahazel is offline sahazel
diyAudio Member
 
Join Date: Mar 2019
Tried switching the 2nd stage and PI to 5678s today (leaving a 6088 in the first stage because it already has plenty of gain to drive the 2nd stage into distortion). I still have some tuning and experimentation to do, but already this is a big improvement: lots more clean volume and punch, as well as more crunch when I crank up the first stage's gain.

The 5678 data sheet calls for a 0v grid bias, but these are also still clipping at cutoff and seem unable to reach saturation. I wonder if I should put a small positive bias on the grid — is that an okay thing to do?

I may also try bringing the plate voltage up to 60v since these can handle a little more than the 6088s.
  Reply With Quote
Old 27th April 2019, 10:56 AM   #9
Thomasha is offline Thomasha
diyAudio Member
 
Join Date: Jun 2012
There are some subminis that work well until +1v grid bias (The russian rod tubes). I'm not sure if it applies, in any case it won't drain current enough to burn. So you can give it a try.

You definetely should try running the 5678 at 60v. In my build I run them between 60v and 40v (first stage)
  Reply With Quote
Old 28th April 2019, 12:24 AM   #10
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: Maine USA
> unable to reach saturation.

Try larger plate resistors. 30K is awful low for any small tube. 100K may be better.
  Reply With Quote

Reply


Submini 6088 and 5672 PP amp, please critiqueHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
High gain DHT submini preamp+amp with the 5678 and 5672 Thomasha Instruments and Amps 9 24th March 2019 10:20 PM
PP 5672 direct heated tube amplifier help needed Thomasha Instruments and Amps 11 27th December 2017 03:38 PM
6088 Amplifier Jean-Paul_2 Tubes / Valves 2 30th January 2014 11:02 AM
i have one 5672 tube (new) east electronics Tubes / Valves 0 9th November 2010 11:30 AM
Submini Headphone Amp Help n00beR Headphone Systems 17 3rd July 2008 12:19 PM


New To Site? Need Help?

All times are GMT. The time now is 10:01 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki