Implications of LOW B+ on the TSE-II?
I tested the 182As I found, and a couple of my 71A stash, in my MONO TSE clone (powered by a Heathkit Regulated P/S), using 5K:8 ohm cheapo Edcor OPTs. Shockingly good, particularly the 71As. I had forgotten how good the cheap Edcors are.
I'd like to build a TSE-II that could use either of these tubes, as they both require ~5K OPTs and have 5V filaments.
I could accommodate the different Plate Voltages (250V max for 182B, 180V max for 71A) with a little legerdemain in the Power Supply, e.g. different rectifier, and/or change C4, and/or add an RC filter stage. I have some ancient Shuggie 5Z4Ps (ST bottle) that the DuncanAmps PSU Designer says will work well, AND they're ST bottles about the same size as the output triodes, so that's nice.
However, not quite sure about what changes might be required to run these VERY LOW Plate Voltages. For example:
(A) Is the ratio between B+ and the Positive MOSFET supply linear all the way down to below 200 Volts B+? Or does R36 need to be changed?
(B) With a B+ of about 190V, will there be enough span in the 5842 cathode pot to get 175V at the 5842 plate? I suppose that if the pot is at ~0 ohms, the cathode is effectively grounded (which I understand the 5842 was designed-for), so I needn't worry about the value of the fixed-value cathode resistors. Is that correct?
(C) The 10M45 CCS is set at about 9mA IIRC (3V/330ohms = 0.009A).
Does this need to be changed? I assume not, but is that right?
(E) Anything else need changing for Vp of 250V or 180V? Worse, are there any deal breakers using Power Drive on 20mA DHTs?
PS the 182B yields ~1.35 Watts out, while the 71A yields ~0.8Watts at essentially the same 5K ohm load.
Both are biased at 20 mA. I assume the difference in output is due to (a) higher Vp and (b) MUCH higher filament current on the 182B. Is that correct?
PPS One alternative to a Low B+ TSE-II is a point-to-point "clone" using (1) DC filaments and (b) the 10M45 CCS on the driver.
(Parenthetically: After tinkering with these circuits for a long while, I think that the fidelity improvements from the Power Drive for DHTs are due to, in order of importance: (a) fixed bias (b) DC filaments (c) CCS and (d) MOSFET on the grid. Discuss.)
I was initially contemplating using a 6SL7 (have some nice ones languishing) with one triode section driving each output tube. I did this with very good results on a mashed-up 6AV5 SE using BATTERY BIAS (!).
HOWEVER, the 10M45 was set at 3v/1750 ohms (matched resistors) = 1.7 mA, which is BELOW the headline lower limit for the 10M45 CCS (although still on the R vs current graph); worked REALLY well, though..
I tested the 182As I found, and a couple of my 71A stash, in my MONO TSE clone (powered by a Heathkit Regulated P/S), using 5K:8 ohm cheapo Edcor OPTs. Shockingly good, particularly the 71As. I had forgotten how good the cheap Edcors are.
I'd like to build a TSE-II that could use either of these tubes, as they both require ~5K OPTs and have 5V filaments.
I could accommodate the different Plate Voltages (250V max for 182B, 180V max for 71A) with a little legerdemain in the Power Supply, e.g. different rectifier, and/or change C4, and/or add an RC filter stage. I have some ancient Shuggie 5Z4Ps (ST bottle) that the DuncanAmps PSU Designer says will work well, AND they're ST bottles about the same size as the output triodes, so that's nice.
However, not quite sure about what changes might be required to run these VERY LOW Plate Voltages. For example:
(A) Is the ratio between B+ and the Positive MOSFET supply linear all the way down to below 200 Volts B+? Or does R36 need to be changed?
(B) With a B+ of about 190V, will there be enough span in the 5842 cathode pot to get 175V at the 5842 plate? I suppose that if the pot is at ~0 ohms, the cathode is effectively grounded (which I understand the 5842 was designed-for), so I needn't worry about the value of the fixed-value cathode resistors. Is that correct?
(C) The 10M45 CCS is set at about 9mA IIRC (3V/330ohms = 0.009A).
Does this need to be changed? I assume not, but is that right?
(E) Anything else need changing for Vp of 250V or 180V? Worse, are there any deal breakers using Power Drive on 20mA DHTs?
PS the 182B yields ~1.35 Watts out, while the 71A yields ~0.8Watts at essentially the same 5K ohm load.
Both are biased at 20 mA. I assume the difference in output is due to (a) higher Vp and (b) MUCH higher filament current on the 182B. Is that correct?
PPS One alternative to a Low B+ TSE-II is a point-to-point "clone" using (1) DC filaments and (b) the 10M45 CCS on the driver.
(Parenthetically: After tinkering with these circuits for a long while, I think that the fidelity improvements from the Power Drive for DHTs are due to, in order of importance: (a) fixed bias (b) DC filaments (c) CCS and (d) MOSFET on the grid. Discuss.)
I was initially contemplating using a 6SL7 (have some nice ones languishing) with one triode section driving each output tube. I did this with very good results on a mashed-up 6AV5 SE using BATTERY BIAS (!).
HOWEVER, the 10M45 was set at 3v/1750 ohms (matched resistors) = 1.7 mA, which is BELOW the headline lower limit for the 10M45 CCS (although still on the R vs current graph); worked REALLY well, though..
Meanwhile,changed my mind COMPLETELY on 2A3s
While waiting for some commentary on low voltage TSE-II Implementations, I resurrected the mono Simple 45, since reconfigured as a TSE-almost II (doesn’t have adjustable bias on 5842s)
Decided to try 2A3s, as I have 2 pairs (Sovtek + Snuggie) and a couple pairs of 3K OPTS.
Only issue was that the Heathkit Regulated power supply only puts-out about -100 V bias voltage. I suppose I could mod the P/S, but figured it was easier to tinker with the bias voltage divider around the bias pot. Spec is 150k resistor over 25K, total 175k ohms. I used 75K top & bottom, total about 150k. With 200K bias pot, total is ~250K ohms.
This succeeded in getting the balance pot closer to center point with the wimpy -100 bias Voltage..
Initially, used the cheap Edcors 5K loaded at 2.5K.
Showed some promise, so switched to a HANDWOUND 3K:8 ohm SE OPT., which are 2+ kg-class OPTs (slightly heavier than 15 W Transcendars.
Wow. I take it all back about 2A3 s being MEH.
The 2A3s are going into the queue .
Back in circa 1981, I heard Alberta Hunter at The Cookery (Greenwich Village IIRC). Put her on the Spotify Premiun, and I’m back sitting at the front table….
While waiting for some commentary on low voltage TSE-II Implementations, I resurrected the mono Simple 45, since reconfigured as a TSE-almost II (doesn’t have adjustable bias on 5842s)
Decided to try 2A3s, as I have 2 pairs (Sovtek + Snuggie) and a couple pairs of 3K OPTS.
Only issue was that the Heathkit Regulated power supply only puts-out about -100 V bias voltage. I suppose I could mod the P/S, but figured it was easier to tinker with the bias voltage divider around the bias pot. Spec is 150k resistor over 25K, total 175k ohms. I used 75K top & bottom, total about 150k. With 200K bias pot, total is ~250K ohms.
This succeeded in getting the balance pot closer to center point with the wimpy -100 bias Voltage..
Initially, used the cheap Edcors 5K loaded at 2.5K.
Showed some promise, so switched to a HANDWOUND 3K:8 ohm SE OPT., which are 2+ kg-class OPTs (slightly heavier than 15 W Transcendars.
Wow. I take it all back about 2A3 s being MEH.
The 2A3s are going into the queue .
Back in circa 1981, I heard Alberta Hunter at The Cookery (Greenwich Village IIRC). Put her on the Spotify Premiun, and I’m back sitting at the front table….
Attachments
a couple of questions on the TSE II 300B. Do I hook up all the PS center taps, except the 5V? I thought I read somewhere not to hook up the 6.3V center tap. Pulled the bottom off my TSE I, 45 amp and 6.3 center tap is hooked up. The other question is would I be better off using a nichicon NS 680 mfd/450 WV Ripple 6.7A or an Aerovox 22.5uf/550V as a replacement for the motor run cap? I thought I had an extra 370V motor run, but I guess I used them all. I could always by another. I'm not sure if the Aerovox is a run or start cap. The big Nichicons may have ben pulls out of something.
Meanwhile,changed my mind COMPLETELY on 2A3s
While waiting for some commentary on low voltage TSE-II Implementations, I resurrected the mono Simple 45, since reconfigured as a TSE-almost II (doesn’t have adjustable bias on 5842s)
Decided to try 2A3s, as I have 2 pairs (Sovtek + Snuggie) and a couple pairs of 3K OPTS.
Only issue was that the Heathkit Regulated power supply only puts-out about -100 V bias voltage. I suppose I could mod the P/S, but figured it was easier to tinker with the bias voltage divider around the bias pot. Spec is 150k resistor over 25K, total 175k ohms. I used 75K top & bottom, total about 150k. With 100K bias pot, total is ~250K ohms.
This succeeded in getting the balance pot closer to center point with the wimpy -100 bias Voltage..
Initially, used the cheap Edcors 5K loaded at 2.5K.
Showed some promise, so switched to a HANDWOUND 3K:8 ohm SE OPT., which are 2+ kg-class OPTs (slightly heavier than 15 W Transcendars.
Wow. I take it all back about 2A3 s being MEH.
The 2A3s are going into the queue .
Back in circa 1981, I heard Alberta Hunter at The Cookery (Greenwich Village IIRC). Put her on the Spotify Premiun, and I’m back sitting at the front table….
While waiting for some commentary on low voltage TSE-II Implementations, I resurrected the mono Simple 45, since reconfigured as a TSE-almost II (doesn’t have adjustable bias on 5842s)
Decided to try 2A3s, as I have 2 pairs (Sovtek + Snuggie) and a couple pairs of 3K OPTS.
Only issue was that the Heathkit Regulated power supply only puts-out about -100 V bias voltage. I suppose I could mod the P/S, but figured it was easier to tinker with the bias voltage divider around the bias pot. Spec is 150k resistor over 25K, total 175k ohms. I used 75K top & bottom, total about 150k. With 100K bias pot, total is ~250K ohms.
This succeeded in getting the balance pot closer to center point with the wimpy -100 bias Voltage..
Initially, used the cheap Edcors 5K loaded at 2.5K.
Showed some promise, so switched to a HANDWOUND 3K:8 ohm SE OPT., which are 2+ kg-class OPTs (slightly heavier than 15 W Transcendars.
Wow. I take it all back about 2A3 s being MEH.
The 2A3s are going into the queue .
Back in circa 1981, I heard Alberta Hunter at The Cookery (Greenwich Village IIRC). Put her on the Spotify Premiun, and I’m back sitting at the front table….
Update: HANDWOUND OPTs
It's too bad that David Lucas DBA "Handwound Transformers" was such a scam artist, cause these particular OPTs are pretty darned good.
Weight comparison:
Handwound SE 3K 2041 grams
Transcendar TT-012-OT 1851 grams
EDCOR GXSE-15-5K 1250 grams
so a nice mid-weight SE OPT; sort of wish EDCOR made something BETWEEN the GXSE and the CXSE, but that's just me.
Of course, I haven't removed the end bells to look inside. Let sleeping dogs lie, I always say. OTOH, here's George, from an 811 amp thread:
My OPT's actually came from a guy who made a living ripping people off. I got these early on when he made some decent sounding OPT's but upon opening one up I see that they were made with paper and package shipping tape. How long they will work at 1100 volts remains a mystery.
If David Lucas hadn't screwed the pooch (pardon my French: baiser le chien) he'd probably be doing quite well nowadays. If wishes were horses...
Maybe the paper & shipping tape is magic..
I SHOULD put a dummy load on the back-end and a square wave on the front end, and a scope on both, but
(a) Let Sleeping Dogs Lie (see above) and
(b) I misplaced my 8 ohm power resistors...gotta remember that for next parts order.
It's too bad that David Lucas DBA "Handwound Transformers" was such a scam artist, cause these particular OPTs are pretty darned good.
Weight comparison:
Handwound SE 3K 2041 grams
Transcendar TT-012-OT 1851 grams
EDCOR GXSE-15-5K 1250 grams
so a nice mid-weight SE OPT; sort of wish EDCOR made something BETWEEN the GXSE and the CXSE, but that's just me.
Of course, I haven't removed the end bells to look inside. Let sleeping dogs lie, I always say. OTOH, here's George, from an 811 amp thread:
My OPT's actually came from a guy who made a living ripping people off. I got these early on when he made some decent sounding OPT's but upon opening one up I see that they were made with paper and package shipping tape. How long they will work at 1100 volts remains a mystery.
If David Lucas hadn't screwed the pooch (pardon my French: baiser le chien) he'd probably be doing quite well nowadays. If wishes were horses...
Maybe the paper & shipping tape is magic..
I SHOULD put a dummy load on the back-end and a square wave on the front end, and a scope on both, but
(a) Let Sleeping Dogs Lie (see above) and
(b) I misplaced my 8 ohm power resistors...gotta remember that for next parts order.
Re: Current Draw of TSE-II
In previous post about 2A3 mono amp powered by Heathkit regulated P/S:
The 2A3 is biased at 60mA (the meter is reading 0.6 volts across the 10 ohm bias resistor), and the 10M45 CCS is set at a nominal 9 mA (330 ohm "setting" resistor, as spec'd).
The Heathkit current meter, however, is reading ~90mA for one channel.
Where is the rest of the current (about 20mA) being dissipated? Seems like a fairly large overhead. I assume it's going to:
(a) MOSFET
(b) OPT
(c) CCS parasitic losses
Remember, there's no rectification or filtering losses in this number.
I ask because I'm scrouging a transformer for a 2A3 stereo TSE-II, and it seems like I need
(1) A PST with a 550V, >200mA secondary, and
(2) a separate ~10A 5V fils transformer, if I'm gonna use a separate fils power supply.
OTOH, the Heathkit meter could be off a bit, although I remember calibrating it last year....
In previous post about 2A3 mono amp powered by Heathkit regulated P/S:
The 2A3 is biased at 60mA (the meter is reading 0.6 volts across the 10 ohm bias resistor), and the 10M45 CCS is set at a nominal 9 mA (330 ohm "setting" resistor, as spec'd).
The Heathkit current meter, however, is reading ~90mA for one channel.
Where is the rest of the current (about 20mA) being dissipated? Seems like a fairly large overhead. I assume it's going to:
(a) MOSFET
(b) OPT
(c) CCS parasitic losses
Remember, there's no rectification or filtering losses in this number.
I ask because I'm scrouging a transformer for a 2A3 stereo TSE-II, and it seems like I need
(1) A PST with a 550V, >200mA secondary, and
(2) a separate ~10A 5V fils transformer, if I'm gonna use a separate fils power supply.
OTOH, the Heathkit meter could be off a bit, although I remember calibrating it last year....
Have look at the schematic; the 6.3 CT is connected to the inboard terminal [labeled T1-GRN-3] next to the two 6.3V leads.
For 2.5V filaments, the 6.3V CT is jumpered to the tail-end of the filament rectifier "bridge".
For 5V filaments, the 6.3V CT is left FLOATING (in which case the terminal block is just a handy place to park it)
Attachments
No, the CT on the 6.3 volt winding must be real in a 2.5 volt situation since the power supply is a FWCT design drawing full current through the CT.
The virtual CT thing made with resistors only works for hum balancing.
The low budget option for a 45 amp in the US is the Allied Electronics 6K56VG. It is a Hammond in disguise. Of course I bought mine back when they were $30 or so. I was using the Allied transformers back in the 90's to make guitar amps. The transformer in my 45 amp has a 2004 date on it.
The virtual CT thing made with resistors only works for hum balancing.
The low budget option for a 45 amp in the US is the Allied Electronics 6K56VG. It is a Hammond in disguise. Of course I bought mine back when they were $30 or so. I was using the Allied transformers back in the 90's to make guitar amps. The transformer in my 45 amp has a 2004 date on it.
I had to look that one up as I have never heard of her. WOW, pretty incredible and right up my alley. Many thanks for the tip!
erick
As far as I can tell, we're about to enter what the economists call an "Inflationary Spiral". My fellow Oldsters may remember this phrase from the Carter Administration (12.4% annual inflation in 1980).
[Insert "Welcome Back Kotter"/Vinny Barbarino joke here.]
SERIOUS inflation has already showed-up in copper prices, which of course will quickly appear in transformer prices, so if you are going to order transformers, y'all might want to stock-up, and be quick about it.
Just my 2 cents. Uh-oh, now FOUR cents.
[Insert "Welcome Back Kotter"/Vinny Barbarino joke here.]
SERIOUS inflation has already showed-up in copper prices, which of course will quickly appear in transformer prices, so if you are going to order transformers, y'all might want to stock-up, and be quick about it.
Just my 2 cents. Uh-oh, now FOUR cents.
You're very welcome. If you haven't yet, read her Bio. It's truly wacky and weirdly inspiring. She was quite the pistol, as we say in Texas.
It turns-out that streaming services will recommend "similar" music, which has opened-up a bunch of rabbit-holes for me to go down. (I use Spotify Premium, 32 bit/384K; not the best, but CHEAP at $15/mo for two accounts).
I discovered Kat Edmonson (local Houston girl) and Blossom Darling (real name) that way. Fascinating.
DAMN! Kat Edmonson was born in the wrong century! I am listening to her Dreamers Do abum - spotify as well.
Alberta Hunter, hers is an incredible story for sure. ./e
A quick question, and I'm sure it's in the last 108 pages somewhere, but I'm looking for a quick answer. Do I run the grnd on the IEC plug to the board? I've run a ground buss from black speaker posts and RCA neg(?) to the IEC grnd. I don't know if I should run a wire to the board or to the metal chassis top. All jacks, switches, fuse, IEC are mounted on the wooden back panel. Thanks
Yes. Tie the board and top panel into your ground buss, preferably with ring lugs and star lock washers. It's the most important connection in the amp as far as safety goes.
The common way is to utilize a "star" ground, typically a ground stud on the chassis. You could use a dedicated stud or perhaps one of the transformer mounting bolts....and again you would tie all of your grounds together there.....preferably with ring lugs and star lock washers.
The common way is to utilize a "star" ground, typically a ground stud on the chassis. You could use a dedicated stud or perhaps one of the transformer mounting bolts....and again you would tie all of your grounds together there.....preferably with ring lugs and star lock washers.
Last edited:
“Star” Groubding
I’ve built three TSE-IIs so far (2 mono 300b, one 45s) using this star groubding scheme:
A. Inputs isolated from chassis, grounded to board.
B. A single “Star” ground point on the chassis (at a transformer bolt) with connections to:
(1) IEC ground
(2) transformer shield
(3) wire to the HV center tap terminal ( on the corner of the board near C4)
Be sure that the Star ground makes good electrical contact with the chassis.
Alternately, you could run the PCB ground to the input ground terminal, but should do one or the other, ie have only ONE point connecting the board to the Star ground to avoid ground loops. I recall that George does that, but maybe I’m thinking of the SSE.
My TSE-IIs are dead quiet with this arrangement.
I’ve built three TSE-IIs so far (2 mono 300b, one 45s) using this star groubding scheme:
A. Inputs isolated from chassis, grounded to board.
B. A single “Star” ground point on the chassis (at a transformer bolt) with connections to:
(1) IEC ground
(2) transformer shield
(3) wire to the HV center tap terminal ( on the corner of the board near C4)
Be sure that the Star ground makes good electrical contact with the chassis.
Alternately, you could run the PCB ground to the input ground terminal, but should do one or the other, ie have only ONE point connecting the board to the Star ground to avoid ground loops. I recall that George does that, but maybe I’m thinking of the SSE.
My TSE-IIs are dead quiet with this arrangement.
^What he said, esp about the ring lugs. Buy 100 next Mouser/Digikey order and you won’t be tempted to substitute.