With those measurements I believe the main issue is that your good, low resistance choke presents too little resistance in a circuit Tubelabs originally had a 150 ohm resistor in place of the choke. Additionally, your power transformer sags less under load than anticipated in George’s original design. Given the variety of diy options builders could use, it is not a surprise that a bit of tweaking is required to get the B+ just right. As George wrote on Tubelabs’ instruction for building the SPP, you should aim for 320-340 Vdc B+. You are not far from there in B+ and your heater voltages are within 3%, so good there. (With a CL-80 Thermistor in the primary your heater voltages will be spot on). Bottom line is you have good magnetics that could be made to work without resorting to much change.
One remaining issue is that with the low resistance secondary in your power transformer your 5AR4 does not see enough resistance to keep the charging spike below its maximum of 750 ma. At 300-0-300 volts a 5AR4 requires a minimum of 75 ohms in each leg, including the transf resistance. Your transformer presents roughly half that. I think this is the cause of most problems with arcing 5AR4s, that the “yellow sheet mod partially tries to prevent. More about minimum required resistance (Rt)in this thread: https://ampgarage.com/forum/viewtopic.php?t=32131
So, it is important to add resistance to each of the two secondary wires before connection to the PCB. I would guess 100 ohm, 3 w resistors, the larger resistance to tame the excess B+. Perhaps @jcalvarez could modify his PSUII simulation to tune the value of these resistors, while keeping the original C1=47 uF in place. I believe 140 ma is too low an estimate for average current draw even if the EL84s are running at 35 ms each, given the 12AT7‘s draw as well as PSU discharge resistor paths.
So, before ordering a new choke, etc. Try the resistors in the leads from the power transformer, as well as a CL-80 thermistor in the primary. Cheap solution and I think with the right values your B+ will be good. You know how to calculate tube plate dissipation, so just make sure you final arrangement stays under the limits. If it is still to high, the next step is to apply more cathode bias (~300 + ohms), but I don’t expect itvwill be necessary. Good luck.
Last edited:
You meant this, right?
Here we are. I "added" DCR to the secondary, because I have no idea how to insert a resistor ion that place with PSU designer. Current increased to 160mA:
Adding a couple of 50 ohm resistors results in 323V B+. Using 100 ohms gives ~300V.
Thank you @jcalvarez! Yes, the resistors physically in the leads as you showed in first sketch, but adding the resistance to the secondary resistance should work fine for the simulation. These results look promising. Did you use 12AX4 deliberately, in stead of 5AR4?
Anyway, @calpe, I think the path forward is to try the 50 Ohm resistor in each of the secondary leads from the power transformer, as well as a CL-80 thermistor in the primary. (I you are ordering the resistors I suggest also get a 75 and 100 ohm pair, just in case PSUDII fibbed a bit)
Anyway, @calpe, I think the path forward is to try the 50 Ohm resistor in each of the secondary leads from the power transformer, as well as a CL-80 thermistor in the primary. (I you are ordering the resistors I suggest also get a 75 and 100 ohm pair, just in case PSUDII fibbed a bit)
Thanks guys.
I checked about getting the newer type choke, but i was surprised a little by the the price, so perhaps the in-series resistors is the avenue to take.
I'll have to order several values, from 50 ohm to ohms 3W, 2 of each.
CL-80 thermistor? Fitted in series with the live mains feed to the primary?
https://uk.farnell.com/ge-sensing-thermometrics/cl-80/thermistor-series-cl/dp/1653470
This would be the CL-80 - Thermistor, ICL NTC, 47 ohm, -25% to +25%, Radial Leaded, CL Series, Order Code - 1653470?
This thread has now got a bit long and unwieldy. I was trying to see which choke you had so I could look up its resistance but could not see from your schematic, and don't want to trawl through 400+ posts.
This is a PP amplifier and that type of amplifier is quite good at cancelling out ripple from the B+ supply. Therefore I would just put the 150R resistor back in instead of the choke and use that for now. Then plan for a choke if you have a hum problem. The choke that George/Tubelab used to promote was the Triad C14-X which is 6H, 200mA and 150R, so plenty of resistance. (It is cheap when you place an order and the total order entitles you to free delivery).
The resistance before the rectifier is important, but there is a magic quadrant in which the rectifier will work. The borders are dictated by the voltage, the capacitance of the 1st capacitor (in a CRC setup) and the DCR of the power transformer. This is a known configuration of rectifier, circuit and transformer, so I would have thought the rectifier should have been tolerant of the transformer as long as the first capacitor capacitance was low enough. As long as you are well inside the quadrant, which is on the tube data sheet, then you ought to be OK, although some tubes are better than others.
Other ways to trim a few volts, apart from the Inrush Current Limiter, are to use a different rectifier tube type with a bigger voltage drop (5V4 for example, chart here ... voltage drop chart) or reduce the first capacitor in value - for instance 22uF. If you have another 47uF capacitor you could put that is series with the other 47uF (it will be messy) and that gives you 23.5 uF.
Following on from Francois G's comments, you ought to be able to get up and running with the components you have. Once you have it working you have a better idea of the fine tuning required. For instance if you have a heater voltage problem AND a B+ problem, or just B+.
This is a PP amplifier and that type of amplifier is quite good at cancelling out ripple from the B+ supply. Therefore I would just put the 150R resistor back in instead of the choke and use that for now. Then plan for a choke if you have a hum problem. The choke that George/Tubelab used to promote was the Triad C14-X which is 6H, 200mA and 150R, so plenty of resistance. (It is cheap when you place an order and the total order entitles you to free delivery).
The resistance before the rectifier is important, but there is a magic quadrant in which the rectifier will work. The borders are dictated by the voltage, the capacitance of the 1st capacitor (in a CRC setup) and the DCR of the power transformer. This is a known configuration of rectifier, circuit and transformer, so I would have thought the rectifier should have been tolerant of the transformer as long as the first capacitor capacitance was low enough. As long as you are well inside the quadrant, which is on the tube data sheet, then you ought to be OK, although some tubes are better than others.
Other ways to trim a few volts, apart from the Inrush Current Limiter, are to use a different rectifier tube type with a bigger voltage drop (5V4 for example, chart here ... voltage drop chart) or reduce the first capacitor in value - for instance 22uF. If you have another 47uF capacitor you could put that is series with the other 47uF (it will be messy) and that gives you 23.5 uF.
Following on from Francois G's comments, you ought to be able to get up and running with the components you have. Once you have it working you have a better idea of the fine tuning required. For instance if you have a heater voltage problem AND a B+ problem, or just B+.
OldHector the choke i've fitted is a 1.5H 200mA.
I won't opt for a different rectifer valve as this will increase costs even further.
I'll now probably opt for the in series resistors in the secondary winding (50-100 Ohms 3W) + the CL-80 thermistor.
These will need to be ordered from the U.K.
Long and wieldy...... well until the amp is up and running with a good B+, i'll continue posting my queries, sorry...., and having spent quite a few £££££
But i am thankful for all the assistance
I won't opt for a different rectifer valve as this will increase costs even further.
I'll now probably opt for the in series resistors in the secondary winding (50-100 Ohms 3W) + the CL-80 thermistor.
These will need to be ordered from the U.K.
Long and wieldy...... well until the amp is up and running with a good B+, i'll continue posting my queries, sorry...., and having spent quite a few £££££
But i am thankful for all the assistance
Last edited:
First of all, let me say that the various Tube Lab amplifiers are excellent designs.
No problem here.
. . . But sometimes the builder uses different parts, because of what they have in their parts bin, or to attempt to save money, or because some part or other has become one of three items: Un-obtainium, Over priced, Or Obsolete.
If you already mounted the choke, then just add a resistor in series with the choke.
It will give better ripple filtering, versus with no added series resistor.
Suppose you need 200 Ohms of total series resistance to get the proper B+ Volts.
Example, if the choke DCR is 50 Ohms, then just put a 150 Ohm resistor in series with the choke.
Push pull
Generalizations when the power supply is designed from the start, to be optimal:
The more series resistance(s) you have in a B+ power supply, the more B+ voltage variation you will have when the push pull amplifier goes from Quiescent current, to maximum current when power output is at maximum.
. . . Resistance, what resistance?
Power transformer primary DCR x step-up ratio of primary to secondary (secondary volts / primary volts = step-up ratio)
Power transformer secondary DCR
Rectifier resistance (voltage drop / current; but that is the quiescent resistance; the resistance also varies as the load current varies [dynamic resistance].
Any other series resistors in the B+ circuit
Choke DCR
A regulated supply, has the Smallest change of B+ voltage from quiescent current to the current needed for maximum signal power output.
That is more than I want to build, I like simple amplifier circuits.
A choke input filter, has a Moderate change of B+ voltage from quiescent current to the current needed for maximum signal power output.
I like that solution.
A capacitor input filter, has the Largest change of B+ voltage from quiescent current to the current needed for maximum signal power output.
I no longer will use that kind of B+ filter, when I build more new amplifiers.
The proper design of a B+ supply, starts with a knowledge of the quiescent current and quiescent voltage required by the amplifier circuitry, and the maximum current required when the amplifier is putting out maximum signal power.
Next, make the decision of whether to use a regulated supply (voltage drop in the pass-regulator); and whether to use a choke input filter, or a capacitor input filter (of course the regulated supply needs to use either a choke input filter, or a capacitor input filter).
From those factors, you can properly choose the power transformer you will need.
Do not pick a power transformer first, and then hope to build an optimum B+ supply from it.
When you figure the quiescent current and maximum current, and the needed B+ voltage, then you need to design the amplifier circuitry first.
And be sure to include the current load of the bleeder resistor.
Bleeder Resistor? Safety First! Be sure to use a Bleeder Resistor(s). Prevent the "Surviving Spouse Syndrome".
A bad experience:
Once, I had a power supply that had a moderate amount of ripple.
The common mode ripple that supplied B+ to the phase splitter driver stage, passed that ripple on to the grids of the output stage.
No hum at the output, because it cancelled out in the push pull transformer.
But . . . it did rob power from the output stage, so the available maximum signal power output was less than expected.
It took me some time to figure out what was going on.
Lesson learned, and that mistake is not going to be repeated in any of my amplifiers.
Have Fun!
Happy Building and Listening
No problem here.
. . . But sometimes the builder uses different parts, because of what they have in their parts bin, or to attempt to save money, or because some part or other has become one of three items: Un-obtainium, Over priced, Or Obsolete.
If you already mounted the choke, then just add a resistor in series with the choke.
It will give better ripple filtering, versus with no added series resistor.
Suppose you need 200 Ohms of total series resistance to get the proper B+ Volts.
Example, if the choke DCR is 50 Ohms, then just put a 150 Ohm resistor in series with the choke.
Push pull
Generalizations when the power supply is designed from the start, to be optimal:
The more series resistance(s) you have in a B+ power supply, the more B+ voltage variation you will have when the push pull amplifier goes from Quiescent current, to maximum current when power output is at maximum.
. . . Resistance, what resistance?
Power transformer primary DCR x step-up ratio of primary to secondary (secondary volts / primary volts = step-up ratio)
Power transformer secondary DCR
Rectifier resistance (voltage drop / current; but that is the quiescent resistance; the resistance also varies as the load current varies [dynamic resistance].
Any other series resistors in the B+ circuit
Choke DCR
A regulated supply, has the Smallest change of B+ voltage from quiescent current to the current needed for maximum signal power output.
That is more than I want to build, I like simple amplifier circuits.
A choke input filter, has a Moderate change of B+ voltage from quiescent current to the current needed for maximum signal power output.
I like that solution.
A capacitor input filter, has the Largest change of B+ voltage from quiescent current to the current needed for maximum signal power output.
I no longer will use that kind of B+ filter, when I build more new amplifiers.
The proper design of a B+ supply, starts with a knowledge of the quiescent current and quiescent voltage required by the amplifier circuitry, and the maximum current required when the amplifier is putting out maximum signal power.
Next, make the decision of whether to use a regulated supply (voltage drop in the pass-regulator); and whether to use a choke input filter, or a capacitor input filter (of course the regulated supply needs to use either a choke input filter, or a capacitor input filter).
From those factors, you can properly choose the power transformer you will need.
Do not pick a power transformer first, and then hope to build an optimum B+ supply from it.
When you figure the quiescent current and maximum current, and the needed B+ voltage, then you need to design the amplifier circuitry first.
And be sure to include the current load of the bleeder resistor.
Bleeder Resistor? Safety First! Be sure to use a Bleeder Resistor(s). Prevent the "Surviving Spouse Syndrome".
A bad experience:
Once, I had a power supply that had a moderate amount of ripple.
The common mode ripple that supplied B+ to the phase splitter driver stage, passed that ripple on to the grids of the output stage.
No hum at the output, because it cancelled out in the push pull transformer.
But . . . it did rob power from the output stage, so the available maximum signal power output was less than expected.
It took me some time to figure out what was going on.
Lesson learned, and that mistake is not going to be repeated in any of my amplifiers.
Have Fun!
Happy Building and Listening
Last edited:
@OldHector and @6A3sUMMER
Did you read #444? Do you disagree?
@calpe,
Based on the PSUDII simulation you will need a 100 ohms (3 watts) in each secondary leg of the the power transformer. A NTC Thermistor in series with the primary will likely knock off the 1-3% excess voltage on the heater supplies, which are already very close. This should bring you B+ to within the acceptable range around 330Vdc.
I would try this first and expect all will be well. If the B+ is still too high we will look at further measures, like increasing the resistor pair to 110 or 120 Ohm (save shipping and order with 100 Ohm pair), or a resistor in series with the choke as previously discussed. Additionally you may consider another Thermistor (CL-140) in series with the power transformer center tap to earth, as George discussed in the linked post. Note the sizing of the Thermistors. George recommends CL-90 for ~240Vac service.
https://www.diyaudio.com/community/...-the-tse-must-die.331038/page-19#post-5829892
Mind the heat dissipation of these resistors and thermistors in placing them in you amp, and keep them away from the “yellow sheet diodes” (which you should keep) and other heat sensitive parts.
Did you read #444? Do you disagree?
@calpe,
Based on the PSUDII simulation you will need a 100 ohms (3 watts) in each secondary leg of the the power transformer. A NTC Thermistor in series with the primary will likely knock off the 1-3% excess voltage on the heater supplies, which are already very close. This should bring you B+ to within the acceptable range around 330Vdc.
I would try this first and expect all will be well. If the B+ is still too high we will look at further measures, like increasing the resistor pair to 110 or 120 Ohm (save shipping and order with 100 Ohm pair), or a resistor in series with the choke as previously discussed. Additionally you may consider another Thermistor (CL-140) in series with the power transformer center tap to earth, as George discussed in the linked post. Note the sizing of the Thermistors. George recommends CL-90 for ~240Vac service.
https://www.diyaudio.com/community/...-the-tse-must-die.331038/page-19#post-5829892
Mind the heat dissipation of these resistors and thermistors in placing them in you amp, and keep them away from the “yellow sheet diodes” (which you should keep) and other heat sensitive parts.
Last edited:
Francois,
I do my own B+ power supply calculations on my HP-11 Calculator.
I do not pay much attention to power supply simulation software, I do not use them.
I see postings on various threads of diyAudio, where the simulation software either does not include all of the factors of the power supply, nor the real amplifier circuit, or at least does not give the same results when an actual amplifier circuit is attached.
Your Mileage May Vary.
I do my own B+ power supply calculations on my HP-11 Calculator.
I do not pay much attention to power supply simulation software, I do not use them.
I see postings on various threads of diyAudio, where the simulation software either does not include all of the factors of the power supply, nor the real amplifier circuit, or at least does not give the same results when an actual amplifier circuit is attached.
Your Mileage May Vary.
My reference to #444 was shorthand for the prior discussion that lead to the conclusion that OP’s power transformer has too low internal resistance in the secondary coils to be used with a 5AR4 rectifier without remedy, and required resistors in each lead to comply.
The spec sheets are clear about the “minimum plate supply resistance per plate” (Rs in GE parlance). See the graph on page 3.
https://frank.pocnet.net/sheets/093/5/5AR4.pdf
The spec sheets are clear about the “minimum plate supply resistance per plate” (Rs in GE parlance). See the graph on page 3.
https://frank.pocnet.net/sheets/093/5/5AR4.pdf
If you are going to use a resistor in each plate lead of the rectifier, you have an alternative:
Example:
A pair of 100 Ohm 3 Watt resistors, one in each plate lead . . .
Or, a single 100 Ohm 10 Watt resistor in the cathode lead.
The power transformer and the rectifier tube will not notice the difference.
Guess which resistor(s) will be coolest in that application.
Just saying
Example:
A pair of 100 Ohm 3 Watt resistors, one in each plate lead . . .
Or, a single 100 Ohm 10 Watt resistor in the cathode lead.
The power transformer and the rectifier tube will not notice the difference.
Guess which resistor(s) will be coolest in that application.
Just saying
@6A3sUMMER,
I have great respect for your calculator, stubby pencil and 1950’s Radio Amateur Handbook. I have read for example: https://www.diyaudio.com/community/...ze-in-lc-input-rectifier.365295/#post-6468883
”Just saying” as you often conclude your posts. Now tell us the answer. What alternative will run cooler and why?
Btw., critical inductance for the choke in this application should be ~420/170 = 2.5 H (but that’s not the route OP is taking)
Last edited:
Numerous tasks to do, but i'll keep posting the results here.
Monday, i'll insert a 150 Ohm 3W, before the choke (if i have one) and check the B+ again
yep, it was my choice to order the Transformers from the U.K. company called Primary Windings. The thought behind it was they were British and seemed suitable! The choke also came from them as earlier thoughts on the SPP build was to replace R1 and filter out any 'noise' on the B+
The choke has been mounted and connected from day one, but like i've said i'll now insert a 150 Ohm 3W before the choke on Monday and measure the B+ again,.
I'll also try your earlier suggestion of the psuedo-choke input filter, i have ordered some spares and should have them in a number of days, including the in series primary thermistor CL80 that was mentioned.
i.e.
1. Disconnect the 47 uF cap that is before the choke.
2. Connect the 47 uF cap to the point right after the choke (in parallel with the 150uF cap), now, the B+ is probably a little to low, so...
3. Connect a 1 uF cap right before the choke, where the 47 uF cap was before (a 0.5 uF, 2 uF, 3 uf cap might be needed to put the B+ at the voltage you want it).
This is one of the first diagrams but with R1 being replaced with a choke 1.5H, and some text additions i've added.
So i'm assuming that the changes needed for a psuedo-choke input filter are those above 1. to 3. ?