I'd like to preface this with a disclaimer that I am relatively new to this and this is my first scratch build after building a number of kits and lower stakes scratch builds (passive preamps, etc.).
I've just finished putting together Nobu Shishido's Loftin White inspired amp and I've attached a schematic below and you can see in blue, that I've made a few small changes to the circuit. The changes were made to reflect parts I have on hand as well as some small safety tweaks. The power supply was modeled - to the best of my ability - in PSUD and seemed to be within spec. Changes below...
I then powered it with a dim bulb tester with just the 5U4GB rectifier. Again, no sparks, etc.
When I power it up without the dim bulb tester with the rectifier and just the 12AX7 driver tubes, I am getting 518V B+ at the first power supply cap and further along only 4V B+ going to the preamp tubes where I should be seeing 300V. These are marked in red on the schematic.
I am still working to understand aspects of the schematic but my questions are as follows:
How much will the 2A3 power tubes drop the B+ once there is more load on the power supply? I ask because I don't want to destroy the tubes.
Is there an explanation for the extremely low B+ voltage at the last preamp filter cap? The driver and power tubes are inextricably linked in these circuits but can't wrap my head around the 2A3s effect on this measurement point.
Thanks in advance!
I've just finished putting together Nobu Shishido's Loftin White inspired amp and I've attached a schematic below and you can see in blue, that I've made a few small changes to the circuit. The changes were made to reflect parts I have on hand as well as some small safety tweaks. The power supply was modeled - to the best of my ability - in PSUD and seemed to be within spec. Changes below...
- The power transformer that I have supplies 420V AC instead of 400V AC.
- Replaced the 5AR4 with a 5U4GB (for more voltage drop to account for the different power transformer).
- Replaced the first 100uF capacitor with a smaller 20uF capacitor after the rectifier tube.
- Added a 100uF filter cap in parallel with the existing 100uF filter cap.
- Added a bleeder resistor.
- There are diodes on the 5U4GB a la the Dynamo rectifier mod I've seen online.
- There is also a NTC thermistor on the 120V AC line
I then powered it with a dim bulb tester with just the 5U4GB rectifier. Again, no sparks, etc.
When I power it up without the dim bulb tester with the rectifier and just the 12AX7 driver tubes, I am getting 518V B+ at the first power supply cap and further along only 4V B+ going to the preamp tubes where I should be seeing 300V. These are marked in red on the schematic.
I am still working to understand aspects of the schematic but my questions are as follows:
How much will the 2A3 power tubes drop the B+ once there is more load on the power supply? I ask because I don't want to destroy the tubes.
Is there an explanation for the extremely low B+ voltage at the last preamp filter cap? The driver and power tubes are inextricably linked in these circuits but can't wrap my head around the 2A3s effect on this measurement point.
Thanks in advance!
Well… it is. Embarrassing and temporarily swapped it for a 220K and that got me closer. Will have to order a 200K tomorrow.
Seems like this will get the B+ in range however I’m still concerned about that first power up with the B+ at 518V. How much will the 2A3 draw that down with goal to have the plate around 445V?
It will probably be fine, or pretty close. But a variac would be a good purchase.
Power Transformer puts out a set amount of power, expressed in Watts or Volt/Amps. If you draw less Amps, you get more Volts. Draw more amps, get less volts. "Voltage at rated load" is the easiest to find but "No Load Voltage" is important too. 2A3s pull 60mA each, plus driver tube, so you're at ~130mA. What's the PT rated for?
Put your 200k (or 220k, or 100k etc.) bleeder resistor right after the choke in your supply (that way you only need to use one).
If possible, remove the 50 Ohm 10W resistor in the power supply (the amp will sound better).
Then (very important) reduce the input cap from 20uF to a lower value, somewhere between 1-5uF. Use a film cap here (nothing with oil in it). A DC Link cap rated at 800V is a good idea.
Who knows, maybe you can even remove this cap. Then you will have a choke input supply (which will definitely sound better). Question: What is the value of the choke? I see on the specs sheet for 5U4GB that max. value for choke input supply is 10 Henry's.
I simulate all my PSU's using Duncan amp's PSU Designer II software. I suggest you download it an learn how to use it. I use the old version (2.04 from 2015). But hey, some people here still do calculations by hand. Nothing wrong with that!
Reducing the input cap will bring down your B+, but the precise value of that input cap is not always easy to determine... You can use smaller value caps in parallel though. No downside to doing that.
Besides that, TV damper diodes are far better than conventional audio tube rectifiers for a slow start-up. However they usually require 6.3V heater voltages instead of 5V.
If possible, remove the 50 Ohm 10W resistor in the power supply (the amp will sound better).
Then (very important) reduce the input cap from 20uF to a lower value, somewhere between 1-5uF. Use a film cap here (nothing with oil in it). A DC Link cap rated at 800V is a good idea.
Who knows, maybe you can even remove this cap. Then you will have a choke input supply (which will definitely sound better). Question: What is the value of the choke? I see on the specs sheet for 5U4GB that max. value for choke input supply is 10 Henry's.
I simulate all my PSU's using Duncan amp's PSU Designer II software. I suggest you download it an learn how to use it. I use the old version (2.04 from 2015). But hey, some people here still do calculations by hand. Nothing wrong with that!
Reducing the input cap will bring down your B+, but the precise value of that input cap is not always easy to determine... You can use smaller value caps in parallel though. No downside to doing that.
Besides that, TV damper diodes are far better than conventional audio tube rectifiers for a slow start-up. However they usually require 6.3V heater voltages instead of 5V.
Last edited:
Just purchased a variac last night on eBay!
The PT is rated for 170mA so it should be fine.
Thanks for the info. I’ll definitely move the bleeder and increase its rating.
As it’s my first scratch build, I don’t want to alter it too much until I confirm that I did everything right… layout, wiring, etc.
Once it’s up and running your suggestions sound worth trying out. I think it will just require messing with some values as you stated. Unfortunately the only choke I have is 15H so it might have to go to with another rectifier like a 5AR4. Removing the 50 ohm resistor would raise the voltage a bit, where is a better place to drop it? And what is the 50 ohm the there for? I usually see resistors in the power supply between filter caps…
The 50R resistor is a Dropping Resistor used to adjust the b+. Since you'll be drawing less than the full rated output of the power transformer, which is a good thing, you'll get a higher AC voltage going into the rectifier, and a higher b+ as a result. The Dropping Resistor gets you to that ideal voltage. After it's built you can measure the voltage drop across it to determine the amount of current, and then adjust that resistor's value, if necessary.
Gotcha. In order to drop the B+ via a resistor, is there an optimal place to insert it? After the first cap? Between the filter caps? Is there any benefit to using two resistors in different positions? Would it even make a difference? As long as the caps are rated for the DC voltage of course.
It goes where the 50ohm resistor is. Between the reservoir cap and the choke. But it's better to have no resistors in the power supply at all, for the PT to be ideally matched to the circuit. But if we must drop the b+ we use one resistor.
Reducing the input cap further is the easiest way to lower your B+. The 50 ohms in series won't drop much, but will reduce the ability of your power supply to respond. You can use it after the choke in an RC filter instead. I would not put it before the choke...
Try 5uF instead of 20uF. What high voltage rated caps do you have on hand? You could buy a few 800V 2uF DC Link caps and use them in parallel as well. Two 2uF caps in parallel = 4uF. Three x 2uF = 6uF...
Just want to let you know the amp is up and running and it sounds great! Voltages are about 5% low but happy with it... for now. Measuring a little under 2mV at the speaker outputs. Is this normal? No noticeable hum but I'm not on my most efficient speakers.
Quickly modeled a choke input and seems doable with what I have on hand. Might eventually go that route and it would free up some room in the chassis for some added relay circuitry for a soft start.
Ill take some pictures of the insides when I'm done listening.
Thanks again!
Quickly modeled a choke input and seems doable with what I have on hand. Might eventually go that route and it would free up some room in the chassis for some added relay circuitry for a soft start.
Ill take some pictures of the insides when I'm done listening.
Thanks again!
Choke input drops a lot of volts across the choke, at least 60v, and the choke has to be built extra sturdy. But that Tango looks like it could do it.
I'll have to check the specs on the choke. Maybe go with a 5AR4 for less voltage drop and a small 'tuning' cap before the choke. Maybe 3uF or something...
There's also a strong magnetic field created by the choke that might not agree with the layout you already have. Choke and cap input supplies are very different animals. Prob not something you can switch back and forth. The small value reservoir cap has been claimed to create something in between, and it does lower the b+.
Hmmm. Is the magnetic field greater with a choke input than a cap input? Is it the chokes proximity to the driver tubes?
In a perhaps over simplistic, new to me kind of way, I would just swap the 5U4GB with a 5AR4, remove the 50 ohm resistor and decrease the value of the 20uF cap to a 3.2uF cap. I modeled it in PSUD and it get's me to the same 450V B+ voltage. Just make sure everything's rated well above what I'm expecting, eg. the 100uF filter caps are rated at 800V. That said, I'm unsure about other factors in a choke input like the magnetic field.
In a perhaps over simplistic, new to me kind of way, I would just swap the 5U4GB with a 5AR4, remove the 50 ohm resistor and decrease the value of the 20uF cap to a 3.2uF cap. I modeled it in PSUD and it get's me to the same 450V B+ voltage. Just make sure everything's rated well above what I'm expecting, eg. the 100uF filter caps are rated at 800V. That said, I'm unsure about other factors in a choke input like the magnetic field.
Reducing that input cap and getting serial resistance out of your power supply are good ideas.
Keep in mind that using PSU II is still simulating... sometimes real measured values are slightly different. This also holds true when you calculate values by hand with a slide-ruler...
Keep in mind that using PSU II is still simulating... sometimes real measured values are slightly different. This also holds true when you calculate values by hand with a slide-ruler...
I could be wrong, but I believe the purpose of the 50 ohm resistor was to provide series resistance after the 5AR4 that allowed for more capacitance before the choke, to reduce B+ ripple noise. Of course you can change it, but you may get more ripple on the plate of the 2A3.
Oh the monkey schematic. It pops up here ever so often. I breadboarded a few variants of it some years ago before realizing that there is just too many compromises in it to make the direct coupling any advantage. Feel free to build a variant if you want. If you are going to use a plate choke, then why not also use transformer coupling? Or constant current source with transformer coupling... 😉
Btw - that 417a is running way down in the knees of the plate curves, with only 95V on the plate, drawing only 6mA, plus only 1.1V headroom... 417a is a nice little triode that really performs well with 160V on the plate at 20mA (bias around -2V).
Btw - that 417a is running way down in the knees of the plate curves, with only 95V on the plate, drawing only 6mA, plus only 1.1V headroom... 417a is a nice little triode that really performs well with 160V on the plate at 20mA (bias around -2V).