Dreams about Monsters

[admin]

When is a Dream about a Monster a Good Dream?

...When you're dreaming about building a Monster Vacuum Tube Amplifier.

As mentioned elsewhere, I provide Professional Sound Services for Car Shows and other events. My dream is to build a vacuum tube power amplifier with sufficient power to use as my main live sound amplifier.

This project is in the early planning/design stages, so don't expect much for now. Here are the basic project requirements:

• The amplifier shall produce approximately 100 Watts (TRMS) per channel into 8 ohm speakers. (Almost 200 Watts total output power).
• The amplifier shall be all tube in the analog signal path.
• It will be a straight Power Amplifier. There will be no volume control or other controls/switches, etc., with the exception of a power switch.
• The amplifier shall provide Balanced inputs via XLR and 1/4" TRS connectors. (Separate or combined connectors). It will also have unbalanced RCA inputs.
• As with my other builds, the Monster Amplifier shall have an internal DAC, and provide USB sound card functionality. An Ethernet connection to support one or more industrial A/V over Ethernet protocols is also a possibility.
• As with my other builds, the Amplifier shall use Switch Mode Power Supplies (SMPS) for wide input voltage tolerance, reliability, and weight reduction.
• Sound Quality, frequency response, and Vacuum Tube "sound stage" are definitely goals, but output power is the primary goal.
• As with my other builds, Cost will be minimized by using Chinese sourced SMPS's, DAC's etc., as well as surplus Russian tubes, but Quality discrete components will be used in the analog signal path.

Implementation -- How will the above goals be realized?:

Components:

One of the largest costs of a build like this will be in the Audio Output Transformers. There is a vendor on AliExpress selling 100Watt rated tube output transformers. These transformers have a 3.6K primary impedance which is appropriate for the planned design, and 4 and 8 ohm outputs. The seller seems knowledgeable, and provided specifications and detailed descriptions of the materials used and why their transformers are good quality. Even if they are not on a par with something from a company like Edcor, it is important to keep in mind the goals of this project. While the transformers cost far less than custom parts from a US company, there are significant costs to ship (2) 100W audio transformers from China.

OPT_100W_sm.jpg (34.31 KiB) Viewed 32077 times

Also to keep costs low, I will be using surplus used or NOS tubes. Obviously, I'll need 8 Power tubes for a Stereo Parallel Push Pull (PPP) amplifier in addition to driver tubes. There will be 2 stages of amplification between the inputs and the driver tubes, so we are talking about at least (4) Dual Triode tubes.

For power, I am up to my old SMPS tricks. While power budget calculations are still in work, Inexpensive wide input, 12VDC output SMPS's are available at plenty of current.
The 12VDC will directly power things like the DAC board, Panel and/or visual effects lights, and any 12VDC filaments, as well as other SMPS's to provide High Voltage B+, filament voltage, and bias voltage.
The B+ SMPS will be made from a Chinese inverter module rated at 1000W. The square wave output will be rectified and filtered. The module I am looking at has taps for 220V-330V-380V and 420V output (along with lower voltages) and provides connections for feedback for voltage regulation and.

1000W_Inverter_Board.jpg (18.34 KiB) Viewed 32072 times

Tube Compliment:

For low cost and at least some sustainability, the amplifier will utilize surplus Soviet military tubes. To achieve the goal of approximately 100 Wrms per channel, the output stage of the amplifier will be of Parallel Push Pull (PPP) design. This means that the amplifier will use (4) 6P3S-E tubes and (2) 6N8S per channel, for a total compliment of 12 tubes. While 6P3S-E are rated at about 250V, and 20W, I have read that they can take much more. My plan is to run them at whatever voltage I end up with from the 330V or 380V configuration of the above mentioned Inverter board (voltage after rectification and Filtering.) bias them conservatively for a Class AB amp, and take the power I get.

In addition to the power stage, there will be an input stage appropriate for either a balanced or unbalance source (I am still researching if an input switch will be needed.), and a driver stage. My initial research/design seems to indicate that 6N8S (6SN7 Equivalent) should be suitable in both input stages. This also has a few other advantages for the intended use of this amplifier:

• Cosmetics -- All "Big Bottle" (Big Base) Tubes
• More secure base/socket than e.g. a 12AT7
• Affordability / Maintenance

The chassis will be built from a surplus 2U rack mount chassis already on hand. As with my other builds, the actual amplifier will be built on the top panel, and the SMPS's and other components will be internal. The chassis will have a "roll cage" or similar to protect the tubes as will as provide a method to make lifting the amplifier more easy.

As with my other builds, there will be some sort of visualization. Right now, I am leaning towards VFD VU meters (as opposed to a VFD spectrum display as in my last build), but real moving coil VU meters, LED VU meters, etc. all remain possibilities at this time. It's also possible that it will contain both DJ - Facing visualization, and Audience facing visualization.

It will also have my "Signature" Lime Green painted output transformers, and Rack Mount handles in the front. In other words, it will follow the design clues of my previous builds -- just much larger.


Status:

It will be well into the next year before I can even make a decision to "Pull the Trigger" on this build.

A schematic is in work, and chassis drawings could start as early as tomorrow, assuming that I can get to the chassis in my crowded shop to take measurements.

I am also working on an informal Bill of Materials/Cost estimate.

More details to follow...

[admin]

"The Monster" Quad 6P3S-E Amplifier Power Budget

As Mentioned, this amplifier will not be based on a conventional 60Hz Transformer power supply.

It will use the "Down-Up" Design that I have been successful with on my previous builds. As explained in the other threads, this means that the AC Lines ("Mains" Voltage") powers only a 12VDC Switch Mode Power Supply. The 12VDC, in turn, Powers separate SMPS's for the High Voltage (B+), Heater Power, and Bias Voltage. To this end, what is below are basic power calculations for each needed voltage. Note that there is multiple levels of conservatism in these calculations. The most obvious is that I round up some of the results. Also max ratings of the tube are used, not average or even the maximum expected loads with Music (signal).

Per Channel Heater Current
4 x 6П3С-Е 4@0.9A each = 3.6A
2 x 6Н8С 2@0.3A each = 0.6A
= 5A per Channel

Per Channel B+ (High Voltage) Demands
4 x 6П3С-Е Plate 4 x 72ma = 288mA
4 x 6П3С-Е Screen 4 x 8ma = 32mA

4 x 6Н8С (ea. element) 4 @ 12mA each =48mA
= 368mA Per Channel

370mA @ 400V = 148W/Channel on B+ Per Channel.

296W total on B+ (Both Channels)


12VDC PS Loads
Heater Bucks
6.3V @ 5A = 31.5W = 2.66A @ 12V Per Channel.
5.33A Total

HV Power Supply
296W/12V = 25 A

Bias Supply (--Deleted. See below.)
Estimate 2A on 12V side

= 32.5A @ 12VDC
= 400W @ 12VDC Power Supply
500W @ 12VDC Power Supply for conservatism.

500W may sound like a lot, but here are a few factors to keep in mind -- Especially when comparing it to a conventional 60Hz transformer and Tube Rectifier power supply:

• 500W 12VDC Switch Mode Power Supplies are readily available.
• Even Name-brand Powers supplies with these ratings can be easily purchased for about $70
• This method provides a regulated voltage, and wide input voltage. Some of the power supplies have a 120/230 input voltage switch, while the better ones will work from 85 to 250 VAC input without needing to be switched.
• The better ones contain an automatically controlled fan that runs when needed to regulate internal temperature. While I would not tolerate a fan in a home Hi-Fi design, A fan is acceptable in an "industrial" amplifier design, and all of my professional (DJ) solid state amplifiers contain cooling fans.
• The power supplies have Power Factor Correction for efficiency, and have AC line filtering in addition to the planned IEC connector with built-in filtering. The amplifier will not put noise back into the AC line.
• Note that the calculations assume 2 separate Buck Converters for the Heaters -- One for each channel.
• The heater Buck Converters provide adjustable current limiting (soft start) and are rated at 200W max, 100W continuous. By using a separate one for each channel, I am limiting their nominal operating power to 31.5W each.
• Note that based on the above calculations, the High Voltage (B+) supply should be able to be provided by the 500W (instead of 1000W) version of the Inverter board -- in other words, the same inverter used in my 7591 Amplifier build. While I need to do a little more homework, the fact that the 500W inverter in the 7591 amplifier does not get hot at all, would tend to support the assumption that the 500W inverter will be adequate, although running closer to it's advertised ratings for this build.

UPDATE: I am considering changing the design so that the +/- 56VDC Power supply is not needed. In my previous (7591) build, I similarly deleted the Bias Boost/Buck Dual Rail converter, and generated negative voltage from an 18VAC winding that is on the HV (B+) Inverter. In that previous build, I used a bridge rectifier to generate sufficient negative voltage to bias the 7591's. For 6П3С-Е's, which have less gain, more (absolute value or negative) bias will be needed, this originally led me to thinking that I could not derive the bias from the 18VAC winding. I then realized that by using a voltage doubler instead of a bridge rectifier, that I should be able to get nearly -36 volts to supply the bias circuit. The additional draw on the HV inverter will be negligible, and no additional active components are used. The generation of the bias will be as fail-safe as any grid bias scheme. The power budget above for "Bias Supply" can be deleted.

Check back periodically, as I have been working on both the schematic, and the layout drawing, and may be ready to share something soon.

[admin]

I have decided to release an Extremely Early Render of this project. Layout and Chassis size are likely to change.

Untitled.png (66.11 KiB) Viewed 31627 times

Notes:

• 500W Inverter shown mounted above the chassis, and centered. This is mostly because it was a simple cut and paste from the 7591 render. However the calculations in the previous post seem to indicate that the 500W inverter may be adequate for this amplifier.
• LED VU Meter cut, pasted and stretched from the Poor Man's Go DJ is shown -- again for ease of creating a quick render. This will not be the VU meter used.
• No input select switch/buttons shown.

Edit:
Render Updated. Cosmetic Issues fixed.

[admin]

A small text-only update for this project:

I have begun ordering parts. Of course no "exotic" project like this will be without some difficulty in parts procurement. I have not received shipping information for the Output Transformers (OPT's), the optimistic interpretation of the information I have is that the Transformers are being manufactured in China.

I ordered tubes from an eBay vendor in Kazakhstan, but as I knew was possible, the tubes are actually shipping from (or at least through) Russia. Given the current situation, I have no idea if they will ever make it to me. Even if sanctions/trade restrictions don't get them, the "Monster" is just a Drone Strike away from being destroyed before it's ever built.

The schematic is maturing, and this has resulted in a change to the physical drawings, An additional 6SN7 equivalent (6Н8С) has been added as a cathode follower buffer on the (-)signal input. -- Essentially, it forms a mixer between the global feedback and the (-) input signal. This will bring the total tube compliment to 13, and mean that one of the 6SN7 equivalents will pass signals from both channels.

I want to point out an article over on Aiken Amplification that has helped me in the design of this amplifier. The linked article is about Blocking Distortion, but the 4th bulleted paragraph describes a "a DC-coupled cathode follower between the phase inverter and the grid of the output tubes, with the cathode follower cathode resistor returned to a high negative voltage, and the grid bias applied to the grid of the cathode follower." While I don't know if my amplifier will suffer from Blocking Distortion or not, I will need the additional drive to run the Parallel output tubes hard enough to get the desired balance between distortion and output power. Whether I will actually be pushing the grids of the output pairs positive or not (Class AB2) remains to be seen, but the referenced paragraph basically told me what to design. This topology eliminates additional coupling capacitors, and thus the associated phase delay, and thus again, problems with feedback phasing. The articles are Aiken Amps are well written, succinct, and very helpful. I have learned a lot from the various articles. Thanks to Aiken Amps for putting this information out there.

[admin]

Cross-post from my thread on AudioKarma.com of the same title (and a more recent update that is only posted here for now.)

"A Non-Update on this topic (notice that I am not calling it a project -- yet).

I can't build a tube amp if I can't get parts.

The AliExpress vendor cancelled my order for the Output Transformers. As usual -- and especially on AliExpress, If it sounds to good to be true, it usually is, and free shipping on a pair of 100W OPT's from China certainly fits in that category.

The story with the tubes I ordered is possibly worse. I ordered Soviet tubes -- specifically avoiding Russia due to sanctions and the current situation. The tubes went though Russia anyways. At first, they quickly moved across half of Russia to one of the Moscow airports... and tracking stopped for 2 weeks. 2 days ago, it pinged from Serbia, but USPS no longer sees the tracking number as valid. Despite the package being destined for "США", it looks like the folks at Почта России did make an effort to get the package to me.

I have the same OPT's in my cart from a different AliExpress vendor (with a significant shipping charge). If I actually get the tubes, or the at least show they are in the USA, I'll try again to place the order for the OPT's.

I am accumulating other small parts such as chassis mount XLR connectors, and I now have several scrap devices from which to pick a rack mount chassis and other parts."

UPDATE:
Почта России now indicates that my order of tubes was delivered from Serbia to the USA. USPS does not show any updates since they were at the Airport SW of Moscow. Once I get additional tracking information confirming that they are moving within the United States (or if a Serbian Postal worker shows up at my front door), I will re-order the OPT's from the other AliExpress vendor.

[admin]

First Actual Pictures of the Components of this Amplifier:


I received a package today:

Package with Russian Post markings

DSC_0030_sm.JPG (182.47 KiB) Viewed 18848 times

This package has taken a long an dangerous journey. It traveled from Kazakhstan, through a significant portion of the southwest quarter of Russia. It then was passed to Serbia as an intermediate, and then to the USA where iT spent almost a week in US Customs in New York before suddenly pinging in Orlando, Florida. It then crossed Florida before arriving at my local post office, and finally at my front door.

Interestingly (and I believe coincidentally ) it arrived one day latter than the estimated delivery time as predicted by eBay at the time of sale.

Current world situations and political views aside, I appreciate everyone from the Seller to the Mailman that handed the package to me.

It was well packed, with each tube wrapped in foam, and additional foam all around:

Tubes packed in Foam

DSC_0031_sm.JPG (204.93 KiB) Viewed 18839 times

The one tube (center bottom) that appears white is fine. That's just a reflection of the camera flash.

A randomly selected 6P3S-E is shown below. It has the Reflektor Logo and the diamond OTK markings:

6P3S-E

DSC_0032_sm.JPG (251.29 KiB) Viewed 18839 times

So far, I have tested one each randomly selected 6N8S and 6P3S-E. Both tested Good on my Simpson 1000 Plate Conductance tester:

6N8S Being Tested on My Simpson 1000

DSC_0035_sm.JPG (224.66 KiB) Viewed 18839 times

As previously mentioned, the next step is to make another attempt to purchase the OPT's from AliExpress. I am also shopping for tube sockets.

[admin]

UPDATES:


1. I have ordered Vintage NOS Bakelite Tube Sockets from a US seller via eBay.
2. So far at least 1 of the 6N8S tubes is bad -- One side heater is failed, and there is audible debris loose inside that side of the tube. While this is a disappointment, it is not an issue for this build since I purchased several extra 6N8S tubes.
3. I did a little work on the drawing -- updating it for the extra tube needed for a negative (-) input buffer. I also included a Chinese Lossless Player module that I *may* use as both a DAC input/Built-in Player/Visualizer. The small switch shown to the left of the module will allow it's line in to be optionally disconnected from the internal input circuitry. This will serve both as a Front/Rear input select switch and a way to isolate the module from the input circuitry to insure that it cannot affect the sound. While the inclusion of a module like this may "cheapen" this build, it does check all of the boxes for the functions mentioned above, and there is also the advantage that I have 2 extra of these modules on hand. It also would add Bluetooth, and a 1/8" stereo TRS input with DSP equalization. The main analog inputs (Rear) will not pass through the digital equalizer. A USB-B socket will be included on the rear to allow a "proper" USB connection. (See the linked article on why the front USB-A connector should not be used as a DAC connection.) The module (and thus the front USB-A Socket) will be powered by a separate internal Buck Converter rated for at least 2A, and thus this amplifier build will represent the most overkill for a phone charger ever conceived.

Render_March_2025.png (120.46 KiB) Viewed 18490 times

[admin]

The Monster Will Live !!! (Hopefully)

I have a major update to this Project:

The output transformers have arrived!

DSC_0075.JPG (244.66 KiB) Viewed 2318 times

...and I have to say they are pretty impressive -- at least by weight. They came well packed -- Basically triple boxed with foam packing in each of the first 2 boxes. Of course I haven't done any testing, but they at least look like 100W Tube Amplifier Output Transformers. An Octal Tube Socket is shown so that readers can get an idea of the size of these things!

One thing that was very interesting is that I was able to track the UPS Cargo plane as it brought my OPT's across the ocean. Once the package was handed off to UPS in Hong Kong, UPS had excellent tracking the rest of the way. The times of departure of the flight shown below corresponded very closely to the time UPS said that it departed HKG airport.

Screenshot_20250318-124830_sm.jpg (143.83 KiB) Viewed 2318 times

Arrival time in Anchorage was also too close be be another UPS flight. I was able to track the same aircraft (a 747-800) to Kentucky, where my package was put on a smaller aircraft for the trip to Florida, which I also tracked in a similar manner -- What a Journey! I've never gotten to ride on a 747-800!

In addition to Kudos to UPS, I thank AliExpress, and the AliExpress Seller. After my order from a different vendor was cancelled, I didn't think these OPT's were "real". They ended up getting here 1 day after AliExpress's latest estimated delivery day, and as a result, I received a small "coupon" from ALiExpress.


Other Parts/Components for this build are now also on hand:

All Significant (Read: "Expensive") components are now on hand. I'm not sure I'm going to set this up again, so here is the official pre-build parts shot for this project:

DSC_0072_low.JPG (201.26 KiB) Viewed 2308 times

Obviously, everything to build an amplifier -- especially a Monster Amplifier is not in this picture. As usual, I do not include hardware, discrete (passive) components or "consumables" such as heat shrink, etc.

Additionally, other significant components are not in this picture, even though some are on hand:

• The high voltage inverter board is not shown. It has not been ordered yet.
• Only 2 speaker terminals are shown. The Monster will have 4 and 8 ohm outputs, so I need 4 sets of terminals. Of course I will have matching terminals, so the ones shown may not be part of the completed project.
• I do not have a pair of balanced 1/4 TRS (Phone) jacks. I pulled a stereo RCA jack set for this picture, but I must have set it down someplace other than in the picture.
• I doubt I will use the shown IEC connector, as it is very cheaply made, but it is the one that fits the cut-out in the back of the shown chassis. I will probably use a better quality IEC connector with an integrated EMI filter even though the shown power supply has EMI filtering in the input.
• I mentioned adding toggle switches to allow isolation of the (shown) DAC/DSP module, but since that design is still "fluid", I did not include any toggle switches in this photo.
• 12V to 6.3V bucks (Heaters) are not purchased yet.

Other Notes:

• Note that I am showing the Chinese Lossless Player module that was mentioned previously. Given that this amplifier is not intended for premium living room listening, and given that I the capability to isolate the module (input and output) will be included, I am leaning toward including this module in the build, It is a way to add a lot of functionality (including visualization) in an easy and cheap way.
• The chassis is going to need a lot of metal work. I need to cut and drill new front and rear panels, I have installed the cover upside down from the way it was in the original device. This places the cooling intakes (not shown) at the bottom front for better convective flow. A piece of metal or possibly a PVC duct will separate intake air for the fan in the low-voltage power supply from convective airflow for the rest of the unit.
• The top of the chassis will be cut, and replaced with a thick piece of aluminum that will form the actual base of the tube amplifier. This is the same basic design as I did for my previous build. I plan to go on an aluminum hunting expedition in the coming weeks.
• Note the "Soviet Switch" as also used in the previous build.
• Tube sockets are vintage under-mount, and were purchased via eBay from a US seller.