• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

Tubelab Universal Driver Board, 2015 version

Perhaps 5000 ohms would be a good compromise.......I just happen to have two excellent sets of 5K OPT's. That's for another test day.........

Yesterday was the "another day."

I have a pair of Lundahl LL1663/PP OPT's that are 5K but are rated for "40 watts at 30Hz." I think I can get more than that from these tubes, so......

I have two different pair of Plitron OPT's that are 1.25K and are rated for "400 watts at 20 Hz." I have previously used one set of them at 5K ohms by connecting the 8 ohm load to the 2 ohm output tap. Maybe I won't get 400 watt capability this way, but I have seen 100+ watts. One pair of the Plitrons are seen in post #5 connected to a pair of transmitting tubes and wired for 2500 ohms (8 ohm load on 4 ohm tap). 250 watts flowed forth from this combination, but I didn't run a frequency response test. I decided to use their bigger brothers for this test! Everything else is as ran in the last test with the GE 6550A's. I just swapped the OPT's.

The first two pictures show the huge OPT. These things cost about $900 per pair when Plitron took them out of their catalog. I paid far less for a used pair. We found out last time that expensive tubes produced better MEASUREMENTS than the budget stuff, now does a $400 OPT do better than a $16 OPT........it better:

The results for a 5000 ohm OPT:

power THD B+ current output stage efficiency

0.1 W 0.086% 90 mA 0.2 %
0.5 W 0.074% 93 mA 1.2 %
1.0 W 0.083% 95mA 2.3 %
2.0 W 0.126% 98 mA 4.5 %
5.0 W 0.273% 111 mA 10.0 %
10 W 0.520 % 130 mA 17.1 %
20 W 0.855 % 160 mA 27.7 %
50 W 1.73 % 220 mA 50.5 %

Maximum tube dissipation is reached somewhere between 20 and 50 watts output, and is about 26 watts. It drops to 24.5 watts at 50 watts output. Maximum tube current is reached at maximum power (50 watts) and is 110 mA. Both of these are conservative for KT88 and 6550.

Conclusions.....Unless you need 70 watts, the 5K load is the optimum spot for triode wired KT88's on 450 volts. The distortion numbers can not be directly compared since the better OPT's obviously made better numbers. The power and efficiency numbers are about right.

So if I had frequency response numbers on the 3300 and 6600 cases, I would have been measuring the OPT's and not the amplifier circuitry. This may still be the case with the Plitrons, but I like the numbers. the distortion numbers put that "first watt" the squeaky clean zone below 1/10th of a percent.

Many amps measure the frequency response at something like 1 watt, but those numbers were beyond the range of my test equipment, so I decided to measure the frequency response at FULL POWER, 50 watts.

The low frequency -3db point was somewhere below 10 Hz, the 8903 is not accurate below 20 Hz. OPT saturation effects became visible on the scope around 8 Hz at 50 watts. The distortion at 20 Hz and 50 watts is 1.83% (picture #3). The frequency response at 20 Hz is 0.16 dB down at 20 Hz compared to 1KHz. (picture #4)

The high frequency -3db point is 78,568 Hz. The distortion at 20 KHz and 50 watts is 2.464% (picture #5). The frequency response at 20 KHz is up 0.07 dB compared to 1KHz. (picture #6)

I had planed to move on and test some different tubes, but these numbers have convinced me to build up the second channel, and maybe a power supply to see what this thing SOUNDS like! I plan to order enough parts to build a few more UD boards.....I still have my dead 300Beast amp that I built about 12 years ago. It's the only HiFi amp I saved from the past, still awaiting a total rebuild.

Progress and posting will be slow and sporadic until after the first of the year. With 4 grandkids, the Christmas season insanity is already out of hand, and Sherri has the next week off, so that we can catch up on life, and all the tasks that have been ignored for the last few weeks.
 

Attachments

  • P1000335_x.jpg
    P1000335_x.jpg
    923.9 KB · Views: 1,049
  • P1000336_x.jpg
    P1000336_x.jpg
    693.9 KB · Views: 1,018
  • P1000340_x.jpg
    P1000340_x.jpg
    691.4 KB · Views: 995
  • P1000342_x.jpg
    P1000342_x.jpg
    637.1 KB · Views: 974
  • P1000339_x.jpg
    P1000339_x.jpg
    677.3 KB · Views: 951
  • P1000343_x.jpg
    P1000343_x.jpg
    595.8 KB · Views: 259
Last edited:
If you look at the pictures in post #23 and many of my other posts for the last 10 years you can see a HP8903A Audio Analyzer (on top of the power supply to the left of the scope). It replaced the old HP33X series in the 1980's (I still have an HP331A). It does the same thing, except for having a better notch filter for a lower distortion floor, and it's somewhat automated. The automation is primitive by today's standards, but tuning and nulling the notch filter is automated.

Those distortion readings were made by simply pushing the distortion button on the front panel. It is only a basic THD + noise measurement, but good for comparisons on things like tubes and OPT's. It will do an automated frequency response sweep and send the output to an analog HP plotter, but you can't recover the data any other way. More automated results require an external computer and some programming.

I worked in the test equipment calibration lab at Motorola when these things came out, and I spent quite a bit of time playing with them. The measurements that make are accurate, and better than the old HP33X series due to the deeper notch filter. The HP8903B series is a bit better, and a lot more reliable. I have repaired this one 4 times in 20 years.

The instrument I have actually came from Motorola when the first large Florida plant closed down and all the stuff was scrapped, auctioned, or donated to the local college.

I use a PC with a 24/192 sound card for FFT measurements (individual harmonics), but I haven't made it work right since I moved 3 years ago. I lost the license dongle for WinMLS, so I will try some free software soon.

The second channel PC board is now stuffed with the same parts that is in the board seen throughout this thread. All this stuff will not fit on the plywood base I have, and I'm not about to head out to Lowes today, so I'm repacking it all to fit the available space. The OPT's will have to move to another piece of wood. I have a pair of 100 watt Edcors that need to be tested too!
 
more time going forward to continue the R&D (followed by good sales!)

Sales of my PC boards, DIY audio, and small project DIY stuff in general have always been cyclical. People tend to do more indoor building during their local winter than in their summer months. Sales are very slow during the summer months, then tend to pick up around October, peak in January, then taper off toward April or May.

That did not happen this year. I had 4 sales in October, 3 in November, and the third for December just came in. Despite an unusual uptick in sales that started late last year, 2017 has been the worst year for board sales since 2007. Tubelab Inc. has never made a lot of money, and lost money in several years which was not a problem when I had a full time engineering job with a fat paycheck.

Now that I'm retired and trying to support two families, Tubelab must pay for itself, which it has barely done in the last 3 years. In order for Tubelab to continue I need to lower costs and generate new revenue.

I plan to try an offshore PC board vendor for a test batch or two to see if I can get a better board for less than I'm paying now.

The cost of shipping continues to go up every year, but I have not raised my rates. That may have to change.

I had planned to offer parts kits for the SSE and restock the parts kits for the SPP, but I'm buying parts from the same places everybody else is (Digikey, Mouser....) so the only way that works is to buy large quantities Unfortunately the payoff for this takes years, and a large cash outlay is needed up front. Not likely to happen soon.

I have several vacuum tube PC boards part way through the design process. These should turn into new vacuum tube products.

There are a few non vacuum tube audio related projects that are a possibility, and maybe even RF (ham radio) projects.

In the short term, I can sell these boards. For now the only construction info is what's right here. The boards are $25 USD plus the usual $8 shipping in the US, and $15 international for up to 4 boards.

I have finished populating the second board. It works great stand alone. Assembly of a stereo test amp is underway, but likely stalled for a few days.....4 grandkids, and 2 parents are scheduled to invade this house tomorrow morning. All objects that could be considered hazardous to 3 and 5 year old kids who think that they are Ninja' in training must be hidden!
 
My career became messed up last year ...and the job I'm at now pays no where near what I made before ...but I had an interview that I hope will be in line with the kind of stuff I do for a salary close to what I was make...

I hope things change for the better for you...
 
My boards can be ordered by sending payment via Paypal to the Tubelab email address (see below). Make sure that Paypal has your correct shipping address on file, or provide it in the comments box. State which board or boards you want, and how many in the comments box, or send a separate email with this information within a few minutes of sending payment, so that I can match it with the order.

The SSE, SPP, and TSE boards are $35 USD each. Each board makes a stereo amp with power supply.

The board described in this thread is called the UD or Universal Driver board. It is $25 due to its smaller size. Each board makes a single channel driver. Two boards are required for stereo, and a power supply for the amp is needed.

A suitable power supply schematic will be posted soon, and a PC board for it should be available within a month or two depending on PCB delivery times. I buy my current boards in batches of 100 with a 4 week delivery time to get an acceptable price. That may not be an option this time.

Shipping for 1 to 4 boards in the USA is $8. International shipment is $15.

I do not post my complete email address in plain text format in an attempt to control spam. Even still I get 20 to 30 useless emails a day. My personal email address that I have had for over 25 years gets over 100 spams a day, many loaded with evil payloads.

The address is "tech AT tubelab.com" remove the AT and replace it with the appropriate symbol without the spaces or quotes.
 
George, after few Tubelab SE built on your boards I am really interested in the amp based on the UD board and I will buy few sets when they will be available. It's great that you will publish the schematic of the suitable power supply, but could you please publish also the schematic of the full implementation for the KT88 tubes too?
 
I've built the 300B SE and am happy with it. Once the PS boards are in I'll be ordering one and a set of the driver boards. I'm about to start a new PP amp build and those boards should do nicely. I sent you an email and a PM recently regarding the 300B SE. I know its probably a stupid question, but since it is wired using a three prong plug, does it have to be earth grounded? The home I bought has no grounded outlets. Please disregard the email if you see this first.
 
what input tubes have you been using primarily during your tests.....Hope you had a Merry Christmas!

We're still cleaning up the mess from the invasion of the grandkids, and its too cold to go outside (freezing temps forecast for the next 10 days), but yes it was fun.

So far all testing has been done with 6CG7/6FQ7's in both tube sockets. The choice of tubes depends on the amount of gain needed and the drive voltage requirements, which influence the supply voltage needed by the board.

The UD board started out in 2007 with 9 pin tubes. I used 12AX7's, 12AT7's, 12AY7's, 5965's, 5963's, 12BH7's, 6DJ8's, 6FQ7's, 6BQ7's and a few others that I can't remember right now. The board got re-spun with octal tubes for the 6L6GC in AB2 thread. I used a 6SL7 or a 6SN7 for the input tube, and a 6SN7 for the second stage in those boards, which I still have and tested recently.

Since going back to 9 pin tubes with this board the only other tube I have tried in the input position was the 6DJ8/6922. I have optimized the board for a pair of 6CG7's. Other tubes may need resistor changes, probably just the plate load though. You don't want to suck 10 mA through a 12AX7....for more than a few seconds. A pair of 12AT7's will work, but may provide too much gain, and may not like supply voltages much over 400 volts. I did use a 12AT7 for the input tube in the old board with a 12BH7 or 5965 for the second stage to get big drive voltages for screen drive in the old board.

It's great that you will publish the schematic of the suitable power supply, but could you please publish also the schematic of the full implementation for the KT88 tubes too?

It's all coming in the next few weeks. I will be leaving town for a few days tomorrow morning. If Windows 10 on my laptop chooses to work (a big if) I may have internet access from the hotel in the evenings. If not, I will be back on Monday and I'm going to plug it in!

What's it? Well, I set out to make a few minor changes to my breadboard, and got a bit carried away. I took it all apart and rebuilt it the way I intended 10 years ago. It's not quite done yet, but it is twice what it was before, and it has shiny new (OK, old but never used) blue things that will allow UL testing.

Still no power supply yet. I have a power transformer that I bought 10 years ago......Maryland Toroid part number 236.5072, serial number 001. It cost me $100 then, but they are stupid money now. It looks like a perfect fit for this amp if the negative voltage supply makes enough voltage. If it does, it's going into my amp, but I will find a more reasonably priced alternative for the rest of the world....I don't think too many people want to spend $300 for a power transformer.

I know its probably a stupid question, but since it is wired using a three prong plug, does it have to be earth grounded?

There are two major shock risks associated with a tube amp, and any DIY electronics in general. A tube amp involves high voltages which under normal conditions are contained inside the amp. A short circuit from one of these voltages to something touchable by a user can cause a serious shock, or death. The likely component failure to cause this type of short is a transformer. It is not common, but I have seen two in undamaged amps, and one in my own amp, but it had gotten wet during a hurricane.

If, say an OPT develops a short that puts B+ on a speaker lead, the amp could continue to play normally if the secondary of the OPT was not grounded. This is why I strongly recommend grounding the OPT secondary's common (black) lead. If ALL metal surfaces are connected to the 3rd (safety ground) prong of the power plug a short from a voltage source to the case or other metal will blow the fuse or trip the line breaker. The amplifier's input ground should be connected to the safety ground as well to avoid a possible voltage on the turntable or CD player.

Operating the amp in a system without a connection to the earth ground removes the path to trip the breaker or blow the fuse if such a short should develop in the primary of the power transformer or its wiring. In this case all of your equipment could be hot, and you would never know it until someone touched one of the components and a grounded object. The best alternative would be to run a wire from the amp chassis to a known good ground and plug all of the equipment into a common power strip. Older houses tend to have copper or iron plumbing which should be grounded somewhere.

In addition to the above, if your house has breakers install a GCFI breaker in the circuit that feeds the audio system. It will trip if there is any current leaving the breaker box that does not return through the breaker box. It will NOT trip if there is a short that puts B+ on the speaker leads or chassis, but a properly built amp should blow its internal fuse if that happens.
 

Attachments

  • P1000454_x.jpg
    P1000454_x.jpg
    868.1 KB · Views: 855
Thanks! As most of my equipment is older, console pulls, a Dynaco ST'70, ect. almost all have two prong plugs. I thought about running a ground from the one outlet to a water pipe, but didn't think having just one outlet in a circuit with several outlets in the room was a good idea. Am I wrong?
 
Having one outlet that has a ground for the most risky of your equipment make sense to me. The equipment built today is far safer than old stuff since insulation materials have improved and they are relatively new. The insulation in that transformer in your old Dynaco is paper. Old paper that may have absorbed some moisture too. A lot of old consoles had the "death cap" too. I would get that stuff grounded.

Look at it this way. If you could have seat belts in only one car, but you had two and used them equally, would you put them in that 66 Mustang with drum brakes, iffy steering, a pole spear steering column, and a steel dash, or in a modern car that's full of soft surfaces and nearly drives itself?