Those interested in UNSET boards should read and respond to the questions at the end of this post.
I'm OK. I did have surgery to remove two separate skin cancers. One was deemed "invasive" so further surgery may be needed. The first 62 years of my life was spent in south Florida, and I was outdoors a lot. I have been having skin cancer issues for over 20 years, and likely will for as long as I live. The surgery and the narcotics sidelined me for a about 10 days.
The winter here was the snowiest that I have seen. It stayed cold enough, long enough that we had snow on the ground for all of February and the beginning of March. The big thaw came about 10 days ago, followed by the inevitable flood. As soon as I was able, I began working outside most of the day to clean up the mess. The last few days have been quite nice, but the rain came today, and snow is in the forecast for Sunday night.
Those who follow my antics know that most of my early projects were made from trash, and I learned the art of dumpster diving at a young age. Well last spring, and again this year I negotiated the acquisition of about a ton of scrap ornamental concrete from a local lumber yard. They gave it to me instead of putting it into a dumpster that they would have to pay for using. The only issue is that I need to go get it.
I live on a creek that floods often, so I use the junk concrete for erosion control. Hauling a ton of concrete stones, squares, and cinder blocks in a mini-van takes a lot of trips. I spent most of this morning partially unloading the last of 4 loads. I took a break due to rain.
The short answer is maybe.
Tubelab has been able to pay its way, and even turn a small profit most every year since it's startup in 2007. Up until 2014 I had an engineering salary, so if Tubelab fell short one year it really didn't matter. That situation changed when I took the buyout and moved out of Florida in 2014. Tubelab had to be self sufficient, and it was for 2014 and 2015. A RF design contract job kept Tubelab alive in 2018 and 2019, but 2020 was the worst financial year ever, both for Tubelab and us personally.
The PC board house that I have been using for 15 years with no issues has screwed up 3 out of the 4 last board orders. Their issues have left me with a lifetime supply of SSE boards, a batch of TSE-II's that are worthless, and an order of SPP's placed 6 weeks ago that I'm still waiting on. After over a month they told me that they never got the order, despite me having an email confirmation from their system. Their ordering system and poor communication is how I got SSE's instead of TSE-II's, but the bad TSE-II's are "because I use an old version of Eagle." I will not use them any longer.
These issues have consumed time, and created frustration. Sherri has been telling me to close Tubelab and walk away for a couple years now, so I agreed to close Tubelab when I run out of product to sell, or at the end of 2021. She knows that I have something new in the works (UNSET) and has been hearing it rattle the basement quite often. She also knows how much time I spend supporting any new product, especially something as unique as the UNSET.
Last night I fired a "hail mary" shot that may delay the inevitable. I'm down to two good TSE-II boards, and three SPP's, but have plenty of UD's and SSE's. 50 SPP's have been on order since January. They may or may not show up.
At the recommendation of several diyAudio users I ordered some TSE-II's from JLCPCB in China. If they arrive without issue, I will build a couple and test them thoroughly, in an abusive manner. It the boards are as good as my current boards, I may get boards from them in the future.
Assuming that the Chinese PCB's are good, and I get some and sell enough of them to cover their cost, I could order some UNSET boards from the same vendor.....
So, what is an UNSET board?
There are two home cooked prototype UNSET boards in existence. Both have been taken to the 40 watt per channel power level to find their weak points. There are three weak spots in that design at the 40 WPC level. The board has provisions for using ANY octal tube in the output stage via jumper selection. A different board would be needed for 12 pin compactrons, or any 9 pin tubes. That is a simple layout change.
The 5AR4 rectifier tube is iffy above 15 WPC. This is also seen in the SSE, but 15 WPC is the limit of a 6550 or KT88 in UL. The current board has silicon diodes in series with the plates like those in all SSE's since 2010. That get's most new production rectifiers to 15 WPC. For test purposes, I removed the 5AR4 tube in my board and put some serious $5 diodes in the board. That works up to 40 WPC with no board mods except a jumper across the 5AR4 tube.
The screen grid "regulator" (buffered zener) in the current UNSET board uses a TO-220 mosfet. It must operate in the linear region. Most mosfets designed for SMPS use will fail due to SOA issues when asked to drop 300 to 400 volts at 50 mA or so. There are plenty of budget mosfets that work fine at B+ voltages in the 450 volt range which will get into the 15 WPC region, but may fail at over 450 volts. Again there are IXYS L2 or even depletion mode fets that work fine, but they are not cheap and need a large heat sink.
The board mounted heat sinks for the output tube cathode mosfets are fine at 15 WPC, especially with music. I need forced airflow to run them at 40 WPC sine waves for several minutes. Something better will be needed for high power, I'm planning on using the chassis.
So, several people have expressed the desire for a single board that is the same size as the TSE-II and makes about 15 WPC or less. The current design will do this with no modifications. If the Chinese PC board vendor checks out, I may send in an order for a very small quantity do do some testing. If the results are promising, the 15 WPC unset could be ready in a month or two.
So, what are the other options?
At the end of 2020 I had started laying out a board with the rectifier tube replaced with silicon, and room for a dropping resistor for the screen regulator. All the other stuff was moved around to make room for more air holes like those in the TSE-II. I never finished it. That board could be pushed quite a bit harder since the weak links are removed. It has not been built or tested, so a longer time frame is needed.
Before finishing that layout, I did the push pull experiment that made stupid power outputs.....250 watts from ONE board. Again, much more work is needed.
Further thought brought me back to the modular approach. Here one small PC board would be used for the power supply. A single board design of about 100 mm square would generate B+, screen, and +/- 150 volts or so for UD board designs. This would be compatible with a modular UNSET or a UD board design. The current UD boards would be used for driving any output tube in a conventional G1 or screen drive amp. There would be a new UNSET driver containing two tubes, capable of driving two output tubes in a stereo SE configuration, or a single channel push pull configuration. Two boards would be needed for stereo. Exact details for this approach are not finalized, and it would happen possibly after the current UNSET board is launched If I ever get that far.
I have two different push pull boards using the same technology. They are not unSET, since they are not SE, but a single board that made 40 WPC or more would neo be hard either. One making 70 WPC already exists.
If Tubelab Inc is to continue, I need a new product that would sell enough for this year to be positive cash flow. Chinese boards will help, I was paying nearly $15 each for TSE-II's.....
So what do YOU guys want?
And one more dumm question, Green solder mask or Black? I like the look of a black board with gold finish, green is the typical conservative choice, all the blue ones I have have faded over the years, and red, well I have some of Pete's red boards, but that's his look.
I'm OK. I did have surgery to remove two separate skin cancers. One was deemed "invasive" so further surgery may be needed. The first 62 years of my life was spent in south Florida, and I was outdoors a lot. I have been having skin cancer issues for over 20 years, and likely will for as long as I live. The surgery and the narcotics sidelined me for a about 10 days.
The winter here was the snowiest that I have seen. It stayed cold enough, long enough that we had snow on the ground for all of February and the beginning of March. The big thaw came about 10 days ago, followed by the inevitable flood. As soon as I was able, I began working outside most of the day to clean up the mess. The last few days have been quite nice, but the rain came today, and snow is in the forecast for Sunday night.
Those who follow my antics know that most of my early projects were made from trash, and I learned the art of dumpster diving at a young age. Well last spring, and again this year I negotiated the acquisition of about a ton of scrap ornamental concrete from a local lumber yard. They gave it to me instead of putting it into a dumpster that they would have to pay for using. The only issue is that I need to go get it.
I live on a creek that floods often, so I use the junk concrete for erosion control. Hauling a ton of concrete stones, squares, and cinder blocks in a mini-van takes a lot of trips. I spent most of this morning partially unloading the last of 4 loads. I took a break due to rain.
The short answer is maybe.
Tubelab has been able to pay its way, and even turn a small profit most every year since it's startup in 2007. Up until 2014 I had an engineering salary, so if Tubelab fell short one year it really didn't matter. That situation changed when I took the buyout and moved out of Florida in 2014. Tubelab had to be self sufficient, and it was for 2014 and 2015. A RF design contract job kept Tubelab alive in 2018 and 2019, but 2020 was the worst financial year ever, both for Tubelab and us personally.
The PC board house that I have been using for 15 years with no issues has screwed up 3 out of the 4 last board orders. Their issues have left me with a lifetime supply of SSE boards, a batch of TSE-II's that are worthless, and an order of SPP's placed 6 weeks ago that I'm still waiting on. After over a month they told me that they never got the order, despite me having an email confirmation from their system. Their ordering system and poor communication is how I got SSE's instead of TSE-II's, but the bad TSE-II's are "because I use an old version of Eagle." I will not use them any longer.
These issues have consumed time, and created frustration. Sherri has been telling me to close Tubelab and walk away for a couple years now, so I agreed to close Tubelab when I run out of product to sell, or at the end of 2021. She knows that I have something new in the works (UNSET) and has been hearing it rattle the basement quite often. She also knows how much time I spend supporting any new product, especially something as unique as the UNSET.
Last night I fired a "hail mary" shot that may delay the inevitable. I'm down to two good TSE-II boards, and three SPP's, but have plenty of UD's and SSE's. 50 SPP's have been on order since January. They may or may not show up.
At the recommendation of several diyAudio users I ordered some TSE-II's from JLCPCB in China. If they arrive without issue, I will build a couple and test them thoroughly, in an abusive manner. It the boards are as good as my current boards, I may get boards from them in the future.
Assuming that the Chinese PCB's are good, and I get some and sell enough of them to cover their cost, I could order some UNSET boards from the same vendor.....
So, what is an UNSET board?
There are two home cooked prototype UNSET boards in existence. Both have been taken to the 40 watt per channel power level to find their weak points. There are three weak spots in that design at the 40 WPC level. The board has provisions for using ANY octal tube in the output stage via jumper selection. A different board would be needed for 12 pin compactrons, or any 9 pin tubes. That is a simple layout change.
The 5AR4 rectifier tube is iffy above 15 WPC. This is also seen in the SSE, but 15 WPC is the limit of a 6550 or KT88 in UL. The current board has silicon diodes in series with the plates like those in all SSE's since 2010. That get's most new production rectifiers to 15 WPC. For test purposes, I removed the 5AR4 tube in my board and put some serious $5 diodes in the board. That works up to 40 WPC with no board mods except a jumper across the 5AR4 tube.
The screen grid "regulator" (buffered zener) in the current UNSET board uses a TO-220 mosfet. It must operate in the linear region. Most mosfets designed for SMPS use will fail due to SOA issues when asked to drop 300 to 400 volts at 50 mA or so. There are plenty of budget mosfets that work fine at B+ voltages in the 450 volt range which will get into the 15 WPC region, but may fail at over 450 volts. Again there are IXYS L2 or even depletion mode fets that work fine, but they are not cheap and need a large heat sink.
The board mounted heat sinks for the output tube cathode mosfets are fine at 15 WPC, especially with music. I need forced airflow to run them at 40 WPC sine waves for several minutes. Something better will be needed for high power, I'm planning on using the chassis.
So, several people have expressed the desire for a single board that is the same size as the TSE-II and makes about 15 WPC or less. The current design will do this with no modifications. If the Chinese PC board vendor checks out, I may send in an order for a very small quantity do do some testing. If the results are promising, the 15 WPC unset could be ready in a month or two.
So, what are the other options?
At the end of 2020 I had started laying out a board with the rectifier tube replaced with silicon, and room for a dropping resistor for the screen regulator. All the other stuff was moved around to make room for more air holes like those in the TSE-II. I never finished it. That board could be pushed quite a bit harder since the weak links are removed. It has not been built or tested, so a longer time frame is needed.
Before finishing that layout, I did the push pull experiment that made stupid power outputs.....250 watts from ONE board. Again, much more work is needed.
Further thought brought me back to the modular approach. Here one small PC board would be used for the power supply. A single board design of about 100 mm square would generate B+, screen, and +/- 150 volts or so for UD board designs. This would be compatible with a modular UNSET or a UD board design. The current UD boards would be used for driving any output tube in a conventional G1 or screen drive amp. There would be a new UNSET driver containing two tubes, capable of driving two output tubes in a stereo SE configuration, or a single channel push pull configuration. Two boards would be needed for stereo. Exact details for this approach are not finalized, and it would happen possibly after the current UNSET board is launched If I ever get that far.
I have two different push pull boards using the same technology. They are not unSET, since they are not SE, but a single board that made 40 WPC or more would neo be hard either. One making 70 WPC already exists.
If Tubelab Inc is to continue, I need a new product that would sell enough for this year to be positive cash flow. Chinese boards will help, I was paying nearly $15 each for TSE-II's.....
So what do YOU guys want?
And one more dumm question, Green solder mask or Black? I like the look of a black board with gold finish, green is the typical conservative choice, all the blue ones I have have faded over the years, and red, well I have some of Pete's red boards, but that's his look.
George,
Good to know you are up and about, and able to move big chunks of concrete effortlessly
Hope your strength will fully recover soon and that the future will be long and enjoyable.
The news about UNSET and the related PP possibilities is exciting! Sorry to hear about the issues with your PCB supplier though.
Regarding what we want. Speaking only for myself, I believe the modular approach is the way to go. Separate power supply and UNSET boards. Size and color not important, except enough room for large coupling caps, etc. That would also help with volume of sales, I believe, because lots of DIYers sometimes need a nice B+, regulated screen supply @ bias for all kinds of projects, including non-Tubelab ones.
Regarding the UNSET board, I think the front tube and MOSFET could be onboard, but having the power tubes off-board will allow flexibly to use a wider variety of tubes, as well as heater voltages. If possible, the big MOSFET should also be off board to allow flexibility of heat management schemes and placement.
Thanks for asking for my 2 cents.
Good to know you are up and about, and able to move big chunks of concrete effortlessly
Hope your strength will fully recover soon and that the future will be long and enjoyable. The news about UNSET and the related PP possibilities is exciting! Sorry to hear about the issues with your PCB supplier though.
Regarding what we want. Speaking only for myself, I believe the modular approach is the way to go. Separate power supply and UNSET boards. Size and color not important, except enough room for large coupling caps, etc. That would also help with volume of sales, I believe, because lots of DIYers sometimes need a nice B+, regulated screen supply @ bias for all kinds of projects, including non-Tubelab ones.
Regarding the UNSET board, I think the front tube and MOSFET could be onboard, but having the power tubes off-board will allow flexibly to use a wider variety of tubes, as well as heater voltages. If possible, the big MOSFET should also be off board to allow flexibility of heat management schemes and placement.
Thanks for asking for my 2 cents.
Part of the reason Pete's big red board was so popular was what I did to it. Scattered throughout that thread were my experiments from getting 50 WPC from Pete's 18 WPC board on the first day I had it, to the day it pushed the readout on my HP8903A to 525 watts! A number that has not been seen on that readout since. I do have a plan for making 1KW however.
525 watts were reached with both channels running in parallel through a common OPT. I got 250 WPC in stereo mode, so I backed it down to a "conservative" 125 WPC, and published all the info needed to get there in the original thread. At least 10 of those 125 WPC versions were build in addition to the three that I made. The 50 WPC monoblock board is Pete's more conservative take on my mods.
The one big board thing gets to be a problem with many PCB vendors. Many will not do it at all today, or will ask you to waiver all flatness specs. Big boards also make it harder for a vendor to use every last square inch on a panel, so the cost goes up.
525 watts were reached with both channels running in parallel through a common OPT. I got 250 WPC in stereo mode, so I backed it down to a "conservative" 125 WPC, and published all the info needed to get there in the original thread. At least 10 of those 125 WPC versions were build in addition to the three that I made. The 50 WPC monoblock board is Pete's more conservative take on my mods.
The one big board thing gets to be a problem with many PCB vendors. Many will not do it at all today, or will ask you to waiver all flatness specs. Big boards also make it harder for a vendor to use every last square inch on a panel, so the cost goes up.
Pete eventually went to the off-board power tubes for the reasons I mentioned.
Engineer's amp 50W monoblock
His Engineer’s amp is popular because it is a good amp using (then) inexpensive tubes, once available from the $1 lists. (But have you looked at current prices for 6HJ5s? on theBay?) And George popularized it with his high power experiments.
Having the one huge board is not convenient to use and as George explained, not economical, that was why Pete went to the monoblock approach. And it is a pain to build with the big heatsink that has to be aligned just so to fit above the chassis and be exactly where the board has the pass transistors on the PCB under the chassis.
Engineer's amp 50W monoblock
His Engineer’s amp is popular because it is a good amp using (then) inexpensive tubes, once available from the $1 lists. (But have you looked at current prices for 6HJ5s? on theBay?) And George popularized it with his high power experiments.
Having the one huge board is not convenient to use and as George explained, not economical, that was why Pete went to the monoblock approach. And it is a pain to build with the big heatsink that has to be aligned just so to fit above the chassis and be exactly where the board has the pass transistors on the PCB under the chassis.
Last edited:
PowerDrive “Coupling” Board!
Great to see you back George!
Good news about West Virginia is that the UV is a tiny fraction of what you get in FL (or here in TX for that matter.) Bad news is that the UV is a tiny fraction, so take your Vitamin D3!
I think that Pete Millet has a pretty good PCB business model. A lot of his boards are quite modest, but pretty good quality and a slightly premium price (which covers a lot of sins... I personally bought several pairs of Filament Regulator boards from Pete for $30/pair. Subsequently there have appeared some shunt regulator boards (Coleman, Salas, plus the usual weird Chinese boards) but Pete’s seem to hit my sweet spot.
That having been said, I ordered a couple of Salas boards for preamp experiments, but have yet to build them (documentation is pretty spotty and lots buried on this forum).
Personally, I think you have a gigantic winner in the Power Drive circuit.
I’ve built TWO hand-wired versions which are sloppy and time consuming, but work great. I would pay an exorbitant amount if I could buy a TubeLab PowerDrive board. Mono is fine, probably preferable.
This would include:
CCS for driver anode, with associated resistors
Coupling cap
Maybe driver cathode circuit (but not necessary for most drivers I would think)
MOSFET power source (tapped-off B+)
MOSFET with heat sink
Bias voltage divider
Maybe bias supply...although pretty easy to do this on tag boards
For MY purposes, I’d like terminal blocks for
(1) coupling cap
(2) CCS regulator resistor (snubber can be soldered down)
Connections to the board (B+, connection to driver anode, connection to Power Tube grid stopper, star ground..) could be either terminal blocks or solder pads for flying leads.
Just my 2 Canadian Dollars (a LOONIE, eh?)
In the meantime, The TubeLab Legion would undoubtedly want to help you “dispose” of those surplus boards, especially if there’s a chance that it will help get us an UNSET. Think of it as a TubeLab GoFundMe. Alternately, a REAL GoFundMe page for the UNSET could work...
What have you got, and what do you want for it? I would take SSE or TSE boards in quite decrepit shape for crazy experiments if I could apply a little McGyver and make’em work. Maybe even some PP boards if I could hack together a screen drive 6AV5 PP....
Heck, a lot of us have been hacking at these boards for a decade...we don’t want to stop now.
Again, great to have you back George!
Great to see you back George!
Good news about West Virginia is that the UV is a tiny fraction of what you get in FL (or here in TX for that matter.) Bad news is that the UV is a tiny fraction, so take your Vitamin D3!
I think that Pete Millet has a pretty good PCB business model. A lot of his boards are quite modest, but pretty good quality and a slightly premium price (which covers a lot of sins... I personally bought several pairs of Filament Regulator boards from Pete for $30/pair. Subsequently there have appeared some shunt regulator boards (Coleman, Salas, plus the usual weird Chinese boards) but Pete’s seem to hit my sweet spot.
That having been said, I ordered a couple of Salas boards for preamp experiments, but have yet to build them (documentation is pretty spotty and lots buried on this forum).
Personally, I think you have a gigantic winner in the Power Drive circuit.
I’ve built TWO hand-wired versions which are sloppy and time consuming, but work great. I would pay an exorbitant amount if I could buy a TubeLab PowerDrive board. Mono is fine, probably preferable.
This would include:
CCS for driver anode, with associated resistors
Coupling cap
Maybe driver cathode circuit (but not necessary for most drivers I would think)
MOSFET power source (tapped-off B+)
MOSFET with heat sink
Bias voltage divider
Maybe bias supply...although pretty easy to do this on tag boards
For MY purposes, I’d like terminal blocks for
(1) coupling cap
(2) CCS regulator resistor (snubber can be soldered down)
Connections to the board (B+, connection to driver anode, connection to Power Tube grid stopper, star ground..) could be either terminal blocks or solder pads for flying leads.
Just my 2 Canadian Dollars (a LOONIE, eh?)
In the meantime, The TubeLab Legion would undoubtedly want to help you “dispose” of those surplus boards, especially if there’s a chance that it will help get us an UNSET. Think of it as a TubeLab GoFundMe. Alternately, a REAL GoFundMe page for the UNSET could work...
What have you got, and what do you want for it? I would take SSE or TSE boards in quite decrepit shape for crazy experiments if I could apply a little McGyver and make’em work. Maybe even some PP boards if I could hack together a screen drive 6AV5 PP....
Heck, a lot of us have been hacking at these boards for a decade...we don’t want to stop now.
Again, great to have you back George!
Last edited:
Agreed.
I agree with all of this. That’s one reason why a PowerDrive board should be MONO. Just my Loonie.
I agree with all of this. That’s one reason why a PowerDrive board should be MONO. Just my Loonie.
Monoblock TSE-Ii
Just an example:
I just received two TSE-I boards from George.
I WAS going to use one for a 45 stereo amp, and sit on the other one while I built 300B mono blocks point-to-point, as I HAVE the iron for mono blocks but NOT for a stereo TSE-II 300B.
I just decided last night to use the two TSE-II boards to build the 300B mono blocks. Will use iron already on-hand, will run lots cooler, and I don’t have to run a stereo amp in my mono system. Plus, lots of real estate for massive iron...
Just an example:
I just received two TSE-I boards from George.
I WAS going to use one for a 45 stereo amp, and sit on the other one while I built 300B mono blocks point-to-point, as I HAVE the iron for mono blocks but NOT for a stereo TSE-II 300B.
I just decided last night to use the two TSE-II boards to build the 300B mono blocks. Will use iron already on-hand, will run lots cooler, and I don’t have to run a stereo amp in my mono system. Plus, lots of real estate for massive iron...
I actually like OFF-Board sockets better.
(A) more mounting flexibility
(B) can use tubes with weird pin-outs
For example, one of my SSE boards has a set of outboard sockets for 6AV5s (although the wires could be flopped around for other pin-outs of course)
My point-to-point FLEXI-amp (powered by Heathkit regulated p/s) has both 4-pin and 5-pin sockets in parallel (well, with one dangling participle, of course) and can take for example 71A, 45, 46, 2A3, 300B etc. by virtue of a variable, regulated filament supply (similar to the TSE, except common-mode choke and deck-mounted pot and test jacks).
In an ideal world, the boards would be 2” wide so that they could be arranged vertically, both for more flexibility AND for better cooling air flow over the heat sink fins...
Just my perspective.
I ordered some boards from JLCPCB in China. They will do up to a 400mm by 500mm board, but they charge extra, and will not guarantee the yield. All boards 200mm X 250 mm or smaller are fully specced. I don't plan on doing anything that big.
I did not know about ALLPCB, and there may have been some issues with PCBWAY.
Good to see you back!
Developing products for sale takes a huge amount of time and effort. In my experience, when a hobby has to pay for itself it becomes a chore.
So, what do you want?
In the interests of flexibility my preference is for boards without power supplies or output tubes. I really like the concept of the Universal Driver Board. I have 1 UDB in a breadboard and another 2 that are going into an amp shortly.
If you had either or both single channel SE and PP UNSET PCBs without output tubes and without power supplies available in the next year I'd be interested. I'd be interested in more complete designs too but they're going to take a lot longer.
Maybe boards without output tubes and power supplies initially followed by more complete designs later.
Given the time and effort required to develop complete designs and offer PCBs for sale there's also the option of publishing the designs and letting those interested play with them. There's options in between too. It really comes down to what you want and can realistically do in a time frame that will be of interest to people.
I actually like OFF-Board sockets better.
(A) more mounting flexibility
(B) can use tubes with weird pin-outs
For example, one of my SSE boards has a set of outboard sockets for 6AV5s (although the wires could be flopped around for other pin-outs of course)
My point-to-point FLEXI-amp (powered by Heathkit regulated p/s) has both 4-pin and 5-pin sockets in parallel (well, with one dangling participle, of course) and can take for example 71A, 45, 46, 2A3, 300B etc. by virtue of a variable, regulated filament supply (similar to the TSE, except common-mode choke and deck-mounted pot and test jacks).
In an ideal world, the boards would be 2” wide so that they could be arranged vertically, both for more flexibility AND for better cooling air flow over the heat sink fins...
Just my perspective.
I'm too deeply mired in the other project to have followed the UNSET as closely as I would like.
I think there is some merit to having the power tube and the MOSFET off board so long as stability is not compromised. If stability is an issue, then maybe a hole in the board and solder pads to put a builders choice tube socket in place.
I'm ambivalent about the modular vs. whole board issue. Personally, I don't have any problems with just using the part of a board that I need and leaving the rest unused. That could reduce the number of boards that need to be stocked. Maybe they could be made where the customer just snaps off the parts they don't need / want at the moment.
I'll be the odd person out and suggest yellow for the board color.
I'm glad you're recovering quickly from the health issues
Just in time for mowing season!
I think there is some merit to having the power tube and the MOSFET off board so long as stability is not compromised. If stability is an issue, then maybe a hole in the board and solder pads to put a builders choice tube socket in place.
I'm ambivalent about the modular vs. whole board issue. Personally, I don't have any problems with just using the part of a board that I need and leaving the rest unused. That could reduce the number of boards that need to be stocked. Maybe they could be made where the customer just snaps off the parts they don't need / want at the moment.
I'll be the odd person out and suggest yellow for the board color.
I'm glad you're recovering quickly from the health issues
Just in time for mowing season!
For most of my engineering career at Motorola, I developed products for sale, either police / military two way radio where lives depend on them working and cost is no object, or cell phones where every penny counts, and you learn to ask questions like "how much does it cost to place a part on a board." That one took nearly a year to answer. I have had years of experience doing it, so much of the "chores" come easy.
In the case of the UNSET, I already have a working board for the "all in one" case. It would require minimal if any clean up to be production ready. It's just not the 40 WPC SE amp that some may think it is since I have been squeezing 40 WPC out of it.
The current all in one board has a pair of Octal sockets mounted on the board with each pin connected to a jumper pad, no other connection. During setup a wire jumper is connected from a row of holes to the proper pin at the tube socket. This allows for easy configuration for any octal tube. It would be relatively easy to make a new board for other sockets. It also allows for connecting off board sockets directly. I may try some 4D32's soon to prove that point.
This option depends on how the Chinese boards work out. My previous vendor charged a $100 fee any time new artwork was needed. 20 working day turn around times and 50 board minimum order size was needed to get the cost per board down to under $15. JLCPCB can beat that easily on a 10 board order in shorter time even with shipping, according to their web site. I should know more in a couple weeks.
It may be possible to separate the power supply, driver, and output stages into separate boards, then optimize each board and add what ever is needed to facilitate its use in a more universal situation. For instance the power supply would also get a simple +/- 150 volt source for UD use, options for tube or solid state rectifier, and provisions for a BIG fat mosfet in the screen supply so that it could be used for B+ in a mid sized conventional amp. That board would be UD sized or smaller.
The cathode drive mosfet sits between the driver tube and the output tube cathode. A decision must be made to put the fet on the driver board, the output board, or both, just populate the one you choose.
Putting the mosfet on the output board allows for using multiple output boards on a single driver board. I need this if I ever build the kilowatt audio amp that I have been talking about for 10 years.
Tubelab is not my first attempt at selling my hobby creations. There have been a few others. I learned back in the late 70's making boards for SS-50 bus computers that snap off boards will be snapped by the USPS in transit. Not an option unless they are part of a kit in a big box where the board can be protected.
41 years at Motorola taught me a lot about having too many stock numbers. That's why I don't offer parts kits any more. With tube amps there are just too many choices.
We also learned a lot about minimizing assembly mistakes in 41 years. The wiring needed to interconnect multiple circuit boards is a BIG opportunity to create defects.
See post #76 here:
Mouser delays
Lucky Yellow
(a) "Just using the part of the board I need and leaving the rest"
I agree, and I'm doing that with a pair of TSE-II boards. HOWEVER, with the big boards it does consume a lot of real estate.
(b) "snap-off" boards; Pete Millett has those for an OpAmp experimenters board, IIRC. I'm not a big fan (nor apparently is UPS), but if it got the cost down significantly for say a PowerDrive board, I'll take it.
(c) "Yellow" My saw-sharpening shop sells all manner of high-end woodworking and lawn/tree care tools. They tell me that Dewalt hit a home-run with their trademark YELLOW color because Amarillo is considered a "lucky color" amongst Latinos.
I have NO IDEA if this is true, but why take chances.
Why Latinos eat grapes and wear yellow underwear on New Year's Eve | CTV News
BTW, decades ago KODAK won a so-called "Trade Dress" intellectual property case against Fuji I think for the yellow color of their FILM boxes. Fun Facts from my earlier life as an Intellectual Property guy.
A nice clean yellow (NOT mustard or baby-puke yellow) would be both distinctive board color AND easier for old eyes to work on.
(a) "Just using the part of the board I need and leaving the rest"
I agree, and I'm doing that with a pair of TSE-II boards. HOWEVER, with the big boards it does consume a lot of real estate.
(b) "snap-off" boards; Pete Millett has those for an OpAmp experimenters board, IIRC. I'm not a big fan (nor apparently is UPS), but if it got the cost down significantly for say a PowerDrive board, I'll take it.
(c) "Yellow" My saw-sharpening shop sells all manner of high-end woodworking and lawn/tree care tools. They tell me that Dewalt hit a home-run with their trademark YELLOW color because Amarillo is considered a "lucky color" amongst Latinos.
I have NO IDEA if this is true, but why take chances.
Why Latinos eat grapes and wear yellow underwear on New Year's Eve | CTV News
BTW, decades ago KODAK won a so-called "Trade Dress" intellectual property case against Fuji I think for the yellow color of their FILM boxes. Fun Facts from my earlier life as an Intellectual Property guy.
A nice clean yellow (NOT mustard or baby-puke yellow) would be both distinctive board color AND easier for old eyes to work on.
I like the idea of modular boards myself. Gives more chassis options and opportunity to get creative. From a practical business decision standpoint if you make an all in one UNSET board I will buy one. If you make a modular UNSET board and power supply I will buy an UNSET and 2 power boards (maybe more) to use with the UD as well.
How about "Kickstarter" for the UNSET?
Hey, George, have you looked into "Kickstarter" to finance the UNSET board?
I have no idea now it works, or whether the overhead cost is too high, but it seems like a good way (at least in theory) to "bring forward" the revenue stream.
Plus, folks will over-pay a bit for the opportunity to be (a) early adopters and (b) have input into the design/product development process; this might be a way to "monetize" (like the kids say) what you're now giving away for free.
The UNSET looks like it might be a good candidate.
Hey, George, have you looked into "Kickstarter" to finance the UNSET board?
I have no idea now it works, or whether the overhead cost is too high, but it seems like a good way (at least in theory) to "bring forward" the revenue stream.
Plus, folks will over-pay a bit for the opportunity to be (a) early adopters and (b) have input into the design/product development process; this might be a way to "monetize" (like the kids say) what you're now giving away for free.
The UNSET looks like it might be a good candidate.