It's back to the drawing board for me.
The 6SF5s I intended to use for the input/driver stage aren't working out. I tried using various topologies and peripheral components, but I don't know enough to fine-tune a circuit to circumvent the shortcomings. ( Thanks, by the way, to the previous posters who explained why the *SF5s were a poor choice. Saves me from wasting any more effort on these tubes at this time ).
So now I'm going to explore a couple of very basic circuits: a 12SC7 pentode driving a 12W6, or a 5964 common-cathode input driving a 6AQ5 output. Either of these should be a viable choice; once they're working, I can begin to experiment with tone-controls and intermediate stages.
I'm beginning to understand just how subjective guitar amp design can be. A 'hi-fi' reproduction amplifier basically aims for two goals: no noise, and no distortion -- the mythical 'wire with gain'.
Instrument amps are intended to color and distort the signal -- but how, when, and how much are black arts subject to much discussion. It's something like the difference between forensic photography and art photography. The whole genre of guitar amps, and the designs that became timeless classics, makes me appreciate the intellect and ear of people like Fender and Baxandall much more than I did previously.
The 6SF5s I intended to use for the input/driver stage aren't working out. I tried using various topologies and peripheral components, but I don't know enough to fine-tune a circuit to circumvent the shortcomings. ( Thanks, by the way, to the previous posters who explained why the *SF5s were a poor choice. Saves me from wasting any more effort on these tubes at this time ).
So now I'm going to explore a couple of very basic circuits: a 12SC7 pentode driving a 12W6, or a 5964 common-cathode input driving a 6AQ5 output. Either of these should be a viable choice; once they're working, I can begin to experiment with tone-controls and intermediate stages.
I'm beginning to understand just how subjective guitar amp design can be. A 'hi-fi' reproduction amplifier basically aims for two goals: no noise, and no distortion -- the mythical 'wire with gain'.
Instrument amps are intended to color and distort the signal -- but how, when, and how much are black arts subject to much discussion. It's something like the difference between forensic photography and art photography. The whole genre of guitar amps, and the designs that became timeless classics, makes me appreciate the intellect and ear of people like Fender and Baxandall much more than I did previously.
I just stumbled onto this web page: Pipsqueak pentode DIY guitar amp
Looks like a neat design, and his links to other successful low-power designs should get me back on track.
Looks like a neat design, and his links to other successful low-power designs should get me back on track.
Attachments
Last edited:
I connected my first prototype amp up to each of the speakers that I have purchased in the past few days. So far I am rather underwhelmed. I don't know if the amp just sucks, of if 2 watts just isn't enough to get any of these speakers to really sing.
There were a few limitations that could have affected the outcome. None of the speakers were mounted, just laying on the carpeted floor. My amp is running on a 130 volt supply, I think I will connect it to a variable supply and try a bit more voltage.
After performing this rather uninspiring experiment I connected the preamp section of my prototype amp into the CD player input of a Tubelab Simple P-P. I connected a speaker to each channel so I could switch between channels easilly. 20 Watts changed the picture. The $20 Peavey Blue Marvel started to really scream. The Jensen Alnico also sounded rather sweet. Both were 8 inch speakers and I could get loud enough to get some feedback going. The 6 inch Jensen Mod didn't impress me much. I got two 6 inch Dayton Audio "PA Drivers". These are heavy well built speakers that claim a 75 watt RMS / 125 Watts max.....so.....I fed them 125 watts courtesy of Petes red board. These speakers are the hot ticket if you want maximum loud for a small $20 speaker.
I think I will try an amp design much like the "two amps wired together" design that I was playing today.
There were a few limitations that could have affected the outcome. None of the speakers were mounted, just laying on the carpeted floor. My amp is running on a 130 volt supply, I think I will connect it to a variable supply and try a bit more voltage.
After performing this rather uninspiring experiment I connected the preamp section of my prototype amp into the CD player input of a Tubelab Simple P-P. I connected a speaker to each channel so I could switch between channels easilly. 20 Watts changed the picture. The $20 Peavey Blue Marvel started to really scream. The Jensen Alnico also sounded rather sweet. Both were 8 inch speakers and I could get loud enough to get some feedback going. The 6 inch Jensen Mod didn't impress me much. I got two 6 inch Dayton Audio "PA Drivers". These are heavy well built speakers that claim a 75 watt RMS / 125 Watts max.....so.....I fed them 125 watts courtesy of Petes red board. These speakers are the hot ticket if you want maximum loud for a small $20 speaker.
I think I will try an amp design much like the "two amps wired together" design that I was playing today.
Hear, hear. Cheap pentodes, PA iron, a cheap 10K isolation transformer (phase splitter) and a pair of mutilated wall warts will be the power stage.
All the $$ will disappear in switches and pots in the preamp.
Too true. So to look at it from another perspective, start with the $100 budget, subtract the costs of pots, switches, knobs, jacks, power cords and receptacles, the fuse holder and fuse, and tube sockets. Then when you see whats left you can start making an amp. The 2 or 3 transformers alone are the deal breaker if you just consult a conventional transformer catalog.
If some creativity is used in the power supply and output stage there should be enough $$ left for a single channel preamp with a volume, bass and treble control, maybe a mid and master gain. The challenge is making an output stage and power supply for low $$ that doesn't suck.
If this is to be a serious design effort, start with a spreadsheet. No matter what you are building it will have at least a power cord, a fuse, an input jack, a power switch, and at least one pot with a knob. You are likely going to need a speaker jack unless you are making a combo. List these and other fixed costs in one section. Then make 3 more sections, output stage, power supply, and preamp stages. There are fixed costs in each section, things like switches, transformers, tube sockets and pots. Start with your best guess at what parts you are going to use and put in the costs. You can make a guess at thinks like caps and resistors. I broke it down to small medium and large for resistors and caps and put guesstimates in for them. The spreadsheet has columns for several different designs so I can play with the numbers. You can then see where your budget is best spent, create a few viable paper designs, then start building. As you test different things you may find that the little multi tapped Chinese power transformer sucks as an OPT, so your $4 OPT goes off the list.
My spreadsheet now has 4 pages for 4 totally different designs. There are several different variations of each design. It looks like the difference in cost from a 2 watt amp to a 20 watt amp isn't much since the fixed costs are more than half the budget.
I expect the same spread sheet will keep growing as I expand it to look into guitar amps with a bigger than $100 budget. I also look at things like PC board area since that can also be a cost factor if the amp ever becomes a product.
My surplus $4.50 OT (filament xfmr) got bumped up to twice that price after I bought some to try. (I'm using two of them in a Crowhurst Twin mode to get the primary Z up) They must not have sold one for the last 20 years and decided the price was too low.
Seems that most of the surplus xfmrs I see are priced at, at-least, 1/2 the new price for single quantity. I even see some priced near full single quantity new price (cheaper to buy ten new ones). Surplus stuff used to be priced low enough for some wiggle room, to work around the mismatch to your application,... not any more. When you factor in the low quantity likely available as surplus, and the lack of specs for our application, these look like a dead end. Too bad, because the $4.50 OTs amazingly measure good from 50 Hz to 30 KHz in Twin mode!
That Speco T-7010 doesn't have any frequency specs given. (and a center tap that is off center) The T-7020 at least says it's "HiFi", but still doesn't give any specs. It would have too low a primary Z to use anyway, but might work in a Crowhurst Twin mode.
The Edcor approach is looking better all the time, except for delivery time. If I get one to try though, I want a better GXPP model so the amplifier can be uprated later, so this breaks the $ limit for me still. Not to mention that it would likely arrive while I'm on vacation. So I'm only in the contest in spirit now.
Seems that most of the surplus xfmrs I see are priced at, at-least, 1/2 the new price for single quantity. I even see some priced near full single quantity new price (cheaper to buy ten new ones). Surplus stuff used to be priced low enough for some wiggle room, to work around the mismatch to your application,... not any more. When you factor in the low quantity likely available as surplus, and the lack of specs for our application, these look like a dead end. Too bad, because the $4.50 OTs amazingly measure good from 50 Hz to 30 KHz in Twin mode!
That Speco T-7010 doesn't have any frequency specs given. (and a center tap that is off center) The T-7020 at least says it's "HiFi", but still doesn't give any specs. It would have too low a primary Z to use anyway, but might work in a Crowhurst Twin mode.
The Edcor approach is looking better all the time, except for delivery time. If I get one to try though, I want a better GXPP model so the amplifier can be uprated later, so this breaks the $ limit for me still. Not to mention that it would likely arrive while I'm on vacation. So I'm only in the contest in spirit now.
Solved the farting out problem. GS resistors were too small.
The Peavey $15 surplus speaker is even worse than the Blue Marvel.
That said, probably some of the problem is that I'm driving it with a fair impedance miss-match since the P-T31 is 5K to 8 Ohm and the small tubes I'm using probably need a 8-10K plate load. Throw in the speaker at 4 ohms and the output is seriously miss-matched.
In addition, I suspect the Blue Marvel needs some break in time with the stiff surround it has.
I May try two op tubes in parallel.
The Magnavox 12" sounds pretty nice. Good efficiency and smooth frequency response (my -3dB low end is at 160Hz and may help with any resonance.
The amp sounds best with the G12M (not surprising considering my bias).
I still need to dig out the Weber VST Sig10F to see how it sounds.
Still chasing the oscillation. It shows up with just the last two stages connected and nothing driving it, except the volume control is cranked up almost all the way.
The Peavey $15 surplus speaker is even worse than the Blue Marvel.
That said, probably some of the problem is that I'm driving it with a fair impedance miss-match since the P-T31 is 5K to 8 Ohm and the small tubes I'm using probably need a 8-10K plate load. Throw in the speaker at 4 ohms and the output is seriously miss-matched.
In addition, I suspect the Blue Marvel needs some break in time with the stiff surround it has.
I May try two op tubes in parallel.
The Magnavox 12" sounds pretty nice. Good efficiency and smooth frequency response (my -3dB low end is at 160Hz and may help with any resonance.
The amp sounds best with the G12M (not surprising considering my bias).
I still need to dig out the Weber VST Sig10F to see how it sounds.
Still chasing the oscillation. It shows up with just the last two stages connected and nothing driving it, except the volume control is cranked up almost all the way.
"I May try two op tubes in parallel."
Try some $1 sweep tubes. They can handle the current for low Zprimary impedance. The 6AQ5 has a DC max rating of only 40 mA. The little 6GF5 has a whopping 160 mA rating. Thats what I'm using, and its pin compatible with 12GE5 and 21HB5 (all $1). I can uprate my little amp to 32 Watts (OT limited, 8 and 16 volt sec windings, otherwise it would go to 75 Watts!) with just a tube change.
The other approach for getting a higher Zpri is either use Circlotron mode or Crowhurst Twin coupled (two OTs needed). These both quadruple the Zpri. But you need a real center tap to do the cross coupled capacitors for the Twin.
By the way, this is how my lowly $4.50 filament OT got its 30 KHz top end. It only measures 8 KHz top end in conventional mode (and badly lopsided freqs for the two primary sides besides, like many OTs actually). But with two of them in Twin mode and cross coupled primary caps, all four primary side windings get used by each tube, quadrupling the high end and matching the two tube sides besides. From junk to real HiFi like magic. Crowhurst's Twin Coupled is very much under-appreciated.
Try some $1 sweep tubes. They can handle the current for low Zprimary impedance. The 6AQ5 has a DC max rating of only 40 mA. The little 6GF5 has a whopping 160 mA rating. Thats what I'm using, and its pin compatible with 12GE5 and 21HB5 (all $1). I can uprate my little amp to 32 Watts (OT limited, 8 and 16 volt sec windings, otherwise it would go to 75 Watts!) with just a tube change.
The other approach for getting a higher Zpri is either use Circlotron mode or Crowhurst Twin coupled (two OTs needed). These both quadruple the Zpri. But you need a real center tap to do the cross coupled capacitors for the Twin.
By the way, this is how my lowly $4.50 filament OT got its 30 KHz top end. It only measures 8 KHz top end in conventional mode (and badly lopsided freqs for the two primary sides besides, like many OTs actually). But with two of them in Twin mode and cross coupled primary caps, all four primary side windings get used by each tube, quadrupling the high end and matching the two tube sides besides. From junk to real HiFi like magic. Crowhurst's Twin Coupled is very much under-appreciated.
Last edited:
I finally worked out a preliminary design using a 12SJ7 and a 12V6. Both tubes are available in quantity, and relatively inexpensively ( under $7 NOS ). I took my hand-drawn schematic to a local tube repair guru for his comments. Ed primarily restores antique radios, but has also done a fair amount of amplifier repair over his 50-year career.
Ed laughed and said, "Well, you've basically got a design that looks a lot like a Gibson GA-5 Les Paul Jr., or a Fender 5C1 Champ." He pulled out the schematics, and sure enough, they were virtually the same.
On one hand, I'm pleased that I was able to 'reinvent' this design on my own. My expertise is in photochemistry, not electronics, and it looks like I've finally reached a fair understanding of basic tube circuits.
On the other hand, I'm chagrined that it took me many hours and a lot of head-scratching to come up with a very simple amplifier circuit which was already done over sixty years ago.
It looks like the best path for me, with my limited knowledge, is to implement an existing design, but still beat the $100 price-point. This may involve cheap alternate tubes with the same characteristics (e.g. 7EY6 instead of 6V6, or 10GK6 instead of 6BQ5), or novel sources and uses of transformers.
For now, I'm going to put aside my datasheets and TubeCad printouts, and make a simple wooden cabinet to house the 8" speaker I plan to use. Once I'm done with the speaker enclosure, then I'll start breadboarding the circuit, and see which tweaks to the parts values sound best with the speaker I'm using.
***********************************************
On another subject: There seems to be a time-crunch for several people involving shipping delays and 'real-life' obligations. If there is general agreement that we should extend the time allowed for completion of the amps, let's do so. My hope is that this thread will ultimately produce a good, inexpensive amp which anyone with basic skills could successfully duplicate.
If there's no objection, how about pushing the deadline out a couple months, perhaps to the end of October, or beyond?
Ed laughed and said, "Well, you've basically got a design that looks a lot like a Gibson GA-5 Les Paul Jr., or a Fender 5C1 Champ." He pulled out the schematics, and sure enough, they were virtually the same.
On one hand, I'm pleased that I was able to 'reinvent' this design on my own. My expertise is in photochemistry, not electronics, and it looks like I've finally reached a fair understanding of basic tube circuits.
On the other hand, I'm chagrined that it took me many hours and a lot of head-scratching to come up with a very simple amplifier circuit which was already done over sixty years ago.
It looks like the best path for me, with my limited knowledge, is to implement an existing design, but still beat the $100 price-point. This may involve cheap alternate tubes with the same characteristics (e.g. 7EY6 instead of 6V6, or 10GK6 instead of 6BQ5), or novel sources and uses of transformers.
For now, I'm going to put aside my datasheets and TubeCad printouts, and make a simple wooden cabinet to house the 8" speaker I plan to use. Once I'm done with the speaker enclosure, then I'll start breadboarding the circuit, and see which tweaks to the parts values sound best with the speaker I'm using.
***********************************************
On another subject: There seems to be a time-crunch for several people involving shipping delays and 'real-life' obligations. If there is general agreement that we should extend the time allowed for completion of the amps, let's do so. My hope is that this thread will ultimately produce a good, inexpensive amp which anyone with basic skills could successfully duplicate.
If there's no objection, how about pushing the deadline out a couple months, perhaps to the end of October, or beyond?
Last edited:
Make sure that you have the metal case of the volume control grounded. Guess what I was chasing when I found that the oscillation stopped when I touched the metal knob.
I put amp 1.0 aside for now. It's kinda dead. Amp 1.0 was based on old radio tubes that run from a 150 volt max supply. It sounded best right before it blew up on 320 volts!
I started building amp 2.0. Amp 2.0 is based on old TV tubes.
The parts in both designs were chosen based on world wide availability and at least 1000 of everything in stock somewhere.
Attachments
The Fender 5C1 is what got me started building guitar amps. That was almost 50 years ago. 6SJ7's were everywhere in dead radios. 6V6's were a bit less common, but 6BQ6's were lying in the dirt in the trash dump since about half the old black and white TV's used them.
The 6AU6 or 6CB6 (little 7 pin tubes) and about a zillion other tubes will work for the 6SJ7. The 6EZ5 is still available in at some places for $1 (vacuumtubes.net). It will plug right into a 6V6 socket in some circuits and may require a little bias tweak in others. I was keeping this one quiet, but I have mentioned it before.
I am about halfway done with my cabinet for an 8 inch speaker. I am not even trying to make this one pretty since I know I will do it over before I am done.
No argument here. Sherri's mother will likely pass away during the next month or two bringing all progress to a halt for a week or 2 or 3. I am hoping to have most of the coming 3 day weekend for science.
It does look like some of us are making progress, and I would like to test out a few more ideas before choosing 1 or 2 designs for contest purposes.
I have already decided that contest or not, I am going to make the guitar preamp that I have started and put aside several times over the last few years. That's why I have been looking into super cheap output stage and power supply designs, so I can make a righteous preamp. I have already figured out how to wire it to a SSE board to make a 50+ watt P-P amp.
Unfortunately the 6EZ5 inventory is all gone from Vacuumtubes.net at least the last time I asked they were out. I wanted to try a few in my SSE.
I was afraid of that. I have access to ESRC's inventory database and it looks like they sold about 1200 of them last year when they were on the $1 list. I got 10 and they do work nice in the SSE. There might be a few left but they are now $4 each. We mention a tube here and some %#&hole buys them all. It has happened too many times. I will not mention any more numbers until it is necessary.
Aug 15 is not close yet. It takes at least 1 month to get a standard catalog transformer from Edcor IF you are in the USA. Even guys in Canada will not have a transformer from Edcor by Aug 15 if it was ordered today. Edcor makes probably the best low cost transformers availabale today, but I didn't even look at them due to the delivery times. They are also one of the few vendors that will use a flat rate USPS box to ship overseas.
I never heard of a 1A2 but Frank says it is a Russian tube. Can you get those here in time to design an amp by Aug. 15?
Maybe 1A2 was meant as a joke. I would have suggested 6BK4, but amazingly those things are priced thru the roof now. Used to get them cheap. Must be some HV DIY project that ate them up.
George, you need to pronounce some beam triode as the ultimate SET tube, and then maybe we'll see them show up by the thousands in some Wally Mart, made in China amps.
George, you need to pronounce some beam triode as the ultimate SET tube, and then maybe we'll see them show up by the thousands in some Wally Mart, made in China amps.
Last edited: