Scott Grammer

I finally put up the pictures you asked for. Check out the "Pictures of your tube amplifiers" section.

In the past, I have built and re-built a number of tube amps, but these two are the ones I currently own. The first two pictures are of an amp built on the chassis and iron of one of the cheap Chinese amps sold on Amazon and FleaBay. I acquired one of these third hand, and like its previous two owners, I was not in love with the sound of it. You can see what I did to it by checking out my YouTube series about it here: https://www.youtube.com/playlist?list=PLUD-UYYvvR59Mrh2ON8Qpkb0aJceN3OL8. (Dear moderators: If I have erred in placing this link, please forgive me, I'm new here.) The second two pictures are of what I have always termed the "baby amp." It's a 2.5WPC, all-triode, class-A, single-ended amp with no negative feedback and about as simple a schematic as possible. It uses a 5AR4 rectifier, a 6SL7G voltage amp, and a 6AS7GT output tube. It has very little gain, but it sounds really good with my Cornwalls, possible due to the very low plate impedance of the output tube. I built it in 1995, and it has not been changed since, save for replacing the original output tube which sadly went to air sometime in the past. For a sense of scale, the width of the chassis is only 10 inches (25.4cm).

Those are gorgeous!

Been there, done that, she got the T-shirt. And my vinyl. And my Luxman preamp. And....

I can make that happen. Thanks!

By the way, I kind of like the blue Edcor transformer....

It's good to see folks still building their own amps. I used to do more of that, 20+ years ago, now I just don't have time. I did recently get hold of one of the cheap Chinese amps they sell on Amazon and eBay, and after stripping it to bare metal, I built it back the way I wanted it. I'm quite happy with it. Single-ended, class-A, EL34 outputs, 6SN7's as driver and gain stage. It does about 7WPC, and I ran my JBL L80t's with it for several months before getting my '68 Cornwalls out of storage. I still have only one of the many amps I built back in the day. A tiny, 2.5 WPC job with just three tubes: a 5AR4, a 6SL7GT, and a 6AS7G. Built it back in 1995. I like it a lot, but it has very little gain and must be driven with a preamp.

My DAC is actually a small audio interface - an AudioBox Go made by PreSonus. Currently, amplification is provided either by my 1976 Nikko NA-850, or by one of two custom-built (by me) SE Class-A tube amps.

Ahh. OK, thanks. A rag dampened with denatured alcohol cleaned them right up, easy-peasy. Stuff was going everywhere!

OK, a bunch of stuff done, and some things figured out. I cleaned the oxidation off the tweeters and other places, checked all six drivers (all good), and added crimp-on connectors to all the wires connecting to the crossovers, to get better, more secure connections. I tried changing what taps on the autoformer the mids connect to, and was not happy, so I put them back. After reassembling the speakers (again), I stopped tinkering with them for a while and sat down with the schematic of the crossover, a calculator, and an ice-cold Coke. A bit of figuring with the old TI-30 led me to the realization that this system's impedance in the mid frequencies must be on the order of 64 ohms (!), and about 16 ohms in the top octave or so. Since the woofer has a DCR of about 3.5 ohms, and the woofer choke adds a tad to that, the low frequencies must average (not counting the two impedance peaks that occur in any vented system) about 6 ohms. That's quite a range of impedances. On an amp with tons of negative feedback, it would not matter much. But my SE tube amp has only 6dB of NFB in the output stage, and that can be dialed back to zero. The result is that it has a high output impedance, and this means that its output would fall off in the bass where the load impedance is 6 ohms, and would go up in the midrange where the load is over 60 ohms. And that, dear readers, is EXACTLY what I've been hearing. A quick hookup this morning to my 60WPC Nikko integrated amp verified my suspicions - the speakers sound very different, and the bass is excellent. So now, I have a decision or two to make....

I agree. I do a LOT of recapping in my job, (mostly electrolytics), and when an inexpensive cap meter shows the capacitance going up, it's often due to leakage fooling the meter into thinking the cap is bigger than it is. I may do that B2 crossover after all, not only for the improvement in sound, but to replace the 55-year-old caps. I did a bit of work on both speakers, I'll post that a bit later. Right now, it's dinner time!

6.4 ohms per my component tester, 6.0 ohms per my ohm meter. That's just the right speaker, I have not re-opened the left speaker yet.

I've opened up the right cabinet (the previous pics were from the left) and found some interesting things. One, a really crummy-looking solder connection on the woofer (see pic). I'll fix that later. Two, the tweeter magnet is covered in a thick layer of yellow oxide which is very loosely attached, and merely touching the tweeter knocks most of it off. The screws that hold the terminal strip to the crossover, as well as one lug on the autotransformer, are similarly oxidized. The drivers all check fine for DCR and inductance, and they were wired properly in terms of phase. The connections to the autoformer match the schematics of the 'Type B" crossover. The caps both test a bit high in value, but their ESR and VLoss are fine. The woofer choke measured 2.36mH with a DCR of 0.6 ohms (about .1 ohms of that is the leads of the meter). Finally, the caps on this crossover, one a Sprague and the other a Mallory, match the crossover in the other speaker. I wonder if these were built from Klipsch components by the company that has stamped their name all over the inside of them? Comments, anyone?

I already did the retightening bit. It was a necessity as when I first hooked them up, one woofer was not even playing. Yes, I'm still in the same place. If you look at my profile page you'll find my website, which has my shop address. Tomorrow and Monday I will be knee deep in these things, taking them mostly to bits to check everything, including the cap checks you suggest.

I don't know if they're original or not. Tomorrow, when I open up both cabinets, I can compare the two crossovers to see if they're at least using the same components. I'm glad you noticed that - I had not.

So after a long day of sweating over a hot soldering iron 😆, I came home to tinker a bit with my Cornwalls. I've been experimenting with a piece of equalization software called "Equalizer APO 1.3." It runs in Windows and gives many options for EQ'ing the digital audio leaving your computer on its way to a DAC. I discovered that a simple 3dB dip, covering almost the entire range of my midrange horns, works a pretty treat. It's not exactly the sound I'm hoping for, but it's a big improvement. Tomorrow, I will be taking things apart, and I may well move the midranges down one notch on the autoformers to see what happens. The EQ software is only for experimentation, not for permanent use. It won't help when I play vinyl, for instance. Attached is the curve I've programmed into the EQ, for your edification and consideration.

Not yet. That's coming up this weekend. I'll be taking the speakers most of the way apart to check things. Pictures and readings will be posted once I've done that. And for me, the weekend is Sunday and Monday.

Will do. Already had to tighten screw terminals on the crossovers, as they were loose and causing trouble as soon as I hooked the speakers up.

Well, I did come here asking for help, so all ideas are appreciated!

This weekend (for me, the weekend = Sunday and Monday), I plan to pull the crossovers and check connections, and I may pull the mid drivers to measure DCR and take some pictures and make sure they're original. So far, the tweeters are the only parts I'm not worried about. They're playing, they're undistorted, and sound pretty much the same in both cabinets. Of course, I can only hear to about 12.5 or 13 kHz now, unfortunately.

I checked the woofers a few days ago, at the cabinet speaker terminals, which f course means that the reading includes the DCR of the woofer chokes. I remember both of them being right at four ohms DCR, but I don't remember the exact figures.

The deep bass is not what I remember it being, and not comparable to the JBL L80t's the Cornwalls replaced. A little overbearing around 400Hz or so. Possibly a panel resonance. Very clean, though, and even on limited power they can play quite loudly, as you would expect.

The midrange is indeed too loud. I have not re-opened the cabinets to check the connections to the autoformer. They may well be wrong. I do know that when I put the speakers back in service recently after having stored them for a very long time, I had to tighten the screws on the terminal strips of the crossovers, as several of the connections there were loose. I plan to actually remove the crossovers this weekend, check the connections to the autoformer taps, and check the capacitors for value, leakage, esr, etc. I'll also check the connections to the drivers for correct polarity, and the mids and tweeters for DCR. All this, I hope, will give me a clue as to what's going on. The speakers definitely sound different to my memory of them (a LOT different), but then I stored them 20 years ago, and my ears are now 59 years old, so....

Any suggestions as to how to test to see if the magnets are weaker than normal, short of disassembling them to put a gauss meter probe in the gap? I don't cherish the thought of spending $300 to find out there was no appreciable change. I don't need the power handling, as I drive them with a 7WPC SE tube amp.... Thanks for all input on this. Maybe I can return the favor someday on a topic I know more about.